PAEDIATRIC FORENSIC MEDICINE AND PATHOLOGY
: dited by
Anthony Busuttil Em eritus Regius Professor of Forensic Medicine, University of Edinburgh; and Medical Director, Fo rensic Medical Services, NHS Lothian, Edinburgh, UK
Jean W Keeling _::ormerly Consultant Paediatric Pathologist, Royal Hospital for Sick Children, ::diilburgh, UK
i, ARNOLD HODDER ?ART OF HACHETTE LIVRE UK
I
CONTENTS
I
Contributors
xiii
Preface
xiv
Acknowledgements
xv
ist of abbreviations used
1
Clinical assessment in suspected child abuse
xvi
1
Helen Hammond
2
Introduction Alerting signs Types of abuse The need for comprehensive assessment Significant harm The interagency context (flow chart of process) Joint working and the complementalY skills of paediatricians and forensic specialists
Legislation Consent and confidentiality The process - joint paediatric/forensic examination Documentation and report writing Interpretation of the findings Formulating an opinion Ongoing health care Involvement in ongoing legal and child -care processes References
17
21
21
22
22
Investigation of suspected sexual abuse
24
Jacqueline YQ Mok
Introduction The colposcope in the medical examination Forensic evidence Skills and experience required Consistent vocabulalY Normal female genital anatomy Perianal findings Acute, healing and healed anogenital trauma Female genital findings in sexual abuse Signs of anal abuse Conditions that mimic abuse Screening for sexually transmitted infections Interpretation of clinical and laboratOlY findings SummalY References
1
2
3
4
4
5
6
7
8
8
16
24
27
27
28
29
29
36
37
38
39
40
41
42
43
44
vi I
Contents
3
Radiology of child abuse Maeve McPhillips
Role of the radiologist Radiological investigations
Skeletal injuries
Head injury Visceral injuries Soft-tissue injury Differential diagnosis Conclusion References
47
Haematological abnormalities that can simulate abuse
76
4
47
48 51
60
68 69 69 73 73
Angela Thomas
Introduction
Primary haemostasis Secondary haemostasis
Laboratory tests
Measurements of primary haemostasis Evaluation of a bleeding patient Patterns of abnormal results
Normal coagulation screen with a normal platelet count Abnormalities of platelet number or morphology Coagulation defects The neonate
Drugs associated with bleeding Bone marrow failure syndromes Systemic disease associated with a bleeding tendency Activation of coagulation Conclusion References
5
Biochemical investigations on post-mortem specimens
76 78
79 81
82
82
86
90
94
96
97
98
99
100
101
101
102
106
Denis R Benjamin
Introduction General evaluation Hypoxia Inflammation
Anaphylaxis Infection
Dehydration and electrolytes Time of death (post-mortem interval) Endocrine disorders
Genetic metabolic disorders presenting as sudden unexpected death Technical considerations at the time of autopsy References
6
106 107 109 109 109
110
110
111
112
114
117
120
125
Ocular involvement in non-accidental injury Harry Willshaw
125 125
Introduction
Scope of ocular and adnexal injury
' .-'
,
-----. .
Contents I
Fundus haemorrhages References
vii
128
134
7 The death scene following the sudden death of a child Anthony Busuttil Introduction Scene management The crime scene manager Sequence of events at the death scene Unclothing the body A good look around Sudden infant death syndrome or non-sudden infant death syndrome External petechiae Bruising Abandoned neonates Deaths from trauma Dyadic and multiple deaths Sudden deaths of older children Sensitivity and stress of the investigation Inquests and inquiries References
137
8 Post-mortem examination in babies and children Jean W Keeling
Introduction Death scene investigation Rectal temperature Medical and family history Other important information Radiological examination Photography Microbiological samples Toxicological investigations Biochemical and metabolic investigations Weights and measurements External examination Estimating blood loss Dissection (infants and older children) Examination of the brain, spinal cord and eye Examination of the newly born Histological samples Retention of organs Exchange of information and multidisciplinary review References
145
9
166
Pathology of neurological abnormality in early life Waney Squier Introduction Clinical manifestations of early brain damage: cerebral palsy Timing of injuries by histology
137
137
138
139
139
139
140
140
140
140
141
141
142
142
143
143
145
145
146
146
146
146
147
148
149
149
150
150
152
152
156
158
162
163
163
164
166
167
167
viii I Contents
Acquired intra-uterine damage Birth-related injury Stroke in the developing brain Metabolic disorders Infections References
169
173
]76
177
177
178
10 Fetal and perinatal death
180
Jean W Keeling
Introduction Definitions The law Background information Concealed pregnancy Unattended delivery Was the baby born alive? Is the baby of sufficient maturity to survive? Is there evidence of prolonged or difficult labour? Are there any significant injuries? Fetal death foJlowing maternal injury Is there a natural cause for death? Can I give a cause of death? Should the intrapartum still birth be a medicolegal autopsy? References
180
180
181
182
182
182
183
187
187
]88
190
193
193
194
195
11 Sudden unexpected death in infancy: sudden infant death
198
syndrome or something else?
Jean W Keeling
Introduction The definition of SIDS Epidemiology Sleeping environment Pathological findings in sum Death certification References
198
199
201
203
205
2]8
219
12 Sudden natural death in infants and children
225
Dick Variend
Introduction Cardiovascular causes of sudden death X-linked hypohidrotic (anhidrotic) ectodermal dysplasia Intracranial haemorrhage, neoplasms and malformations Gastrointestinal causes Fatal anaphylaxis Sickle cell disease Haemorrhage as a cause of sudden death Respiratory causes of sudden death Epilepsy and sudden death Deaths from acute asthma
' -
.
------
225
226
235
236
239
239
240
240
240
242
243
.... .
Contents I
Diabetes mellitus
Genetic metabolic disorders
Other bacterial infections
Deaths related to obstetric events and premature birth
Miscellaneous causes of sudden natural death
Sudden unexplained death in older children
Sudden natural death in the early neonatal period
Sudden death associated with 'intermediate' pathology
References
13 Recent advances in paediatric toxicology
ix
243 244 247 248 248 249 249 249 250 256
Patrice Mangin and Christian Giroud
Scope of the problem Specificity of paediatric toxicology Techniques used in drug testing Special techniques for analysis of volatile substances Alternative specimens for drug testing Pitfalls and limitations of drug screens Specific applications The importance of paediatric toxicology in specific cases Conclusions and future considerations in forensic paediatric toxicology References
256 258 259 262 262 267 271 274 274
275
14 Head and neck injuries Robert A Minns and TY Milly Lo Definition
Epidemiology
Non-accidental head injury
Traumatic birth injury
Primary mechanisms of injury to the brain
Secondary mechanisms of brain injury
Injury to the cervical spinal cord
Genetic influence on recovery from traumatic brain injury
References
282
15 Heat-induced injury or death
318
282 283 294 300 302 307 311
312 313
Anthony Busuttil
Introduction
House fire deaths
The pathologist's role
References
318 318 319 327
16 Asphyxial deaths in children
329
Anthony Busuttil
Petechiae
Scene of death
Traumatic asphyxia in children
Entrapment asphyxia
Foreign body inhalation
329 330 330 330 330
xI
17
Contents
Plastic bag asphyxia Overlaying and wedging Strangulation Hanging by a ligature Drowning and near drowning Imposed airways obstruction Abuse of inhalants (solvent abuse) Reverse suspension Chemical asphyxia Prevention References
331
331
331
332
332
332
333
333
333
333
334
Accidental injuries in children
336
Anthony Busuttil
18
Overview of paediatric trauma Bicycle helmets Falls Playground injuries Sports injuries on snow and ice Riding injuries Agricultural injuries Prevention Older children and substance abuse Accidental poisoning Hypersensitivity References
336
338
339
340
340
341
341
341
341
342
342
342
Drowning and near drowning
345
John Pearn
19
Introduction The causes of childhood drowning: a perspective The drowning medium The pathophysiology of drowning Forensic immersion syndromes References
345
345
346
349
351
359
Sudden death of children in hospital
362
Jem Berry
Introduction Definition and frequency Deaths due to natural disease Deaths due to failure to monitor Therapeutic misadventures Deaths due to dmg treatment Deaths due to medical devices and procedures Deaths in the dental chair Sudden death in newborn babies Accidents Suicide Filicide and homicide in hospital
362
362
363
365
366
366
368
371
371
372
373
373
,~------~-- . ,
.
Contents I
Investigation of sudden unexpected death of children in hospital References 20 Road traffic accidents in children
xi
375
377
385
Anthony Busuttil
Road traffic fatalities Investigation of a fatal road traffic collision Vehicular collisions Other supervening problems in collisions Pedestrian injuries Child cyclists Diffuse axonal injury Whiplash injuries Injuries to children in utero Other vehicular accidents References
385
387
390
391
392
392
393
393
393
393
394
21 Forensic DNA profiling in cases involving children Alex M Graham and David J Harrison
Introduction Inheritance of genetic material Forensic DNA analysis: history and techniques Sample collection and processing DNA evidence and child sexual offence Y chromosome short tandem repeat typing Mixed samples Additional sample problems and solutions Mitochondrial DNA Paternity testing Identification of body remains and missing persons Identification of the 'abandoned baby' or fetal material and avenues for identifying the source of an unknown profile
DNA databases References
395
22 The dentist's role in child abuse and neglect
420
395
395
398
402
403
405
406
407
407
408
411
413
414
414
David Whittaker
Introduction Dental neglect Facial and oral pathology The dentist accused of child abuse Bite marks References 23 Paediatric dental identification
G Howard Moody
Introduction Comparison Facial reconstruction and dental profiling
420
422
422
425
425
432
435
435
435
440
xii I
Contents
24
Age estimation References
441
444
The expert witness and expert testimony
447
Anthony Busuttil
447
448
449
449
450
450
450
452
452
453
453
454
454
454
454
455
Introduction Mission statement of the expert Claim to expertise Professional witnesses Opinions Yes or no? Admissibility of expert evidence Communications from the expert witness Declaration by the expert in the report In the witness stand or box Pre-trial communication Conflict of interest Rules of evidence Conclusion Recent developments References
457
471
485
Appendix A: Child protection examination forms Appendix B: Tables of standard measurements Index
~ - - - - - - - - - - - - - - - - -- -~
.
------
.
I
CONTRIBUTORS
I
Denis R Benja min Department of Laboratories Cook Children's Medica l Cen ter Fo rt Worth, TX, USA
Maeve McPhillips
Department of Rad iology
Royal Hospital for Sick Children
Edinburgh, UK
Jem Berry Formerly Professor in Paediatric Pathology Directorate of Pathol ogy St Michael's Hospital Bristol, UK
Robert A Minns Professor of Paed iatric Neurol ogy Dep artment of Child Life and Hea lth The University of Edinburgh Edinburgh, UK
Anthony Busutt il Formerly Emeritus Regius Professo r of Forensic Medicin e Un iversity of Edinburgh; and Med ical Director forens ic Medical Services 0i HS Lo thian Edin burgh, UK
Jacqueline YQ Mok Consultant Paed iatrician Department of Community Child Health Roya l Hospi tal for Sick Children Edinburgh, UK
Christian Giraud Institut Universitaire de Medicin e Legale La usann e, Switzerland .-\lex M Graham Division of Pathology (Fore nsic Medicine) University of Edinb urgh Ed inbu rgh, UK Helen Hammond Consultant Paedia tri can (Community) Department of Community Child Hea lth St John's Hospital Livingston, UK David J Harrison Professor of Pathology Division of Pathol ogy (Forensic Medicine) The University of Edinburgh Edin burgh, UK Jean W Keeling fo rmerly Consultant Pa ediatric Pathologist ~oy al Hospital for Sick Children :::din burgh, UK :y Milly Lo
Cl inical Research Fellow )epartment of Ch ild Life and Health - e University of Edinburgh ~ inburgh, UK ~at rice
Ma ngin :nsritut Universitaire de Medici ne Legale _ausanne, Switzerl and
G Howard Moody Consultant in Oral Pathology Edinburgh Denta l Institute Edinburgh, UK John Pearn Professor of Paediatrics and Child Health The University of Queensland Royal Children'S Hospita l Herston, Queensland. Australia Waney Squier Consul tant Neuropathologist Radcliffe Infirmary Oxford, UK Angela Thom as Consultant Paediatric HaematoJogist Roya l Hospi ta l for Sick Children Edinburgh, UK Dick Variend Consultant Paediatric Pathologist (retd) The Children'S Hospital Sheffield, UK David Whittaker Emeritus Professor in Forensic Dentistry University of Wales Cardiff, UK Harry Willshaw Consultant Paediatri c Ophthalmologist Th e Birmingham Child ren's Hospital Birmingham, UK
I
I
PREFACE
The possibility that a child may have been injured, abused, neglected or otherwise ill treated rightly raises the indigna tion and anxiety of the caring professions and involves law enforcement agencies. However, in the interests ofjustice a nd fairness, a person accused of such injury or negl ect is entitled to appropriate legal representatio n at any hearing and is deemed to be innocent until proven guilty. Quite frequently, dubiety and uncertainty linger about whether or not, in any specific situation, observations made, clinically or pathologi cally, can be interpreted solely as a manifestation of inflicted injury or neglect, or whether there is a possibility that the observed findings could have come about in other, non criminal, circumstances. These matters necessitate advice from those with expelience and expertise in this field. The aim of this book is to furnish an authoritative, com prehensive tex t to assist practitioners of medicine and the law dealing with such cases in the approp ri ate interpreta tion of these matters and to enable clinical and pathological findings to be presented in an unbiased and dispassio nate manner so that the co urts are able to better evaluate the specialist evidence put before them. The investigation and interpretation of findin gs of alleged ill treatment of infants and children requires a multidisciplinary app roach , centred on the child, his or her well-being in both the short term and longer term, as well as that of any siblings within the same environment. All of the avai lable information about any in cident must be care full y collected, collated and evalu ated. Laboratol), data, both clinical and forensic, the results of radiological investigatio ns and information from the examinat.ion of the scene where any incident took place sho uld be carefully sought and evaluated against the clinical findings. A team approach is essential, with close collaboration of famil y physicians, paediatricians involved in both community and hospital practice, the clinical fo rensic medical examiner and specialist pa tho logists, together with police and social welfare serv ices. No incident should be looked at in isola tion but rather in the context of the child's development and interaction with his or her family, environment and peer group. The survivors of inflicted injul)' or neglect in childhood must be carefully followed up, protected and their family unit supported.
In this book, some of the top ics covered here are rele vant specifically to maltreatment in ea rly life, beginning with the examination of an infant or child for whom ab use is suspected, incorporating the family environment and set against criteria for normal deve lopment. The difficult prob lem of suspected sexual abuse of children is considered separately. The extensive clinical experience of the authors of the opening chapters is readily apparent, highlighting, as they do, the pitfalls of incomplete investigation and iIl considered interpretation. The ap propri ate level of investi gation of specific findings, interpretation of investigations and consideration of differential diagnoses are addressed in chapters contributed by a paediatric radiologist, a haema tolo gist and a clinical pathologist, respectively. Those areas requiring specialist clinical expertise and experience - the eyes, mouth a nd central nervous system - are considered by specialists in those fields with ex tensive paedi atric experience. The examination of the scene of death or injury is discussed as a backg round to post-mortem examination of the very yo ung. The interpretation of cerebral pathology in the newborn, the investigation of sudden or suspicious perinatal death and sudden death in both infa nts and older children are addressed by experienced practitioners. Sepa rate consideration is given to sudden or suspicious deaths that occur in hospital. In subsequent chapters, more general areas of forensic pathology, including asphyxia and thermal injury, drown ing, injury to road users and oth er accidents are add ressed from a paediatric viewpoint. A similar approach is evident in the chapters covering toxicological investigation, DNA profiling and dental identification. The book concludes with consideration of the role of th e expert witness in criminal judicial cases and the provision of reports in the civil medicolegal context. Although the majority of contributors to this te xt are UK based , the subject matter is presented, as far as possible, wit hout national or geographic bias, so that the contents have in te rnation al releva nce. Anthony Busuttil Jean W Keelin g January 2008
~-- ----~- -------
.
.
I
ACKNOWLEDGEMENTS
We would like to thank our contributing authors for their hard work and for their patience and ready responses in the li g ht of requests for updates and answers to specific quelies. Colleagues in Edinburgh and elsewhere in the UK have read the Editors' contribu tions and made useful sug ges tions. JWK thanks Dr Roger Malcomson for his assistance
I
and expertise in the preparation of the illustrations for her chapters. We would like to thank the many staff at Hodder Arnold with whom we have been involved fo r their expert ise and encouragement ; in particular, Philip Shaw and our Project Editor Amy Mulick for their most helpful sugges tions in the late stages of manuscript completion.
I
LIST OF ABBREVIATIONS USED
AIDS ALTE APOE aPTT ARDS ARVD ATP AV AvD0 2
amino acid accident and emergency alternating current a cid-citra te-de xtrose adrenocorticotrop hic hormone antidiuretic hormon e adenosine diphosph ate autosoma l dominant polycystic kidney disease acquired immune deficiency syn drome appa rent life-threatening events apolipoprotein E activated partial thromboplastin time adu lt respiratory distress synd rome arrthythmogenic right ventricular dyspla sia adenosine triphosphate atrioventricular arteriovenous ditTerence of oxygen
0APP BPA BPNA BUN
beta amyloid precursor protein British Paediatric Association British Paediatric Neurology Association blood urea nitrogen
CACTO CAP CAPMI CASK CBF CD CESDI SUDI
carn itine acyJcarnitine translocase deficiency common approach pathway computer-assisted post-mortem identification ca rer-associated serial kill ing cerebral blood flow compac t disc Confidential Enquiry into Stillbirth and Deaths in Infancy Sudden Unexpected Death in In fancy ce rebra l function analys in g monitor cystic fibro sis transmembrane conductance regulator congenital heart disease Canadian Hospital Injury Reporting and Prevention Program confidence interval cytokeratin I cerebral metabolic rate for oxygen cytomega lovirus central nelVOUS syste m copy number variation ca rboxyhaemoglobin Combined DNA Index System combined paternity index
AA ARE AC ACD ACTH ADH ADP ADPKD
CFAM CFIR CHD CHIRPP CI CKI CMR0 2 CMV CNS CNV COHB CaDIS CPI
CSF CSM CT CVP CVR CZE
cerebral perfusion pressure cardiopulmonary resuscitation cumulative pressure-time index carnitine palmitoyltransferase type " carnitine palmitoyltransferase type 1 deficiency carnitine plamitoyltransferase type 2 deficiency cerebrospinal fluid crime scene manager computerized tomography central venous pressure cerebrovascular resistance cap illary zone electrophoresis
DAB DAI DAVlD DC DIC DMF DNA DRVvr DVD
DNA Advisory Board diffuse axonal injury disaster and victim identification direct current disseminated in travascular coagulation decayed , missing and filled teeth deoxyribonucleic acid dilute Russell's viper venom time digital versatile disc
ECF ECG EDH EDTA EEG EFE ELISA EM EMIT EPP EPS EPU ERG ESR EVG
extracelluar fluid electrocardiogram extradural haemorrhage ethylenediamine tetra acetic acid electroencephalography endocardial fibroelastosis enzyme- linked immunosorbent assay electron microscopy enzyme-multiplied immunoassay technique polypropylene expanded polystyrene expanded polyurethane electro retinograp hy erythrocyte sedimentation rate elastic van Geison stain
FAa FBI FOP FHM FIl FLAIR FPIA
fatty acid oxidation Federal Bureau of Investigation fibrinogen degradation product familial hemipl.egic migraine fabricated or induced illness fluid-attenuated inversion recovery fluoresce nt polarization immunoassay
CPP CPR CPT CPT 11 CPTlD CPT2D
~ ~----------
.
I
. .
List of abbreviations used I
FPP
fitness to practice panel
GAS GC-MS GCS GDP GI GMC GMD GOS GSD
group A streptococcal infection gas chromatography-mass spectrometry Glasgow Com a Scale general dental practitioner gastroin tes tin a I General Medical Council genetic metabolic disorder Glasgow Outcome Score glycogen storage disease
HbF HCM HDN HE HELLP
fetal haemoglobin hypertrophic cardiomyopathy haemorrhagic disease of the newbo rn haematoxylin and eosin (e.g. HE stain) haemolysis, elevated liver enzymes, low platelets hypoxic-ischaemic encephalopathy hypoxic-ischaemic injury human immunod eficiency virus haemophagocytic Iymphohistiocytosis hi gh-molecular-weight kininogens high-performance liquid chroma tography human papillomavirus herp es simplex virus
HIE Hll HIV HL H HMWKS HPLC HPV HSV ICD JCH ICP ICU lDDM IgA IgE IPH ISS ITP lTU
IUGR IVF rVB
LCAD LCN LCHAD LQTS LM LoC LSD \-1A DD \,IlAP .\1 CA \ilCAD
International C1a ssificatiol1 oj Diseases intracrania l haemorrhage intracra nial pressure intensive care unit insulin-dependent diabetes mellitus immunoglobulin A immunoglobulin E idiop athic pulmonary haemosiderosis inUlY severity score idiopathic thrombocytopenic purpura intensive therapy unit intrau terine growth restriction in vitro Ferti Iization intraventricu lar haemorrhage
long-ch ain acyl-CoA deficiency low copy num ber long-chain 3-hydroxyacyl-CoA dehydrogenase long QT syndrome laser microdissec tion loss of consciousness lysergic acid diethylamide multipl e acyl-CoA dehydrogenase deficiency mean arterial pressure middle cerebral artery medium chain acyl CoA dehydrogenase deficiency
MDA MDMA MECC MELAS MERRF MfV MPS MRI mtDNA MS/MS MSUD MVC MVF NAAT NAHI NAI NAIT NEC NEQAS NICHD NKH NSPCC
OECD OJ
OR OTS OXPHOS
Pac0 2
xvii
3,4-methylenedioxyamph etamine 3,4-methylenedioxymethamphetamine micellar electrokinetic capillary chromatograp hy mitochondrial encephalomyopathy myoclonic epilepsy with ragged red fibres mean flow volume mucopolysaccharide magnetic resonance imaging mitochondrial DNA tandem mass spectrometry maple syrup urine disease motor vehicle collision mean flow volume nucleic acid amp lification test non-accidental head injury non-accidental injury neonatal alloimmune thrombocytopenia necrotizing enterocolitis National External Quality Assessment Scheme National Institute of Child Health and Development non-ketotic hyperglycinaemia National Society for the Preven tion of Cruelty to Children Organisation for Economic Co-operation and Development osteogenesis imperfecta odds ratio ornithine transcarba milase oxidat ive phosphorylation
PSA PT PTA
partial pressure of arterial carbon dioxide plasminogen activator inhibitor-I Perls ' Prussian blue reactio n polym erase chain reaction pyruvate dehydrogenase posit ron emission tomography posterior inferior cerebellar alielY paediatric intensive care unit prekallikrein phospho lipid post-m ortem examination plasma membrane carnitine transpolier deficiency phosph ate-specific antigen prothrombin time post-traumat ic amnesia
RBC RIA RCPCH RFLP
red blood cell radioimmunoassay Royal College of Paediatrics and Child Health restriction fragment length polymorphism
PAI-I PBR PCR PDH PET PICA PICU PK PL PM PMCTD
I
I
CHAPTER 2
INVESTIGATION OF SUSPECTED
SEXUAL ABUSE Jacqueline YQ Mok
Introduction The colposcope in the medical examination Forensic evidence Skills and experience required Consistent vocabulary Normal female genital anatomy Perianal findings Acute, healing and healed anogenital trauma
24 27 27 28 29 29 36 37
INTRODUCTION
Sexual abuse has been defined as the involvement of depend ent, developmentally immature children and adolescents in sexual activity that they do not fully comprehend and to which they are unable to give informed consent or that vio late the social taboos of family roles. l In this situation, there is an imbalance of power between abuser and abused, and an element of control of the child by a trusted adult. Sexual abuse ranges from violent assault to gentle seduction. The activities include all forms of digital-genital , oral-genital, genital-genital contact between abuser and child, as well as non-contact abuse such as exhibitionism and use of the child in the production of pornographic material. A diagnosis of sexual abuse has both civil and criminal implications. The medical profession 's early involvement with child sexual abuse was limited to psychiatrists, who were interested in the behavioural manifestations following sexual abuse. The subject received scant paediatric recogni tion until publications by Jaffe 2 and Kempe3 changed the willingness and ability of paediatricians to recognize and deal with the problem. Paediatricians must have a high index of suspicion and be willing to consider sexual abuse as a dif ferential diagnosis when a child presents with behavioural problems or somatic symptoms that suggest that sexual abuse might have occurred. All physician s should act accord ing to their local child protection procedures, with the welfare and protection of the child as paramount considerations.
-
~
-----
-
--
- -----
--
Female genital findings in sexual abuse Signs of anal abuse Conditions that mimic abuse Screening for sexually transmitted infections Interpretation of clinical and laboratory findings Summary References
38 39 40 41 42 43 44
The widespread occurrence of child sexual abuse has been known for many years, although the exact prevalence is diffi cult to define. Prevalence studies have been canied out by interviewing adults about childhood events. In a summary of 19 studies conducted in the USA or Canada between J 980 and 1994, Finkelhor4 found that the rate of sexual abuse reported by women varied from 2 per cent to 62 per cent, with an approximate prevalence of 20 per cent. The ra tes reported by men varied from 3 per cent to J 6 per cent, with a reasonable estimate of approximately 10 per cent. However, studies based on interviews of adults provide limited information owing to differences in study design, the response rate, methods of data collection, definition of sexual abuse, definition of a child, and the accuracy of recall of events that might be traumatizing. Data from studies of reported incidences of abuse pro vide information about the number of children recognized annually, usually at individual centres. It does appear that increasing numbers of children are referred because of suspicions of sexual abuse, owing to either a t ru e increase in the occurrence or better recognition of the problem. In the USA it is estimated that between 120 and 150 per 10000 children have been subjected to sex ual abuse. More recently, Jones et a1 5 identified a significant reduction in the incidence of substan tiated child sexual abuse in the USA, and a population-based study in Australia has also provided evidence of a decline in the underlying rate of child sexual abuse.6 Such findings may indicate the effect iveness of personal safety progra mmes for young children.
---,------
Introduction I
An earlier paper from one city in the UK 7 described 51 children in 1985 who presented after they disclosed abuse, a number that rose to 79 in the subsequent year. Thirty eight per cent of the children were aged less than 5 years of age at diagnosis and the mean age was 8 years. Of the alleged abusers, 60 per cent were related to the child, and one-half of these were natural fathers. Because the vast majority of children are abused by someone they know and trust, force and restraint are not commonly used when abusers engage children in sexually inappropriate activities. There is usually little intent to harm the child, and bribes or threats are often used to prevent the child from reporting the abuse. Children are ideal victims for sexual exploitation and abuse, as they are vulnerable and see adults as trusted and powerful. The 'child sexual abuse accommodation syn drome's describes the stages of traumatic sexualization, feelings of betrayal, powerlessness and entrapment that lead to the child's accommodation of the abuse. As a result, very few children disclose abuse immediately following the event, making the retrospective interpretation of healed trauma a great diagnostic challenge.
The Medical Evaluation Child sexual abuse should rarely, if ever, be diagnosed on the basis of physical signs alone. A clear statement from the child is the single most important factor towards mak ing a diagnosis. The medical evaluation of children sus pected to have been sexually abused should be part of a multidisciplinary process that spans the investigative, diagnostic and therapeutic needs of the child and family. The medical examination should be a comprehensive health assessment that should aim to: • establish any need for immediate treatment; • provide background information that mayor may not support the diagnosis; • provide information or evidence to sustain criminal proceedings and/or care plans; • plan or coordinate ongoing care; • reassure the child and family. A minimum of physical examinations should be con ducted as is necessary, and if several medical specialists are to be involved it is desirable that they should examine the child together. Joint examinations can be performed by a paediatrician and a forensic medical examiner (police sur geon) to encompass, in a single examination, the child's medical needs with the legal requirements for evidence. 9 Although a carefully structured approach to the examin ation is required for legal purposes, the medical assessment should be compassionate and thorough, resulting in a thera peutic experience for the child and family.lO.11 The examin ers must be both familiar and comfortable with normal childhood behaviour and development, genital and anal anatomy, as well as physical findings of abuse. Core skills
25
and case-dependent skills for any paediatrician or forensic medical examiner who undertakes such examinations have been defined by the Royal College of Paediatrics and Child Health and the Faculty of Forensic and Legal Medicine (October 2007).12 Core skills include: the ability to commu nicate with children and their carers about sensitive issues; an understanding of the child's developmental, social and emotional needs; a knowledge of consent and confidential ity issues as they relate to children; an understanding of the range of normal genital and anal anatomy for the age and gender of the child to be examined; an ability to docu ment and interpret the clinical findings; competence in the production of a report; a willingness to communicate and co-operate with other agencies; and the aptitude to present the evidence and be cross-examined in civil and criminal proceedings.
History As with all medical consultations, the starting point is in the history, taking care to avoid asking leading questions. Investigative interviews are usually conducted by desig nated agencies (police and social services) to avoid repeti tive questioning of the child. This should not preclude physicians asking relevant questions that are essential to the medical examination. In eliciting the history, the clin ician should approach the child in the same manner as he or she would in any other paediatric condition. Details should be obtained about the child's birth and development, past medical history, and family and social background. The systematic enquiry should cover symptoms and signs relat ing to the genitourinary and gastrointestinal systems. Specific questions should be directed to the presence of pain, itch, rash, discharge or inflammation 'down below', episodes of bleeding (on pants or on toilet paper), and fre quency and consistency of bowel movement. When appro priate, girls should be asked the menstrual history, type of sanitary protection used, sexual history and previous gynaecological examinations. The child's terminology for the various body parts should be documented, and any statement made by the child recorded verbatim. 13 The importance of the psychological aspects of sexual abuse, both in the short and long term, has been high lighted. 14 ,15 Emotional difficulties include anxiety, sadness, anger, behaviour problems, school refusal, sleeplessness, withdrawal and sexualized behaviour. Somatic complaints that have been reported include: eating disorders, abdom inal pain and headaches, as well as loss of bladder and bowel control. The medical evaluation should include an enquiry into the child's emotional status and general well-being, as the carers may not associate behavioural difficulties with a past history of abuse. The carers may even be unaware that sexual abuse has occurred. Child sexual abuse is a diagno sis that, like other paediatric diagnoses, requires a consider ation of the history, physical examination and supportive
26 I
Investigation of suspected sexual abuse
laboratory tests when appropriate. 16 Although the medical history qualifies as 'hearsay evidence', many states in the USA permit an exception to the hearsay rule, for medical histories obtained by physici ans,u
Technique of the Examination Th e exa mination should be carried out in the presence of a trusted adult, usually the child's mother. The whole child should be examined, and this includes measuremen t of hei ght and weight, assessment of the general appearance, developmental milestones, demeanour and behaviour. As part of the physical examination, general signs associ ated with trauma should be sought, such as distribution a nd pattern of bruises, grip marks, ' love bites', teeth marks and scratches, as well as injuries within the mouth. In most situ ations, the disclosure involves past abuse and th e exam in at ion can be planned to suit the child and family. Delay should be minimized in the following circumstances, when: • the a buse has occurred within the previous 72 hours; • there is a history of acute trauma; • there is a possibility of pregnancy resulting from the abuse, so that post-coital contraception can be prescribed. Considerable reassurance will be required and to avoid further distress no force or restraint should be used, no mat ter how well intentioned. An explanation ofwhat the exam ination entails should be given, with clear emphasis on 'a health check' and 'taking a look'. Wi th calm reassurance and an unhurried approach, most children can be examined without the use of any sedation. The recommended position for the female genital examination is the supine 'frog legged' position, with the hips flex ed and abducted, the soles of the feet touching. Very young children can be examined on their mothers' laps. Sometimes the adult can assist by sitting astride the examining couch, cradling the child as she leans back against the adult's body. In the supine 'frog-legged' position , the external geni talia should be inspected for signs of injury. Gentle lateral parting of the labia majora with the examiner's middle and index fingers (labial separation) allows visualization of the posterior fourchette, vestibule, perihymen al regions and urethra; the hymen may be visible at this stage. In order to visualize th e ma rgins, configuration and size of the hymenal openin g, labial traction is applied . This involves grasping the posterior ends of the labia majora between thumb and index finger and pulling gently up an d out wards. The vaginal walls may be demonstrated by this pro cedure; however, digital examination of the vagina is rarely indicated in the prepubertal child. Any sign observed should be described and documented, using the clock face to denote the location of the finding; in the anatomical position, 12 o'clock is anterior while 6 o'clock denotes the posterior position .
~
----~
-
--
-~
-
--
-
-
-
Sometimes the hymen can be difficult to visualize because of sticky or redundant folds. If labial traction does not cause the hymen to open, douching with warm water will result in 'floating of the hymen', Further definition of the anatomy can be obtained with the use of a cotton bud, applied behind the hymen to 'tease out' the folds, This technique should not be used in prepubertal children, in whom the hymen is extremely sensitive, unless preceded by the gentle application of a local anaesthetic gel. In ado lescent girls, the hymenal edge can also be examined using the Foley cath eter stretch technique,18 A 14-gauge Foley catheter is inserted through the hymenal orifice into the vaginal vault, and infl ated with 40 mL of air. Gentle pulling toward the hymen results in stretching and displaying of hymenal tissue over the surface of the balloon,19,20 The toluidine blue dye test was deve loped by Lauber and Souma 21 to help detect lacerations in adult rape victims. Toluidine blue is a nuclear stain that will bind to nuclei in the deeper dermis when exposed by lacerations in the skin. In a study of girls "vho alleged sexual abuse, the applica tion of a I per cent aqueous solution of toluidine blue dye was found to increase the detection rate of posterior fourchette lacerations from 4 per cent to 28 per cent in adolescents, and from 16.5 per cent to 33 per cent in the paediatric population .22 The prone knee-chest position is recommended if there is difficulty in visualizing the posterior hymenal rim. With t he child resting on her knees and elbows, the buttocks are parted upwards and outwards, using the examiner's palms. This technique usually exposes the vestibule, perihymenal structures and hymen. Any suspicious findin gs in the pos terior hymenal rim can be clarified , as the effect of gravity causes the posterior rim to stretch out. The perianal area is traditionally examined in the left lat eral position. With use of the colposcope, some examiners prefer to continue with the examination in the supine pos ition, by asking the child to 'curl up into a little ball'. This is achieved by flexion of the child's hips and knees against the abdomen. The buttocks are gently separated using the palms of both hands, and the perianal area inspected for signs of abuse. The external anal sphincter usually relaxes during this procedure, making it easier to visualize fissures in the anal margin. Sometimes the ana l canal opens to reveal a clear view of the rectum. A digi ta l examination of the anus is rarely necessary and is unhelpful in assessing anal tone. There is also no evidence to suggest that a digital examin ation provides useful clinical information. If there are suspi cions of injuries or pathology higher up the anal canal or rectum, referral should be made to a paedi atlic surgeon for an examination under anaesthesia. Most units have a dedicated suite of rooms where chil dren can be examined in a child- fr iendly environment. The minimum requirement is an area that is integrated into a children's outpatient department to allow access to labora tories a nd other investigative facilities but which offers some privacy. Other important considera tio ns include the
------ - - ---
Forensic evidence I
specific needs of adolescents, the gender of the examiner and the availability of follow-up for sexual health coun selling and therapeutic support.
Examination of Boys The same sensitive and age-appropriate techniques should be used when examining boys. As with girls, any evidence of trauma such as bruises, swellings, scratches and bites should be sought for and documented . The genital examination should focus on the groin, penis, urethra, scrotum and testes while the ch ild is supine. The presence of discharge or warts should also be noted. "'lhen examining the pelianal region, various positions can be used, depending on the age of the boy. The young child may be examined in the supine knee-chest position, immediately after examination of the penis and scrotum. Older boys may be more comfortable in the left lateral position, with either the right knee or both knees pulled up to the chest. The anal examination usually involves inspection of the soft tissues only. An assessment of the anal to ne can be made, with the buttocks parted. Find ings should be documented either by line drawings or photo graphs. The colposcope can be used as a source of light and magnification, as well as photographic documentation.
THE COLPOSCOPE IN THE MEDICAL EXAMINATION The medical examination of children reporting sexual abuse evolved from a search for ways to improve lighting and visualization of the young child's genitalia. Methods for recording observations were also important, to assist in the interpretation of findings. In the early 1980s when physicians began to examine the genitalia of children, instruments such as a hand-held magnifying glass and an auriscope were used to enhance visualization. In 1925, the colposcope was invented by Hinselman , to assist in exam ination of the female cervix. Teixeira 23 used the colposcope to examine victims of alleged sexua l assault. Five years later, Woodling and Heger24 promoted the use of the colpo scope in the assessment of chi ldren who alleged sexual ab use. The colposcope is a no n- invasive magnifying instrument with a built-in light source, which allows mag nification of the external genitalia from 2 to 25 times. It contains a binocular system of lenses of varying strengths, co upl ed to an integral light source. The instrument can either be mounted on the wall, the examination couch, or on a caster system to enable mobility. Light intensity can be varied to cope with changes in magnification, and most instruments incorporate a green filter to improve the visu alization of abnormal vasculature or scars. Tei xei ra reported that an additional 11.8 per cent of sus pected cases were corroborated than would have been when conventional examination techniques were used. In
27
the prospective study by Muram and Elias,25 130 prepubertal girls (mean age 5.5 years) were evaluated bo th by an unaided examination and by colposcopy. Overall, 92 girls were found to have abnormal findings, the majority of which (96 per cent) were observed during the unaided examination. Of the four patients in whom findings were detected initially by colposcopy, these were observed dur ing the repeat unaided examination. In only one patient were the findings observed only by colposcopic examin ation alone. The authors concluded therefore that una id ed examination by an experienced cl ini cian is adequate for the evaluation of most victims of sexual abuse. The most obvious advantage of the colposcope is the integral photographic facility. Documentation of all visible findings in abuse is increasingly expected as the standard of good practice. Images produced by the colposcope can be converted into slides or photographs using a single-lens reflex camera or a Polaroid camera. Modern video technol ogy allows the image to be recorded onto a videotape, compact disc (CD) or digital versatile disc (DVD) for imme diate viewing, and has the advantage over still photog raphy as it shows the dynamic variability of anogenital anatomy as the examination proceeds. The colposcope is now standard and acceptab le equip ment in the examination of sexua l abuse. With photo doc umentation, the most obvious benefit to the child is that there is no need for repeated examinations. Many examin ers do not position their eyes in the binocular lenses of the colposcope; rather they view the image produced on the monitor at a more comf0l1able position and distance from the ch ild . The child is also able to observe the examination on the monitor, and this helps to achi eve a sense of control and participation in the examination. The instrument is usually in trod uced to the child to gain cooperation, and many enjoy playing with it prior to the examination. The use of high technology is both accepted and expected in the modern health care system, and is preferable to attempting to peer between a ch ild's legs wit h an auriscope. The ability to capture images, either as slides, photographs, videotapes, CDs or DVDs enhances undergraduate and postgraduate physician training. 26 - 3o
FORENSIC EVIDENCE Verba l consent is usually adequate for both the examin ation and photo-documentation. This should be obtained from the person with parental responsibility and from the child of sufficient maturity to understand the nature and consequences of the examination. If the child is the subject of legal proceedings, the consent of the court is required. The person obtainin g consent must record the procedure in the case notes. Written consent for each component of the examination can be documented on a standard form and on appropriate forms.
28 I
Investigation of suspected sexual abuse
Correct procedures must be followed for the collection of forensic sampJes and evidential material, recognizing the principle of the unbroken chain of evidence. This legal concept requires that the origin and history of any exhibit to be presented in a court of law must be clearly demon strated to have followed an unbroken chain from its source, through its examination and to the court. A note of the persons handling the sample, time, date and place where the sample was obtained, along with the places and condi tions of storage must be documented. In an acute assault (within 72 hours), the clothing worn during the assault should be coI1ected individually in paper bags and handed to the investigating police officer. The skin should be inspected for a ny stains, and swabs taken to examine for saliva, lubricant or semen. If judged appropriate, a sample of blood should be taken for DNA analysis. Stains on cloth ing are best preserved by drying and storing at cool room temperature. Blood or semen collected on swabs is best preserved by frozen storage. The presence of lubricant or spermatozoa can be sought from swabs in and around the mouth (taking care to swab in the labial-gingival sulci) and from saliva. Young children will tolerate external and internal anal a nd vaginal swabs, if the a ppropriate size of swabs is used, moistened with water. Spermatozoa can be detected for up to 14 hours in saliva J I 3 days in anal sam ples J 2 and 6 days in vaginal samples. JJ The time limits for detection of seminal fluid are 3 hours in the anus and J 2- J 8 hours in the vagina. Improved DNA diagnostic tech niques allow detection of useful evidence from unwashed or partially washed clothing, bedding and other items used during the assault. General guidelines for the collection of forensic evidence in cases of acute sexual assault a re not well suited for prepubertal children. In a review of 273 child victims of sexual assault aged less than 10 years, the majority (64 per cent) of forensic evidence was found on linen or clothing. Over 90 per cent of children with positive forensic evidence were seen within 24 hours of the assault. After 24 hours, all evidence, with the exception of one pubic hair on a child, was recovered from clothing or linens. The authors suggested that any collection of foren sic evidence from a prepubertal child 's body may not be necessary more than 2 days after the assault. J4 In cases of chronic abuse, when the last episode of contact was more than a week previously, collection of evidence consists of a careful histolY, thorough examination and screening for sexually transmitted infections (STls) . Although a highJy trained and experienced examiner is not likely to miss any abnormalities during an unaided examin ation, the colposcope allows a review of photographs and video recordings when findings thought to be suspicious during the examination might be interpreted on review as normal or non-specific. Permanent documentation of visible findings protects the examining practitioner by providing objective evidence, thus only the interpretation of the find ings can be challenged. Still photographs, videotapes or CDs of the injuries often provide key information for assessment
--
~--------
-
-
----
--
and evidence for prosecution in child abuse cases. Tradi tional systems for photographing abused children using a single-lens reflex camera still provides the highest quality image. The use of Hi-8 and Super VHS video cameras offers significant improvements in image quality and resolution over the traditional videotapes. More recent developments in computer imaging technology have resulted in greater ease in transforming analogue images to a digital format for computer storage and conversion into slides or prints. [mages can also be converted into a computer file, which assures preservation without degradation of the initial image. A video capture card, a scanner and/or a digital cam era, allow a computerized database to be developed. With appropriate software products, images can be transmitted over the Internet, enabling a second opinion to be sought and thereby brings telemedicine into remote areas. J 5,J6 Photographs, video, CD and DVD recordings must be properly verified and relevant, they must bear the patient identifier, date and time of recording, and must be signed by both medical and forensic examiners. Images are neces sary to explain or illustrate adequately the complexity of the injuries, and are therefore relevant for courtroom pro duction. However, the production of sensitive images and photographs of children's genitalia in the courtroom is not common practice and should be discouraged. An alterna tive line diagram can always be used to illustrate the injuries. All recorded material should be available to any medical expert instructed by defence solicitors.
SKILLS AND EXPERIENCE REQUIRED Although medical findings are not necessary for the legal conclusion of sexual abuse, health-care professionals who examine children for signs of sexual abuse are often asked to render an opinion as to whether their examination was nor mal, non-specific, suggestive or indicative of abuse. Despite this important role, the literature contains little information regarding examiner competence in assessing children's geni talia. Undergraduate and postgraduate training on the subject of child sexual abuse is woefully inadequate. The medical investigation of children in whom abuse and neglect is suspected has become a complex and technical specialty. A new group of physicians has emerged, who are specialists in the wide ramifications of child abuse and neglect. A sound knowledge is expected in general, behavioural and developmental paediatrics, as well as in . gynaecology and infectious diseases. The child abuse specialist also needs to be familiar with forensic examinations, civil and criminal laws, child advocacy and public policy, and must be comfortable when testifying in court. For these reasons, structured train ing programmes must be developed to improve the quality of care delivered to abuse children and their families.37 Brayden et al J8 evaluated the interobserver reliability of clinicians rating colposcopic photographs, and examined
---
Normal female genital anatomy I
correlates of reliable interpretations. Seventy-one physici ans and two nurse practitioners were asked to assess blindly six photographs of prepubertal female genitalia, five of which were taken from children who had given histories of sexual abuse. Experts in the field of child sexual abuse assessment made significantly more accurate assessments than paedia tricians, paediatric and family practice residents, and trainee physicians. In another study, Adams and Wells 39 sought to determine how well medical examiners agreed on the significance of certain anogenital findings in children, by showing colposcopic photographs of 16 patients to 170 med ical examiners who were blinded to the history of each patient. The agreement between the participants and the experts on the abnormal cases (mean 81 per cent) was signif icantly higher than on the normal cases (mean 71 per cent, p = < 0.001). There was also higher agreement on genital findings (78 per cent) than on anal findings (63 per cent, p = 0.000). In the participants, higher experience level and use of a colposcope were associated with higher overall agreement with the experts (74 per cent versus 44 per cent, p = < 0.0001). The history was also found to influence physicians' interpretation of girls' genital findings. Paradise et al 40 conducted a questionnaire survey of 1387 randomly selected fellows of the American Academy of Pediatrics and all 802 members of child abuse professional groups, using seven simulated cases, in 6 of which the histories were changed at a second mailing 4 months later. The proportion of changed or reversed opinions varied from 0 per cent to 5.6 per cent amongst experienced examiners; 1.6-19.8 per cent amongst moderately experienced examiners and 3.6-27.2 per cent where the examiners were inexperienced. The like lihood of an interpretation being changed was influenced by a diagnostic expectation (change in history) and ambiguity of the photographs.
CONSISTENT VOCABULARY Medical examiners are often asked to determine whether a child has been sexually abused and whether penetration has occurred, as well as how often a child has been abused. More often than not, the physical appearances will be nor mal or non-specific. In preparing the medical report, care must be taken to avoid terms that can be misinterpreted and lead to confusion. Subjective descriptions such as a 'lax' sphincter or a 'gaping' hymen are unhelpful unless sup ported by measurements. It is important that clinicians are aware of the recommended terminology used to describe the genitalia, as well as interpretation of anogenital findings, so that those who examine children for alleged sex ual abuse can understand each other's descriptions. 41 - 43 Practice guidelines are now available and should be widely dissemi nated. 44 - 47 In an excellent review of hymenal morphology and non-specific findings in girls selected for non-abuse,
Heger et al 48 called for a thorough understanding of normal
29
studies and a consistent application of established terminol ogy that can prevent the misinterpretation of non-specific or congenital findings as post-traumatic changes.
NORMAL FEMALE GENITAL ANATOMY Figure 2.1 is a diagrammatic representation of the prepu bertal female genitalia, with the parts labelled using recom mended terminology. The most consistent landmark is the clitoris, which is usually prominent in young girls because of the lack of subcutaneous fat in the surrounding tissues. The vestibule is the area enclosed within the labia minora, and includes the urethral opening which can be pinpoint or patulous. Bilateral lines (Hart's lines) drawn medially from the posterior ends of the labia minora converge at the pos terior fourchette. The introitus refers to the hymen and hymenal opening or orifice. Sometimes the vaginal walls can be seen through the hymenal opening. The area between the posterior fourchette and inferior edge of the hymen is referred to as the fossa navicularis. Between the posterior fourchette and the anus, deep within subcuta neous tissues, is the perineal body. In order to identify the signs of sexual abuse, the clinician must have a sound knowledge of normal anogenital anatomy and its variants. Contemporary textbooks provide little insight into the characteristics of hymenal anatomy or the changes of the hymen in response to oestrogen levels during the childhood years. The appearances vary considerably according to the age of the child. In the newborn, the effects
Labium-
majorum
Clitoris - - - - H - t -
Labium - - i - t ' minorum
(}L-1--\;\---\---'.-- Ure th ra I
opening
Hy menal}2 orifice Hymen .E
'e
-.....;:::7'--+-7--.1----
--tt
Fossa ----''<---'..--~navicularis ...... - _ _~,-'--_ Posterior
P"~,~:
Figure 2.1
Prepubertal female genitalia.
"",,h""
Table 2.1
Cross-sectional studies of normal female genital anatomy
Authors
Number of subjects
Age
Summary of common findings
McCann et al 1990 52
93
10 month s-lO years
Berenson et al 1991 57
468
Ne wbo rn s
Beren son et al 1992 53
211
Erythema of vestibule 56% Periurethral bands 51% Labial adhesions 39% Hymenal mounds or projections 34% Longitudinal intravag inal ridges 90% Intravaginal rugae 89% Annular hymen 80 0/0 Fimbriated hymen 19% Hymenal clefts (anterior segment only) 34% External hymenal ridges 87 0/0 Longitudinal intravag in al ridges 53% Hymenal tags 13 0/0 Vest ibular bands 98 0/0 < 12 months, fimbriated hymen 46% > 24 months, cresce ntic hymen 51% Transverse hymenal diameter increa sed with age Increased vascularity 44% Midline avascular area s 27%
1 month-7 years, mean 21 ::!: 20.6 months
Gardner 1992 55
79
3 month s-11.5 years, mean 5.3 years
Yo~an a ndYo~an 1992~
168
7-17 years
Emans et al 1994 119
300
Median 18 yea rs
Beren so n et al 2000 56
200
3-8 years, case - co ntrols for abu sed childr en
Heger et al 2002 81
147
Mean 63 (::!:38) months
Ragged posterior fourchette/notch in posterior fourchette 28% Peri hymenal teth ers 14% Hymena I bu mps 11 % Effects of puberty - labia minora increased in size and pigmentation; hymenal tissue increased in thickness; increa sed vaginal se cretion s Complete hymenal clefts seen significantly more frequently in girls who were sexually active; not related to participation in sports, prior pel vic examination or tampon use Median hymenal diameter increased significantly with sexual activity Vulvar features: erythema 37%; partial agglutination 7010; linea vestibularis 7%; vaginal disch arge 4%; friability 4%; perineal depression 2% Hymen al features: periurethral band s 95%; longitudinal intravaginal ridg es 87%; vestibular bands 60%; bumps 46%; external ridges 80/0; prominent vessels 7%; superficial notches 5%; tags 5% Hymenal configurations: annular 53%; crescentic 29.2%; sleeve like 14.9%; sep tate 2%; other < 1% Non-spec ific findings: perihymenal bands 91.8%; longitudinal intravaginal ridges 93.8%; hymenal edge irregular 51.7 0/0 , thickened 45.5%, narrowed 22.4%; erythema 48.9%; vascular changes 37.4%; hymenal bumps/mounds 34% ; hymenal opening size > 4 mm 30.6%; posterior hymenal concavity 29.5%; partial posterior hymenal notch/cleft 18.3%; labial ad hesion s 15.6%; failure of midline fu sion 0.6%
Normal female genital anatomy I
of maternal oestrogens result in promin ence of the labia and clitoris, a redundant and fleshy hymen and copious vaginal secretions. In early infancy, the labia majora gra dually fold across the introitus to form a protective pad. As the oestrogen effects wear off, the labia majora appear flattened and the labia minora are seen as thin folds. The inner margins of the labia minora may be set deeply and easily mistaken for the hymenal orifice. Although there are several cross-sectional studies on genital appea rances throughout childhood, a methodo logical flaw of most of these studies is the inclusion of girls who may have been sexually abused, as the authors have relied on parental history alone to exclude abuse. These cross-sectional studies have included girls over multipl e age groups, making it difficult to eva luate findings by age. Table 2.1 summarizes some studies that have attempted to define normal genital anatomy in non-abused girls. Although cross-sectional studies cannot assess the forma tion of new features such as bumps or notches, they have been useful in clarifying the range of normal findings, as well as helped to define hymenal configuration with age. Longitudinal studies have been useful in documenting anatomical changes over time. Reports of longitudinal studies have mainly come from one centre in the USA 49,50 and are summarized in Table 2.2.
Labial, Posterior Fourchette and Vestibular Findings Yordan and Yordan 51 have documented changes in the labia minora with age and onset of puberty [see Table 2.2). In the absence of oestrogens, the labia minora appear small, thin and without pigment. The epidermis of the vestibule is thin and smooth. The fossa navicularis is characterized by a network of Table 2.2
fine submucosal blood vessels that extend onto the edge of the hymen . The hymen itself is thin and almost translucent. With the onset of early puberty, there is a reduction in the superfi cial vascular prominence in the vestibule, fossa navicularis and hymen. As puberty progresses, the labia majora and minora become pigmented. Secretions are abundant in the vestibule, which becomes darker and more textured in its appearance. Subcutaneous deposits of adipose tissue may be seen. The hymen becomes fleshy and redundant, making detailed examination of the rim difficult. Labial adhesions, either partial or extensive [Fig. 2.2), have been reported in 5-39 per cent of girls studied. 52 ,53 The presence of labial adhesions does not imply that sexual abuse has occurred, as they could result from poor hygiene, inflammation or recurrent infection, especially in a young child still in nappies.47 Increased friability (Fig. 2.3) is defined as a smal l dehis cence of the tissues of the posterior fourchette, with or with out bleeding. This is usually seen towards the end of the examination and has been reported in 4.7 per cent of girls examined in the cohort reported by McCann. 52 In a minolity of girls abnormalities that had previously been documented in sexually abused girls were noted. These included increased vascularity, mid-line avascular areas, as well as urethral dilatation (Fig. 2.4). Although it is likely that sexu ally abused girls may have been inadvertently included in these cohorts, these studies highlight the dangers of over interpreting minor anatom ic findings that may lie within the range of normal variation. 54 .55 In a case-control study of anatomic changes resulting from sexual abuse, no difference was observed in the percentage of abused versus non-abused children with regard to labial adhesions, increased vascular ity, linea vestibularis, friability, a perineal depression, or a hymenal bump, tag, longitudinal intravaginal ridge, external ridge, band or a superficial notch. Furthermore, the mean
Longitudinal studies on normal female genital anatomy
Authors
31
Number of subjects
Ages examined
Summary of findings
Berenson 1993 49
62
Birth and 1 year
Berenson 1995 50
134
<2 months and about 3 years; 42 also examined at about 1 year
Decrease in hymenal tissue seen in 33 out of 57 Hymenal notches seen for the first time in nine girls; resolved in five Hyme nal tags seen for first time in four girls; resolved in two Posterior hymenal notches not observed Change from annular/fimbriated to crescentic hymen in 87 out of 134 External hymenal ridges resolved by 3 years Tags decreased by 1 year; resolved by 3 years Longitudinal intravaginal ridges and periurethral bands more easily seen at 3 years Mean horizontal and vertical transverse hymenal diameters increa sed with age Posterior hymenal rim did not va ry with age; > 2 mm Racial differences in hymenal anatomy
32 I
Investigation of suspected sexual abuse
number of each of these features per child did not differ between the two groups of children. Vaginal discharge was seen more frequently in abused children (11 per cent versus 4 per cent, p = 0.01) especially if the abuse took place within the last 7 days, involved penile penetration and was reported on three or more occasions. 56 This finding is not pathogno monic for abuse, however, as infection, allergy or the pres ence of a foreign body can cause vaginal discharge.
Hymenal Configuration
Figure 2.2 Extensive adhesions of the labi a majora in a l -year old girl, obscuring introitus. The labia minora are seen at the anterior end of the adh esions.
In the prepubertal chi ld, the hymen can be fimbriated (with a ru ffled edge). crescentic (absence of tissue anteriorly), annular (tissue present all round, 360°), imp erforate, sep tate or cribriform. At birth, the annular sleeve-like hymen is th e most commonY A decrease in hym enal t issue is observed by the age of 1 year in approximately two-thirds of the children49 and by th e age of 3 years, a crescentic hymen is th e most predomin ant finding.5o Variations of hymenal morphology have been reported consistently in all studies of norm al anatomy and can be explained by the lack of consistency in terminology, as some authors c1as si fy annular hymens with ventral clefts as crescentic. Exam ples of an imperforate, an annular and a crescentic hymen are shown in Figs 2.4-2.6.
Hymenal Ridges, Bumps, Tags and Bands
Figure 2.3 Friable posterior fourchette in an 8-year-old girl, with sp litting and bleeding at the end of the examination. Notch seen at 5 o'clock position of hymen.
Figure 2.4 Imperforate hymen in a 10-year-old girl, with a dilated, patulous urethra that could be mistaken for hymenal orifice.
-
-----
~---------
-
-
-
--
External hymenal rid ges are usually seen between the external surface of the hymen and the vestibular wall, either in the 6 o'clock pos ition, or in the sub-urethral region. This congenital feature has been observed in up to 80 per cent of newborn in fan tsY There are no published data on the fre quency of external ridges in children older than four years
Figure 2.5 Annular hymen in a 3-year-old girl, with peri hymenal bands at 10 and 2 o'clock.
- - -----
-
Normal female genital anatomy I
Figure 2.6 Crescentic hymen in a 6-year-old girl. Note the absence of hymenal tissue between 11 and 1 o'clock and the translucent smooth-edged posterior rim.
old, but it is likely that the majority of these ridges resolve by three years 01d. 50 External ridges may persist in some individuals and should not be interpreted as a sign of abuse. Longitudinal intravaginal ridges that extend from the vagin a onto the internal surface of the hymen have been described in all four quadrants of the hymen. Although first described in a bused children,58 they have also been found in cohorts of non-abused girls and are likely to be congenital findings, occurring in 56 per cent of newborn females. 57 Beyond the neonatal period, longitudinal intrav aginal ridges have been reported in 25-94 per cent of girls (see Table 2.1, p. 30), and these differences may reflect vari ations in definitions or terminology. Interestingly, there may be racial differences as intravaginal ridges were noted to occur more frequently in black girls than in Caucasian or Hispanic girls.50 A hymenal bump represents a localized, rounded, thick ened mound of tissue on the edge of the hymen, is wider than it is long and can be seen anywhere on the edge (Fig. 2.7). This feature has been reported in 7 - 34 per cent of girls without a history of abuse, and in older studies, hymenal bumps or mounds were thought to represent an area adja cent to a healed tear in the hymen. 58 More recent studies on children selected for non-abuse suggest that this is a common and non-specific fe ature, and may represent an area where a longitudinal intravaginal ridge ends at the hymen. 48 ,52,53,56 A tag is an elongated projection of tissue above the hymen, and is usua lly longer than it is wide. It can some times be seen to extrude from the vaginal canal onto the hymen. Tags are commonly found in the midline (Fig, 2.8), and have been observed in 13 per cent of newborns,57 Although most resolve in the first few months of life, many persist for years. 48 - 50,56 Bands can extend from the hymen to the vestibule, and can be found in all four quadrants. They have been
Figure 2,7 hymen.
33
Bump or mound of tissue at 6 o'clock on a crescentic
Figure 2.8 Fleshy tag, arising from between 10 and 11 o'clock on the hymen, and resting on right labium minus.
reported in up to 98 per cent of newborn in fants57 and 92 per cent of prepubertal girls. 4B Perihymenal or vestibular bands are often symmetrical, and create a semilunar shaped space between the bands, which ca n be mistaken as the hymenal orifice (see Fig. 2.9).
Hymenal Notches, Clefts and Transections A U-shaped, concave indentation on the edge of the hymen, cau sing a dip beneath the baseline or a break in the continuity of the membrane, is usua lly referred to as a notch, while a cleft is V-shaped and angular. The concav ity in the hymenal tissue is persistent, independent of the examination technique and posi tion of the child. The extent of a notch/cleft can be classified as superficial or partial, when its depth is less than or equal to one-half of
34 I
--
~
Investigation of suspected sexual abuse
Figure 2.9 Symmetrical perihymenal bands in a 4-year-old girl. The semilunar shaped spaces created by the bands could be mistaken for the hymenal opening.
Figure 2.10 Transection of the hymen in the 5 o'clock position. The hymen also shows ea rly oestrogen effects, appearing thickened and redundant.
the width of the hymenal membrane, or deep, when it is more than one-half of the width. Notches or clefts are seen commonly at the 12 o'clock position of annular or redun dant hymens. They should not be reco rded as pathological findings in the fimbriated hy men (because of its fringed nature) or between 11 a nd 1 o'clock in a crescentic hymen , whe re the re is a natural absence of hymenal tissue. 56 The sign ificance of a hymenal notch or cleft depends on the location and ex tent of the defect. Anterior and latera l notches, between the 9 and 3 o'clock positions, have been observed in 35 per cent of newborn girls 57 are usually caused by asymmetrical insertions of a crescentic hymen, and should be considered normal. Anterior notches tend to widen to form crescentic hymens, whereas some notches disappear as hymena l tissue becomes less redundant. 49 Although the observation of a latera l notch in the first 3 years of life has been documented as a new finding, none of the children followed up longitudinally developed a new posterior notch. 50 Care must be taken to distinguish a true concavity from an appa rent hollo wing in the hymenal rim caused by an adjacent mound or tag. Until recently it was thought that notches in the posterior portion of the hymen were rarely observed in non-abused children. 53 ,5 7 With a more consistent application of terminology, two groups of examiners have established that superficial notches /partial clefts can be found in the posterior hymen in 5-18 per cent of girls who had been screened for non-abuse, and have proposed that this is a non-specific finding. 48 ,56 A recent study 59 suggested that some partial tears of the hymen heal to leave shallow notches. Figure 2.3 demonstrates a notch in the 5 o'clock posi tio n of the hymen, with friability of the posterior fourchette. A hymenal transection describes a complete deficiency in hymenal tissue that ex tends to the junction between the hymen and vestibule and, when seen in any location, is associated with trauma. 56 ,60 Other terms used to describe this a ngular concavity include complete notch/cleft, tear,
healed laceration, rupture and gap. It is widely agreed that a deep/complete cleft or transection located between the 4 o'clock and 8 o 'clock positions in the hy men is consistent with abuse (Fig. 2.10), as these have never been reported in non-abused children .47 ,48,56
---
-
~- -
-
-
-----
-
---
Gaping/Narrowing/Attenuation of the Hymen The hymenal orifice has been described as gaping w hen, with the child in the supine frog-legged position, a view is obtained into the vagina, with thigh abduction and without the exam iner manipulating the labia majora (Hobbs, Wynne and Thomas, 1995).61 This subjective term is best avoided. A narrow ing of hymenal tissue is referred to as attenuation of the hymen. The narrowed rim is usually found at the poster ior edge and should be confirmed with the child in the knee-chest position. Although usually associated wit h chronic abuse, this term is comparative and implies that the examiner has prior knowledge of the hy menal anatomy. Unless the child has been examined previously it is not pos sible to determine w hether the hy menal rim is narrowed. Heger et al 48 defined a narrow hymenal membrane as the w id th of th e membra ne viewed in the coronal plane (i.e. from the edge of the hymen to the muscular portion of the vaginal introitus) of less th an 2 mm . A 'thickened edge' was a term used to describe the relative amount of tissue between the internal and external surfaces of the hymenal membrane. Using these terms, the hymenal rim was found to be thickened in 45.5 per cent; irregular in 51.7 per cent and narrowed in 22.4 per cent of 147 prepubertal girls screened for non-abuse. Over three-quarters (78 .7 per cent) of girls with a narrowed hym enal rim were found to be overweight (> 75th centile for weight). The posterior hymenal width cannot be measured accu rately, especially when the rim is narrow. However, a com plete absence of the posterior hymen (confirmed in the
----- ---------
Normal fem ale genital anatomy I
35
knee-chest position) has not been reported in new borns or in prepubertal girls selected for non-abuseY
Hymenal Measurements Owing to the absence of a single diagnosti c test for ch ild sexual abuse, physicians have attempted to establish sta nd ards of dia gnosis by develo ping objective criteri a, such as the measurement of the hyme nal open ing. Much of the early literature concentrated on small differences in the diameter of th e hymenal opening. One criterion freq uently cited as dia gnosti c of child sexual abuse was the presence of a hymenal openin g in excess of 4 mm. 62,63 Later studies have show n that the hymenal ap pea ran ce and diameter vary with the age of the child, 52,53 as well as wi t h the position of the child during examin ation and the examination technique. 54 In a cohort of non-abused pre-pubertal children, about one third had a hymenal openi ng size of g reate r than 4mm.48 The kn ee-chest position and th e supin e labial trac ti on method were sup erio r to the supine lab ial separation tech nique in opening the vaginal introi tus. The largest vertica l transhymenal diameter was produced in the kn ee-chest position, whereas the greatest transverse horizonta l diameter was produced by the supine lab ial traction procedure. Other fa cto rs that influence the hymenal diameter include th e hymenal configuration, state of relaxatio n of the child an d the skill and ex perience of the examiner. Examiners should expect to vis ualize the intravaginal contents during labial traction, especially when the child is relaxed. Measurements obtained und er general an aesthesia will be greater tha n those taken when the child is awake. The hymen is an elas tic and dy namic piece of tissue, wit h its ap pearance and size changing du ring normal physiological activities su ch as breathin g and coughing. The effects of examination tech niqu e, as well as genera l anaesthesia, on the configuration and size of th e hymenal orifice are shown in Figs 2.11-2.14.
Figure 2.11 Labial separation shows a flesh y annular hymen in a 5-year-old girl.
Figure 2.12 Same child as shown in Fi g. 2.11. Labial traction reveals an enlarged hymenal opening, allowing a cle ar view of the vag inal wa lls.
Figure 2.13 Labial tra ction showing an annular hymen with perih ymenal bands in a 2-year-old girl.
~ Figure 2.14 Same ch ild as in Fig. 2.13, examined under general anaesthesia, using the same technique. Note the dilated appearance of th e hymen .
F'
36 I
Investigation of suspected sexual abuse
With t he onset of puberty, hymenal elasti city increases to such a n extent that penile penetration can occur without any sig ns of trauma. Meas urement of the hymenal dia meter then becomes an unreliable a nd useless exercise. Hymenal measurements in non-abused children have provided data to refin e a normal database against which to comp are the anatomy of suspected child abuse victims. However, a nar row focus on the hymenal diameter as a criterion for sexua l abuse has the potential to do more harm t han good. 54 - 55 There is subst antial overlap of mean transverse hymenal diameters between prepubertal sex ually abused girls and t hose selected for non-abuse. The hymenal diameter is non-discriminatory for sex ual abuseY Figure 2.15
A visi bly relaxed anus, with
venous
pooling seen at
the end of an examination in the supine knee-chest position .
PERIANAL FINDINGS The skin of the anal verge and epithelium of the anal canal exhibit regular folds when the anus is closed. Midline anterior or posterior smooth areas in the anal verge are common find ings, where the defe ct is lined by unbroken skin and repre sents a gap in the fibres of the superficial part of the external anal sphincter. The significance of lateral defects in the anal verge is unclear. The perianal skin may look red and moist, due to poor hygiene, irritation, inflammation, infections or infestations. Caution must be exercised when interpreting perianal hyperpigmentation in non-white children. The plexus of veins underlying the peri anal skin can be seen to distend if examination is prolonged, producing a purple-blue discolouration around the anus (venous poolin g). This is a normal finding, caused when buttock traction and increased intra-abdominal pressure occludes venous drainage (Fi g. 2. 15). Haemorrhoids in children are extremely rare. Intermittent cont ractions of the external a nal sphincter can sometimes be seen (twitching or winking) , a sign which has been observed in normal children. This must be differen tiated from reflex anal dilatation, when the anus opens widely after a brief period of buttock separation. Reflex a nal dilatation is an incompletely understood phenomenon that must be interpreted with caution a nd in contex t. It is best demonstrated by the buttock separation test in the left lateral position, when the buttocks are parted gently and observa tion takes place for 30 seconds. A positive result is obtained when the ex ternal anal sphincter contracts briefly, follow ed by dramatic rel axation of both the external and intern al sphincters to reveal a cylindrical opening into the rectum (Figs 2.16-2.18). Although first described as indicative of anal abuse,57 the significance of this sign rema ins controver si al. Dilatation of up to 10 mm has been reported in up to 49 per cent of 267 normal children. 58 Other authors have reported a prevalence of the sign in 4-14 per cent of non abused children. 59 ,7o Reflex anal dilatation is observed more freq uently when the exam ination is prolonged, in the knee-chest position, and in children who have a history of constipation. 7I ,72 Chronic straining at stool may lead to weakness of the anal sphincter, and a 'visibly relaxed ' anus
..-:-
-----
-
-
-~
-
-
-------
~-
Figure 2.16
Gross 'reflex anal dilatation', allowing a clear view
of the anal canal. The child was examined in the supine knee-chest position.
Figure 2.17
Same child as in Fig. 2.16. Appearances following
anal dilatation, showing perianal erythema and loss of anal folds.
- - - -----
Acute, healing and healed anogenital trauma I
37
significance of medical findings; venous engorgement was observed in 52 per cent of the children, after 2 minutes in the knee-chest position; some dilatation of the anus was observed in 49 per cent of the children. The frequency of these findings were confirmed in a cohort of 89 children of less than 18 months 01d. 74 Findings infrequently noted in the absence of abuse included fissures (in the absence of a recent history of constipation), skin tags outside the mid-line, and irregularity of the anal orifice without dilatation. Anal/ perianal bruising, anal lacerations, tears and scars have not been observed in the absence of abuse. 47
ACUTE, HEALING AND HEALED ANOGENITAL TRAUMA Figure 2.18 Same child as in Figs 2.16 and 2.17. Follow-up examination 3 weeks later. Scar tissue is seen in the 12 o'clock position. (Fig. 2.15) has been described when stool may be present in the rectum. Other situations when anal dilatation is observed include the period following defecation, neurogenic bowel disorders, use of musc]e relaxants during assisted ventila tion, general anaesthesia and during post-mortem examina tions. Under these circumstances, the sphincters are relaxed and the anal dilatation is passive. They should not be misdi agnosed as reflex anal dilatation, which is a dynamic response. Reflex anal dilatation remains a controversial, highly variable and non-specific physical sign. Early reports on children who had been anally abused described erythema, swelling, venous engorgement, skin tags, wedge-shaped smooth areas and anal dilatation. 67 .?3 Many of the physical signs, however, are minor and non specific but can be used to corroborate the child's history. They should not be used in isolation, as they have also been reported in cohorts of children who have not reported abuse (Table 2.3). In a study of 267 prepubertal children selected for non-abuse, McCann et al 68 encountered a relatively high incidence of perianal soft tissue changes, and urged exam iners to exercise caution when rendering an opinion on the Table 2.3
Controversy remains about the frequency of abnormal find ings in children examined for sexual abuse. Different study populations and the accessibility of clinics to the children who disclose abuse when the injuries can be documented can explain some of the differences. Public and professional awareness of child sexual abuse will lead to earlier disclos ures in children and therefore result in more abnormal findings. The conduct and timing of the medical examin ation, interest, experience and expertise of personnel involved also differ between centres, and depend on the presenting complaint. Children and adolescents may pre sent with a variety of symptoms and signs to health profes sionals at busy emergency departments or outpatient clinics, where the diagnosis of sexual abuse is not sus pected. Disclosures may be made to staff at educational establishments, social services or law enforcement agencies. When appropriate child protection procedures are followed, the medical examination is planned and forms part of the investigative process. However, the medical examination on its own must not be relied upon to diagnose sexual abuse. With the introduction of standard terminology and photo-documentation, practice and research have become more reliable. The rates of 'abnormal findings' have dropped as previously reported abnormal findings have
Perianal findings in non-abused children
Author
Number of subjects
Age
Summary of findings
McCann etal1989 68
161 girls, 106 boys
2 months-11 yea rs
Berenson et al 1993 74
89 girls
Mean 10.7 months
Erythema 41 0/0 Hyperpigmentation 30% Venous engorgement after 2 minutes of examination 52% Midline wedge-shaped smooth areas 26% Anal skin tags/folds 11% Anal dilatation 49% Smooth areas 26% Hyper/hypopigmentation 10% Erythema 7% Skin tags 3% Venous pooling 1%
-
--------~~
-
-
--
=-
--
~
38 I
Investigation of suspected sexual abuse
been recognized as non-specific findings, present also in control subjects who have been carefully selected for non abuse. It is also important to remember that some forms of child sexual abuse will leave no physical signs. Other fac tors that influence physical findings include the position of the child in relation to the abuser, the size of the object inserted relative to the size of the orifice, the degree of force used, the use of lubricant, the frequency and duration of abuse, as well as the period of time that has elapsed between the last episode of abuse and the examination. Erythema and superficial abrasions may disappear within 2 days, and bruises fade after a few days. A single episode of penetrative anal abuse may result in perianal svvelling and a gaping anus, signs that will resolve within 1 week. There is now good documentation of norm a l variants in anoge nital anatomy, but literature on the healing process of the traumatized anogenital region is scant. In general, hea ling is rapid and occurs through rege neration and repa ir. Wound healing by regeneration is complete within 48-72 hours, and normal appearances are restored by 6 weeks. Deeper injuries heal by repair, a process involving granulation and scarring that takes approximately 2 months. Because of the frequent delay between the alleged abuse, disclosure and medical examination, there is a need to understand the heal ing process, how an injury might affect the genital anatomy of the prepubertal girl and whether signs of trauma persist after the onset of pubeliy. Finkel 75 reported on the findings of seven children who experienced acute genital and anal trauma, and were fol lowed up until their injuries were healed . All of the chil dren had acute signs of injury documented after the incident, most of which were superficial lacerations. In only one child, who had a deep lacerated penetrating wound, was there any sca r formation that distorted the "hymenal and perihymenal tissues. Obvious acute superfi cial trauma healed without residua, by a process of regen eration. Healing was rapid, and by the second examination (3-13 days later) t here was little apparent sign of injury. McCann et al76 followed the resolu tion of acute genital injuries in three children who had been sexually assaulted. Initial signs of injury included erythema, oedema, submu cosal haemorrhages, abrasions, lacerations and hy menal transections. Acute tissue changes resolved quickly, with ery thema and oedema disappearing by 11 days, whereas submu cosal haemorrhages resolved within 27 days. The shaJ1l jagged edges of the damaged hymen gradually smoothed out as healing took place, and there was relatively little scarring in the hym en or posterior fourchette. With the onset of puberty, the hymenal changes in one gi rl were obscured by hypertrophy of the hymen. This highlights the importance of using techniques to enable a thorough examination of the fimbriated, redundant hymen seen in pubertal girls. The natural history of perianal injuries has also been studied in four children. Acute signs seen included ery thema and perianal oedema, which caused the folds to thicken or flatten . Venous congestion was present in three
-
-
~
--
-
- -----
-
-- - - - -
-
children, and two children had slight anal dilatation. Va ri ous degrees of laceration were noted in all the children. By the eighth day, acute signs of trauma were present in only one child. Second-degree lacerations healed over 5 weeks, leaving na rrow bands of scar tissue; however, deeper lacer ations took 12-14 months to hea l. Even deep injuries faded into surrounding tissues with time, making detection of periana l injury increasingly difficult. 77 More recently, the healing patterns of anogenital injuries in 94 prepubertal children were described in a longitudinal study over 10 years. 59 The authors confirmed that anogenital trauma healed quickly, often without residua. Of the 171 acute irUuries see n, diagnostic changes remained in only 25 (14.6 per cent). All trauma to the labia majora or minora healed without residua. Injuries to the posterior fourchette and fossa navicularis (24 out of 47) healed with non-specific residua (labial fusion, vascular changes and scarring) . Hymena l tran sections did not heal unless repaired surgically, whereas 5 out of 37 partial tears of the hymen healed with residual sha llow notches. Only J out of 31 anal injuries healed with anatomic changes (tag, scarring and hypeJ1ligmentation).
FEMALE GENITAL FINDINGS IN SEXUAL ABUSE Over the past decade a great deal of information has been gathered by medical exam iners who see children who have been sexually abused, and a wide range of findings have been reported. It is well documented tha t ab normalities of the genital tract associated with abuse may be minor and non-specific, occurrin g a lso as a result of inflammation, infection or poor hygiene. Although Emans et a l58 found that a number of genital findings were significantly more common in sexually ab used children than in asy mptomatic girls, many of the findings were also present in girls who were examined bec ause of symptoms relating to vaginitis, vulvitis, bleeding or dysuria. Debate continues regarding the frequency of ab normal findings in sexually abused children because of difficulties in reaching an agreement on what constitutes an abnormality and the lack of a true go ld standard diagnostic test for sexual abuse. There may be no physical findings in more than one-half of the chil dren examined for suspected sexual abuse. The absence of physical signs neither confirms nor negates a diagnosis of abuse. An earlier study by Muram 78 on 31 girls whose per petrators confessed abuse reported that the medical exam ination failed to detect any abnormalitY in 29 per cent of the girls. In a case review of 236 children whose peJ1letra tor had received conviction for sexual abuse, Adams et al 79 found that the genital examination was normal in 28 per cent, non-sp ecific in 49 per cent, suspicious in 9 per cent and abnormal in 14 per cent of cases. Abnormal anal find ings were found in only 1 per cent of patients. Using dis criminating ana lYSis, the two factors that significantly correlated with the presence of abnormal genital findin gs
- - - -----
---
--
Signs of anal abuse I in girls were (1) the time since the last incident and (2) a history of bleeding being reported at the time of ab use. Among the clinical findings that require diagnostic interpretations are concave va ri ation s of the hymenal rim. These have been referred to as clefts, notches, tears, fis sures, transections, healed laceratio ns or hymenal attenua tions. The cha racteristics and clinical correlates of concave hymena l variations were described in a large cohort of 13 83 prepubertal and adolescent girls who alleged sexual abuse. Concavities in the hymen were found in 174 patients (12 .6 per cent), of whi ch 3 5 per cent were anterior. Of these, two -thirds were smooth and curved, wh ereas one third were angular and /or irregular. Posterior or lateral concavi ties were found in another 57 per cent of the girls, of whom one-third were curved and smooth, and two thirds were angular and/or irregu lar. Multiple hymenal concavities were found in 13 girls. Each hymenal concav ity was categorized by location, contour (curved or angu lar) and smoothn ess (smooth or irregular). The concavities were fUlther sorted into one of five categories, reflecting the presence or absence of the clinical impression of anogenital tra uma: benign, benign with reservations, uncertain, suggestive of trauma , and trauma. Poste ri or and/or lateral location of the concavity, angular contour and rim irregu larity of the concavities we re found to be significantly associated with trauma .60 Abnormal genital findings are more likely to be documen ted when the girls report bleeding at the time of the assault, or when the examina tion occurred within 72 hours of the last episod e of abuse. so The vulvovaginal signs of ab use have been summarized by a working party of the Royal College of Physicians of London,12 with emphasis on the type of abuse and t he timing of the examination (Table 2.4). The
Table
2.4
Vulvovaginal signs of abuse'
Non-specific signs when seen within 72 hours of abuse Erythema Bruising Superficial laceration/a brasion Oedema Signs supportive of abuse Notch/cleft in posterio r hymena l rim, which may be associated with scarring Scar in posterior fourchette Hym enal opening> 15 mm due to attenuat ion of hymen Signs diagnostic of penetrating injury Fresh laceration of hymen Old tear in hymen that may have healed with scarring and interruption of hymenal rim Attenuation of hymen/disappearance of hymenal rim, usually posteriorly
'Adapted from ref.
46.
39
1997 publication has been replaced by an evidence-based review of physica l signs of child sex ual ab use.47 These signs are currently subject to further scrutiny in the light of more recent studies that suggest that the geni tal examination of the abused child rarely differs from that of the non-abused child. Using a case-control study design, 192 prepubertal children who alleged penetratio n were examined using a colposcope, and the photographic records we re compa red with 200 control subj ects who were matched for age and ethnicity, and who were found not to have been abused after careful screenin g. Abnormal find ings (hymen al transection, perforation or deep notch) were seen in only 2.5 per cent of subjects or four children, aLi of whom gave a history of penile or digita l penetration. How ever, only 17 ab used children were examined within 1 week of the abuse.56 A prospective study of 2384 children referred to a tertiary centre from 1985 to 1990 for possible sexual abuse found that 96.3 per cent of all children had a normal medical exa min ation. Even with a history of vaginal or anal penetration, the rate of abnormal medical signs was only 5.5 per cent, seen in 61 out of 444 (6 per cent) girls and 2 out of 251 (1 per cent) boys. Most children in this study were examined within 1 week of abuse. s1 In a study of female sexua l assault victims , 94 per cent of 200 women were found to have trauma at one or more locations - posterior fourchette, labia minora, hymen and fo ssa navicularis. Injuries were found to vary by site, with tears appearing most often on the posterior fourc hette and fossa navicularis, abrasions on the labia, and ecchymosis on the hymen. s2 However, the mean age of the cohort was 24 years (range 11-85 years) and the majority were exam ined within 24 hours of the assa ult, factors that a re atyp ical of children who disclose sexual abuse. Eighty-four (88 per cent) women returned for follow-up examinations within an average of 25 days (range 4- 50 days). All injuries had resolved , with no evidence of t rauma in 71 (87 per cent) women. Tears to the hymen and fossa navicularis did not reunite, and no scarring was observed at follow-up. Kellogg et al s3 conduc ted a retrospective case review of 36 pregnant ado lescent gi rls (age range 12.3- 17 .8 yea rs) who presented for sexual abuse exa minations. Despite definitive evidence of sexua l contact (pregnancy), evidence of penetrative trauma (clefts in the posterior, one-half of which extended to the base of the hymen) was found in only 2 of the 36 girls, whereas four had suggestive findings (deep notches in the poste rior half that did not extend to the base of the hymen, or visible scars) .
SIGNS OF ANAL ABUSE Non-specific signs that may be seen in the acute stages include erythema, bruising and oedema, which vary in exte nt depending on the degree of friction and force involved. Bruising may result in a haematoma formation at the anal verge, caused by sheari ng of the subcutaneous
40 I
Investigation of suspected sexual abuse
venous plexus. Fissures or tears may be single or multiple, seen as superficial or deep clefts in the perianal skin or mucosa, which, in the acute situation, are painful. The fis sure is usually triangular or fan-shaped, with its apex point ing towards the anal canal and the lower edges pulled apart by the fibres of the external sphincter. There may be bl eed ing or anal spasm associated with the fissure, or on its own. Deep fissures are likely to heal with scarring and there may be an associated skin tag caused by skin overgrowth during healing. Opinions differ as to the significance of anal fissures in relation to abuse, and the presence of a fissure gives no information on its aetiology. Fissures resulting from consti pation cannot be distinguished from those caused by anal abuse. They have been reported in normal cohorts of infants and children,68,74 and also in children with constipation .lI ,72 Chronic trauma may lead to thickening of the anal canal folds. There may be squamous metaplasia of the columnar epithelium. The perianal tissues look and feel calloused, and hypertrophy of the external sp hincter results in a prominent ring of tissue surrounding the anal orifice. This appeara nce has been referred to as the 'tyre sig n',67 a nd is illustrated in Fig. 2.19. An isolated finding of reflex anal dilatation is of no diagnostic significance unless it is > 15 mm in size and reproducible. When seen in association with a history of anal abuse, especially in the presence of other signs such as fissures, it supports the diagnosis. In the abse nce of a dis closure from the child, follow-up is justifi ed. A summary of the perianal signs of abuse is shown in Tab le 2. 5. These signs, from the 1997 publication by the Royal College of Physicians, have been replaced by an evidence-based review of physical signs of child sexual abuse .47
CONDITIONS THAT IVIIIVIIC ABUSE General paediatlicians are often asked to see children because a physical finding may raise concerns of sexual abuse. A lack
Figure 2.1 9 'Tyre' sign, showing thickened anal folds with venous congestion.
-.:,,1
--~-
--~-
---
-
----
--~
-~
Table 2. 5
Perianal signs af abuse'
Non-spe cific acute changes
Erythema Oedema Fi ssures Venou s co ngestion Bru isi ng Signs supportive of abuse
Anal laxity without other explanation Reproducib le reflex anal dilatation> 15 mm Chronic changes, i.e., thickening of anal skin verge, increased elasticity and reduced anal sphincter tone Bite marks Signs diagnostic of blunt force penetrating trauma
Fresh lace ration Healed sca r extending beyond anal margin on to perianal sk in in the absence of a reasonable alternative explanation -Ada pted from ref. 46.
of fa miliarity with the normal female genital anatomy, as well as the manifestation of infection, common childhood conditions and accidental trauma, will cause physicians to misdiagnose sexual abuse. In the atopic individual the vulval skin can be moist and inflamed, as a result of contact derma titis from use of cosmetics or washing material (bubble bath, biological soap powders). Vulvovaginitis is the most common gynaecological complaint in childhood, and major causes include poor hygiene, threadworms and infec tion with Candida or group A beta-haemolytic streptococcus.8486 Soreness and itch lead to scratching or rubbing, and bleeding may be a presenting complaint. A prolonged history of 'vulvovaginitis' may be a pre sentation of lichen sclerosus, often misd iagnosed as recur rent thrush. This is a disease of unknown aetio logy and with an unpredictable course. Affected areas include the external genitalia and perianal skin. In boys, th e prepuce and glans penis may be more variably involved. The child may present with bleeding associated with pain and itch, and on exa mination the perineal skin is thin and friable, with white shiny macules. There may be vascula r or pur puric areas, sup erfici al abrasions, haemorrhagic bullae, erosion and ulceration . Usua lly the affected skin is sharply demarcated from the surrounding normal skin, rarely extendin g beyo nd a figure-of-eight distribution around the labia and anus. Lichen scl erosis has been mistaken for sex ual abuse 87-89 Submucosa l haemorrhages in lichen sclero sus are seen in Fig. 2.20. Physicians in the children's emerge ncy department are sometimes asked to see children who present with perineal injuries, and the question of sexual abuse may arise. Unin tentional injuries to the perineum in children are usually accompanied by a witnessed account of the event and are usually superficial. Hymenal injuries are rarely the result of
----------
Screening for sexually transmitted infections
Figure 2.20 Lichen sclerosis. Cutaneous and submucosal haemorrhages seen in the posterior fourchette and vestibule. The patient was a 6-year-old girl who presented with pain and bleeding 'down below:
accidental injury, when the labium minomm is the most frequent structure involved. Deeper injuries tend to be anterior or lateral to the hymen. In a multicentre study of 56 prepubertal girls who sustained unintentional perineal injuries, the hymen was involved in only one child who fell in a park with her legs abducted. A pinpoint abraded area was found on the hymen at the 3 o'clock position. 9o Congenital abnormalities such as haemangiomas and failure of fusion of the median raphe raise concerns of abuse when they first present. The raphe is a flesh-coloured. slightly raised, linear structure in the perineum, and when failure of fusion occurs the opposing edges may split and appear like a traumatic lesion, with bleeding. Midline avas cular areas in the posterior part of the vestibule or posterior fourchette may be seen as a white line (linea vestibularis) or a white spot (partial linea vestibularis), and is easily mis taken for scar tissue. They have been observed in 25 per cent of newborn girls,91 and can vary in size as well as con figuration in the first year of life. 92 In contrast with scar tis sue the linea vestibularis is generally in the midline, without accompanying disturbance in vascularization. Normal vari ants such as bumps and notches between 3 and 9 o'clock on the hymen can be mistaken as residua of sexual abuse. Urethral prolapse is a circular eversion of the mucosa that usually occurs at the urethral meatus without accompanying symptoms. Some bleeding may occur following straining, and assumed by carers to have come from the vagina. 9 ] Care ful examination reveals an oedematous area anterior to the hymen, through which the urethral opening may be identi fied (Fig. 2.21). Crohn's disease can involve any part of the alimentary system, and one presentation consists of extra intestinal manifestations. There may be perianal and vulval ulceration with oedema, suggesting trauma. 94 ,9 5 Clinicians may also have to sort through histories given by adults who have misinterpreted nomlal childhood masturbation or
-
I
41
~ Figure 2.21 A 10-year-old girl with a 4-year history of 'vaginal bleeding'. An oedematous and erythematous 'polyp' is seen anterior to the hymenal opening. The appearances suggest a prolapse of the urethra.
sexual play between children, and by parents engaged in custody disputes. The normal, age-appropriate sexual exploitative behaviour must be differentiated from disturbed behaviour arising from child abuse.96.97
SCREENING FOR SEXUALLY TRANSMITIED INFECTIONS Debate continues as to whether the screening for STls should be routine in the evaluation of sexual abuse. The presence of a sexually acquired organism can indicate prior sexual contact in a child, and when accompanied by other indicators of sexual abuse supports the diagnosis. Non-sex ual transmission of STls is rarely an issue in adults, but when the same diseases are found in children, there is a tendency to attribute them to an asexual mode of transmis sion. 98 The immature anogenital tract is more vulnerable to infection, especially if there is a breach of the mucosal lin ing following traumatic abuse. The low numbers of children reported to have acquired STIs from sexual abuse may rep resent the lack of systematic screening, or the non-recogni tion by clinicians that children can be infected through sexual abuse. However, other routes of transmission that must be considered, although difficult to exclude are: • perinatal acquisition from an infected mother who may be asymptomatic (Chlamydia trachoma tis, Neisseria gonorrhoeae, Trichomonas vagina/is, herpes simplex vims, human papillomavirus [HPV], human immunodeficiency virus [HNJ); • non-sexual adult to child contact (HPV, herpes simplex vims); • auto-inoculation (HPV, herpes simplex virus). The risk of a child or young person acquiring an STI depends on the prevalence of STis within the local population: maternal STI during pregnancy that might lead to vertical
42 I
Investigation of suspected sexual abuse
transmission; the type of sexual contact during abuse ; injuries to the genital tract; the sexual maturity of the victim; and whether a condom was used during abuse. Transmission from mother to child is welJ documented for most STIs and the organism can lie dormant for up to 2 years (and possibly 3 years). In dealing with a child who is less than 3 years old, it is therefore important to establish if the parents are infected and to screen them if possible. Sexual abuse should be sus pected when an infection is diagnosed after infancy and before sexual activity occurs in the older child. Under these circumstances, gonorrhoea and syphilis would be diagnostic of sexual abuse, whereas infection with Chlamydia, Tri chomonas, herpes simplex virus and HPV would be suggestive of sexual abuse. 12,44,99,100 A prospective study of 1538 children who were examined for possible sexual abuse found the overall prevalence of STIs to be 6 per cent. The diagnosis of a STI was highly cor related with a history of sex ual contact (alleged by 49 per cent of the children) and the presence of a discharge. l01 The authors had attempted to exclude perinatal acquisition as a source of infection by selecting verbal children who could understand questions regarding sexual contact. Although the prevalence of STIs was only 3.7 per cent in a cohort of 159 girls who had been sexuaJly abused, Robinson et al 102 also found a significant association of sexually transmitted organisms with the presence of a vaginal discharge. The authors recommended that the presence of vaginal discharge in sexuaJly abused girls was an indication for STI screening. The use of an algorithm to assess the risk of gonococcal and chlamydial infections was found to reduce the cost and trauma of unnecessary sampling. Major factors in the risk assessment were genital-genital or genital-anal contact or penetration, suspicious anogenital findings, genital dis charge and suspicion of an STI in the perpetrator, although vaginitis and referral for suspected abuse in a sibling were designated minor factors. Restricting testing to children with at least one major or two minor factors in their study population of 3040, 45 per cent would have been tested and all known cases of gonorrhoea or chlamydia infection would have been identified. Testing would have been avoided in 51 per cent of girls and 72 per cent of boys. 103 More recently, the Association of Genitourinary Medicine and the Medical Society for the Study of Venereal Disease in the UK produced guidelines for the management of sus pected STls in children and young people. The recommen dations included the need for sta Ff worki ng in genitourinary medicine to be: alert to the possibility of child abuse and neglect ; aware of local guidelines; and trained in child pro tection procedures and protocols. Screening should be con sidered in all young people who may have been sexually abused or who have been found to have an STI. 104 Human papillomavirus infection is one of the most common STIs, and is estimated to affect 10 per cent of the adult population. Subclinical infection is common, and the increasing incidence of condyloma in children probably reflects the increased prevalence of HPV disease in the
~
---
-
'-
-
-
-
-
---~-
adult population. Results from studies of children with anogenital warts suggest that the majority of children do not acquire these sexually. Vertical transmission is an important source of infection in young children, and household members of children with anogenital warts should be routinely screened for the presence of non-geni tal as well as anogenital warts. 105.106 At present, there is no reliable method to determine the mode of acquisition in a child with HPV infection. Children who disclose sexual abuse should be screened for STIs under the following circumstances: • presence of symptoms and signs at the time of assault or subsequently (bleeding, discha rge, dysuria); • suspicious anogenital findings; • history of genital-genital or genital-anal contact or penetration; • alleged abuser known to have, or at high risk of acquiring, STI; • STI prevalent in the community; • request from child or young person. The forensic science laboratory will not examine samples obtained for screening of STls. For each organism, sampling techniques and transport media should be individually tai lored and must meet the criteria for maintaining the chain of evidence. Swabs must be taken by a doctor with appropriate training and experience. lt is therefore important to establish good liaison with the diagnostic laboratories, as well as physicians in genitourinary medicine. Fine, wire-mounted urethral swabs can be passed through the hymen, even in young children, to enable cultures for STls. 107 A vaginal wash procedure, using stelile saline introduced through tubing from a butterfly needle, has also been found to be acceptable as a diagnostic test for STIs in prepubertal girls.!08 Testing for infection with HlV involves a blood test for HlV antibody, obtained at the time of the initial examination and at follow up. Newer, more sophisticated tests for HfV-RNA or proviral DNA need only be done in a young child to exclude vertical transmission. 109 Serological testing for infection with Tre ponema pallidum, hepatitis B or hepatitis C virus should be considered on an individual basis. The child or young person may require active and passive immunization against hepati tis B infection as well as antiretroviral therapy. Expert opin ion should be sought from an infectious diseases specialist. In general, there have been relatively few studies where children with a particular STI have been evaluated for the possibility of child sexual abuse. This has resulted in a lim ited evidence base to determine whether a particular STI is a marker for sexual abuse 47
INTERPRETATION OF CLINICAL AND LABORATORY FINDINGS Many changes have occurred in the medica l evaluation of children suspected of having been sexually abused, and
-----
Summary I
how physical findings are interpreted. A comprehensive list of the clinical and laboratory findings seen in abused and non-abused children was first drawn up by Adams et al. 43 Sometimes known as the Adams Classification System, it has been developed over the years and the latest revision 45 is the result of examining published data and consensus amongst 18 experienced physicians. The classification sys tem can be summarized as: • findings documented in newborns, or commonly seen in non-abused children: - normal variants; - findings commonly caused by other medical conditions. • indeterminate findings (owing to insufficient or conflicting data): - physical examination fmdings, for example deep notches or clefts in the hymen; smooth, uninterrupted rim of hymen of < 1 mm wide; immediate dilatation of the anus> 2 cm; - Lesions with indeterminate specificity for sexual transmission: for example, anogenital condyloma accuminata, anogenital herpes. • fi ndings diagnostic of trauma and/or sexual contact: - acute trauma to external genital/anal tissues, for example fresh laceration or extensive bruising; - residual (healing) injuries, for example perianal scar; - injuries indicative of blunt force penetrating trauma: for example acute laceration, extensive bruising, healed hymenal transection; - presence of infection that confirms sexual
contact, for exam pIe gonorrhoea;
- diagnostic of sexual contact: for example, pregnancy or sperm identified in specimens taken directly from the child's body. Until a better evidence base is available, this system provides a useful tool that both assists physicians in inter preting clinical findings and helps to achieve some consist ency in terminology.
SUMMARY Whenever abuse of a child is suspected the clinician must refer to local multi-agency child protection procedures for appropriate action. Siblings and close friends of the victim must also be interviewed and examined if appropriate, as it is likely that they may also have been exposed to the alleged perpetrator. 110 There may be an association between different types of abuse, and following a diagnosis of any form of abuse a medical assessment for sexual abuse should also be considered. In the investigation of sexual abuse of children, clinicians should remember that the medical examination is only a part of the jigsaw. Physical evidence is neither essential for, nor
43
predictive of, conviction. The child's history is the single most important factor in the accurate diagnosis of most cases of sexual abuse. III ,11 2 In a review of child sexual abuse criminal court cases, those involving the youngest victims were found to have a significantly lower conviction rate. Successful prosecution, particularly in the youngest chil dren, depended on the quality of the verbal evidence and the effectiveness of the child's testimony. II ) The literature reports a marked variability of medical findings in children examined for sexual abuse and the diagnosis of sexual abuse cannot be made on the basis of physical findings alone. It has been suggested that a model could be developed for the prediction of anatomic findings, based on an inventory of variables that would include the age of the child, type of contact, whether penetration occurred, and a history of pain or b1eeding.1l4 Expert medical testimony may help to interpret the presence or absence of physical signs, but the overall effect on the legal outcome is unknown. Increased communica tion between prosecutors and paediatricians may improve the outcome, especially if the paediatrician is told whether she is a witness to fact, or an expert witness. The distinc tion is important to allow adequate preparation and to avoid an injustice to the child. A medical expelt should be able to demonstrate relevant training or experience in child abuse cases that are similar to ones in which he or she has been called on to provide expert testimony. Irresponsible medical testimony must be avoided; this includes use of unique theories of causation, unusual interpretations of medical findings, alleging non-existent physical signs, deliberate omission of pertinent facts or knowledge and misquoting of medical literature H5 Medical , social and legal professionals have relied too heavily on the medical examination in diagnosing sexual abuse. Normal findings are consistent with abuse, and the examiner must document all signs, positive or negative, whenever any child is examined. There is now a consensus on terminology and interpretation of findings in child sex ual abuse, 12,44,45,47,48,56 but these guidelines are only helpful if examiners are meticulous in documenting their findings. There needs to be an agreement on the minimum standards for the training of medical professionals who perform sex ual abuse examinations. 9 ,37,116,117 Medical care for children and young people who disclose abuse has evolved into a complex and technical specialty. Expert medical testimony can be crucial to the legal outcome of a case of alleged sex ual abuse. The evaluation of child sexual abuse is not rou tinely taught in undergraduate or postgraduate training programmes. Specialized training programmes must be developed to ensure better research as well as the dissemin ation of knowledge and expertise.37, 11 8 In September 2006, the American Academy of Pediatrics recognized child abuse as a new pediatric subspecialty. It is hoped that the accred itation of trained specialists will benefit child victims of sexual abuse, families, health-care professionals and those in social services and law enforcement.
44 I
Investigation of s us pected sexual abu s e
REFERENCES
2
3 4 5
6
7 8 9
10
11
12
13 J4
15
16
17 18
19
20
21 22
23
, --
24
Sch echter IVID, Roberge L. Sexual ex ploitatio n. In Hel fe r RE, Kempe CH (ed s) Child A buse and Neglect: Th e Family and th e Community. Ca mbridge, MA: America n Academy o f Pedia tri cs, 1978. Ja FFe AC, Dynn eson L, ten Bensel RW. Sex ual a bu se of children: a n epi demiological study. Am J Dis Child 197 5; 129 :689 - 92. Kemp e CH. Sexu al a buse, a nother hidd en pedi a tri c probl em. Pedia trics 1978; 62 :382-9 . Finkelhor D. Current in formation on the scope and na tu re of child sexual abuse. The Future ofCIlildren J994; 4:3 1-5 3. Jones LM, Finkelhor D, Ko pie c K. Why is sexual a bu se declin ing? A survey o f state chil d pro tecti on a dministra tors. Child A buse Et Neg lect 2001 ; 25 : 11 39- 58. Dunne MP, Purdi e DM, Cook MD et al. Is child sex ua l abuse declining? Evid ence Fro m a popul a tio n-based survey o f men and wo me n in Australi a. Child A buse & Neg lect 2003 ; 27:141- 52. Hobbs CJ , Wy nne JM. Child sexual abu se: an in creasing ra te of di ag nos is. Lancet 1987; ii: 837- 4 5. Summit RC. Th e child sexual abuse acco mmodati on sy ndrome. Child A buse & Neglect 1983; 7 : 177-93. Mo k JYQ, Busuttil A, Hamm ond HF. The j o in t paedia tric fo rens ic exa min ati on in child a buse. Child A buse ReIJ 1998 ; 7 :194-203. Levitt CJ. Sex ua l abu se in children : A compassiona te ye t thorough approach to eva lua tion. Postg rad IVIed 198 6 ; 80 :201-15. De San Lazaro C. Ma king paedi atric assessment in susp ected sex u al abus e a therap euti c ex perience. Arch Dis Childh 199 5; 73 : 174- 6. Th e Royal CoJl ege o f Pa edia trics and Child Hea lth and Fac ul ty of Foren sic a nd Legal Medi ci ne. Guida nce all pa ediatri c
25
26
27
28
29
30
31 32 33 34
35
36
f orensic examina tions in relation to poss ib le child sexu.al abuse. September 2004. www.rcpch .ac. uk/doc.aspx? id_
37
Resource=3 409 Roya l College of Paediatri cs and Child Hea lth . Child protection companion. Lond on, RCPCH. Kendall-Tacke tt KA, Meye r Williams L, Finkelhor D. Impa ct of sexu al abuse on children: A review and synthesis of rec ent empirical st udi es. Psychol Bull 1993; )] 3 : 164 -8 0. Be rliner L, Elli ott DM. Sexual a bu se of children. In Briere J, Be rliner L, Bulkley C et al (eds) The APSAC handbook on child maltreatment. Thousa nd Oaks, CA: Sage, pp. 51-71. Lev itt CJ. The medica l ex amin a tio n in child sexua l abuse: A ba lance betw ee n history a nd exam. J Child Sex Abuse 199 2; 1:11 3-21. Myers JEB. The ro le of the physici an in presen tin g verbal evid ence of child abuse. J Pediatr 19 86; 109 :409 -11. Ferrell J. Foley cath eter balloon technique fo r visu ali sing th e hym en in fe male ad olescent sexu al a buse vic tims. J f merg Nurs 1995 ; 21 :585-6. Starling SP, Jenny C. Fo ren s ic exa mina tion of adolescen t fema le genitalia: th e Foley cathete r technique. A rch Pedia tr Adolesc Med 1997 ; 151 : 10 2- 3. Persa ud Dl, Squires JE, Rubin-R emer D. Use of Foley ca theter to examin e estrogenised hy mens for evidence of sexual abu se. J Pediatr Adolesc GYl1ecol 1997; 10: 83-5. Lauber A, Souma G. Use o f toluidine blue fo r docum enta ti on of trauma tic intercourse. Obs tet Gyneeol 1982 ; 60:64 4-8. McC au ley J, Gorma n RL, Guzins ki G. Toluid in e blue in the detecti on o f pe rineal la cera ti o ns in pedi a tri c a nd adolescen t sex ual a buse victims. Pediatri cs 1986; 8 6: 1039-43. Tei xei ra WR . Hy menal col poscopi c ex amina tion in sex ual oFfences. A m J Forensic Mcd 19 81; 2 :209-14.
38
-
-
---- - -
--
-
-
-
39
40
41 42
43
44
45
46 47
Woo dling BA, Heger A. The use of the colpos cope in the diag nos is o f sex ua l a bu se in th e pedi a tri c age group. Child A buse Et Neglect 1986 ; 10: J11 - 14. Mura m D, Eli as S. Ch ild sex ual abus e: geni ta l tra ct findin gs in prep ubertal girls . n. Comp arison of colp oscopi c and unaided examinati ons. Am J Obste tr Gy J1 ecol 1989; 160 :3 33 - 5. Ada ms JA, Phillips P, Ahm ad M. The usefulness o f coJ pos copic pho to graph s in the evaluat io n o f suspected child sexual abuse. Adolesc Pedia tr Gyneco /1 990; 3:75-8 2. Mura m D, J ones CE oTh e use o f vid eo- colposcopy in the gynecologic exa minati o n of infa nts, children, and yo ung adolescents. Ado lesc Pediatr Gym'col 1993 ; 6: 154-6. McCann Jl. The colposcop ic ge nita l exa min atio n o f the sex ua lly abu sed prep ub erta l fe male. Adolesc Pedi atr GYl1 ecol 1993 ; 6:123 - 8. Hobb s C, Wy nn e J. Use of the colpo scope in examin ation for sex ual abuse. Arch iues of Diseas e in Childhood 1996 ; 75 :53 9- 42. Heger AH. Twe nty years in the ev aluati on of the sex ually a bused child : h as medi cin e helped or hUlt th e child a nd the fa mily? Child Abuse Et Neglect 1996; 20:893 -7. Will ott GM, Cross MA. Th e de tec tion o f sperm a tozo a in the mo uth. J Fo rel/sic Sci Soc 1986 ; 26 :12 5-8. Willot t GM, All ard JE. Sp erma tozo a - their persi sten ce a Fter sex ual in terco urse. ForeJ1 sic Sci Int 1982; 19:13 3-5 4. Davi es A, Wilson EM . Th e persi sten ce of se minal constitu ents in the hum a n vagina. Forensic Sc ience 1974; 3 :45-55. Christ ian CW, Lavelle JM, De Jong AR et al. Fore nsi c ev id ence fi ndi ngs in prepub erta l v ictims of sex ual assault. Ped iatrics 2000 ; 106 : 100-4. Fin kel MA , Ricci LR. Docum enta ti on and prese rva ti o n o f visua l evid ence in ch ild a buse. Ch ild Ma ltreatment 1997; 2 :322-30. Finkel !VIA. Techni cal condu ct of the CSA medic al exa min at ion. Child Abuse & Neglect 1998 ; 22 :555-66. Bl ock RW, Palusci V1. Ch ild Abuse Pedia trics: A new pedi atri c subs peci alty. J Pediatr 2006 ; 148:711- 12. Brayden RM, Altemeier WA IJ], Yeager T, Muram D. Interp reta tio ns o f coJposcopi c photogra ph s: Evidence fo r com peten ce in assess ing sex ua l a buse. Child Abuse Et Neglect 1991 ;15:69-7 6. Adam s J A, Well s R. Normal versus a bno rma l gen ital findings in ch ildren: how well do exa min ers agree? Child Abuse & Neglect 1993 ; 17:663-7 5. Paradi se l E, Win ter MP, Finkel IVIA et al. Influ ence of hi story on physici ans ' interp ret ation of girls ' genit al finding s. Pediatrics J999 ; 10 3 :980- 6. Bu tle r-Sloss E. Report of the inqui ry il1 to child abuse ill Clellelan d 19 87. London: HMSO. Adams JA. Gra nd round s : medica l iss ues and case s tudi es. Sign ifican ce of medi cal findin gs in s uspec ted sex ual a buse: mo vin g towards con sensus. J Child Sex Abuse 1992 ; 1 :9 1-9. Ada ms J A, Ha rper K, Kn ud son S. A proposed syste m for the cla ss ifi cati on of anogenital findings in children with susp ected sex ual abus e. Adolesc Pediatr Gy neco l199 2; 5:73-5. Ameri can Aca demy of Pediatrics Committee on Child Abuse a nd Neg lect. Gonorrhea in prepu berta l children. Pedia trics 1998; 101: 13 4- 5. Ada ms J A. App ro ach to the in terpreta ti on of medica l and labora tory findings in sus pected child sex ual abu se: a 2005 revisio n. Th e APSAC Advisor 200 5; Summer:7-1 3. Royal College of Physi Cia ns of Lond on. Physical Sig ns of Se.rual A bus e in Child re n. Lond on: Lav enham Press , 1997. Ro yal College o f Paediatrics a nd Child He alth. Th e physical
signs of child sexu al abuse. A n elJ ide nce-based relJ iew and gu idance fo r best pra ctice. Lo nd on, RCPCH.
--- ------
-----
-----
-
References I
48 Heger AH, Ticson L, Guerra L et al. Appearance of the genitalia in girls selected for non-abuse: review of hymenal morphology and non-specific findings. ] Pediatr Adolesc
Gynecol2002; 15:27-35. 49 50 51 52
53 54
55 56
57 58
59
60
61
62
63
64 65
66 67 68
69 70 71
72
-
Berenson AB. Appearance of the hymen at birth and one year
of age: A longitudinal study. Pediatrics 1993; 91:820-5.
Berenson AB. A longitudinal study of hymenal morphology in
the first 3 years of life. Pediatrics 1995; 95:490-6. Yordan EE, Yordan RA. The hymen and Tanner staging of the breast. Adolesc Pediair Gynecol1992; 5:76-9. McCann J, Wells R, Simon M, Voris J. Genital findings in prepubel1al girls selected for nonabuse: a descriptive study. Pediatrics 1990; 86:428-39. Berenson AB, Heger, AH, Hayes, JM et al. Appearance of the hymen in prepubertal girls. Pediatrics 1992; 89:387-94. McCann J, Voris J, Simon M, Wells R. Comparison of genital examination techniques in prepubertal girls. Pediatrics 1990; 85:182-7. Gardner JJ. Descriptive study of genital variation in healthy, nonabused premenarchal girls.] Pediatr 1992; 12:251-7. Berenson AB, Chacko MR, Wiemann CM et al. A case-control study of anatomic changes resulting from sexual abuse. Am ] Obstet Gynecol 2000; 182:820-34. Berenson A, Heger, A, Andrews, S. Appearance of the hymen in newborns. Pediatrics 1991; 87:458-65. Emans SJ, Woods ER, Flagg NT, Freeman A. Genital findings in sexually abused, symptomatic and asymptomatic girls. Pediatrics 1987: 79 :778-85. Heppenstall-Heger A, McConnell G, Ticson L et al. Healing patterns in anogenital injuries: a longitudinal study of injuries associated with sexual abuse, accidental injuries, or genital surgery in the preadolescent child. Pediatrics 2003; 112:829-37. Kerns DL, Ritter ML, Thomas RG. Concave hymenal variations in suspected child sexual abuse victims. Pediatrics 1992; 90:265-72. Hobbs CJ, Wynne JM, Thomas AJ. Colposcopic genital findings in prepubertal girls assessed for sexual abuse. Arch Dis Child 1995; 73:465-9. Cantwell H. Vaginal inspection as it relates to child sexual abuse in girls uncler thirteen. Ch ild Abuse Et Neglect 1983; 7:171-6. White ST, Ingram DL, Lyna PRo Vaginal introital diameter in the evaluation of sexual abuse. Child Abuse Et Neglect 1989; 13:217-24. Kerns DL. Cool science for a hot topic. Child Abuse Et Neglect 1989; 13:177-8. Paradise JE. Predictive accuracy and the diagnosis of sexual abuse: a big issue about a little tissue. Child Abuse Et Neglect 1989; 13:169-76. Heger A, Emans SJ. Introital diameter as the criterion for sexual abuse. Pediatrics 1990; 85:222-3. Hobbs CJ, Wynne JM. Buggery in childhood: a common syndrome of child abuse. Lancet 1986; ii:792-6. McCann J, Voris J, Simon M, Wells R. Perianal findings in prepubeltal children se.iected for nonabuse: a descriptive study. Child Abuse Et Neglect 1989; 13:179-93. Priestley B. Reflex anal dilatation and abuse. Lancet 1987; ii:1396. Stanton A, Sunderland R. Prevalence of reflex anal dilatation in 200 children. EM] 1989; 298:802-3. Clayden G. Reflex anal dilatation associated with severe chronic constipation in children. Arch Dis Childh 1988; 63 :832-6. Agnarsson U, Warde C, McCarthy G, Evans N. Perianal appearances in childhood constipation. Arch Dis Childh 1990; 65:1231-4.
45
73 Hobbs CJ, Wynne JM. Sexual abuse of English boys and girls: the importance of anal examination. Child Abuse Et Neglect 1989; 13:195-210. 74 Berenson AB, Somma-Garcia, A, Barnett, S. Perianal findings in infants 18 months of age and younger. Pediatrics 1993; 91 :838-40. 75 Finkel MA. Anogenital trauma in sexually abused children. Pediatrics 1989; 84:317-22. 76 McCann J, Voris J, Simon M. Genital injuries resulting from sexual abuse: a longitudinal study. Pediatrics 1992; 89:307-17. 77 McCann J, Voris J. Perianal injuries resulting from sexual abuse: a longitudinal study. Pediatrics 1993; 91:390-3. 78 Muram D. Child sexual abuse: Relationship between sexual acts and genital findings. Child Abuse Et Neglect 1989; 13:211-16. 79 Adams J, Harper K, Knudson S, Revilla J. Examination findings in legally confirmed child sexual abuse: it's normal to be normal. Pediatrics 1994; 94:310-17. 80 Adams JA, Knudson, S. Genital findings in adolescent girls referred for suspected sexual abuse. Arch Pediatr Adolesc Med 1996; 150:850-7. 81 Heger A, Ticson L, Valesquez 0, Bernier R. Children referred for possible sexual abuse: medical findings in 2384 children. Child Abuse Et Neglect 2002; 26:645-59. 82 Slaughter L, Brown CRV, Crowley S, Peck R. Patterns of genital injuIY in female sexual assault victims. Am ] Obstetr Gynecol 1997; 176: 609-16. 83 Kellogg ND, Menard SW, Santos. A. Genital anatomy in pregnant adolescents: 'Normal' does not mean 'Nothing happened'. Pediatrics 2004; 113 :e67-9. 84 Paradise JE, Campos JM, Friedman HM, Frishmuth G. Vulvovaginitis in premenarchal girls: clinical features and diagnostic evaluation. Pediatrics 1999; 70:193-8. 85 Straumanis JP, Bocchini JA. Group A beta-hemolytic streptococcal vulvovaginitis in prepubertal girls: a case report and review of the past twenty years. Pediatric Inject Dis] 1990; 9:845-8. 86 Pierce AM, Hart CA. Vulvovaginitis: causes and management. Arch Dis Childh 1992; 67:509-12. 87 Jenny C, Kirby P, Fuquay D. Genital lichen sclerosus mistaken for child sexual abuse. Pediatrics 1981; 83:597. 88 Handfield-Jones SE, Hinde FR, Kennedy CT. Lichen sclerosus et atrophicus in children misdiagnosed as sexual abuse. EM] 1987; 294: 1404-5. 89 Muhlendahl KE. Suspected sexual abuse in a 1O-year-old girl. Lancet 1996; 348:30. 90 Bond GR, Dowd MD, Landsman I, Rimsza M. Unintentional perineal injury in prepubescent girls: A multicenter, prospective report of 56 girls. Pediatrics 1995; 95:628-31. 91 Kellogg ND, Parra JM. Linea vestibularis: a previously undescribed normal genital structure in female neonates. Pediatrics 1991; 87 :926-9. 92 Kellogg ND, Parra JM. Linea vestibularis: follow-up of a normal genital structure. Pediatrics 1993; 92:453-6. 93 Johnson CF. Prolapse of the urethra: Confusion of clinical and anatomic characteristics with sexual abuse. Pediatrics 1991; 87 :722-5. 94 Wallis SM, Walker-Smith J. Case report: an unusual case of Crohn's disease in a West Indian child. Acta Paediatr Scand 1976; 65:749-51. 95 Sellman SPB, Hupertz VF, Reece RM. Crohn's disease presenting as suspected abuse. Pediairics 1996; 97:272-4. 96 Friedrich WN, Grambsch P, Broughton D et al. Normative sexual behavior in children. Pediatrics 1991; 88:456-64. 97 Friedrich WN, Fisher J, Broughton D et al. Nonnative sexual behavior in children: a contemporary sample. Pediatrics 101 :e9.
46 I
Investigation of suspected sexual abuse
98 Neinstein LS, Goldenring J, Carpenter S. Nonsexual transmission of sexually transmitted infections: an infrequent occurrence. Pediatrics 1984; 74 :2 17-25. 99 American Academy of Pediatrics. Sexually tra nsmitted diseases. In Pickering LK (ed .) Rep ort oj th e Co mmittee on InJectious Diseases, 26th ed n. Elk Grove Village, lL: Ballinger, 2003, pp. 157-67. 100 Am erica n Aca demy of Pediatrics Co mmi ttee on Child Abuse and Neglect. The eva luation of sexua l abuse of children. Pediatrics 2005; 11 6:506-12. 101 In gra m DL, Everet t D, Lyna PR et al. Epidemi ology of adult sexually transmitted disease agents in children being evaluated for sexual abuse. Pediatr Inject Dis J 1992 ; 11 :9 45-50. 102 Robinson AJ , Watkeys JEM, Ridgway GI. Sexually tra nsmitted organisms in sexually abused ch ildren. Arch Dis Childh 199 8 ; 79: 356-8. 103 In gram DM, Miller WC, Schoenbach VJ et al. Risk assessment for gono co cca l an d ch lamy dial infec tions in young children underg oing evalua tion for sexua l ab use. Pediatrics 2001: 107 :e73-80. 104 Th o mas A, Forster G, Robinson A, Rogsta d K for the Clinical Effectiveness Group (Association of Genitourinary Medicine and the Medical Society for the Study o f Venereal Diseases). Nation al guideline for th e management of suspected sexu ally transmitted infections in children a nd yo ung people. Sex TrailS Inject 2002; 78:324-31 . 105 Handl ey J, Dinsmore W, Maw R et aJ. Anogenital warts in children; sexual abuse or not? Int J STl ft AIDS 1993 ; 4:271-9. 106 Si eg fri ed E, Rasnick-Conley J, Co ok S et al. Human pap ill o mavirus screening in pediatric victims of sexual abuse. Ped iatrics 1998 ; 101 :43- 7. 107 Steele AM , de Sa n Lazaro C. Transhymenal cultures for sexually transmissible organisms. Arch Dis Childh 1994; 71 :423-7.
J. - - - - -
108 Embree JE, Lindsay D, William s T et al. M. Acceptability a nd use fulness of vaginal washes in premenarch eal girls as a diagnostic procedure for sexua lly transmitted infections. Pediatr Inject Dis J 1996; 15 :651-66. 109 Mok JYQ. Routine Hrv testing after child sexual abuse? Child Abuse ReI! 1998; 7:6 3-9. 110 Muram D, Spe ck PM, Gold SS. Genital abnorma lities in female siblings and friends of child victims of sexual abuse. Child Abuse ft Neglect 1991 ; 15: 105-10. III Bays J, Ch adwic k, D. Med ical diagnosis of the sexually ab used child. Ch ild Abuse ft Neglect 199 3; 17:9 1-11 0. 112 Muram D. Child sex ual ab use. Curl' Opin Obstetr GYllecol 1993 ; 5:784-90. 11 3 De Jong AR, Rose M. Legal proof of child sex ual abuse in the absence of physical evidence. Pediatrics 199 1; 88:506-11. 114 Kerns DL. Triage and referral s for child sexual a buse medical examinations: Which children are likely to have positive medical findings? Ch ild Abuse ft Neglect 1998 ; 22:515-18. 115 Chadwick DL, Krous HF. Irresp ons ibl e testimony by medical ex perts in cases involv ing th e physica l abuse and neglect of ch ildren. Child Maltreatment 1997; 2: 313 -21. 116 Adams JA. Th e role of photo do cu mentati on of genital findings in medical evaluations of suspected child sexual abuse. Child Maltreatment 1997; 2:341-7. 117 Mok JYQ, Busu ttil A. Medical exa minati ons for Child Sex ual Abuse in Scotla nd: good enough practice? Child Abuse Revi ew 2004; 2004; 13:324-37. 118 Jen ny C. Pediatric fellowships in child abuse and neglect: the development o f a new subspecialty. Child Maltreatment 1997; 2:356-61. 119 Emans SJ, Woods ER, Allred EN, Grace E. Hymena l findings in adolescent wo men: impact of tampon use an d consensual sexual activity. J Pedlatr 199 4; 125: 153-60.
-
----
--
-
I
CHAPTER 3
I
RADIOLOGY OF CHILD ABUSE Maeve McPhillips
Role of the radiologist Radiological investigations Skeletal injuries Head injury Visceral injuries
47 48 51 60 68
Child abuse has been a concem of paediatric radiologists since Caffey's description of multiple fractures in children with a subdural haematoma in 1946. He describes fractures of the long bones as a 'common complication of infantile subdural haemato ma' and, although suggesting that there is a trau matic origin, states that 'the causal mechanism remains obscure'.! In 1971 , Guthkelch, a British neurosurgeon, first proposed shaking as a cause of infantile subdural haematoma, and the following year Caffey published his paper 'On the theory and practice of shaking infants'. 2,J Since then, non accidental injury has been a subject of great importance to those who have responsibility for the care, diagnosis and investigation of children who may have been victims of abuse. Teams of clinicians work closely together, aware that only when they are in possession of all the information available about the child can the correct diagnosis be reached.
ROLE OF THE RADIOLOGIST Paediatric radiologists have expertise in the interpretation of imaging of in fants and children and a familiarity with the appearances of the normal brain and ske leton. The appeara nces of the skeleton and brain followin g accidental trauma will be well known to them, and they will have an understanding of possible mechanisms of injury, and so be able to judge whether the given history is appropriate for the injury. The first role of the radiolo gist is to be alert for injuries that, in the light of the history supplied, may not have
Soft-tissue injury Differential diagnosis Conclusion References
69 69 73 73
occurred accidentally, or have been identified as incidental findings on a radiograph performed for another purpose. Such cases w ill need to be discussed with the referring clinician . When referred a case of possible non-accidental injury, the radiologist should review the presenting injury, and discuss the patient with the clinician. A knowledge of nor mal skeletal development and variations in ossification can prevent unn ecessary investigation. 4 - 7 There shou ld be a recognized departme ntal protocol for a skeletal survey for suspected non - accidental injury and all im ages should be reviewed by the radiologist before the child leaves the department. Th is is to ensure high-quality images and to allow for any necessary supplementary views. An assess ment should be made of the age of any injuries. The possi bility of an underlying bone disease or other condition, together w it h the need for appropriate im aging, should be considered at this stage. The report should be commun icated verb ally to the responsible clin ician as soon as is practicable and a formal written report issued promptly. If there is doubt as to the presence or significance of a lesion, this should be clearly stated together w ith a plan of investigation to clarify the findings. The need for further imaging, in particular neuro imaging, should be discussed with the clinician at th is stage. The radiologist should be available for any discus sions with clinicia ns and members of the chi ld protection team, including the police. Because of the possibility of child protection or other legal proceedings, all images and reports should be retain ed in a secure place.
48 I
Radiology of child abuse
RADIOLOGICAL INVESTIGATIONS
A separate exposure should be obtained of each anatomical area to optimize image quality, reduce geomet ric distortion and allow for the detection of subtle abnor malities. At least two views of the skull should be obtained. Both oblique views of the ribs should be obtained routinely. Two perpendicular views should be obtained of any focal injury. Coned views of the metaphyses, in anteroposterior (AP) and lateral projections, may be helpful to confirm or exclude classic metaphyseal lesions. The skeletal survey is not an emergency investigation and should be performed during the standard working day. It requires two radiographers and a lo t of time. It should be performed in the radiology department unless the child is critically ill , in which case it may be performed using mobile radiography equipment in the paediatric intensive care unit. Image quali ty may be compromised in this situation. Most radiol ogy departments no longer use hard-copy radiographs for reporting. Soft-copy reporting has been shown to be superior to hard copy owing to its abil ity to vary the grey -scale settings and to magnify sections of the image selectively. 13
Skeletal Survey The radiographic skeletal survey is the mainstay of investi gation of non-acc idental injury in yo ung children and infants. There has been much variability in the quality of exa minations 8,9 a nd national stand ards have been intro duced by the American Academy of Paediatrics, JO the American College of Radiology,l1 the Royal College of Radiology and the Royal College of Paediatrics and Child Health. 12 Table 3.1 shows a suggested protocol. A ' babygram' of the whole body with a si ngle exposure or coverage of the whole body using th ree or four films is inadequate. High-quality radiographs must be obtained with optimum exposure factors and good coning. High resolution cassettes should be used with no grid. The patient name, side marker, and date and time of the examina tio n must be clearly visib le on the radiograph . The names of the radiographers must be recorded: they should wo rk in pairs. Not only does this allow for increased efficiency in what can be a prolonged investigation, but it provides legal safe guards for both the patient and the ra diographers. The radiograp hers should obtain positive identifi cation of the patient from the accompanying staff or carer and the identity should be checked on the na me band. It is impor tant that the responsible clinician has discussed the need for the skeletal survey with the parents or caregivers, as they should be allowed to accompany the child during the investigation. It may also be appropriate that a member of the nursing staff is in attendance.
Table 3.1
Follow-up Skeletal Survey If there is ongoing clinical concern, the skeletal survey should be repeated in 10-14 days, apart from the skull radiographs. This is particularly relevant if the presentation has been with head injury and the initial skeletal survey has been normal. If there are areas of radiological concern, repeat radiographs should be obtained of these areas, also at lO-14 days (Fig. 3.1). For patients in whom there is no particular diagnostic unceltainty, it is still important to obtain repeat chest radio graphs, including obliques, and views of the id entified bony abnormalities. It has been shown that repeat skeletal surveys can identify a greater number of fractures, particularly of ribs, and can confirm suspected fra ctures. 14-16 Lack of change over time can confirm a normal valiant. Radio grap hs taken on two separate occasions can also help more accurate dating of injury and show evidence of different ages of injuries.
Suggested protocol for initial skeletal survey
Skull
Spine Chest
Abdomen Limbs
Supplementary
AP and lateral views Towne's view if occipital injury suggested clinically Lateral views of cervical, thoracic and lumbosacra l sp ine AP to include clavicles Lateral Oblique views of both ribs - each to include whole chest AP whole abdomen, to include pelvis and hips AP both upper arms AP both forearms PA both hands AP both femora AP both lower legs AP both feet Additional views of any questionable areas AP and la tera l coned vie ws of metaphyses
Post-mortem Skeletal Survey The same high standards should be applied to the examina tion of the deceased child as to the live child. A 'ba bygram' should not be performed . Individual exposures of each anatomical area should be obtained, with particular atten tion paid to the presence of COITect identifiers on each image. The skeletal survey must be obtained prior to the autopsy (Fig. 3.2). If necessary, the autopsy should be delayed to allow this. The reporting radiologist should co mmunicate his or her findings to the pathologist as soon as possible. A verbal report will suffice, as long as it is documented in the notes and in the final radiological report. Detailed radiography of removed specimens of bone can be very
AP, anteroposterior; PA, posteroanterior.
-
- -
~
-
-
-
----
---
--
-
-
-
Radiological investigations I
49
helpful (Figs 3.3 and 8.1, p. 147). Perpendicular views shou ld be obtained of all bones. For the ribs this mean s that the additional view is an axial , or supero-inferior image. 17, 18
Ultrasound Ul trasound is not used routinely in the assessment of non accidental musculoskeletal injury, mainly because of its
Figure 3.2 Skeletal survey following a post- mortem showing bowel in the chest. Known rib fra ctures are obscured by bowe l.
Figure 3.1 (a) Admission film. Healing fracture s of the necks of the seventh , eighth, ninth and tenth ribs can be seen on the left as we ll as a mid-shaft fra cture of the clavicle. (b) Follo w-up fi lm 12 days later. Now visible are healing fra ctures of the left clavicle and of the necks of the fifth to eighth ri bs on the left and of the sixth to ninth ribs on ri ght. The left tenth rib has compl etely remodelled in the 12-day interval and the ninth rib is identifiable only by minimal irregul arity.
- ~ "'- -----~--
Figure 3.3 Subtle metaphysea l fra cture identified on post mortem skeletal survey prior to autopsy. High-definition specimen radiograph following stripping of soft tissues. The irregu larity in the distal femoral metaphysis was co nfirmed by histo logy to be a fracture. It was not palpable.
50 I
Radiology of child abuse
Figure 3.4
Transfontanellar high-resolution ultrasound image.
Cystic cavities (arrows) at the junction of the grey and white matter are tears caused by shearing.
dynamic, operator-dependent nature. Positive findin gs will need to be documented using a nother imaging modality. Ultrasound can be used, particularly in the very young child, to assess cartilaginous epiphyseal fracture-separation, which may be underdiagnosed radiographically. 19,20 Its use has also been described in the detection of rib fractures. 21 The use of transfontanelJar ultrasound is more esta blished although it is still a subjective examination and not useful i~ a legal setting, as experts find it difficult to review another operator's images. It is easily performed in the clitically ill infant at the bedside and has a definite role in the manage ment of an infant with head injUly, to monitor extra-axial collections, ventricular dilatation and intracranial blood flow. High-resolution ultrasound using high-frequency probes (10 MHz) can differentiate subarachnoid from subdural fluid, particularly when colour-flow imaging is used, Con tusional tears at the grey-white matter junction can be demonstrated elegantly and may not be visualized on com puted tomography (CT) (Fig. 3.4). Fresh tears may contain a clot and older tears appear cystic; these cysts may later col lapse. 22 Collapsed tears may be overlooked at autopsy unless specifically targeted. Fluid is well visualized by ultrasound as free intraperi toneal or pleural fluid. The liver, spleen and kidneys are easily assessed by ultrasound, with an accuracy of 94 per cent,23 but CT is more widely used in th e assessment of blunt abdominal trauma. 24
Computed Tomography Prior to the development of CT for neuroimaging, subdural haemorrhage was identified indirectly by cranial angiography
-
-
-- -
-
-
-
or air encep halography, or at post-mortem examination. In accidental and inflicted head injury, CT shows different patterns of injury.25.26 It is the imaging modality of choice for suspected brain injury. lt is recommended as part of the investigation of all suspected non-accidental injuries in children less than 1 year of age. 12 There is some controversy about this, particularly in a child with no neuroloo'ical '" abnormality, regarding the radiation burden. If neuroimag ing is not performed, the rea son should be documented in the patient's notes. After presentation of a non-accidental injury, CT scan ning should be performed as soon as possible, without intravenous contrast. This method of imaging is widely available, easily performed, and fast. With newe r scanners the need for sedation or general anaesthesia in the restless child is less. Standard brain and bone windows should be provided. Acute haemorrhage is readily visualized, even in the subarachnoid space. Lesions requiring neurosurgical intervention are reliably seen. Changes of brain oedema can be subtle and may be missed. Known fractures, visible on skull radiography, may not be visible on CT images, as they may lie in the plane of the imageY Three-dimensional (3D) volume rendering may be helpful in identifying frac tures or, indeed, in differentiating normal variants, such as parietal fissures, accessory sutures and synchondroses, from suspected fractures. 28 This imaging modality can be used to assess the bony integrity of th e cervical spine and fractures of the facial bones. 28 Suspected visceral trauma may also be investigated using contrast-enhanced CT. Although not used routinely, post-mortem CT scanning may be useful, particularly as the normal limitations to dose no longer apply and a high-resolution volumetric scan can be obtained in a very short tim e. Owing to the amount of data produced, reporting such an investigation will be time consuming for th e radiologist but may give considerable further information about metaphyseal and rib fractures.
Magnetic Resonance Imaging Magnetic resonance imaging (l'v1RI) is complementary to CT scanning. Although l'v1RI is less sensitive for acute haemor rhage, especially in the subarachnoid space, it elegantly demonstrates subacute subdural haemorrhage, contusions and oedema (Fig. 3.5). Images should be obtained in transverse, coronal and sagittal planes. The recommended sequences are T1-weighted, T2-weighted, FLAIR (fluid-attenuated inversion recovery), gradient echo, which is sensitive to blood products and detects small areas of haemorrhage, and diffusion weighted imaging/apparent diffusion coefficient (OWl/AD C), which demonstra tes areas of evolving brain injury, in parti cular areas of hypoxia-ischaemia. After presentation, an l'v1RI should be ca rried out as soon as possible. Because the signal from the damaged tissue normalizes after 7 days, OWl is most useful in the first week. 29 It may be appropriate to assess vas cul ar structures with mag netic resonance (MR) arteriography
---
----
-
--
- -
-
--
Skeletal injuries
Figure 3.5 Magnetic resonance scan, Tl-weighted image, showing a fresh small linear subdural in the right posterior parietal region and high signal as a result of fresh haemorrhage (arrows). Overlying this is a scalp haematoma. A skull fracture was not visible on magnetic resonance imaging. and venography, and to use transverse Tl-weighted fat saturated sequences to search specifically for inj ury to the major vessels in the neck. 28 Spinal cord injury is best assessed with MR, using short tau inversion recovery (STIR) sequences. Post-mortem MR scanning has been shown to be more sensitive than autopsy alone in the detection of a shearing injury, cortical haemorrhage and mastoid fluid, and to be equal to autopsy in the demonstration of cerebral oedema, focal contusions and subfalcine herniation. 3D
Scintigraphy Isotope bone scanning, using technetium-99-labelled methylenediphosphonate (99Tc-MDPl, can show bone pathology. In one series,3l 10 per cent of fractures were seen only on scintigraphy, and Mandelstam has shown that 50 per cent of rib fractures may be visible only on isotope bone scanning 32 (Fig. 3.6). The yield from radiography may have improved since the introduction of oblique views of the ribs at presentation. Many fractures also become appar ent on follow-up radiographs. Scintigraphy has a Jow sen sitivity for skull fractures. Because of the high activity in normal growth plates, metaphyseal fractures may not be apparent on isotope bone scanning, particularly if bilateral (Fig. 3.7). Fractures may show increased activity on scintigraphy for up to a year after injury. Any areas of increased activity detected using scintigraphy should
I
51
Figure 3.6 Isotope bone scan of the upper body show ing multiple areas of increased uptake in the posterior ribs, representing fractures of the posterior shafts at their necks. A spi ral fracture of the left humeral shaft is identifiable by the asymmetrical uptake in the humeri. therefore be further assessed by radiography. Bone scintig raphy is complementary to radiographic skeletal survey, and may be useful in individual patients.
SKELETAL INJURIES Fractures in child abuse are most common in infants and children under 2 years of age. It is convenient to separate fractures into two groups: (a) those that are seen fre quently, but are not very specific for non-accidental injury; and (b) those that are highly specific for inflicted injury but are less commonly seen (Tables 3.2 and 3.3).33
Periosteal New Bone Fractures repair by the laying down of subperiosteal new bone. With no obvious fracture, the presence of sub periosteal bone may be due to injury to the periosteum by rough handl ing, or by acceleration-deceleration forces during shaking. Physiological periosteal reaction is seen in normal infants as young as 8 weeks, up to about 8 months of age. It is smooth, rarely more than 2 mm in thickness and is seen along the diaphyses of the long bones. It is usu ally symmetrical, though may be more obvious on one side.
Long-bone Fractures Diaphyseal fractures of long bones are common in non accidental injury; some authors have found them to be four
52 I
Radiology of child abuse
(a)
PO :;. T
LT
Table 3.2
Fractures that are frequent but with low specificity for non-accidental injury
Midclavicular fractures Simple linear skull fractures Single diaphyseal fractures Reproduced with pe rmission from Cart y33
Table 3.3
Fractures considered to have a high specificity for
child abuse
Meta physeal fractures Rib fractu res Scapular fractures Fractures of the outer end of the clavicle Vertebral fractures or subluxations Finger injuries in non-ambulant children Fractures of different ages Bilateral fractures Complex skull fractures (b) Reproduced wi th permission from Carty33
Figure 3.7 Isotope bone scans of lower body. (a) Posteroanterior view shows metaphyseal fracture of the left proximal tibia (arrow). (b) Lateral view of lower limbs. The fracture cannot be seen. A posterior rib fracture is visible in the lower left chest (arrow). times more common than the more specific metaphyseal fracture. J4 It is important to correlate the appearance of the fracture with the clinical history (Fig. 3.8). Suspicion of abuse increases when there is evidence of healing, indicating a delay in seeking medical care for a child in pain. Excessive callus formation is seen owing to movement at a fracture site, causing repetitive subperiosteal bleeding (Fig. 3.9). Multiple fractures, especially in different stages of healing, without an obvious history of significant trauma, suggest at least poor parenting. In the presence of a
.
.-
--
Figure 3.8 A 6-week-old infant presented to the accident and emergency department. The mother gave a story of the child waking up and not using her arm, saying she must have caught it in the cot bars. The story is inappropriate for this fracture, which was almost certainly caused by the child being roughly lifted by this arm.
-
--
--
-
- - --
Skeletal injuries I
53
Figure 3.10 A 14-year-old boy who died as a result of abuse. Fractures of shafts of rad ius and ulna. These are typical defensive fractures sustained when the arm is held up to ward off a blow from an assailant.
Figure 3.9 Toddler presenting with severe head injury. The radiograph shows a new buckle fracture of the distal radius and ulna and an older, untreated supracondylar humeral fracture, with resulting extensive subperiosteal reaction. fracture of high specificity, a diaphyseal fracture takes on increased significance. An isolated shaft fracture becomes more suspicious when there is other evidence of physical abuse, such as bruising. Bilateral forearm fractures in infants and toddlers are worrying beca use young children do not ex hibit the normal protective instinct to break their fall using their outstretched arms .34 There are welJ-recognized mechanisms for typical spiral, oblique and transverse fractures. The described mechanism given in the history should be assessed for each individual fracture. It should be remembered that non-accidental injury may occur when the limb is used to lift or drag the child, or if the child is thrown, or be the result of direct injury. In par ticular, transverse fractures of the forearm bones can be sus tained as defensive injuries to ward off a blow (Fig. 3.10). It should be remembered that infants who cannot yet roll over are unlikely to fall off an elevated surface and that reported falls in children who are not yet cnlising should raise suspi cion. 35 It is, therefore, important to know the degree of mobility of the individual child. The spiral fracture of the
~------
tibia, the 'toddler's fracture', is very commo n as children start to walk. Impaction fractures in the lower limbs appear as torus or 'buckle' fractures, and are caused by transmission of force vertically up the shaft of the bone when the child is forcibly thumped down onto a hard surface in the standing position (Fig. 3.11).
Metaphyseal Fractures Although these are highly specific for non-accidental injury in infants less than 1 year of age, they are seen in only 11-39 per cent of children surveyed. 36 .37 Their detection depends on the quality of the radiography and they may be seen as cor ner or 'chip' fractures, suggesting local avulsion, or as 'bucket-handl e' fractures depending on the projection. They are more commonly seen in the lower limb, but are also seen in the upper limb. They can be caused by shaking, but, when seen in a single limb, a twisting, pushing or pulling mecha nism may be responsible. This description is supported by the report of classic metaphyseal fractures in a group of children being treated for clubfoot, in whom the foot was pushed and twisted into forced eversion and dorsiflexion before being put into a cast. 38 Similar fractures are recognized rarely following Caesarean section,39 a breech delivery or an armling delivery. Metaphyseal corner fractures do not require any specific treatment. They are not palpable.
54 I
Radiology of child abuse
Figure 3.11 Healing impaction fracture of the distal femoral shaft with disruption of the normal contour anteriorly and subperiosteal new bone (arrows). The apparent defect in the anterior tibia is not a further fracture but is the site of the unossified tibial tubercle.
These classic metaphyseal lesions are planar fractures through the primary spongiosa of the metaphysis, with tra becular disruption giving rise to a disc of bone and cartilage. At the periphery, adjacent to the periosteal collar, there are more trabeculae so the fragment is denser there. The increased visibility of the peripheral bone gives rise to the radiographic appearance of a comer fracture, sometimes with a subtle metaphyseal lucency. An oblique view of the metaphysis may demonstrate the disc-like fracture fragment. Some bones, such as the tibia, are more likely to show a 'bucket-handle' appear ance, as the metaphysis is relatively straight (Fig. 3.12). The comer fracture appearance is more likely to be seen in meta physes with a more complex appearance, such as the proximal humerus or distal femur 40 (Fig. 3.13). The fracture may be
~
-
Figure 3.12 (a) Anteroposterior and (b) lateral views of typical bucket handle metaphyseal fracture of the distal tibia (arrows).
incomplete and not pass through the whole metaphysis. Volu metric CT imaging of metaphyseal fractures may demonstrate the disc-like fragment, but would appear to be unwalTanted except in specific cases, or for research purposes. Appearances during healing are variable. The fracture may become more
---
----
-
--
-
-
---
Skeletal injuries I
Figure 3.13 Typical metaphyseal corner fracture seen at the posterior aspect of the distal femur (arrow).
55
Figure 3.15 Metaphyseal fractures of the distal femur and proximal tibia with periosteal new bone along the lateral aspect of the proximal tibial metaphysis. Periostea l new bqne along the medial tibia, lateral fibula and femur is confined to the diaphysis and is probably physiological. Reproduced with permission from Carty33
Figure 3.14 Coned view of the ankle showing a healing corner fracture of the posterior tibia and a fresh corner fracture anteriorly. The healing posterior fracture shows periosteal new bone.
apparent, maximal at 10-14 days. Frequently, there is no other evidence of healing as the periosteum may not be disrupted. If it has been, there may be faint subperiosteal new bone forma tion, although more extensive new bone formation can result in thickening of the adjacent cortex and a squaring off of the metaphyseal contour40 (Figs 3.14-3.16). Healing is usually complete within 6 weeks (Fig. 3.17). Extensive periosteal
Figure 3 .16 Healed metaphyseal fracture of the right proximal humerus with subtle irregularity and sclerosis. No evidence of subper iostea l new bone formation. Healing fractures of two posterior ribs are visible (asterisks).
56 I
Radiology of child abuse
Figure 3.17 Close-up view of the knee. The lucent lines on the distal femur and medial half of the proximal tibia are healed metaphyseal fractures. reaction extending up the shaft of the bone can suggest local trauma, perhaps a result of forceful gripping of the bone (Figs 3.18-3.21).
Epiphyseal Plate Injury An epiphyseal plate injury may be identified using a radi ographic skeletal survey but, without a metaphyseal com ponent, when the epiphysis is un ossified they may be missed. Ultrasound or MR scanning can show displacement of the epiphysis and confirm the presence of a fracture/sep aration injury, and should be considered in a child who does not use a limb and with no obvious explanation on the skeletal survey.19
Rib Fractures In infants and young children, rib fractures are highly spe cific for non-accidental injury. They can be seen in signifi cant accidental trauma, such as motor vehicle accidents, or in infants with bone disease, such as rickets or bone disease of prematurity.41,42 They are recognized as occurring rarely as a result of birth trauma, presumably because of pressure from
--
-----
Figure 3.18 Healed metaphyseal fracture of the distal tibia, lateral view, showing squared contour and slight irregularity. There is irregular periosteal new bone extending proximally up two-thirds of the shaft of the tibia, a disproportionate reaction for a relatively minor fracture. the maternal symphysis pubis, in which case there will prob ably be a history of a large baby and a difficult deJivery.43.44 Rib fractures are splinted by adjacent soft tissue and neighbouring ribs. They may result in mild respiratory dis tress. If several ribs are fractured in more than one loca tion, this will give rise to a flail segment of chest wall, with more severe respiratory consequences. In child abuse, although ribs may be fractured by direct trauma and impact, or by compression by kneeling on the chest, the usual mechanism of injury is squeezing of the chest by the perpetrator's hands. The infant is held facing the adult, with the fingers on the back of the chest, the palms at the side and the thumbs in front. Pressure causes anteroposterior compression of the rib cage. Although fractures can be seen anywhere along the rib arc, the most characteristic fracture site involves the posterior rib, at the rib head and the costovertebral junction45 (Figs 3.1, 3.6 and 3.7, pp. 51-52). This is due to leverage of the poste rior rib on the fulcrum of the transverse process of the verte bra, with an adjacent fracture. Similar fractures have also been demonstrated in a cadaver study simulating median sternotomy with rib retraction. 46 In the same paper, a study was done comparing sternal compression and AP manual compression in rabbits. No rib fractures were seen following
----
-
--
Skeletal injuries I
Figure 3.20
57
Healing metaphyseal fracture of the distal humerus
with gross periosteal new bone extending almost to the proximal meta physis.
Figure 3.19
(al Admission radiograph. Tiny metaphyseal
fractures of the distal tibia and fibula (arrows). There is soft-tissue swelling of the calf. (bl Follow-up film shows massive healing subperiosteal new bone, mainly due to healing of a subperiosteal haematoma, the extent of which was totally inappropriate for the tiny fracture. This represents very extensive trauma to the leg.
sternal compression on a firm surface and the posterior ribs showed no change in their relationship to the spine; however, fractures did occur with manual compression. Lateral rib fractures are due to the compressive force on the ribs and are
-...
.
,
-
Figure 3.21
Healing bucket handle metaphyseal fractures of the
distal tibia bilaterally with disproportionate periosteal new bone formation. Irregular periosteal new bone on the fibular diaphyses bilaterally. The appearances probably represent a direct gripping injury.
~
58 I
Radiology of child abuse
shown at autopsy to have distraction of their outer surface and impaction of the inner surface 47 (Figs 3.22 and 3.23). Fractures of the anterior ends of the ribs involve the costo chondral junction. These show disruption of the posterior surface, which may be a result of direct pressure from the thumbs. Rib fractures may be difficult to see unless there is some displacement. Only 36 per cent of fractures identified at autopsy were visible on the skeletal survey in Kleinman 's study.47 Lateral rib fractures may show adjacent extrapleural opacity, owing to focal haemorrhage and pleural reaction. Posterior lib fractures may also be obscured by the trans verse processes. Oblique views of the ribs are useful in detec tion of both posterior rib fractures and fractures at the costochondral junction. However, many rib fractures do not become visible until the follow-up radiograph shows callus formation. Volumetric CT scanning of the thorax may reveal rib fractures, but the radiation dose associated would pre clude its routine use for this purpose. Specimen radiography of resected ribs should be per formed in both the frontal and the axial projections. There have been studies of children which have shown no radiographic evidence of lib fractures resulting from car diopulmonary resuscitation (CPR)48.49 A recent studyS° found autopsy evidence of rib fractures in 11 per cent of 70 infants who received CPR immediately before death. The fractures were all anterolateral, linear and often bilateral. There was little, if any, associated blood and no reactive change.
Figure 3.22 Child admitted moribund. Healed fractures right sixth and seventh ribs postero'atera"y (arrows) plus fresh corner fracture proximal right humerus (white arrow). The healed left clavicular fracture could be a birth injury.
.
Spinal Trauma
-----
Severe spinal trauma is much less common than other frac tures in non-accidental injury. It may be underdiagnosed, but routine lateral radiographs of the whole spine (Fig. 3.24) and follow-up skeletal surveys may increase its identification. Vertebral body fractures are thought to be due to hyper flexion. This may be during a shaking episode, holding the child by the chest or the shoulders, or by slamming the child down onto a hard surface on the buttocks, transmit ting the force vertically up the spinal column. These may be anterior compression fractures, fractures extending to the superior end plate, or a combination of the two. The superior end plate fracture is thought to be similar to the classic metaphyseal fracture.5l Fractures may also occur of the spinous process, with a similar mechanism. The carti lage tip of the spinous process may be avulsed with only a small fragment of bone, which may be recognized on the lateral spine radiographs. If diagnosis is delayed, the mar gins of the avulsed fragment may be irregular, no longer matching the defect in the spinous process. This is due to further growth of the avulsed fragment. 52 Fracture disloca tion of the spine may occur, which can be associated with cord compression and long-term neurological damage. It is described at any level in the spine, including the sacrococ cygeal region. It may present as an unexplained kypho sis. 53-56 Traumatic spondylolysis of C2 , the hangman's fracture, has been described with anterior subluxation of C2 on C3. 57 Full visualization of the vertebral injury may
Figure 3.23 There are multiple healing rib fractures in both axillae with periosteal new bone around the fractures.
--
-
--
Skeletal injuries I
require CT, but MRI will be necessalY to exclude a ny asso ciated cord injury.
Digital Injuries Injuries to the metacarpals, metatarsals and phal a nges are uncommon, but have a high specificity for abuse. They are usually torus fractures, and the mechanism is thou ght to be twisting, bending or hyperextension 58 (Fig. 3.25). Crush fractures may also be seen.
59
fracture healing. A radiologist whose practice involves paedi atric trauma will have co nsiderable experience in the report ing of acute and healing fractures, when the timing of the injUly is precisely known . This can then be extrapolated to the appearance of fractures for which there is an inad equate history. Despite the lack of published data, there is remark able agreement between radiologists. Approximate dating of
Other Bony Injuries Fracture of the outer end of the clavicle usuall y resu lts from a fall. A midclavicular frac ture may be as a res ult of birth trauma and an appropriate histOlY should be sought. Fractures of the acrom ion and the body of the scapula 56 ,59 (Fig. 3.2 6) have been described, as have sternal fractures. GO Pelvic fractures usually involve the superior pubic rami. 61
Dating of Injuries Precise dating is impossible. Accepted g uidelines ha ve been published (Table 3.4),62,63 but there have been few studies of Figure 3.25
Healing fractures of the bases of the proximal
phalanges, great and little toes, due to hyp erexte nsion injuries.
Figure 3.24 Lateral view of the lumbosac ral spine show in g compression fracture of L2 (arrow) with a mild lum bar kyphosis.
~
~-------
-
--
~
-
,
-
-
--~--
Figure 3.26 Ossifying haematoma projected over scapula in a child who had been severe ly beaten. Reproduced with permission from Ca rty,s9
60 I
Radiology of child abuse
Table 3.4 Dating of fractures (adapted from O'Con~or .IF, Cohen J In: Diagnostic imaging of child abuse. Baltimore: Williams and Wilkins, 1987112)
Soft-tissue resolution Early periosteal new bone Loss of fracture line definition Soft callus Hard callus Remodelling
Time
Peak
2-10 days 4-21 days 10-21 days
4-10 days 10-14 days 14-21 days
10-21 days 14-90 days 3 months to 2 years
14-21 days 21-42 days
Reproduced with permission from Carty] 3 Figure 3.27 Posterior linear parietal skull fracture in an abused child. The only indication that this was caused by abuse was the la ck of clinical history for the fracture.
fractures is always given as a range, with the limits becoming wider as time elapses from the injury. The appearance of sub periosteal new bone formation is the earliest sign of healing. It has been described as being visible as early as 4 days after injury and seen in 50 per cent within 2 weeks.63
HEAD INJURY In non-accidental injury, the infant and young child may suffer injury to the scalp, skull and face, suffer an intracra nial or brain injury, or a combination of the two. Brain injury is the leading cause of mortality and of significant neurological impairment in survivors. 64 - 67 The infant with brain injury may present acutely as encephalopathy or seizures, less acutely as irritability or vomiting, or as a relatively well child with macrocrania or failure to thrive. The initial diagnosis may be of meningitis, apnoea or collapse of unknown cause, or a search may be made for an infectious or gastrointestinal cause for vomiting. Findings suggestive of non-accidental injury may be seen on other investigations, such as incidental healing rib fractures on chest radiographs, or subdural collections on cranial ultra sound. These will then prompt further investigations.
Figure 3.28 Wide diastatic skull fracture, allegedly caused by falling off a sofa. In addition, there is a further extensive linear fracture across the lower part of the skull. bilateral fractures, complex or stellate fractures and grow ing fractures are more common in non-accidental injury74 (Figs 3.28, 3.29 and 14.6, p. 299). Bilateral fractures can be seen in accidental trauma with a fall onto the occiput or the vertex, with symmetrical transmission of force. No sku ll fracture type is characteristic, and in any individual patient the appearance of the fracture alone does not allow ajudgement as to its cause.
Skull Fractures It is important to remember that a child may sustain a skull fracture with no associated brain injury, or may have severe brain injury without a fracture.68 A sk ull fracture requires direct impact or compression. 69 There have been many stud ies of falls, including population studies/ o- 73 showing that skull fractures are unusual in low falls and that a fall from about 1 m is usually required for a skull to fracture. This is the approximate height of a fall from an adult's arms. A typical accidental skull fracture is a unilateral linear parieta l fracture, which is also the most common fracture seen in non-accidental head injury (Fig. 3.27). However
-
-
-
-
-
.
Dating of Skull Fractures Skull fractures cannot be dated radiographically. Some indication of age may be given by the presence of adjacent
-
---
--
--
-
-
-
Head injury I
Figure 3.29 Symmetrical horizontal fractures, which meet in the midline, and an additional linear parietal fracture on the left. There was no clinical history to explain how these fractures were sustained.
soft-tissue swelling, which resolves gradually according to the size of the haematoma. However, care should be taken, as a swelling may appear after some delay as a result of seepage of cerebrospinal fluid (CSF) through a fracture, or enlarge ment of a scalp haematoma during resolution. The clinician should also realize that focal scalp swelling may occur in the absence of an underlying fracture (Figs 3.30 75 and 3.31).
Extradural Haematoma Extradural haematoma is unusual in children, and rare in non-accidental head injury. Radiologically, it is a lentiform collection overlying the brain, tending to compress the underlying brain substance. There may be an associated fracture, with typical tearing of the middle meningeal artery, although venous bleeding can cause an extradural haematoma.
Subdural Haematoma Subdural haemorrhage is caused by bleeding from bridging veins crossing the subdural space, which are stretched and torn when the brain moves excessively relative to the over lying dura. A degree of trauma is required to tear these veins, though the minimum degree is not known. Mechanisms of injury in inflicted trauma are discussed in Chapter 14. Subdural haemorrhage is seen in birth trauma 76 but has been shown to be present in 8-17 per cent of asymptomatic neonates, including 6-26 per cent of spontaneous vaginal
61
deliveries. 77 ,78 Subdural blood was seen most frequently in the posterior fossa, and no interhemispheric blood was seen in either series, of 199 babies in total. Delivery by Caesarean section may still be associated with subdural haemorrhage, as there may be some difficulty in disimpact ing the engaged head. In Whitby's cohort of 111 babies, fol lowed for 2 years, the subdural haemorrhages had all cleared by 4 weeks?7 There were no recurrent bleeds. Accidental trauma can result in subdural haemorrhage, most commonly over the cerebral convexities and at a single site, localized to or opposite the impact, although occasionally, when there has been major trauma, the haemorrhage may spread over the convexity. Interhemispheric haematoma is rare, but is described, and is therefore not specific for inflicted head injury?9 Homogeneous hyperdense haematoma is seen more often in accidental injury, though this changes in the days after presentation in 25 per cent. 79 Subdural haemor rhage is, however, more common in inflicted head injUly. In a series of 100 children admitted as a result of head injury, Duhaime et al 80 identified 76 as accidental and 24 as inflicted. Only three children with accidental head injury had subdural haemorrhage, all of whom had been involved in road traffic accidents. By contrast, ) 3 of the 24 children with non accidental head injury showed subdural haemorrhage, with either no history of injury or history of a low fall (~l m) in 22 patients, and admitted assault in two. Ewing-Cobbs et al 81 studied 40 children with a head injury, 20 accidental, 20 non accidental. Subdural haemorrhages were seen in 16 of the non-accidental group, but in only nine of the accidental group, most of whom had been involved in motor vehicle accidents and none of whom had fallen. Hymel et al S2 com pared CT scans of 39 children with inflicted head injury with scans from 39 control patients with accidental head injury. Subdural haemorrhage was found in 17 in the non-accidental group, but in only four in the accidental group; interhemi spheric falx haemorrhage was seen in 17 and 2 respectively. In non-accidental head injury, subdural haemorrhage is usually bilateral, and almost always interhemispheric (Figs 3.30 and 3.32); it spreads over the cerebral convexity and may be seen in the middle cranial fossa (Fig. 3.33, p. 64). Although it is often obvious, it may be a shallow layer of blood, easily overlooked. It is uncommon for subdural haematomas at presentation to cause significant mass effect on the underlying brain (Fig. 3.30), though blood may continue to accumulate with time. S) Mixed density haematoma is more common at presentation in non accidental injury, but has been described within 2 days79 and 1 week of accidental injury.s4 Precise dating of subdural haemorrhage is difficult on both CT and MR scans, unlike the dating of intracerebral haemorrhage in adults. There seems to be considerable variation in density and intensity, which may be related to the initial volume of haemorrhage, the patient's haemoglo bin level, dilution by CSF, the oxygen tension within the CSF, and by a layering out of blood within the haemor rhage, the 'haematocrit effect' (Fig. 3.34, p. 65). Intervention,
62 I
Radiology of child abuse
Figure 3.30 Magnetic resonance scans of an infant. (a) Fluid-attenuated inversion recovery [FLAIR) coronal, (b) Tl-weighted sagittal at presentation and (c) T1 sagittal 10 weeks later. High-intensity interhemispheric subdural haemorrhage is indicated by arrowheads. The low-intensity posterior fossa subdural collections [stars) are presumed to be older. Secondary herniation of the craniocervical junction (white arrows) resolved along with the subdural. An ill-defined, subependymal, high-intensity signal suggests a shearing injury [arrow). There is also a large subgaleal fluid collection [asterisk). Adapted with permission from McPh il lips.75
-
4.:
-
-
--
---
~
-
~
Head
injury I
63
such as the insertion of a pressure monitoring device or a subdural tap, will also have an effect. Appearances on sequential scans and comparison of initial CT and MR scans may be helpful.
Subarachnoid Haemorrhage Spontaneous subarachnoid bleeds can occur due to rupture of an arteriovenous malformation or an aneurysm, which is thought to be rare under 1 year of age. The distribution of blood differs from that in non-accidental injury.76 Acute subarachnoid blood is more clearly seen on CT than on MR (Fig. 3.31). In inflicted brain injury subarachnoid blood is not unusual and may result, acutely, in arterial vasospasm and later in secondary hydrocephalus.
Parenchymal Brain Injury
Figure 3.31 Computerized tomography scan shows a fresh, high-density scalp haematoma. Subarachnoid haemorrhage is visible in the sulci, mainly over the left parietal and occipita l lobes.
Focal areas of parenchymal contusion and haematoma are caused by impact of the brain against the adjacent skull or dura, such as the falx or tentorium, and are most common in the cortical grey matter of the frontal and anterior tem poral lobes. It is not clear if contusional injUly can be caused by shaking alone, or whether impact against a firm surface is required. While these injuries may be seen on CT
Figure 3.32 (a) Acute computerized tomography scan showing high-density fresh interhemispheric subdural haemorrhage poster iorly. There is some loss of the normal grey-white matter different iation due to mild cerebral oedema . (b) Follow-up scan 4 weeks later. There is mild generalized cerebral atrophy.
64 I
Radiology of child abuse
Figure 3.33 (a) Fluid-attenuated inve rsion recovery (FLAIR) coronal MR image at presentation shows a thin layer of subdural blood over the convexity on the right and in the subtemporal region bilaterally (arrows). T2-weighted transverse images. (b) At presentation there is loss of grey-white matter differentiation on the right. with sl ight swelling and subtle midline shift. (e) Ten days later there is increased intensity in the right hemisphere and left frontal lobe. with prominent sulci due to early atrophy. (d) Nine months later there is extensive atrophy on the right. sparing only the basal ganglia. The left hemisphere is normal at this level. Adapted with permission from McPhillips.75
-
-
--
Head injury I
scanning, MR scanning, particularly in the coronal plane, demonstrates them well, and gradient echo (GRE) images are very sensitive for blood products, which are seen as low intensity. They may be subtle at autopsy, and knowledge of the MR findings will direct attention to the appropriate areas 30 Shearing injury is unusual in head injury but, when seen in the absence of a history of a high-velocity injury, such as a road traffic accident, it is highly suggestive of non-accidental injury. The most common site for focal shearing injury is at
the grey-white matter junction in the frontal or frontopari etal lobes or in the corpus callosum (Figs 3.30 and 8.17, p. 157). At the grey-white matter junction it may actually be seen as a tear, which is readily visualized using high frequency ultrasound scanning, and is seen as a low echogenicity focus. 22 Tears may be seen on CT, but as they lie close to the vertex they are not easily characterized. On MR they may be seen as haemorrhagic (Fig. 3.34) tears initially, later becoming cystic. Follow-up imaging shows that some persist as cysts 85 (Fig. 3.35), while some show focal gliosis
Figure 3.34 Sagittal Tl-weighted magnetic resonance images. The presentation scan (al shows increased intensity posteriorly, in keeping with layering in a subdural haemorrhage (arrowheads), and a focal area of high intensity (arrow) over the frontoparietal convexity, suggestive of focal clot formation. Six days later (bl, an ill-defined, high-intensity signal can be seen in the frontoparietal subcortical white matter (arrow), representing a tear. Two months later (cl, there is residual focal atrophy (white arrows). Adapted with permission from McPhilli ps l5
..
-------
65
66 I
Radiology of child abuse
Figure 3.35 Coronal fluid-attenuated inversion recovery (FLAIR) magnetic resonance im ages at (al 3 days, (bl 3 weeks and (cl13 months after presentation. (a) III-defined, high-inten sity signal over the left convexity (arrow) and subtle irregularity inthe underlying parenchyma. (b) Cortical tear with low-intensity cyst formation (arrow) . (c) The tear now appears as a slit-like lesion (arrowl with adjacent gliosis (arrowheads) and subtle atrophy. Adapted from with permission from McPhilli ps 85
and others collapse with associated white matter loss 86 (Fig. 3.3 4) . Sometimes shearing injury may be seen as tiny focal haemorrhagic areas on MR scanning, and knowledge of these findings may direct the pathologist to a subtle lesion. 3D
0-
_
Oedema or swelling of the brain may be as a direct result of trauma, or may be a result of hypoxic-ischaemic damage or hyperaemia. Traumatic oedema is usually associated with contusion or parenchymal haemorrh age. Hypoxic-ischaemic
-
- -
- -
Head
injury I
67
Sequelae of Head Injury Some children who have suffered non-acciden ta l hea d injury will have no residu al radiological abnormality on follow-up several years later. 64 - 67 Others show mild cere bral atrophy with prominent ve ntricles and CSF spaces (Fig. 3.35). Focal injury may be seen as areas of focal infarction, while some who have had extens ive hypoxic-ischaemic injury may have widespread multicys tic encephalomalacia and severe atrop hy (Fig. 3.33). Hydrocephalus may be seen secondary to haemo rrhagic arachnoiditis. Some children with sub dural haemorrhage who have also developed cerebral atrophy may have chronic subdural collections (Figs 14.9 and 14.10, pp. 300- 30 1).
Spinal Cord Injury
~
Radiological appearances of spina l inju ry in non-accidental injury are not well described. There are reports of fracture-subluxation of the vertebrae with associated cord compression.53 . 54 For eva luation of suspected injuries to the spinal cord, MR is necessary as injlllies may be present in the absence of other rad iographic abnorma lity (spinal cord injury with out radiographic abnorma lity, SCIWORA). Under lying abnormalities, such as atlan to- ax ial instability or block vertebra, which would predispose to cord injury, should also be sought. 28
Figure 3.36 CT scan showing an acu te reversal sign with low density brain, loss of normal grey-white matter differentiation and relative sparing of the thal am i an d basal ganglia, which are dense by comparison. Also visible are multiple areas of intra cerebra I petech ia I haemorrhage; bi lateraI ch ron ic subd ura I haemorrhages with fresh interhemi spheric subdural blood. There is an old linear infarct in the left occipital lobe.
Cerebrospinal Fluid Spaces
mJury is the most common radiological manifestation of non-accidental head injury and is often quite extensive. In severe cases, there is sparing only of the cerebellum and pos sibly the basal ganglia and th alami , giving rise to the 'rever sal sign' on CT scanning, in which the cerebral hemispheres are hypodense compared with these structures 87 (Fig. 3.36). Except for the occasional documented vascular dissection , the underlying causes are not well established. There are many theories abou t the seconda ry mechanisms of brain injury and cerebral oedema, wh ich are discussed further in Chapter 14. In hypoxic-ischaemic damage there is loss of clarity of the grey- white matter interface on all scanning techniques (Fig. 3.33). On CT scanning, the affected area may show abnormally low density (Fig. 3.32, p. 63), although on MR sequences the in tensity will reflect the increased water content, low on Tl-weighted sequences and high on T2 weighted sequences. Th e combination of diffusion-weighted imaging (OWl) and ap parent diffusion coefficient (ADC) mapping will show restricted diffusion, and is more sensitive than other MR sequences; however, this is not specific and may be seen in metabolic disorders, following seizures, and in encephalitis. It may be useful in predicting the prognosis.
The depth of CSF fluid spaces surrounding the brain changes in the first 2 years of life. Paediatric radiologists who scan young children are familiar with the healthy child who has widened CSF spaces and minim al, if any, ventricular dilatation. It is considered a developmental variant. Kleinman et al 88 reviewed the CT scans of 34 nor mal children, finding that the extraventricular subarach noid space is increased in children under 2 years of age, normalizing at aro und 2 years and becoming 'monoto nously uni form ' by 3 yea rs. They postulated theories for this transient increase in CSF spaces, suggesting that it was a transient alteration in CSF dynamics, perhaps associa ted with a response to the grow ing brain. 88 Kapila et al 89 stud ied CSF dynamics using nuclear cisternography a nd found no evidence of communicating hydrocephalus. Libicher and Trbger90 scanned 89 infants to determin e an upper limit of normal, based on the 95th percentile. Wilkinso n et al 91 obtained similar results, with some variation of depth with position. Fessell et al 92 reviewed cranial ultraso und scans of 38 patients with macrocrania, an occipitofrontal head circumference of 95 per cent or greater for age, who had been followed up for a mean period of 55 weeks. They showed that a CSF depth of < 10 mm had a 94-100 per cent
-~ -
68 I
Radiology of child abuse
negative predictive value for the development of neurolog ical abnormality. These [SF spaces should co ntain fluid with the same characteristics as the intraventricular [SF. If the density on [T images, intensity on MR images or echogenicity on ultrasound scanning is different, then this raises the possi bility of inflammation or haemorrhage, and requires fur ther investigation. Ultrasound scanning, using colour flow imaging to demonstrate subarachnoid vessels, can demon strate which comp artment is involved.
VISCERAL INJURIES Almost any organ can be injured as a result of inflicted trauma, bu t visceral injury is very unusual in child abuse, when compared with the incidence of skeletal and head injury. The typical abdominal and chest injuries result from kicking or punching the abdomen and chest, or from kneel ing or standing on the child. The child may present acutely owing to peritonitis and blood loss, or with late complica tions of injury, such as su bacute gastrointestinal obstruc tion resulting from stricture forma tion or pancreatic pseudocyst format ion. The investigation of visceral injury should be the same as for other blunt abdominal trauma.
Liver, Spleen, Kidneys and Adrenals The mechanism of abusive injury of these organs does not differ fro m accidental blunt trauma . Direct compression can cause laceration, rupture and haemorrhage. Associated rib fractures may be seen and the presence of other non accidental injury and the lack of a specific history of injury may suggest abuse. 93 95
Pancreas Pancreatic trauma can result in acute pancreatitis, often haemorrhagic. The presence of associated injuries, particu larly to nearby bowel, may suggest the traumatic aetiology. Pseudocyst formation can occur with associated mass effect. For the assessment of the degree of injury and asso ciated abnormality it may be necessary to use [T scann ing with intravenous and oral contrast. 96 -9S
Bowel Injury Duod ena l and jejunal injury can occur from direct trauma , compressing the duod en um against the spin e, but may also result from deceleration forces with shearing injury to the root of the mesentery.99 Duod enal hae matoma may be seen or transection of the j ej unum in the region of the ligament of Treitz. Ultrasound is not usually useful for ep igastric
--
-
Figure 3.37 Barium meal in a child who presented wi th vomiting, dehydration and weight loss, showing proximal duodenal dilatation secondary to duodenal haematoma causing an apparent stricture. The jejunal mucosa is thickened secondary to haemorrhage. trauma. Upp er gastrointestinal contrast studies may be use ful if the child is not acutely ill (Fig. 3.37), bu t abdominal [T (Fig. 3.38) with intravenous and oral contrast may be most helpful. Mesenteric tears with associated vascular injury may present late with multiple strictures. 94 Gastric rupture is a rare occurrence, and is likel y to be due to com pression of a distended stom ach. 95 Pneumatosis of the gas tric wall has been described. 100 Rectal perforation is usually the result of penetrating injury, associated with sexual ab use. 94 • 10 1
Visceral Chest Trauma Injury to the lungs, heart and mediastinum is unusual, a nd likely to be a result of direct compression injury. This can result in diaphragmatic rupture, pneumothorax or pneu mopericardium. Airway obstruction may cause a pneumo mediastinum. Rib fractures may cause a haemothorax (Fig. 3.39). Secondary inflammatory change .may extend to the mediastinum from adjacent structures, most commonly being involved with pancreatitis.
Penetrating Trauma Although pharyngeal perforation is a well-recognized iatro genic injury, it is a rare manifestation of non-accidental
-
-
-
-
--
Differential diagnosis
I
69
Figure 3.39 Healing rib fractures in both axillae with a right haemothorax. Recent angulated fracture of right humerus. The raised right hemidiaphragm is due to liver contusion.
SOFT-TISSUE INJURY It is not unusual for soft-tissue injury to be sufficient to
Figure 3.38 Abdominal computerized tomography with intravenous contrast. (a) Fluid in dilated duodenal loop (asterisks) with denser haematoma medially. Also visible is free fluid around the liver and spleen. (b) Caudal image of the same child also shows free intraperitoneal fluid and intense enhancement of bowel wall due to 'shock bowel'.
injury, but has been described. In infants it is due to the insertion of a finger or a sharp object. IO l-1 03 In older chil dren, penetration may be related to sexual abuse. Resulting retropharyngeal abscess formation may compromise the airway and inflammation may extend into the mediastinum. Ng et al 94 described two cases of penetration by multi ple needles. In one case, the needle marks were visible and CT scanning showed one needle to lie in close proximity to the carotid artery. In another case, the needles had been inserted through the umbilicus and probably per rectum, with abscess formation in the abdomen and pelvis. Foreign bodies seen on radiographs must not be assumed to be external to the patient. 94
cause obliteration of fat planes on radiographs, and such changes can also be seen on ultrasound and MR imaging. But, in general, radiology cannot reliably detect bruiSing. Older children may be beaten, with focal haemorrhage in muscle and soft tissue. These may ossify and appear radi ographically as heterotop ic new bone formation. 104 The mechanism is similar to post-traumatic myositis ossificans. Carty has described soft-tissue calcification of a neck lace distribution in the neck. It is thought that this may represent focal fat necrosis and ischaemia. 105
DIFFERENTIAL DIAGNOSIS To miss a diagnosis of non-accidental injury and to fail to safeguard the child may result in further injury or death. To misdiagnose non-accidental injury, when there is another cause for the child's condition and radiological appear ances, is to cause heartbreak and devastation to a family. It is therefore important that all paediatric radiologists have a knowledge of the differential diagnosis of both head injuries and individual skeletal abnormalities.
Head Injury The major differential for scalp swelling and skull fracture is accidenta l injury. The history should be sufficient to explain the findings and the degree of injury. Bony skull defects with bruising or the appearance of periorbital haematoma are recognized presentations of leukaemia and metastatic neuroblastoma.
70 I
Radiology of child abuse
Subdural Haematoma Even with a bleeding diathesis, some trauma is required to cause a subdural bleed, but it may be relatively minor in nature. Haematological investigation is appropriate in unexplained or suspicious cases of subdural haemorrhage. Rarely, an aneurysm or arteriovenous malformation may bleed, causing subarachnoid and subdural haemorrhage. Subdural effusions, which may be abnormal when com pared with intraventricular CSF on scanning, can be seen in meningitis. This can be confirmed clinically and by lab oratory testing. Herpes simplex encephalitis can be haem orrhagic; again clinical assessment and virological investigation should clarify this. Glutaric aciduria is uncommon, but may present with macrocrania. It can be recognized by the combination of cerebral atrophy, with particular widening of the Sylvian fissure and the CSF space a nterior to the temporal lobes, and abnormal signal in the basal ganglia and periventricular white matter. 106
Physiological Periosteal Reaction Physiological periosteal reaction is seen in normal infants between the ages of 8 weeks and about 8 months (Fig. 3.40). It is rarely more than 2 mm in thickness, smooth and seen along the diaphyses of the long bones. It does not extend to the epiphyseal plate. It is usually symmetrical, though may be more obvious on one side and, if identified on the tibia, should be visible on the femur (Fig. 3.15, p. 55).
important to be aware of the patient's stage of develop ment than their actual age, can they roll, do they stand, are they cruising? This information, together with the history, can often clarify a worrying scenario (Figs 3.41 and 3.42).
Normal Variants The paediatric radiologist should be familiar with the appearances of nutrient foramina, which may be mistaken for fractures by the unwary.107 There are numerous variants of ossification, particularly around metaphyses. These may be spurs, beaks and the 'step-off appearance, which resemble metaphyseal fracture. But close inspection, using coned views and magnification of high-quality radio graphs if necessary, will show that these structures are in continuity with normal bone and do not represent frac tures. 4 The acromion may also show irregular ossification. 6 Irregularity of the anterior cortex of the proximal tibia is also a normal variant,4 as is irregularity of the radial shaft. There are numerous sutures and synchondroses in the skull, most of which are easily recognized. The parietal fissure, or 'strip suture', is frequently misdiagnosed as a fracture. It is a short tapering lucency extending from the skull vertex into the parietal bone. They are no wider than
Accidential Injury In accidental injury, the fracture should be compatible with the mechanism described in the history. It is more
Figure 3.40
Simple physiological subperiosteal new bone seen
along the diaphysis of both femora , both medially and laterally.
~--
Figure 3.41
Typical innocent toddler's spiral fracture.
-
--
Differential diagnosis I
71
2 mm, and are not assoc iated with overlying soft-tissue swelling.
described, and it should be remembered that it can be detected for up to 4 weeks following delivery.77,78
Birth Trauma
Osteomyelitis
The most common fracture sustained during delivery is of the midshaft of the clavicle. It may not be recogni zed at birth but it either presents at 2-3 weeks of age, with a palpable lump, or is seen as an incidental finding on a chest radiograph. Most h ave healed by 12 weeks and are undetectable. A few fail to unite, leaving a pseudoal1hrosis of the clavicle, which may be mistaken for a new fracture at a later date. Other diaphyseal fractures are usually recognized owing to reduced movement of the limb, humeral fractures in shoulder dystocia and femoral fractures in extended breech delivery. Rib fractures, resulting from impaction against the maternal symphysis pubis, are usually unilateral. 43 ,44 They may present as a result of crepitus or respiratOlY distress, or be seen as an incidental finding on a chest radiograph . Metaphyseal fractures may rarely be due to birth trauma, when a limb has been twisted or pulled, in the course of Caesarean section,39 an armling deliv ery or a breech delivery. The finding of subdural haematoma in a significant number of asymptomatic neonates has been
Periosteal reaction and minor metaphysea l or cortical irregularity may be see n on ultrasound or, sev era l days later, radiographically. In the absence of a histolY of trauma or clinical evidence of infection this may be ascri bed to non-accidental injury.
Metabolic Bone Disease Several conditions can present with reluctance to use a limb, in the absence of trauma. Radiographs may show a metaphyseal lucent line. Careful inspection of the radi ograph will show that this is present at every metaphys is, though more obvious at some. There is also no visible frac ture. This is a well-recognized presentation of leukae mia and neuroblastoma, and the child is likely to be systemi cally unwell. Langerhans cell histiocytosis may h ave a sim ilar presentatio n but the bone lesions are unlikely to be symmetrical and generalized. Meta physeal fractures are well recognized in scurvy, Menke syndrome and metabolic disease of the newborn. Scurvy has typical radiological appeara nces, with osteopenia, blurred metaphyses and poor trabeculation. Clinical appea r ances and biochemical investigations will clarify the diagno sis in most cases. Copper deficiency is rare, but well described. 108 Prema turity, mal nutrition, malabsorption and dietary defic iency are predisposing factors. Copper has been added to modern milk formulae and parenteral feeds to prevent its defi ciency in preterm infants. Fractures are seen in copp er defi ciency only in the presence of widespread symmetrical skeletal chan ges, showing periosteal reaction, metaphyseal spurs and cupping and fraying of the metaphyses, visible especially at the wrists, knees and costochondral junctions. Biochemical investigations will confirm this disorder. Rickets is characterized by metaphyseal fraying and cup ping, with osteopenia and poor trabeculation. Fractures may occur, but only in the presence of obvious bony abnormal ity, confirmed by biochemical abnorm alities (Fig. 3.43). Congenital syp hilis can be confirmed by clinical features and serology. Metaph yseal in'egularity is seen, but the appearances are those of undermining of the metaphysis.
Osteogenesis Imperfecta Figure 3.42 (a) Simple torus fracture of the distal tibia with slight change in the normally smooth contour of the anterior cortex. This fracture is technically metaphyseal, but this is a simple accidental injury, not non-accidental injury. (b) Healing with sclerosis at the fracture site.
Osteogenesis imperfecta is a disorder of synthesis of type I collagen, with resulting bone and connective tissue abnor malities. It is an inherited condition, with four major types described by Sillence. 109
72 I
Radiology of child abuse
Figure 3.43 Immigrant child with nutritional rickets and severe osteopenia. (a) Left lower limb shows healing fractures of the tibia and fibula and a fresh corner fracture of the proximal tibia. (b) Upper limb shows a healing fracture of the radius and abnormal periosteal new bone extending to the humeral metaphysis. Typical metaphyseal fraying and cupping best seen in the distal ulna. The faint metaphyseal sclerosis is due to healing, following treatment. A diagnosis of non-accidental injury was not made because of the rickets, although there were grave concerns about the family.
• Type I is the most prevalent form, with blue sclerae, a family histolY of deafness, and sometimes abnormal dental development. Ten per cent of people with osteogenesis imperfecta show fractures at birth. This is an autosomal dominant condition but new mutations do occur. • Type II can be recognized in utero and is usually lethal in the fetal or perinatal period. It is thought to be transmitted as an autosomal recessive condition. • Type III is a rare autosomal recessive form of osteogenesis imperfecta, resulting in deformity of the limbs, spine and skull. Fractures are often present at birth. • Type IV is regarded as a rare form of osteogenesis imperfecta. The sclerae are normal in appearance. The bone disease is very variable in severity. This group may be subdivided, according to the presence of dentinogenesis imperfecta. In a mild case, particularly because of the normal sclerae, it may be difficult to
..
-
distinguish osteogenesis type IV from non-accidental injury.
It should be remembered that children with osteogenesis imperfecta susta in fractures similar to fractures of other children from accidental trauma, just with a lesser degree of force. They w ill therefore have fractures that are similar in distribution to those in accidental trauma, and would not be expected to show metaphyseal corner fractures. Simila rly, rib fractures and skull fractures are not commonly seen in osteogenesis imperfecta. Because fractures are painful and cause the child distress, there should not be any significant delay in presentation to medical attention; therefore, previ ously unidentified fractures in various stages of healing would not be seen. Subdural haemorrhage is a rare feature of osteogenesis imperfecta, recognized in severe cases, in which the diagnosis is not in doubt. 110 Brain injury is not a feature of osteogenes is imperfecta.
--
-
References I
There are rep orts of suspected non-accidental injury that have later been proven to be osteogenesis imperfecta . Review of the described histories would sugges t that fea tures of osteogenesis imperfecta or clinical features unusual in non- accidental injury were overlooked.IIL.112
Temporary Brittle Bone Disease There have been two hypotheses proposed for conditions that wo uld render a child more susceptible to fractures for a short period of time, with spontan eo us resolution of the abno rmali ty and with no biochemical or pathological abnormalities at the time of diagnosis. Paterson et al sug ges t tha t this is caused by a temporary copper deficiency, resulting in a temporary bony fragility w hich does not resolve until the copper levels have returned to normal. liJ His cohort included patients with rib, metaphyseal and skuJl fractures. Although metaphyseal fractures are recog nized in established copper deficiency, bone changes are acco mpanied by abnormal biochemistry and by anaemia: skull fractures are not seen. 108 A review by Chapman and Hall has co nsidered Paterson's proposed condition in detail. I14 Another group suggest that reduced bone density due to lack of fetal intrauterine movement could cause abnormally low bone density and a higher propensity to fracture. 115 The validity of this data is questioned 115 and contradicted by other authors. 117
CONCLUSION It is the du ty of the paediatric radiologist to be an advocate for the child, who should be protected fro m abuse. Natural conditions which could mimic aspects of non-accidenta l injury, or underlying conditions which could predispose to fra cture or subdural haemorrhage, must be identified where present. It must be remembered that, although rarely, such a child may also be abused. IIB.119 A mistaken diagnosis of child abuse, when the appear ances have been caused by another disorder, can be devastat ing for the child and family. However, if we fail to recognize inflicted injury, the child may be the victim of further assault, which may result in life-long disability, if not death.
REFERENCES
2 3
4
Caffey J. Multiple Fractures in the long bones of infants suffering from subdural haematoma. A]R 1946; 56:163-73. Guthkelch AN. Infanti le s ubdu ra l haematoma a nd its relationship to whiplash injuries. EM] 1971; 2:430-1. Caffey J. On the theory and practice of shaking infants. Its potential residual effec ts of permanent brain damage and mental retardation. Am] Dis Child 1972; 124:161-9. Kleinman PK. Belan ger PL, KareJlas A, Spevak MR. Normal metaphyseal radiologic variants not to be confused with ftndings of inFant abuse. A]R 1991 ; 156:781-3.
73
5 Kleinman PK. Schm id-like metaphyseal chondrodysplasia
Sim ul a tin g child abuse. A]R )991; 156: 576-8.
6 Kleinman PK, Spevak MR. Variations in acromial ossificatio n Simulating infant abuse in victims of sudden infant death syndro me. Radiologv 199 1; 180:185- 7. 7 Sbapiro R. Anoma lous parietal sutures and the bipartite parietal bon e. A m] Ro el1tgenol Radium Ther Nucllvled 1972; 115:569-77. 8 Omah AC, Hall CM. Observational study of skeleta l surveys in suspected non-accidental. injuly. Clinical Radiology 2003; 58: 70 2-5. 9 James SU, Halliday K, Somers J, Broderick N. A survey of
non-accidental injury im aging in England, Scotland and
Wales. Ciin Radio/2003; 58:696-701.
10 Sane SM et al for American Academy of Pediatrics. Dia gnost ic imagi ng of child abuse. Pediatrics 2000; 105:1345-8. 11 American College of Radiology. ACR Practice Guidelinefor Skeletal Surveys in Children (Res. 47, 17,35). Reston, VA: ACR Standards American College of Radiology, 2001, pp. 145-9. 12 Royal Co llege of Radiolo gists and Royal College of Paediatrics and Child He alth. Standards for Radiological hll!es tigatiolls of Su spected Non-accidental h~j!II)!. London: Royal College of Paediat rics and Child Health, 2008. 13 Omah AC, Moon L, Hall CM, Todd-Pokropek A. Diagnostic accuracy of fracture detection in suspecte d non- acc idental injury: the effect of edge enhancement and digital display on observer performance. Ciill Radiol 2006; 61: 163-73. 14 Zimm erman S, Mako roFF K, Ca re M, et al. Utility of Follow-up skeleta l surveys in suspected ch ild physical abuse evaluations. Child Abuse Et Neglecr, 2005; 29: 1075-83. 15 Kleinman PK, Nimkin K, Spevak MR, et a!. Follow-up skeletal surveys in susp ec ted child abuse. A]R 1996; 167: 893- 6. 16 Anilkumar A, Fender LJ, Broderick NJ, et a1. The role of the follow-up chest radiograph in suspected non-accidental injury. Pediatl' Radio12006; 36:216-18. 17 Kleinm an PK, Marks SC, Adams VI, Blackbourne BD. Factors affecting visualization of posterior rib fractures in abused infa nts. A]R 1988 ; 150:635-8. 18 Kleinman PK, Marks Sc. Spevak MR, Richmond JM. Fra ctures of the rib head in abused infants. Radiology 1992; 185: 119-23. 19 Nimkin K, Kleinman PK, Teeger S, Spevak MR. Distal humeral physeal injuries in chi ld abuse: MR imaging and ultrasonography findings. Pedlatr Radio/ 1995; 25:562-5. 20 Dias JJ , Lamont AC, Jones JM. Ultrasoni c diagnosis o f neonatal separatio n of th e distal humeral ep iphysis. ] BOHe ]Oil1t Surg (Er) 1988; 70-B:825-8. 21 Smeets AJ, Robb en SG, Meradji M. Sono grap hically detected costo-chondra l dislocation in an abused chi ld. Pediatr Radiol 1990; 20:566-7. 22 Jaspan T, Narborough G, Punt JAG, Lowe J. Cerebral contusional tears as a marker of child abuse - detection by cran ial sonography. Pediat,. Radiol 1992; 22:237-45. 23 Filiatrault D, Garel L. Commentary: Pediatric blunt abdominal trauma - to sound or not to so und? Pedia tr Radiol )995; 25:329 - 31. 24 Sivit CJ, Kaufman RA . Commentary: Sonography in the evaluation of children following blunt trauma: is it to be or not to be? Pediatr Radio 11995 ; 25: 326-8. 25 Ellison PH, Tsai FY, Largent JA. Computed tomography in c hild abuse and cerebra l contusion. Pediatrics 197 8; 62:151-4. 26 Zimmerman RA, Silaniuk LT, Bruce D, et al. Computed tomography of craniocerebral inju ry in the abused chi ld. Radiology J 979; 130:687 -90. 27 Saulsbury FT and Alford SA. Intracranial bleeding from child abuse: the value of skull radiographs. Perliatr Radiol 1982; 12: 175 -8.
74 I
Radiology of child abuse
28 Barnes PD, Krasnokutsky M. Im aging of the central nervous system in suspected or alleged nona ccid ental injury, including the mimics. Topics MR Imaging 2007 ; 18:53-74. 29 Suh DY, Dav is PC , Hopkins KL, et a1. Nonaccidental pediatric head injUly: diffusion-weighted imaging findings. Neurosurgery 2001; 49 :309-20. 30 Hart BL, Dudley MH, Zumwalt RE. Post mortem cra nial MR T and autopsy correlation in suspected child abuse. Am )
51
52 53 54
Foren sic NJed Patho/1996; 17: 2 17-24. 31 Sty JR, Starsh a k R1. The role of bon e sci ntigraphy in the eva luation of the sus pec ted abused child. Radiology 1983; 146:369-7 5. 32 Ma nd elstam SA, Cook D, Fitzgera ld M, Ditchfielcl MR. Complementary use of radiolog ical skel etal survey and bo ne scintigraphy in detection of bony injuri es in su spected child abuse. Arch Dis Child 2003; 88:387-90. 33 Carty HML. Fractures ca used by child abuse. ) Bo ne Joint Surg Br 1993; 75B:849-57. 34 Rao P, Carty H. Non-accidental injury: review of the radiology. Clin Radiol 1999; 54: 11-24. 35 Taitz J , Moran K, O'Mea ra M. Long bone fractures in children under 3 years o f age: is abuse being missed in emergency department presentatiolls? ) Paediatr Child Health 2004; 40:170-4. 36 Worlock P, Sto wer M, Barbor P. Pattern s of fractures in accidental and non-acc id ental injulY ill children: a comparative study. Br Med) Ciin Res Ed 1986; 293:100-2. 37 Kleinman PK, Marks Sc, Richmond 1M, Blackbo urne BD. lnflicted skeletal injury: a postmortem radiologic histopathologic study in 31 infants. AJR 1995; 165: 647-650. 38 Grayev AM, Boal DKB, Wallach DM, Segal LS. Metaphyseal fractures mimicking abuse durin g treatment for clubfoot. Pedin tr Radiol 2001; 3 1 :559-63. 39 O'Connell AM, Do noghue VB. Can classic metaphysea l lesions fol low un co mpli ca ted caesarean section? Pediarr Radiol 2007; 3 7:488-91. 40 Kleinman PK, Marks SC, Bl ackbourne B. The metaphyseal lesion in a bu sed infants: a radiologic-histopathologic study. AJR 1986 ; 146:895-905. 41 Bulloch B, Schubert CJ, Brophy PD, et a1. Cause and clinical character isti cs of rib fr ac tures in infants. Pediatrics 2000; )05:E48. 42 Cadzow SP, Armstrong KL. Rib fr ac tu res in infants: red al ert! The clini cal features, investig ations and child protection ou tco mes. ) Paediarr Child HealrlJ 2000; 36 :322-6. 43 Durani Y, DePiero AD. Images in internal medicine. An n Em erg Med 2006; 47:210-15. 44 Rizzolo PJ, Coleman PR. Neonatal rib fracture: birth trauma or ch ild abu se 7 ) Family Practice 1989; 29:561-3. 45 Kleinman PK , Marks SC, Adams VI, Blackbourne BD. Factors affecting visualization of po sterior rib fractures in abused infants. A)R 1988; 150:635-8. 46 Kleinman PK, Schlesinger AE. Mechanical fa c tors associated with posterior rib fractures: laboratolY and case studies. Pediotr Radiol 1997; 27:87-91. 47 Kleinman PK, Marks Sc, Nimkin K, et a1. Rib fra ctu res in 3 1 abused infants: postmortem radiologic-histopathologic study. Radiology 1996; 200:807-10. 48 Feldman KW, Brewer DK. Child abuse, cardiopulmonary resuscitation and rib fra ctures. Pediarrics 1984; 73:339-42. 49 Sp evak MR , Kl ei nm a n PK, Belanger PL, et a1. Cardiopulmonary resuscitation an d rib fractures in infan ts . A post- mortem radiologic-pathologic st ud y. )AMA 1994 ; 272:617-18. 50 Dolinak, D. Rib fractures in infants due to cardiopulmon ary resuscitation efforts. Am J Forensic iVied Parlwl 2007; 28:107-10.
...
-
--
55
56
57 58
59
60
61 62 63
64 65
66 67
68
69 70
71
72 73
74 75
76
Kleinman PK, Marks Sc. Vertebral body fra ctures in child abuse: radiologic-histopathologic correlates. In vestigative Radiology 1992; 27:715-2 2. Kleinman PK, Zito JL. Avulsion of the s pinous processes caused by infant ab use. Radiology 1984; 151:389-91. Rooks VJ , Sisler C, Burton B. Cervical spine injury in child abuse. Report of 2 cases. Pediatr Radiol 1998; 28:193-5. Cullen Jc. Spinal lesions in battered babies. ) Bone Joint Surg Br 1975; 57B:364-6. Lonergan GJ , Baker AM, Morey MK, Boos Sc. Child abuse: radiologic-pathologic correlation. Radiograph ics 2003; 23:811-45. Kogut! MS, Swischuk LE, Fagan CJ. Patterns of injury and significance of uncommon fractures in th e battered child syndrome. Am) Roencgenol Radium Ther Nucl Med 1974; 121:143-149. Kleinman PK, Shelton YA. Hangman's fracture in an abused infant: imaging features. Pedian' Radio11997; 27:776 -7. Nimkin K, Spevak MR, Kleinman PK. Fractures of the hands a nd feet in child abuse: Imaging and pathologic features. Radiology 1997; 203 :233- 6. Carty H. Th e non-skeletal injuries of child abuse: part II the body. In The Year Book of Paediatric Radiology, Borsod County Teach ing Hosp ital, Miskolc, 1991 ; 3:25-34. Hechter S, Huyer D, Manson D. Sternal fractures as a manifestation of abusive injury in children . Pediatr Radiol 2002 ; 32 :902-6. Ablin DS, Greenspan A, Reinhart MA. Pelvic injuries in child abuse. Pediatr Radiol 1922; 22 :4 54-7. O'Connor JF, Cohen 1. In Kleinman PK (ed.) Diagnostic Im aging of Child Abuse. Baltimore: Wi lli ams and Wilkins, 1987, p. 112. Prosser I, Maguire S, Harrison SK, et al. How old is this fracture? Radiologic dating of fractures in children: a systematic review. A JR 200 5; 184: 1282 -6. Haviland J, Ross Ru sse ll RI. Outcome after severe non accidelltal head injury. Arch Dis Child 1997; 77:504-7. Jayawant S, Rawlinson A, Gibbon F, et al. Subdural haemorrhages in infants: population ba sed s tudy. BM) 1998 ; 317:1 558-61. King SJ, Boothroyd AE. Cranial trauma following bi rth in term illfa nts. Br J Radiol 1998 ; 71 :233-8. Lo TYM, McPhillips M, Minns RA, Gibson RJ. Cerebral atrophy following shaken impact syndrome and other non-accidental head injury. Pediatric Rehabilita tion 2003; 6:47-55. Lloyd DA , Carty H, Patterson M, et al. Predictive value of skull radio g raphy for intracranial injury in children with blunt head injUly. Lancet 1997; 349:821-4. Billmire ME, Myers PA. Serious head injury in infants: accident or abuse? Pediatrics 1985; 75:340- 2. Warrington SA, Wright CM , ALSPAC Study Team. Accidents and reS Ulting injuries in premobile infants: data from tb e AlSPAC stu dy. Arch Dis Child 200 1; 85: 104-7. Rob erton DM, Barbor P, Hull D. Unusual injUly? Re cent injury in normal children and children with suspected non accidenta l injury. BMJ 1982; 285:1399-401. Helfer RE, SJovis TL, Black M. Injuries resu lting when small children fallout of bed. Ped iatrics 1977;60:53 3-5. Kravitz H, Driessen G, Gomberg R and Korach A. Accid ental fa lls from eleva ted surfaces in infants from birth to o ne yea r of age. Pediatri cs 1969; 44 [Supp1.):869-76. Stoodley N. Neuroima ging in non-accidental head injury: if, when, why and how 7 Ciin Radial 2005; 60 :22-30. McPhillips M. Non-acc identa l injulY in the yOUllg infant. ln Rutherford M [ed.) MRJ of the Neonatal Bmin. London: WB Saunders, 2002 , pp. 261 -71. Newton RW. Inrra cranial haemorrhage and non-accidental injury. Arch Dis Child 19 89; 64:188-190.
-
References I
77
78
79
80
81
82
83
84
85
86
87
88
89
90 91
92
93
94
95 96
Whitby EH, Griffiths PO, Rutter S, et al. Frequency and natural history of subdural haemorrhage in babies and relation to obstetric factors. Lancet 2004; 363:846-851. Looney CB, Smith JK, Merck LH, et al. Intracranial hemorrhage in asymptomatic neonates: prevalence on MR images and relationship to obstetric and neonatal risk factors. Radiology 2007; 242:535-541. Tung GA, Kumar M, Richardson RC, et al. Comparison of accidenta l and nonaccidental traumatic head injury in children o n noncontrast computed tomography. Pediatrics 2006; 118:626 - 33. Duhaime AC, AJario AJ, Lewand er WJ, et al. Head injury in very young children: mechanisms, injury types and ophtha lmologic findings in 100 hospitalized patients younger than 2 years of age. Pediatrics 199 2; 90: 179-85. Ew ing-Cobbs L, Kramer L, Prasad M, et a l. Neuroimaging, physica l, and developmenta l findings after inflicted and noninflicted traumatic brain injury in young children. Pediatrics 1998; 102 :3 00-7. Hy mel KP, Makoroff Kl, Laskey AL, et al. Mechanisms, clinica l prese ntations, injuries, and outcomes from infli cted ve rsus noninflicted head trauma during infancy: resu lts of a prospecti ve, multicente red, co mp arative study. Pedia trics 2007; 119:922-9. Case ME, Graham MA, Handy TC, et aJ. Position paper on fata l abusive head injuries in infants and yo ung children. Am J Forens Med Path 2001 ; 22: 112 - 22. We lls RG, Vetter C, La ud P. Traumati c low attenuation subdural fluid co llec tions in chi ldren younger than 3 years. Arch Pediatr Adolesc Med 2003; 157: J005 -1 0. McPhillip s M. Initial and sequential MRI in non-accidental hea d injUiy. In Minns RA, Brown JK (eds.) Shaking and other Non-a ccidental Head Injuries ill Children. London: MacKeith Press, 2005, pp. 262-70. Ordia IJ. Strand R, Gilles F, Welch K. Computed tomography of contusional clefts in the white matter in infants. J Neurosurg 1981; 54:696-8. Han BK, Towbin RB, de Courten-Myers G, et al. Rev ersa l sign on CT: Effect of anoxic ischae mi c cereb ra l injury in ch ildren. Am J Neuroradiol 1989; 10: 1191-8. Kleinman PK, Zito lL, David so n RI, Raptopoulos V. Th e subarachnoid spaces in ch ildren: normal variations in size. Radiology 1983; 147:455-7. Kapila A, Trice J, Spies WG, et a!. Enl a rged ce reb rosp inal fluid spaces in infants with subdural hematomas. Radiology 1982; 142:669-72. Libicher M, Trager J. US meas urements of the subarac hnoid space in infants: normal values. Radiology 1992; 184:749-51. Wilkinson AG, Cooke R, Tallur KK, et al. Pericereb ra l space measurements in infants: sonographic determination. Paper presented at Hydrocephalus 2005 Internati ona l Co nsensus Meeting, Queenstown, New Zealand, August 2005. Fessell DP, Frankel DA, Wolfson WP. Sonography of extraaxial fluid in neurolo gica lly normal infants w it h head circumference greater than or equal to the 95th percentile for age. J Ultrasound !vIed 2000; 19:443-7. Nimkin K. Teeger S, Wa lla ch MT, et a!. Adrenal hemorrhage in abused children: imaging and post mortem findings. AJR 1994; 162: 661-3. Ng CS , Hall CM, Shaw DG. The ran ge of v isce ral manifestations of non-accidental injury. Arch Dis Child 1997 ; 77:167-74. Barnes PM, Norton CM , Dunstan FD , et al. Abdomina l injury due to chi ld abuse. Lancet 2005; 366:234-5. Touloukian RJ. Abdomina l visceral injuries in battered ch ildren. Pediatrics 1968 ; 42:642-6.
75
97 Siovis TL, Berdon WE , Haller 10, et at. Pancreatitis and the battered child syndrome: Report of 2 cases with skeleta l involvement. Am J Roentgenol Radium Tit er Nucl iVIed 1975; 125:456-61. 98 Sivit Cl, Ingram 10, Taylor GA , et at. Post traumatic adrena l hemorrhage in children. CT findings in 34 patients. AJR 1992; 158;1299-302. 99 KJeinman PK, Brill PW, Winchester P. Resolving duodenal jejunal hematoma in abused chi ldren. Radiology 1986 ; 160:747-50. 100 Fulcher AS, Das Narla L, Brewer WH . Gastric hematoma and pneumatosis in chi ld abuse. Am J RO(,lltgenol 1990; 155: 1283-4. 101 Ablin OS, Reinhart MA. Esophagea l perforation by a tooth in ch ild abuse. Pediatr Radial 1992; 22:339-4 1. 102 McDowell HP, Fielding Ow. Traumatic perforation of the
hypopharynx: an unusual form of abuse. Arch Dis CiJild
1984; 59:888-9.
103 KJeinman PK, Spevak MR, Hansen M. Mediastinal pseudocyst ca used by pharyngeal perforation during ch ild abuse. Am J Roentgenol 1992; 158:1111-1 3. 104 Ablin OS, Greenspan A, Reinhart MA. Pelvic injuries in child abuse. Pediatr Radial 1922; 22:454-7. 105 Carty H. Case report; ch ild abuse - necklace calcification - a sign of strangulation? Br J Radio11993; 66:1186-8. 106 Twomey EL, Naughten ER, Donoghue VB, Ryan S. Neuroimaging findings in glutaric aciduria type I. Pedia tr
Radio/2003; 33:823-30. 107 108 109 110
III 112
113
114 115
116 117
118 II 9
Hartley LM, Khwaja OS, Verity CM. Glutaric aciduria type I and nonaccidental head injury. Pediatrics 2001; 107; 174-5. Shaw JCL. Copper deficiency and non-accidental injury. Arch Dis Child 1988; 63: 448-455. Sillence D. Osteogenesis imperfecta. An expanding panorama of variants. Clin Orthop 1981; 159:11-25. Pozzati E, Poppi M, Gaist G. Acute bilateral extradural hematomas in a case of osteogenesis imperfecta congenital. Neurosurgel)' 1983; 13:66-8. Gahagan S, Rimsza ME. Child abuse or osteogenesis imperfecta: how can we tell? Pediatrics 1991; 88:987-92. Wardinsky TO, Vizcarrondo FE, Cruz BK. The mistaken diagnosis of child abuse: a three-year USAF medical center analysis and literature review. Mil Med 1995; 160:15-20. Paterson CR, Burns J, f\IlcAJlion SJ. Osteogenesis imperfecta: the distinction from child abuse and the recognition of a variant form. Am J Med Gellet 1993; 45:J87-92. Chapman S, Hall CM. Non-accidental injury or brittle bones. Pediatr Radiol 1997; 27: 106-10. Miller MN, Hangartner TN. Temporary brittle bone disease: association with decreased fetal movement and osteopenia. Calc Tiss Int, 1999; 64:137-43. Mendels on Kl. Critical review of 'temporary brittle bone
disease'. Pediatr Radiol 2005; 35:1036-40.
Rodriguez Jl, Palacios J, Ruiz A. et al. Morphological cha nges in long bone development in fetal akinesia deformation sequ ence: An experimenta l study in curarized rat fetu ses. Teratology 1992; 45:213-21. Knight 01 , Bennet Gc. Non-accidental injury in osteogenesis imperfecta: a case report. J Pediatr Orthop '1990; 10:542-4. Dunca n AA, Chandy J. Ca se report: multiple neonatal fractures - dietary or deliberate? Ciin Radiol 1993; 48:137-9.
Further Reading Kleinman PK (ed.) Diagnostic Imaging of Child Abuse, 2nd edn.
Ba ltim ore: Mosby, 1998.
I
CHAPTER 4
I
HAEMATOlOG ICAL ABNORMALITIES THAT CAN SIMULATE ABUSE Angela Thomas
Introduction Primary haemostasis Secondary haemostasis Laboratory tests Measurements of primary haemostasis Evaluation of a bleeding patient Patterns of abnormal results Normal coagulation screen with a normal platelet count Abnormalities of platelet number or morphology
76
78 79 81
82 82 86 90 94
INTRODUCTION
Medical, nursing and other personnel who care fo r children have a responsibility to be aware of signs and symptoms that are suggestive of child abuse, including non-accidental inju ry (NAI) . .Equally, however, they must also recognize that medical and physical conditions may simu late abuse and that appro priate measures must be taken to confirm or rule out these conditions. Cutaneous lesions are by far the most common presenting manifestations of child abuse, I,2 and of the conditions that may simulate ab use haematol ogical abnormalities are manifest usually as bruising or other bleeding into the skin or mucosal mem branes. Over-reporting of natural disease as NAI will occur, especially if a full assessment of the child is not carried out. An incorrect diag nosis of ch ild abuse, a lthough at times unavoidable, can be devastating for the family a nd child 3,4 and, exceptionally, has led to parental suicide .s Another consequence of such misdiagnosis is that a condition such as haemophilia will go unrecognized and therefore untreated. This in itself may lead to morbidity and even mortality. If the child is revealed to have a genetic or serious blood diso rd er, regaining the trust of that patient and their family may be impossible. The mistaken diagnosis of child abuse has been described as a form of medical abuse. 6 However, it is important to remem ber that diagnosis of a medical condition that can simulate
-
-
Coagulation defects The neonate Drugs associated with bleeding Bone marrow failure syndromes Systemic disease associated with a bleeding tendency Activation of coagulation Conclusion References
96 97
98 99 100 101 101 102
abuse does not rule out abuse. 7 When the child has bruising in a recognizable pattern, such as a belt or hand, then sus pected abuse must be repo rted regardless of the results of laboratory tests. B•9 In addition, these children may be at greater risk of serious injUly secondary to bleeding/,ID as well as greater risk of bleeding secondary to an abusive injury.l1 The physi cian who evaluates the bruised child for a disorder of coagulation sh ould always assess the history and physical examination to determine whether the bruises were acqui red spontaneously, accidentally or as the result of abuse. The pattern of bruising or associated findings may yield important information. Knight and Bennett J2 describe a case of abuse in a child with osteogenesis imperfecta, which was diagnosed from the pattern of the injuries; in this case a periorbital haematoma in association with a spiral fracture of the humerus. Unusual diseases may mimic abuse - misinterpreta tion of ' usual' by the inexperienced, for example a Mongolian blue spot (see Chapter 8, Fig. 8.7, p. 152), can a lso lead to a mis diagnosis. 6 Advice from a paediatric haematologist, paediatric dermatologist, paediatrician or haematologist may be crucial in reaching the right diagnosis. Caution must be exercised when fo llowing guidance (frequently given) that abuse shou ld be suspected if there is significant bruising or bleed ing with no history of trauma or a history inconsistent with the severity of the injury.13-16 In a ch ild with a bleeding
-
-
Introduction I
diathesis, the severity of the bruising may be disproportion ate to the injury sustained or there may be a denial of any significant injury by the parent or caregiver ; a paediatrician who is not familiar with haematological disorders may mis takenly interpret this as evidence of abuse. 17 Clinical identification of the age of a bruise compared with the age given from the history has been used to assess the credibility of the history an d to identify whom the child was with at the time the bruise occurred. However, there is an increasing amount of evidence to suggest that the age of a bruise cannot be estimated reliably.18 A bruise is caused by blood that has escaped from damaged or leaky blood vessels, usually capillaries, into the interstitia l tiss ues, but the appearance of the bruise depends on several factors. The amount of blood that escapes depend s upon the integrity of the coagulation system, the force of the injury and the integrity of the vessels. The location of the bruise also determines its characteristics; for instance, periorbital and genital bruises will appear sooner than bruises on the extremities because the tissues are loose and the vessels poorly supported. 19 The clinical ageing of bruises is most frequ ently based upon colour of the bruise when compared with an estab lished chart. 2o However, the colour of a bruise depends not only on age, but also on amount of blood present, location beneath skin, skin colour and amb ient light. 21 Stephenson and Bialas 22 conducted a study on chil dren with bruises of known ages; photographs of the bruise were taken at different time intervals and aged by a 'blind' observer. They concluded that several different colours could be present at the same time within anyon e bruise and that bruises change colour at very different rates, even when sustained at the same time in the same child. Ageing of bruises from photographs is much less precise than many textbooks imply.21 ,22 However, assessing the age of a bruise on clinical examination may be easier, as other clues such as tissue swelling or abrasion might be presentY This was not the finding in a study by Munang and col leagues, 2J in which marked variability in colour descript ion between observers of the same bruises, and thus estimation of the age of the bruise, was demonstrated; less than one third of descriptions between two observers tallied either in vivo or from clinical photographs. Wide variability in bruise development and healing urges caution in ageing bruises and therefore this should never be used as the sole criterion for child abuse;21 the pattern of distribution of the bruising is a key factor that must be linked to the child 's history and stage of development. Bruises found in atypi cal areas in toddlers, such as the trunk, hands or buttocks, are of concern, as are bruises in normal infants under 9 months who are not yet mobile. 24 Sibert and colleagues l8 ,25 have develop ed a scoring system to evaluate these factors and such tools may be useful for discriminating between abused and non-abused children.18,25 In a prospective analysis of bruising in children with a nd without an inher ited bleeding disorder it has been shown that non-mobile babies without a bleeding disorder do not bruise, but once they start to roll they develop a few bruises that increase in
77
number as they begin to crawl and walk. How ever, children with severe bleeding disorders did develop bruising when non-mobile and had larger bruises than those with mild bleeding disorders, who in turn had larger bruises than those wit h no bleeding disorder. These observations indi cate that children with severe bleeding disorders develop bruises before they are mobile and bruise more frequently than a control popul ation. 26 Although the finding of bruises of different ages coupled with an inconsistent his tory has been cited as a hallmark of child abuse,9 not only is it difficult to age a bruise accurately, but also it is impor tant to note that a child w ith a bleeding diathesis may pres ent, quite normally, with bruising of differen t ages plus an inconsistent history. Haematological investigation of a brui sed child is mandatory in all cases when the bruising is une xplained or implausible, and in cases where some explanatio n is given or found but the bleeding that results is disproportionate to the injUly sustained. In a child who may poss ibl y have been ab used, it is essential that investigations are as atrau matic as possible and yield the maximum information. In the case of a bruised or bleeding child under these circum stances, whether or not a diagnosis of NAl is enteliained, a set of screening investigations should be performed on blood taken from a single ve nepuncture, with the labora tory set up and alerted to analyse the tests with the min imum of delay, using sma ll plasma samples. Further invasive tests, such as bleeding time, should usually be avoid ed at this stage. Initial screening and investigation is aimed at the diagnosis of the commoner causes of bleeding, and exclusion or confirmation of some of the rarer causes for the safety and management of the child. Further inves tigations might be needed if no explana tion of the bleeding is found or no admission of NAJ is made. In order to be able to interpret the results of coagul ation testing, the fundamentals of primary and secondary haemostasis must be understood. Artefact can significantly distort resul ts and lead to misdiagnosis; utmost care must be exercised in the way the specimens are taken and han dled prior to processing. Haemostasis depends upon the integrity of the tissues and vasculature, the n umber and function of the platelets (pri mary haemostasisl and the formation of fibrin (secondary haemostasis). The first of these is outwith the scope of this chapter and will not be dealt with in any detail, exceptions to this being Henoch-Schbnlein purpura, which is not uncommon in children and may be mistaken for NAl, and vita min C deficiency, which will be discussed briefly. A basic scheme of the haemostatic mechanism is shown in Figure 4.1. When a vessel is injured, the subendothelial collagen is exposed, resulting in platelet adhesion, activation and aggregation forming the primary haemostatic plug. Vascular injury also leads to vasoconstriction and slowing of blood flow, aiding this process. At the same time, tissue factor is expressed on subendothelial fibroblasts and monocytes ini tiating the coagulation process, by which the inactive plasma coagulation factors are converted to their active
78 I
Haematological abnormalities that can simulate abuse
Coll age n exposure
1
Vasoconstriction
Platelet ....- - acti vat ion
1
Reduced blood flow
Plait pug
Ti ssue thrombop lastin
1
Blood coagulation
~ Th"~bi"
1
Clot
Figure 4.1
1
and fibronectin, which result in the spreadi ng of the platelets over the vascular surface. In areas of low flow or low shear stress, this may be sufficient to keep the platelets attached . However, in areas of high shear stress, such as those found in capillalY beds and the arterial circul ation, further interaction is required between the GpIb-lX receptor and von Willebrand factor (VWF), a large, adhesive pl asma protein. Conditions where either of these receptors is absent, for example Glan zmann's thrombasthenia (Gpllb-lIIa) or Bernard-Soulier syn drome (GpIb-lX), or where there is a lack of VWF, are characterized by abnormal bleeding of variable severity and will be discussed later in the chapter.
Platelet Activation
Fibrin
Haemostasis after injury. Activation of the
coagulation system and platelets is necessary for formation of a stable fibrin plug.
counterparts by cleavage of peptide bonds. Thrombin is generated, converting fibrino gen to fibrin , which then sta bilizes the pl atelet plug. Activation offactor XIII by thrombin allows cross-linking of the fibrin strands and further stabil ization. Thrombin also activates the protein C pathway, which is inhibitory to coagulation and profibrinolysis, allowing dissolution of the fibrin clot as healing progresses.
PRIMARY HAEMOSTASIS Platelets must be present in ad equate numbers and function normally in order for the primary haemostatic plug to form effectively. Platelets are small, disc- like cells that remain in the circulation for 7-10 days. In their quiescent state, they ex press predominantly neutral phospholipids on their outer membrane, which are haemostatically inert. However, when activated, negatively charged phospholipids, predominantly phosphatidyl serine, are exposed, providing the membrane surface required for the coagulation reactions. On the cell surface there are also specific protein receptors for platelet agonists and adhesive glycoproteins, which are essential for effective haemostasis. Storage gran ules are also contained within the platelet cytoplasm, ex granules and dense (6) granules, whose contents are released by fusion of their membrane with that of the platelet.
Platelet activation rap idly foll ows adhesion and results in a change in shape of the platelets, generation of active medi ators an d secretion of the g ranul ar conten ts. The activation process is hi g hly specific owing to the presence of platelet recepto rs that recognize appropriate ago nists, the most important of these being thrombin and ad enosine diphos phate (ADP). Thrombin binds to the platelet at several si tes, including the Gplb-IX receptor as well as a sp ecific throm bin receptor, which is the most importa nt for activation. The thrombin recepto r has been identified as a large seven domai n transmembrane protein that binds proteolytically active thrombin, which is essential for the induction of platelet activation. This results in a change in shape of the platelets, platelet aggregation and secretio n of granule contents. A second agon ist, ADP, is much weaker than thrombin in vitro but is important in vivo as both red blood cells and vascular tissues release ADP in response to damage and lead to ADP-induced platelet activation and aggregation.
Platelet Aggregation Once platelets are activated, they aggregate, linking to each other to form a haemosta tic plug. The binding of fibrino gen to the GpUb-lIla complex is essential for bridging between acUacent platelets and is crucial to platelet plug formation. Fibrin is also bound by this receptor, helping to stabilize the platelet plug.
Platelet Secretion Platelet Adhesion The initial event in haemostasis is the attachment or ad hesion of platelets to the vascula r subendothelium that is exposed after injury to the non-thrombogenic vascular endothelial lin ing. Collagen receptors on the platelets, thought most likely to be the glycoprotein (Gp)la-IJa complex, bind to the exposed subendothelial coliagenY After adhesion, the platelets undergo further interactions involving the GpIlli-llia receptor
When platelets are activated , in addition to aggregation, a release reaction occurs in which the contents of the ex gran ul es and 6 granules are secreted. The constituents so released contribute to the process of primary and second ary haem ostasis as well as wound repair and vascular remodelling. The a granules release, amongst other things, VWF (synthesi zed by the pl ate let-forming megakaryocytes) and fib rinogen (synthesized by the liver and taken up into
Secondary haemostasis the platelet). It is possible that these pro tei ns bind to the platelet surface and participate in platelet adhesion and aggregate formation. This may be particularly important in platelet-rich plugs in which th ere is limited access to plasma proteins. Factor V is also contained in the ex gran ules and is important for formation, along with factor Xa, of the prothombinase complex on the ex posed phospho lipid surface of the activated platel et.
I
79
receptor that binds fibrino gen but also VWF, facilitating aggregation. VWF is non-covalently bound to factor VIIl and protects it from protein C proteolysis. Low levels ofVWF result in correspondingly low levels of factor VIII, which, if severe, can result in defects of secondalY haemostasis.
SECONDARY HAEMOSTASIS
Initiation and Amplification of Coagulation
Platelet Enhancement of Coagulation Activated platelets accelerate coagulation several thousand fold. As well as releasing procoagul ant microparticles an d factor V, activation of the platelet revea ls anionic phospho lipid sites on the platelet surface that are required as cofac tors. Two complexes of clotting fac tors, the tenase (X ase) complex (factors lXa and VIIIa) and the prothombinase complex (factors Xa and Va) are bound to the platelet sur face, thereby increasing their effective concentration and bringing them into closer proximity. Once bound to the platelet surface, coagulation factors are protected from coagulation inhibitors such as antithrom bin and activated protein C.
Von Willebrand Factor Von Willebrand factor (VWF) is a multifunctional adhesive protein that plays an important role in primary haemostasis. It is secreted from endothelial cells and is essential for stable platelet adhesion, paliicularly at high shear rates. On release, it binds to platelet GpIb-IX, allowing adhesion and subse quent activation of the platelet. This exposes the GpIIb-I1Ia
Ex posure of tissue factor on damage to the vascular endothelium results in the binding of both factor VII and factor VIla, each of which are present in the circulation in the quiescent state. Once factor VlJ is bound to tissue factor, it is very rapidly activated by the tissu e factor-factor VIla complex. This results in the activation of both factor X to factor Xa and factor IX to fac tor IXa. Some of the factor Xa thus formed will also bind to and activate the tissue factor/factor VII complex but, once bound, is rapidly inacti vated by tiss ue fac tor pathway inhibitor (TFPI). Unbound factor Xa stays on the cell surface and activates factor V to factor Va, which, together wi th factor Xa, generates a small amount of thrombin (IJa). This thrombin burst is sufficient to activate platelets, activate fuliher factor V, activate factor VIII by cleavage from VWF and perhaps activa te plasma factor XI to factor XIa. The generation of thrombin by this pathway, however, is insufficient to sustain adequate fibrin formation because of the ina ctiva tio n of the tissue factor-factor VIla-factor Xa complex by the binding of TFPI (Fig. 4.2). Amplification of the coagulation pathway comes from the sequence of reactions following the gener ation of factor IXa and the generation of factor XIa (Fig. 4.3). Factor IXa finds its way to the surface of an activated
x
TFPl
.~
1
TF
l.I
II
.---'----,-_--,-.L---,--_,...~ _ ___,_vv_~-~a--,-r---. Xa - Va ~ Endothelial cell or monocyte
Vila
Pl ate let
i TG~\NF
XI_Xla
"\
Va IX
Villa
IXa
Activated platelet
Figure 4.2
Tissue factor (TF) is exposed after injury and binds both factor VII and activated factor VII (Vila), which both activates factor IX and genera tes a small amount of thrombin from prothrombin (II ) by activating factor X. The thrombin activates factors XI, Vand VIII by cleavage from von Willebrand factor (VWF). Platelets are also activa ted and surface phosphol ipid (PL) is exposed. Tissue factor pathway inhibitor (TFPI) binds the tissue factor-Vlla-Xa complex and inactivates it, switching off the ini tia ting pathway.
80 I
Haematological abnormalities that can simulate abuse
Endothelial cell or monocyte Insoluble fibrin
Vila
~
IX
XIII
• Xilia
Figure 4.4 Conversion of fibrinogen to fibrin by thrombin and thrombin activation of factor XIII forming an insoluble clot by cross- lin kage.
Figure 4.3 Amplification of the coagulation system occurs after the initial thrombin burst when factors V, VIII and XI are activated along with platelets. Activated factor VII (Vila) by binding to tissue factor (TF) has already generated a small amount of activated factor IX (lXa), which binds to the exposed phospholipid (PL) on the activated platelets along with activated factor VIII (Villa) forming the 'tenase' complex. This activates free factor X, which binds to the platelet surface, along with activated factor V, forming the 'prothrombinase' complex. This converts factor II (prothrombin) to thrombin sufficient to cause clot formation. Generation of thrombin is sustained by continued activation of factor XI by thrombin activating more factor IX in turn.
platelet. where it binds to the exposed anionic phospholipid in association with its cofactor, factor V1Ila, which, generated by the initial burst of thrombin, is already bound to the platelet surface. This 'tenase' complex converts free plasma factor X to factor Xa. Still bound to the platelet surface, fac tor Xa and factor Va, again generated by the initial burst of thrombin , form the prothrombinase complex, which con verts plasma prothrombin to thrombin in amounts suffi cient to cause clot formation. Generation of thrombin is sustained by the activation of factor XI to factor Xla by thrombin, which activates more plasma factor IX to factor IXa in addition to the factor IXa initially formed by the tis sue factor-factor V1la complex. Thrombin finally converts fibrinogen to fibrin and through the activation of factor XIlI promotes cross-linkage of the fibrin to form a stable clot (Fig. 4.4).29 Thrombin also activates a proenzyme, thrombin activatable fibrinolysis inhibitor (TAFI) , which downregu lates fibrinolysis thus slowing clot lysis.29
Classical Coagulation Pathway Until recently, the process of coagulation has been described as a biochemical cascade consisting of an extrinsic and an
-
.
-
intrinsic pathway coinciding at the activation of factor X to form the final common pathway (Fig. 4.5).30.31 The intrinsic system is predicated on the assumption that factor XI is acti vated by the contact factors factor XII , prekallikrein and high-molecular-weight kininogens (HMWKs). Factor Xla then, as described above, activates factor IX in association with its cofactor, factor V1Ila, which then activates factor X and so on. In the extrinsic system, it is assumed that the release of tissue factor and factor V11 by damaged vessels activates factor X directly. The integrity of these assumed pathways is measured in vitro using the activated partial thromboplastin time (aPTI), which is based on contact acti vation of factor XII for the intrinsic pathway and the pro thrombin time (PT) for the extrinsic pathway. This scheme sti ll works well for the diagnosis of clinical bleeding prob lems, although it is inadequate to explain what happens dur ing in vivo haemostasis. For instance, individuals with factor XII deficiency do not bleed, those with factor XI deficiency have a mild to moderate bleeding disorder, and those with factor V11I or IX defiCiency have a severe bleeding disorder. Patients with factor V11 deficiency have bleeding problems despite having an intact intrinsic system, but for the most part only if the level of FV1I is less than 0.10 U/mL. J2 These anomalies are much better explained by the revised coagu lation process depicted in Figs 4.2-4.4. It can be seen that because in vitro testing is used to elu cidate the complex in vivo system, en'ors of interpretation can occur where prolonged clotting times may not be associated with a bleeding diathesis. Examples of this are factor XII defi ciency and deficiencies of prekallikrein and HMvVKs. They are not necessary for normal haemostasis in vivo, as they are required neither for initiation nor amplification of coagula tion . The aPTT can also be prolonged if a lupus anticoagulant is present. Phospholipid is added exogenously to the in vitro coagulation reaction and can be easily inhibited by antiphos pholipid antibodies - the so-called lupus anticoagul ant. In vivo, no such inhibition occurs because the phopholipid is provided by the activated platelet and is protected from the circulating antibody.
Laboratory tests I
Intrinsic pathway
81
Extrinsic path way
XII_Xlla
1
1
XI_Xla
aPTI
IX-IXa
VII_Vila
Vllla~~
PT
Com,," p",hw",
X II -
}-:: Throm bin - - -
Fibrinogen _ Figure 4.5
1
Fibrin
The classical pathway of co agulation: it is proposed that the intrinsic system is activated by cont act with an activating
surface and the extrinsic system by tissue factor released from damaged vessels or tissue. 80th systems activate factor X which via the final common path way results in the formation of fibrin. The prothrombin time (PT) reflects the acti vity of the extrinsic and common 8ath w ays. The acti vated partia l thromboplastin time (aPTI) is most sensitive to changes in the intrinsic pathway. II, factor II (prothrombin); Va, activated factor V; Vila, activated factor VII; Villa, activated factor VIII; IXa, activated factor IX; Xa, activated factor X; XI a, activated factor XI; Xlla, activated factor XII.
LABORATORY TESTS A full blood count will yield important information regard ing platelet number, and some au tomated counters will give the mean platelet volume. However, there will be no evaluation of function, an important parameter in primary haemostasis. Haemoglobin and white cell count may also indicate an underlying haematological disorder such as leukaemia or aplastic anaemia. Examination of the blood film will reveal morphological abnormalities of platelets such as size (Wiskott-AJdrich, Bern a rd-Souli er, May Hegglin, idiopathic thrombocytopenic purpura) or abnormal granulation (Chediak-Higashi and Hermansky-Pudlak syndromes). The basic screening tests available for coagu lation are the PT, the aPIT (written also as PITK) and fibrinogen measu rement and/or thrombin time (IT) . These laboratory investigations are designed to test the integrity of both the extrinsic and intrinsic pathw ays of the classical scheme of coagulation (Fig. 4.5) and remain invaluabl e for the understanding of clot formation in vitro. The PT reflects the integrity of the extrinsic pathw ay (fac t ors V1I and X), as well as factors II (prothrombin) and V of the common pathw ay. The aPIT tests the intrinsic pathway and is prolonged by deficiencies of factors XII, XI , IX, V11I and X as well as prekaJJikrein and HMWKs. Thrombin time is a functional test of fibrinogen, by which mrombin is added to test plasma and the time to clotting is measured. It is affected by the concentration of fibrinogen and abnormalities of fibrinogen (dysfibrinogenaemia). Inhibitors of the reaction include fibrinogen- and fibrin-degradation products
(FOPs) and heparin. The clotting time and the appearance of the clot are informative. A prolonged IT is seen in hypofibrino genaemia, both congenital and acquired, as in, for example, disseminated intravascular coagulation (DIC), in which there is consumption and inhibition by FOPs and liver disease (raised FOPs) . A prolonged IT is also seen with hypoalbuminaemia. Heparin causes extreme prolongation of the IT, but a 'reptilase' time, which utilizes a thrombin-like enzyme obtained from a snake venom that is unaffected by heparin and relatively insensitive to FOPs, is normal. It is useful therefore in differen tiating contamination by heparin from fiblinogen deficiency and also in indicating if a dysfibrinogenaemia is pre-sent when it is generally more prolonged than the thrombin time. In the latter case, the nature of the clot is often abnormal , being transparent and bulky as a result of abnormal fib rin polymer ization. This can be seen in liver disease. Fibrinogen is measured in a variety of ways; some auto mated machines derive the value from the prothrombin time, which is quick and simple. However, PT-derived fibrinogen levels appear to show a false elevation in a variety of clinical settings. Conversely, if the PT is very prolonged , a fal sely low fibrinogen measurem ent can result. The Clauss fibrinogen assay is a functional assay, based on the princip al that the thrombin clotting time is inversely proportional to the fibrinogen concentration )) and is the most reliable me thod for general use in clinical laborator ies.)4 The use of diluted pl as ma and a relatively high con centration of thrombin results in little interference by FOPs or heparin, which are known to influence the technique. If a PT-derived method is used, fibrinogen levels of <2.0 gil
82 I
Haematological abn ormalitie s that ca n simulate abu se
or >6 .0 gil should be rechecked using the Clauss method. 35 The go ld standard method is measurement of clot weight but this is time-consuming and not practi ca l as a screening test. VelY lo w levels of fibrino gen «0,8 giL) wiiJ begin to prolong both the PT and t he aPTT, especially when auto mated analysers are used,
MEASUREMENTS OF PRIMARY HAEMOSTASIS In paediatrics, defects of prima ry haemostasis are more com mon than coagu]opathies. 36 Primary haemostasis is depend ent upon both number and function of platelets. Other physiolog ical variables such as temperature,37 stress,38 anaemia,39 leu copenia 40,41 and the integrity of the connective tissues and vessels also influence primary haemostasis. Many of the va ri ables that might cause abnonnal primary haemostasis such as cardiovascular or renal disease, cardiopulmonary bypass sur gery or ingestion of aspirin-like drugs, will be obvious from the history and examination of the child, or from the results of initial blood tests such as full blood count or plas ma urea level. However, platelet function and platelet-vessel wall interaction are not eluci dated by these methods. The bleeding time is an in vivo test and dependent upon both haemostatic and other physiological variables outlined above. For instance, it is prolonged in patients with the con nective tissue disease Ehlers-Danlos syndrome, generalized vasculitis and scurvy (vitamin C deficiency), patients who have a low body temperature (as can occur in prolonged sur gical procedures) and in those with leukaemia or uraemia. AJthough the bleeding time can be a useful screening test to di agnose heredita ry bleeding defects that involve platelet endothelium interaction, the test requires a highly motivated and experienced operator who und erstan ds the many vari ables influencing the result and can inform interpretation of the test. 36 The most widely used method is the Ivy bleeding time 42 using a template. The test itself is invasive; small inci sions of a standard width and depth are made in the forean}] which can cause scarring and sign ificant distress to the patient, although there are specially designed templates for children minimizing th is problem. 43-46 The norma l ranges in children of different ages vary from those in adults and older ch ildren and between bleeding devi ces. Two further methods are available to screen for platelet dysfunction: both use small volumes of whole blood, from 2 flL to 1.5 mL, a nd require special instrumentation. The first method uses flow cyto metlY and utilizes the change in expression of platelet membrane proteins as an indicator of platelet activation,47 Upon activation, P-selectin is translo cated to the platelet surface membrane due to degranulation of the a granules and there is increased expression of the GpJlb-IIIa complex. Conversely, as platelets are activated, the GpIb-IX complex is internalized and results in decreased expression. These changes in expression ca n be detected using different monoclonal antibodies both on resting
platelets and on platelets after stimulation by agonists in controlled conditions.48 Thus platelet activation and reactiv ity can be determined. The use of whole blood in the second method allows for the contribution of red and white blood cells to the haemostatic process . This method measures aggrega tion and the release reaction of pl atelets in whole blood using the PFA-lOO system 49 ,50 Using this system, whole, anticoagul ated blood is drawn under a constant vac uum into a collagen/ADP-coated or collage n/adrena line coated membrane, which has a small aperture in it. A platelet plug is formed and obstructs the fl ow through the apertu re and both the max imum velocity of flow an d closure time are recorded. The test can be performed both on resting platelets and after stimul ation by agonists in controlled con ditions. The PFA-lOO is sensitive in detecting classi ca l defects resulting in maj or platelet dysfunction, such as Glanzmann's thrombasthenia and Bernard-Soulier syn drome and also von Willebran d's disease (VWD), although a full blood count and film will demonstrate a macrothrombo cytopenia in Bernard-Soulier syndrome and abnormal val ues of VWF antigen and/or activity will be seen in VWD, if specifically tested as is recommended in investigation of NAI (see beloW). False-negative results occur with the PFA-lOO in milder platelet defects, such as storage pool disorder and release defects, which are not detectable by t he routine labor ato ry tests 50 and it is not sensitive to vascular-col lagen dis orders. Although it may be useful as a screening tool for a bleeding diathesis in children,51 its use in identifying those who have a bleeding diathesis in cases of possible NAI has not been tested. Clinicians should always perform a full range of platelet function tests when clinical susp ic ion is strong and when exclusion of a platelet defect is essenti al. If pl atelet dysfunction is suspected from the pattern of bleeding or needs to be excl uded, formal platelet function testing can be performed. This requires larger volumes of blood and arrangements need to be made with the labora tory to discuss the extent and range of tests. Screening tests would include aggregation response to the agonists ADP, adrenaline (epinephrine), ristocetin and collagen and 5-hydroxytryptamine release, indicating the secretory resp onse. If abnormal ities are fo und on screeni ng tests, fur ther testing could include t hromboxane genera tion, platelet nucleotides, t10w cytometry or electron microscopy to define t he disorder more precisely. There are, however, problems with reproducibility and the overall haemostatic condition of the patient cannot be fully assessed by this method as vessel-wall interaction is not taken into account. It ca n be difficult to determine the contributio n to bleeding made by minor abnormalities of platel et fun ction testing.
EVALUATION OF A BLEEDING PATIENT When a child presents with bruising or bleeding, the main differential diagnoses are physiological or accidental bleeding. NAI or a bleeding diathesis, The sex of the
Evaluation of a bleeding patient I
patient, age, clinical presentation, past histoty and family histoty are all important in helping distinguish these dif ferent diagnoses, directing investigations and informing interpretation of the basic screening tests. Sex is obviously important in determining the likeli hood of an X-linked disease such as haemophilia A or B. These disorders can occur in girls but, unless consanguin ity or Turner's syndrome is present, are vety rare. Extreme lyonization of X inactivation can also result in girls being affected and, although such diagnoses are unlikely, they should be tested for full evaluation. The age at presentation influences the likelihood of a particular cause of a bleeding diathesis; for instance, a neonate with a purpuric rash will have a different set of differential diagnoses from an older child. A patient with a severe congenital bleeding diathesis is unlikely to present for the first time in the adolescent years. In addition, the plasma concentration of many of the coagulation and fibrinolytic proteins is age dependent, so normal ranges of screening tests are age dependent. It is important for laboratories to establish normal ranges for age using their own reagents and methods. 52
Clinical History and Presentation The type of bleeding at presentation and the cunent history may indicate the nature of the bleeding problem. If a child has a bleeding diathesis, there is often a lack of history of trauma or the bleeding is disproportionate to the injury. For example, a boy who has bitten his tongue, which then bleeds, stops and rebleeds on a continuing basis should be investi gated for haemophilia or factor XIII deficiency if initial tests are normal. On the other hand, a child who bleeds intermit tently from the same nostril that can be easily controlled, may well have a local cause, such as a superficial vessel in Little's area. Active children, particularly as they learn to walk, often have bruises on their shins but not usually on non-exposed areas. If such bruising is associated with painful joints or reluctance to use a limb, haemophilia may be the cause. Alternatively, if the child is constitutionally unwell, the underlying disease may be malignancy, with bone mar row infiltration such as leukaemia or neuroblastoma, both of which can cause painful limbs and joints in addition to thrombocytopenia. If a child is acutely unwell and shocked with widespread petechial haemorrhages then DIC secondary to sepsis, particularly meningococcal, should be considered. Persistent mucocutaneous bleeding such as gum bleeding, epistaxis (often bilateral) or heavy menstrual bleeding indi cates a platelet disorder, either of number or function, or VWD. Drug histoty is also important, in particular whether the child is on warfarin, heparin, sodium valproate or a non steroidal anti-inflammatory drug. Past history is important in both determining whether there has been a continued bleeding problem, whether the problem is of more recent onset and whether the child has had any significant haemostatic challenges. Congenital
Figure 4.6
83
Haematoma formation 2 hours after a dose of
intramuscular vitamin K in a neonate with haemophilia.
bleeding problems such as severe haemophilia tend to pre sent early, either at birth with cephalhaematoma, intracra nial haemorrhage or after an intramuscular injection of vitamin K (Fig. 4.6) or before the first birthday as the child becomes more active and starts to crawl and toddle. Con genital thrombocytopenia presents early with bruising and petechial haemorrhages in the neonatal period, although in some conditions the initial platelet count is preserved but deteriorates with time. Significant haemostatic challenges include surgery, for example circumcision, tonsillectomy or removal of teeth. Bleeding response to injuty such as biting the tongue, sustaining a fracture or involvement in a road traffic accident can yield important information. In girls, increased menstrual loss may be an indicator of a bleeding diathesis. Concomitant disease such as hepatocellular dys function, renal disease, malabsorption or a connective tis sue disease such as Ehlers-Danlos syndrome may result in abnormal bleeding.] ·53.54 Most of the clotting proteins are synthesized by hepatocytes, and the vitamin K-dependent factors II (prothrombin), V1I, IX and X) are the most sensi tive to liver dysfunction ; however, with increasing damage other factors such as factor V and fibrinogen are affected. The vitamin K-dependent factors may also become signifi cantly depleted in malabsorption syndromes. Renal failure and associated uraemia can contribute to a bleeding diathesis as certain accumulating metabolites interfere with platelet function and accompanying anaemia results in the loss of red cells transporting centrally flowing platelets to the vessel wall, promoting adhesion to the subendothelium. 55 Paradoxically, nephrotic syndrome is associated with venous thrombosis due to an imbalance of the haemostatic system. Acute phase reactions or other mechanisms cause high VWF and fibrinogen in conjunc tion with low levels of antithrombin, an endogenous anti coagulant, secondaty to urinaty loss. Cyanotic congenital
84 I
Haematological abnormalities that can simulate abuse
heart disease can lead to thrombocytopenia, secondary to shortened platelet survival, and hypofibrinogenaemia, sec ondary to poor liver function, with reduced synthesis and clearance of clotting factor intermediates leading to low grade disseminated intravascular dissemination. Family history is clearly essential to detennine the like lihood of an inherited disease. Children with X-linked dis ease such as haemophilia may give a family history of affected males carried through the female line. However, the family history is not always helpful as about 30 per cent of those with haemophilia A are due to spontaneous mutations. 56 VWD is an autosomal dominant condition with variable penetrance, which often gives a positive fam ily history of bleeding but sometimes appears to skip a generation owing to VWD modifier genes that have no association with the VWF gene. 57 Autosomal recessive conditions, such as factor XIII deficiency, or conditions in which the heterozygote may only have a mild or no bleed ing tendency such as factor XI deficiency, usually do not have a positive family history unless there is consanguin ity within the family. For example, factor XI deficiency is found most frequently in those of Ashkenazi Jewish descent (in whom the carrier rate is 8 per cent) and who tend to many within a small community, resulting in a restricted gene pool. 58
Examination of the Child The general health and state of the child should be assessed. In a bruised child, particular points to look for are the distri bution, size and age of the bruises; whether they are in any recognizable pattern such as a hand or belt mark; and whether there is a haematoma indicating the extent of bleed ing. It should be noted that fingertip bruising is not infre quently found in children with a bleeding diathesis and is not pathognomonic ofNAl (Fig. 4.7). Distribution of bruising can be the key to the diagnosis of Henoch-Schonlein purpura, which is present on extensor surfaces and is due to a vasculi tis rather than a coagulopathy. Coagulation screening and a full blood count will be nonnal and thus confusion with NAl may arise (Figs 4.8 and 4.9).16.59.60 Presence or absence of petechial haemorrhages will help to differentiate disorders associated with thrombocytopenia. Such haemorrhages also occur in situations when there is raised intrathoracic pressure in the absence of a bleeding diathesis, such as a severe bout of coughing or vomiting or in strangulation. In these situ ations, the petechiae will be seen in the distribution of the superior vena cava area or around the eyes or mouth 61 (Fig. 4.10). A swollen, tender joint may indicate bleeding into the joint, as is seen in haemophilia or, much more rarely, VWD, but tender joints may also be seen in Henoch-Schonlein pur pura, acute leukaemia or neuroblastoma. Neuroblastoma may present with bilateral black eyes as a result of tumour infiltra tion; comparatively minor injury can cause the same in haemophilia (Figs 4.11 and 4.12). Systemic upset and
Figure 4.7 Fingertip bruising in a child with Glanzmann's thrombasthenia.
Figure 4.8 Symmetrical bruising on extensor surfaces seen in Henoch-Schbnlein purpura.
enlarged lymph nodes, liver or spleen usually accompany bruising and petechial haemorrhages secondary to acute leukaemia . A limp, with or without bruising, can also be seen at presentation in this disease. A rare disease, haemophago cytic Iymphohistiocytosis (HlH) presents with systemic signs and symptoms, such as pancytopenia, coagulopathy and organomegaly, but may also present with central nervous
Evaluation of a bleeding patient I
Figure 4.9
85
Buttock bruising in Henoch-Schonlein purpura . Figure 4.11 Bilateral black eyes after minor trauma in a chi ld with haemophilia.
Fig ure 4.12 Black eye and mild proptosis in a child with neuroblastoma. (Courtesy of the late Dr J. Pritchard, Edinburgh.)
Figure 4.10 Subconjunctival haemorrhage secondary to coughing in a child with no haemostatic abnormality.
Haematological Investigation system (eNS) manifestations. These can range from irritabil ity to encephalopathy and coma. Retinal and intracranial haemorrhages can be found in these circumstances and this condition has been mistaken for abuse. 62 Full examination and laboratory evaluation, however, will indicate the correct diagnosis. Disseminated intravascular coagulation with secondary bleeding is generally seen in an ill child with signs and symptoms of infection.
-
- -- - - - - -
FIRST-LINE INVESTIGATIONS
Once history-taking and examination have been done, inves tigations can proceed. The way blood is taken and processed is of paramount importance, as activation of the coagulation system by a difficult venepuncture and contamination with tissue fluids or by the presence of air bubbles will invalidate the result. Shortening of the PT or aPIT through activation of a specimen may mask a significant bleeding disorder such as
86 I
Haematological abnormalities that can simulate abuse
Table 4.1
Common pitfalls in specimen collection and processing
, Problem
Process
Result
Poor ve nepuncture technique
Activation of sample
Specimen taken from heparinized line Over- or underfilling specimen tube Severe polycythaemia (reduced plasma vo lume)
Heparin contamination Ratio of citrate-plasma not 1:9 Ratio of citrate-plasma> 1:9 Activation of sample Loss of factor activity
Prolongation or shortening of clotting times Thrombocytopenia Prolongation of apn
Inappropriate storage/transport
haemophilia. Conversely, if blood is taken through a cannula that has been kept patent with heparin then contamination frequently results, giving abnormal values for many of the coagulation tests. Venepuncture of children should be done by an experienced operator wherever possible, at a time when the blood can be rapidly dealt with by the laboratOly. Blood should be taken with the minimum of venous stasis and sub sequent handling. Spinning, freezing and thawing blood can cause al1efact; for example, factor XI levels may rise.6) Trans port to a laboratory some distance away, even if the specimen is transported on ice, can also cause artefact, usually a lower ing of clotting protein values. Some common pitfalls in speci men collection and processing are shown in Table 4. I. Initial tests include a coagulation screen, PT, aPIT, IT and fibrinogen as outlined above and a full blood count and film. A fa ctor VlU, factor IX level and VWF antigen and activity are recom mended in all cases of suspected NAI, as a normal or margin ally prolonged aPIT can be associated with a significant decrease in factor VIII or IX levels.64 Factor XIll may also be measured in neonates with intracranial haemonhage. If the blood is flowing well , a few millilitres extra can save a second venepuncture if an abnormality is found, for example a pro longed aPTT, and further testing is required. Bleeding time is an invasive test and although it will demonstrate the integrity of the platelet-vessel wall interaction, is usually unnecessary in the early stages of investigation. It is highly operator dependent. One of the newer tests (flow cytometry or the PFA 100) may be helpful in the future but there is no published work in this clinical setting and false-negative results may be misleading. It is sometimes helpful to investigate parents, especially if results are equivocal or subsequent testing for clarifica tion of a ny abnormality requires large volumes of blood. Identification of a child's natural parents is not always cer tain, or they may not be available or readily located, mak ing this approach impossible.
PATIERNS OF ABNORMAL RESULTS The pattern of abnormalities obtained using first-line tests along with the clinical presentation may indica te an und er lying defect/disorder. The extended testing discussed
Sho rtened or prolonged clotting times Prolonged clotting times Shortened or prolonged clotting times Prolonged clotting times
above, which includes VWF antigen and activity, and levels of factors VIII and IX , will add further information a nd minimize del ay. The results of the tests will inform the choice of fu11her tests to determine the precise diag nosis.
Isolated and Prolonged aPTI (PT Normal, TI Normal, Fibrinogen Normal, Platelets Normal) Prolongation of the aPIT is probably the commonest abnor mality found on pelforming the coagulation screen. This may be by only a second or two but can signal a significant bleed ing diathesis. Referring to the classical scheme of coagulation (Fig. 4.5, p. 81), which is a good model for in vitro coagul a tion, this indicates a defect in the intrinsic pathway. This pattern is fo und in: 1. con genital deficiency of factors VIII, IX, XI and XII, as well as prekallikrein and HMWK ; 2. VWD, as this may result in low levels of factor VIII; 3. circulating inhibitors, e.g. lupus anticoagulant; 4. mild deficiencies of factors II (prothrombin) , V and X may prolong the aPIT, whereas the PT rema ins norm al (reagent dep endent); 5. heparin, either as a contaminant or a therapeutic agent, is a common cause of a prolonged aPIT; the IT is very sensitive to heparin and will be prolonged; a reptilase time will be normal. Thus the prolongation of the aPIT is due either to a clotting factor deficiency or an inhibitor of coagulation. To differentiate between the two, a 50:50 mix of test plasma with normal plasma can be performed . If there is a deficiency there should be a correction of the clotting time to within a few seconds of normal and at least less tha n 50 per cent of the difference between the control and test plasmas. If there is no significant correction then the presence of an inhibitor is suspected. This might be heparin contamination or therapy, an acquired inhibitor of a clotting factor (velY rare, especially in children) or the lupus anticoagulant (Fig. 4.13). Therapeutic heparin administration or heparin contamin ation of the specimen can be ruled out from the history, by checking that tlle specimen was not taken from a heparinized
Patterns of abnormal results I
87
PT, aPIT, Fibrinogen, IT
... Isolated
.
I
PT
Isolated
.. I
aPIT
t Unlikely
t
t
'+
Correction
.
i
Likely
t ?
Unlikely
(I IT.
t
t
'+ Unlikely
t
t
K warfarin liver disease
111, V, X lv+vlll LA + 111
I PT, I aPn, 1 fibrinogen
Normal (including platelet count)
t
i
DIC
Severe liver
disease Afibrinogenaemia Dysfibrino genaemia
Haematological
t
Glanzmann's thrombasthenia Platelet SPD 1 XIII
1 cx,antiplasmin
1 PAI-l
Tissuelvascular
t HSP 1 vitamin C Ehler's-Danlos
? LA'
reagent
dependent:
111, V, X
HM\AlK' PK'
hw
i
aPn
t
1 XI
v\AlD 1 VIII llx
heparin Rep time N)
I
Likely
1 vit
No correction
'+
PT,
.
i
..
due to: Cong 1 VII Reagent early warfarin dependent: therapy LA' early 1 vit K early liver mild 111, V, X disease
Figure 4.13
I
50:50 plasma mix
Likely
1 VII
t
...
(I DRVVT)
t ?
factor VIII, IX or
XI inhibitor (DRVVT N)
Algorithm for initial investigation and interpretation of results in a child with a possible bleeding diathesis.
II, factor II; V, factor V; VII, factor VII; VIII, factor VIII; IX, factor IX; XI, factor XI; XII, factor XII; XIII, factor XIII; V+VIII, combined factor Vand VIII deficiency; HMWK, high molecular weight kinonogen; PK, prekallikreth; cong, congenital; LA, lupus anticoagulant; HSP, Henoch- Schbnlein purpura; SPD, storage pool disorder; DIC, disseminated intravascular coagu lation; vWD, von Willebrand Disease; PAI-l, plaminogen activator inhibitor-l ; H thrombin time; Rep, Reptilase; DRWT, dilute Russell viper venom time. 'Not associated with bleeding Table 4.2
Prolonged aPTT:differentiation of an inhibitor from a deficiency
apn
50:50 mix
DRVVT
Platelet neutralization
Reptilase time
PT
t t
Correction No correction n/a
N N N
N N N
n/a n/a No correction No correction
N N N N
N N
Inhibitor
Lupus anticoagulant Heparin Factor VIII and IX antibody
t t t
t t t
t t
Correction Correction Correction Correction
N
Deficiency
Factors VIII, IX, XI and XII VWF Factor II Factors V, X
or N sl t or N
t
N N
t t
'May be normal if deficiency mild.
aPH partial thromboplastin time; DRWT, dilute Russell's viper venom time; n/a, not applicable; N, normal; PT, prothrombin time; Willebrand Factor.
Note: Specific patterns of bleeding may be seen with specific diagnoses (Table 4.3).
line or put into a bottle containing lithium heparin before being transferred to a citra ted one. It is also possible that heparin may have been given inadvertently or with intent to cause bleeding.55 If there is uncertainty, a reptilase time can be performed, which, by activating fibrinogen directly, is nor mal in the presence of heparin. If the pattern of correction is one of an inhibitor, and heparin has been excluded, the most likely cause is a lupus anticoagulant (Table 4.2). Lupus anticoagulant is a phospholipid antibody that interferes with the phospholipid added to the in uitra test ing model. It affects the aPTT rather than the PT, as the phospholipid in the former test is more dilute and usually more sensitive than that used in the PI. FUliher tests, such
t· t <
51,
slightly; VWF, von
as the dilute Russell's viper venom time (DRVVT) can dif ferentiate this from other factor inhibitors. Russell's viper venom causes clotting by direct activation of factor X. This is inhibited by the lupus anticoagulant, but is not affected by deficiencies or inhibitors of factor VIII or factor IX. A platelet correction test can then be performed, whereby the DRVVT is repeated but using platelet-associated phospho lipid rather than thromboplastin. Platelet-associated phos pholipid is protected from the action of the antibody and the test revelis to normal with a lupus anticoagulant. Defi ciencies of factors X, V and II will also prolong the DRVVT but these will usually give a prolonged PT and will not cor rect with platelet-associated phospholipid (Table 4.2).
88 I
Haematological abnormalities that can simulate abuse
If the pattern of correction is one of a deficiency FVIII and FIX should be assay ed if not already done. If these fac tors are normal then FXI and XII should be assayed. If the only defect is a deficiency of FXII then this is not associ ated with a bleeding diathesis and will not account for any haemorrhagic symptoms or signs. 66 Prekallikrein a nd HMWK deficiency, which are rare, will also prolong the aPTI but, again, are not associated with a bleeding diath esis. Von Willebrand factor antigen and activity shou ld also be checked. Factor VIII is non-covale ntly bound to VWF, which protects it from proteolysis in the plasma. Deficien cies of VWF may therefore be associated with reduced levels of factor VIIl, such that the aPTT is prolonged; however, this is not necessarily seen (Table 4.2). The pattern of bleeding that is identified on clinical examination may predict the most likely cause of the abnormal test result (Table 4.3). Mucocutaneous bleeding and/or ecchymoses are a common presentation of VWD. Factor Xl deficiency also presents with this pattern of bleed ing but is very rare excep t within certain population groups such as Ashkenazi Jews, for whom the gene fre quency is around 8 per cent 67 Factor VIII and factor IX deficiencies do not necessarily present with a history of painful or sw oll en joints and, although spontaneous muco cutaneous bleeding is rare, bruising is common. As well as ecchymoses with or without joint swelling and bleed ing, factor VIII or factor IX deficiencies can present with inter mit te nt but persistent bleeding from mucocuta neous injury, such as a bitten tongue or lip . Mucocutaneous bleed ing and/or ecchymoses in a child whose aPTI does not correct, may be secondary to an acquired inhibitor, such as a factor VIII inhibitor, which
Table 4.3
can result in serious bleeding, or the lupus anticoagulant, which is likely to be an incidental finding and unconnected with the bleeding and bnlising. The extra blood taken at ini tial venepuncture ca n be used to carry out further test in g such as the DRVVT with platelet correction as described above. If the pattern is not that of a lupus anticoagulant, a specific inhibitor such as to factor VIII or very rarely factor XI should be considered. In these circumstances, when the plasma is serially diluted, the clotting times shorten , rather than lengthen , as the inhibitor is diluted out. The aeti ology of factor VIII inhibitors in children is uncertain but can be associated with malig nancy or possibly recent use of high dose penicilli n 68 .G9 Factor XI inhibitors are very rare and have been described foll owing viral infections, resolving within a few weeks 70 ,71 There are only a handful of reported cases of acquired coagulation factor inhibitors in children.
Isolated and Prolonged PT (aPTT Normal. TT Normal. Fibrinogen Normal. Platelets Normal) Referring again to the classical scheme of coagulation (Fig. 4.5, p. 81), this indicates a defect in the extrinsic pathway, that is factor VII. This pattern is found in: 1. factor VII deficiency, either congen ital or secondary to liver disease or vitamin K deficiency; 2. early warfarin therapy; 3. mild deficienci es of factors 11 (prothrombin), V and X may prolong the PT, wheras the aPTI remains within the normal range, depending on the reagents used;
Patterns of coagulation results and patterns of bleeding: possible diagnoses
Coagulation results
Prolonged aPTT PT/fibrinogen/platelets normal Prolonged PT aPTT/fibrinogen/plate lets normal Prolonged PT and aPTT Fibrinogen/platelets normal
Prolonged PT and aPTT Decreased fibrinogen Normal or low platelets
Mucocutaneous bleeding
Ecchymoses with or without
and ecchymoses
joint bleeding
WJD' Factor XI deficiency Fa ctor VIII/IX inhibitor Warfarin ingestion Early vitamin K deficiency Early Iiver dysfunction Over warfarinization Severe vitamin K deficiency Over heparinization Factor X, Factor V or prothromb in deficiency Acquired inhibitors Severe liver dysfunction Dysfib ri nag ena emia/afibri nogena em ia Die (including meningococcal sepsis)
Factor VIII deficiency Factor IX deficiency VWD (2N) Factor VII deficiency
Fa ctor X deficiency Factor V deficiency Prothrombin deficiency
Dysfibrinogenaem ia Afibrinogenaemia
'VI/VD subtype 28 is associated with reduced platelets.
APTI, partial thromboplastin time; Ole, disseminated intravascular coagulation; PT, prothrombin time; VWD, von Willebrand's disease;
2N, subtype 2N.
Patterns of abnormal results I
4. lupus anticoagulant (rarely); it can be more sensitive to reagents used in the PT than those used in the aPTT; a mixing test will not correct. Both warfarin and vitamin K deficiency can cause this picture. Although they affect the carboxylation of all the vitamin K-dependent coagulation factors (factors ll, VII, IX and X), factor VII has the shortest half-life and is the most sensitive to failure of that step. However, with increasing dosage of warfarin or severity of vitamin K deficiency, the aPTT wiH become prolonged, due to the involvement of other factors (Figure 4.13). Mucocutaneous bleeding and ecchymoses in association with a prolonged PT suggest warfarin ingestion, early liver dysfunction or early vitamin K defiCiency (Tabl e 4.3)' although none is necessarily associated with bleeding. Vita min K deficiency is a recognized hazard in the neonate, especially in those who are excl usively breast fed, and is corrected by parenteral vitami n K at birth or repeated oral vitamin K at birth and the early neonatal period. Malab sorption syndromes are also associated with vitamin K deficiency. More severe deficiency ho wever, results in a prolongation of both the PT and aPTT (see beloW). A trial dose of oral or intravenous vitamin K can be given, which shou ld begin to correct abnormalities within 4-12 hours if deficiency is the cause. Prothrombin defici ency is rare but usually presents with mild mucocutaneous bleeding or with post-traumatic haemorrhage. The aPTT may be slightly prolonged along with the prolonged PT, but fibrin ogen is normal. Dyspro thrombinaemia has also been described with a similar pre sentation and prolonged PT.72 Ecchymoses and bleeding into joints is seen with con genital factor VII deficiency, and thus a factor VII assay can be performed once warfarin ingestion and vitamin K deficiency have been reasonably excluded and liver func tion checked. A trial of vitamin K can be given as above.
Prolonged PT and apn (n Normal, Fibrinogen Normal, Platelets Normal) This pattern is found in: 1. vitamin K deficiency, when the PT is usually more prolonged tha n the aPTT; 2. oral anticoagulants, when, again, the PT is usually more prolonged than the aPTT; 3. liver disease, giving rise to multiple factor deficiencies; fibrinogen may be abnormal in severe disease ; 4. deficiencies of factors II, V, X or combined V and VIII; 5. lupus anticoagu lant, with acquired prothrombin deficiency. For both of these screening tests to be prolonged either there are multiple defects in the coagulation pathway or there is a deficiency in the final common pathway of the
89
classical scheme (Fig. 4.5, p. 81). Inhibitors can also pro duce this pattern; in paliicular, over-heparinization can lead to a markedly long aPTT, but with only a minor pro longation of the PI. Mixing experiments using the PT may be useful if there is no history 0 f anticoagulant use or obvi ous cause of vitamin K deficiency, for example parenteral nutrition or prolonged antibiotic use. Inhibi tors of coagu lation factors may be specific, for example to factor V or factor II (prothrombin). or may affect more than one factor. However, they all have the characteristic that as the plasma is serially diluted, the clotting times shorten rather than lengthen as would be seen in a simple deficiency. Mucocutaneous bleeding and ecchymoses are seen in severe vitamin K deficiency, over-warfarinization and over-heparinization. Severe liver dysfunction is associated with a reduced fibrinogen level. Acquired inhibitors of coagulation can cause very severe bruising, muscle bleed ing, and gastrointestinal or urogenital bleeding, but are very rare. Factor X, factor V and prothrombin deficiencies, which are also rare, can all present with mucocutaneous bleeding and bruising of varying severity. Factor X defi ciency can also present with haemalihroses, as can factor V and prothrombin deficiency (less commonly). Specific factor assays will need to be performed to differentiate between them. This pattern of coagul ation tests can be seen in associ ation with the lupus anticoagulant and acquired prothrom bin deficiency. However, sponta neous bleeding is rare and mixing tests are indicative of lupus anticoagul ant.
Prolonged PT and apn, Long n with Low Fibrinogen (Platelets Normal) This pattern is found with: 1. high level s of heparin (with lower levels, TT is long but fibrinogen is normal); 2. hypo- and afibrinogenaemia; dysfibrinogenaemia; 3. systemic hyperfibrinolysis; 4. some cases of liver disease. When the fibrinogen level falls below 0.8 giL (measured by a functional assay), as well as a prolonged TT, the PT, and to a lesser extent the aPTT, also become prolonged. This occurs with hypo- and afibrinogenaemi a and with the majority of dysfibrinogenaemias. The reptilase time, which is based on clotting induced by release of the fibrinopeptide A fragmen t from fibrinogen, is also prolonged, especially with the dysfibrinogena emias. The reptilase time is not affected by heparin and so is useful in determining whether there is contamination by this anticoagulant or not. The platelet count is normal. Bleeding in those with afibrinogenaemia and fibrinogen levels of less than abo ut 0.5 giL is lifelong and varies in severity between patients. Ecchymoses, muco cutaneous haemorrhage and haemarthroses have all been described. Patients with dysfibrinogenaemi a are usu ally
90 I
Haematological abnormalities that can simu late abuse
asymptomatic with bleeding (when it do es occur) often limited to epistaxis, menorrhagia and mild-to-moderate post traumatic bleeding, including surgely.
Prolonged PT and aPIT, Long IT with Low Fibrinogen and Platelets
PAl-l and Ctrantipl asmin assays or platelet function studies may be necessa ry. Abnormalities of collagen vascular dis eases and vascular integrity such as Henoch-Schonl ein pur pura will also have normal investigations (see Chapter 4, p.94).
Von Willebrand Disease This pattern is found in: 1. disseminated intravascu lar coagulation ; 2. so me cases of severe liver disease. Children with DIe secondary for instance to infection , such as meningococcal septicaemia, are usually clinically sick, with evidence of in fection in addition to ecchymoses and mucocutaneous haemorrhage. Th e coagulation scree n will be deranged bu t the platelets will be low. Sequential measurements of coagulation and full blood count may show a continuing fall in both platelet count a nd fibrino gen concentration. Evidence of increased fibrin degrad ation will support the diagnosis a nd meas urem ent of the D-dimer level is probably th e most reliable. D-Dimers a re formed followin g the di gestion of cross-linked fibrin by plasmi n and are raised in DIe. Fibrinogen degradation produ cts (FDPs) can also be measured but are not specific to cross-linked fibrin and can be raised in other conditions. Severe liver dis eas e may also give a pattern of prolonged PT, aPTT with a low fibrinogen and lo w pl atelets secondary to hype rsplenism, a lthough these can be normal. Clinical exa mination will di fferentiate the problem as well as add itional tests of liver fun ction .
Prolonged PT and aPIT, Platelets Low (IT I\lormal, Fibrinogen Normal) This pattern is fo und in: 1. mass ive transfu s ion ; 2. some cases of chronic liver disease, for exampl e cirrhosis. The clinical findings and liver function tests will clarify the cause.
NORMAL COAGULATION SCREEN WITH A NORMAL PLATELET COUNT Several conditions need to be considered ifthere is Significant bleeding and screenin g investigations are nom1al. The com monest is VvVD but factor XIII deficiency, platelet storage pool disorder and Glanzmann's thrombasthenia can produce these results, although very rarely. Disorders of fibrinolysis such as cxra ntiplasm in deficiency and plasminogen activator inhibitor-l (PAl- 1) defiCiency may also predispose to bleed ing. In these rare instances, clot solubility assays, factor XIII,
Von Willebrand disease is a bleeding dis order secondary to either a quantitative or qualitative abnormality of VVI/F. Diagnosis hin ges on demonstrating these defects. Von Willebrand factor is a multimeric, high-molecular-weight glycoprotein that is synthesized both by endoth elial cells an d megakaryocytes, and plays an impol1ant paI1 in pri mary hae mostasis. It is essential for platelet adhesion to the vessel wall, which, having been secreted from endothel ial cells, it initiates by bindi ng to the GpIb-IX receptor. This ex poses the GpIlb-llla receptor complex to wh ich it binds a lon g with fibrinogen, facilitating platelet aggregation , es pecially in high-shear conditions. It acts as a carrier pro tein for circulating FVI llc to wh ich it is non-covalently bound, protecting it from proteolysis by protein C and thu s its rapid clearance from the plasma. Decreased levels of VWF or reduced binding of V\I\fF to factor VIlle are accom panied by correspondingly low levels of circulating factor VIlIc. It can be seen therefore that both qu antitative and qualitative defe cts of VWF may affect both primary haemostasis (resulting in mucocutaneous bleeding) and secondary hae mostasis (resulting in haema to ma formation a nd , more rarely, hae marthroses). Several types and sub types of VWD have been described; however, the id entifi cation of the latter requires specialized testing. Di agnosis of the specific subtype, although importan t for correct treatment, is not necessary for the purposes of identifying if a chUd has a propensity to bl eed. The importan ce of screening for VWD lies in the fa ct that, excluding type 3 VWD, it is the commonest inherited bleeding disorder, with a prevalence of between 0.1 per cent and 2 per cent 73 75 and that the coagulation screen and full blood count can be normal. Levels of VWF vary with environmental factors, such as stress and exercise. To minimize the risk of misdiag nosis, VWF antigen and function must be measured in samples obtained on at least two occasions with consistent resu lts. 76.77 However, from a pragmatic point of view, ini tial tests will guide the clinician in early management but should be repeated at a later stage w here possible. Type 1 V\I\fD is the commonest form of the disease (80 per cent of cases) an d is cha racteri zed by a quantitative defect. It is inherited as an autosomal dominant but with marked variability of both phenotypic penetrance an d exp ressivity. This may be because th ere are VWD modifier genes unassociated with the VWF locus as well as the VWF gene S7 .78 A bleeding diathesis secon dary to V\I\fD can usu ally be id entified from evidence of a fa mily history of bleeding, a clinical histOIY and la boratory demonstratio n of VWF
Normal coagulation screen with a normal platelet count I
deficiency. However, without documentation of all three of these features, a diagnosis of type 1 VWD can be more dif ficult. Type 2 VWD is inherited as an autosomal dominant condition when there is a qualitative defect in the VWF. For both type I and type 2 diseases, the clinical picture is one of mucocutan eous bleeding, most commonly epistaxis, gum bleeding and bleeding from superficial cuts and wounds. Petechi ae are rarely a feature, with none reported in a study of 1257 patients 79 ,Bo but in 11.5 per cent of patie nts reported in a study by Silwer. BI The haemorrhagic tendency is velY vatiab le and depends On the type and severity of the disease. In many patients with type 1 or 2 disease there may be no history of bleeding and caution should be exercised when attributing bruising in a child with suspected NAI automatically to VWD on the basis of laboratory testing.? VWD may induce increased bleeding wi th trauma, but not precipitate spontaneous bleeding. Diagnosis of a bleeding diathesis does not exclude NAl. and the histolY and clinical findings remain important contributors to determining the cause of bleeding7 The results of the screening tests usually reveal a normal platelet count, although mild thrombocytopenia may be found in subtype 2B or platelet-type-pseudo VWD (see beloW). The PT is normal, whereas the partial thromboplas tin time may be mildly prolonged and dependent upon the plasma level of factor VIllc (Table 4.5). In mild type 1 dis ease, t he bleeding time may be normal or prolonged and thus may not be helpful in the diagnosis and should be avoided,n However, an in vitro bleeding time using the platelet function analyser PFA-lOO shows good sensitivity to the defect in primary haemostasis that occurs in VWD and may be helpful if the device is available. 50 ,5 1 Tests used for the primary diagnosis of VWD include a factor VIlIc level and VWF antigen (VWF:Ag). Factor VIllc half-life is regulated by VWF and is frequently reduced in VWD, although factor VJlIc levels are not necessarily reduced. 77
Table 4.4
VWF:Ag is decreased in type 1 VWO and decreased or nor mal in type 2. More than 80 per cent of all patients with VWD will have an abnormal v'vVF:Ag result BO but, if done alone, this will miss a proportion of patients with normal antige n levels but abnorm al function, i.e. type 2 VWD, In addition, in some cases the levels will be borderline and may require repeat testing. 77 In those patients for whom the diagnosis of VWD is strongly suspected or needs to be excluded, results of several different tests shou ld be analysed to y ield the maximum information from which to draw conclus ions. Further tests include VVllF function, measured using a platelet-based ristocetin cofactor (RiCo F) and VWF collagen binding assays. These assays improve the abi lity to detect type 2 variants and more clearly define type 1 VWD. Other assays including agarose gel elec trophoresis of the VWF multimer pattern, further platelet agglutination studies and VWF-factor VIllc binding assays are required to subclassify VWD but are not always neces salY to establish the basic diagnosis of VW0 77 (Table 4.5). An important point to note is that VWF levels vary accord ing to blood gro up and also in crease as part of the acute phas e response. The mean plasma VWF:Ag level is about 30 per cent lower for those with blood group 0 than those with blood group A, and 39 per cent lower for those of blood group AB.B2 Whether the ABO group has any effect on the specific activity ofVWF is unresolved. Variation due to the acute phase response means that those with VWD may have intermittently normal levels and , therefore, a singl e normal level does not exclude the diagnosis. There is no evidence that venepuncture alone results in raised lev els of VWF, but it is very likely that venepuncture per formed under stressful circumstances can result in an increase in factor VIIIc and VWF. 77 Unfortunately, children are often stressed when blood is taken and, if this occurs, results shou ld be interpreted with this in mind and the fact recorded in the notes, As mentioned above, VWF:Ag and
Causes of bleeding in a well child Coagulation screen and platelets Normal
Common
Abnormal
VWD
ITP
Henoch-Schiinlein purpura
Ha emophi lia A or B Vitamin K deficiency Wa rfarin or heparin
Uncommon
91
Glanzmann's thromba sthe nia
Congenital platelet abnormality
Platele t storage pool disorder
Deficiencies of factors II, V, VII, X, XI
(excluding Glanzmann's thrombasthenia) Fa ctor XIII deficiency
Dysfibri nogenaem ia
PAI-l deficiency
Afibri nogenaem ia
a:2-ant iplasmin deficien cy ITP, idiopathic th rombocytopenic purpura: PAI-l, plasminogen activator inhibitor-l ; VWD, von Wi llebrand's disease.
92 I
Haematological abnormalities that can simulate abuse
Table 4.5
Laboratory testing of von Willebrand's disease
VWD
apn
Type 1 Type 2A Type 2B
Nor Nor N or
Type 2N
l'
Type 2M Type 3
VIII:c
l' l' l'
N or N or N or
t t t
Platelets
VWF:Ag
N N
N or
t t
RiCoF
Further tests
t t
Multimers normal Multimers abnormal Multimers abnormal Platelet aggregation with low-dose ristocetin Multimers normal Abnormal factor VIII binding Multimers normal Virtually absent multimers
t
t
N or
t
N
N
N
N
N
N
l' l'
t t
N N
t t 1
t t
t
N, normal; RiCoF, ristocetin cofactor; VWD, von Willebrand's disease; VWF:Ag, von Willebrand factor antigen.
function must be measured in samples obtained on at least two occasions. One subtype, type 2B, is characterized by an increased affinity of the abnormal VWF for the platelet. This often results in a mildly reduced platelet count with large forms seen on the blood film 83 (Table 4.5). The platelet counterpart to this can be seen in platelet-type-pseudo VWD, in which the platelet Gplb-IX complex has an increased affinity for normal VWF and produces the same clinical and laboratory picture B4 Differentiation between the two can be made by mixing experiments. A variant of VWD, type 2N, can also present in this way with a prolonged aPTT on testing. Factor VIII is unable to bind to the abnormal VWF and factor VIII levels are reduced (Table 4.5). The condition can be confused with mild haemophilia A, particularly as VWF levels and activity can be normal. However, the differentiation between mild haemophilia and VWD type 2N is not necessary for the immediate management of the child . It is inherited as an autosomal dominant and can be confirmed at a later stage once family history and investigation has taken place. A fur ther variant, type 2M, has a type 2 pattern of response, i.e. normal VvVF:Ag and reduced activity but, unlike the other type 2 variants, it has a normal multimer pattern (Table 4.5). Type 3 VWD has an autosomal recessive inheritance and is very rare. It results in a severe bleeding disorder with markedly decreased or undetectable VWF/Ag and activity and has factor VIII levels in the region of 0.01-0.05 IU/mL. Acquired von Willebrand syndrome has been described, although very rare, and has similar laboratory findings to congenital disease. 8 5 The severity of the bleeding tendency varies from mild to severe and has been found in associ ation with several disease states including systemic lupus erythematosus 86 and congenital cardiac defects.8o
secret ion mechanism. They are often associated with rela tively mild bleeding states and diagnosis may be delayed until later in childhood, or even in adult life. Common pre sentations include easy bruising and possibly prolonged bleeding from superficial cuts, epistaxis and menorrhagia. Screening tests are normal, therefore specific platelet func tion testing, including platelet aggregation and secretion, are required to confirm the diagnosis. Again, however, the diag nosis of platelet storage pool disorder does not exclude NAI, and the history and clinical findings are very important. Specific syndromes such as Grey platelet, Hermansky Pudlak, Chediak-Higashi and Wiskott-AIdrich syndromes, which are associated with other defects, and specific blood film appearances in addition to the st orage pool disorder are described under separate headings. Deficiencies can occur in either the 0. granules (Grey platelet syndrome) or the 6 granules or both. Abnormalities of platelet aggregation may be detected but can be normal in milder cases. The definitive diagnosis depends on the demon stration of the deficiency either biochemically (using tests of platelet secretion), by electron microscopy or both. Failure of the secretion mechanism occurs in aspirin ingestion and cyclooxygenase deficiency. Defective aggre gation and secretion is seen to specific agonists. Most patients with congenital cyclooxygenase deficiency do not have an accompanying bleeding diathesis, perhaps because the defect is balanced by impairment of prostaglandin gen eration in the vascular endothelial celIs.87 Defects have been shown in platelet testing t o weak agonists only and described as weak agonist response defect (WARD) . The clinical significance of this is very uncertain especially since defective responses to weak agon ists such as epinephrine are found in normal platelets and results should be interpreted with caution.
Platelet Storage Pool Disorders Factor XIII Deficiency These are a collection of disorders that are characterized by failure of secretion of the contents of the platelet granules on stimulation. This may be due to either absence of the gran ule contents (storage pool deficiency) or to failure of the
Factor XIII is responsible for the stabilization of the fibrin clot by polymerization 88 and also inhibition of fibrino lysis by the binding of <:X2-plasmin inhibitor to the fibrin. 89
Normal coagulation screen with a normal platel et count I
Deficiency of factor XIII results in reduced clot stability and is associated with delayed haemorrhage and poor heal ing. A classical presentation is at birth, with delayed sepa ration of the umbilica l stump, sometimes taking over 4 weeks to separate, and continued bleedin g from the site. 90 There is delayed and repeated bleeding from superficial wounds when a clot is seen to form normally at the site of the wound only to break down 24 hours later with resumed bleeding. Intracranial haemorrhage is a considerable risk, often resulting in death. 91 To screen for factor XIII defi ciency, solubility of the clot is tested in urea or monochlo racetic acid. If soluble, further immunochemical tests can confirm and quantitate the deficiency.
Antiplasmin Deficiency The enzyme plasmin digests fibrin clot as healing takes place and is regulated by Ctrantiplasmin. Defici ency of this enzyme results in a severe bleeding disorder that can be inherited. 92 ,93 Patients present with symptoms similar to factor XIII deficiency, with delayed bleeding after trauma and mucocutaneous bleeding as well as joint haemor rhages. Coagulation screening tests are norm a l but tests measuring fibrinolytic activity, such as the euglobulin lysis time, are usually abnormally shortened and C\:2-antiplasmin levels are reduced. Figure 4.14
Multiple bruises on the trunk and arms of a child
with Glanzmann's thrombasthenia.
Plasminogen Activator Inhibitor-l Deficiency Deficiency of PAI-l can cause a hereditary bleedin g dis order, with haemorrhage occurring most usually after sur gery or trauma. 94 - 96 The coagulation screening tests are normal with a shortened euglobulin lysis time. Levels of PAI-I antigen and activity are reduced.
Glanzmann's Thrombasthenia In Glanzmann's thrombasthenia, the platelets lack an intact Gpl/b-lIIa complex, the fibrinogen receptor essential for aggregation. 97 ,9B The clinical features of this condition are those typical of primary haemostatic bleeding: petechiae, purpura (Figs 4.14 and 4.1 5) and mucous membrane bleed ing, including epistaxis which can be life-threatening, men orrhagi a and gastrointestinal haemorrhage. 99 There is also significant haemorrhage in response to minor trauma (Fig. 4.16) and major trauma is life-threatening. It is inherited as an autosomal recessive and there is usually no family history of bleeding but the child may have a history of easy bleeding and bruising from birth. It cannot be distinguished on clini cal grounds from other severe platelet defects or from severe VWD. Screening tests are normal and the diagnosis can be made on platelet function tests that show lack of aggregation to all agonists including ADP, thrombin and collagen. The PFA-lOO shows very prolonged closure times with both
Figure 4.15
Multiple bruises on the buttocks and thigh s of a
child w ith Glanzmann's thrombasthen ia.
93
94 I
Haematologica l abnormalities that can simu late abuse
in peri follicular haemorrhage, subperiosteal, orb ital or sub dural haemorrhage. Purpl e, spongy swelling of the gums where teeth have erupted may occur, and there may be melaena and haematuria. Bone disease is also manifest with osteoporosis, metaphyseal white lin es, submetaphy sea l lucency and subperiosteal haemorrhage. 103 The diag nosis is made on clinical and radiographic gro unds, by assessmen t of vitamin C intake and by measurement of leucocyte ascorbic acid levels. Treatment with asco rb ic acid stops the bleeding manifestations.
ABNORMALITIES OF PLATELET NUMBER OR MORPHOLOGY Congenital
Figure 4.16
Bleeding after a sma ll scratch in a child w ith
Glanzmann's thrombastheni a.
cartridges. Ad hesio n to connective tissue occurs normally. The bleed ing time is also markedly prolonged. Several conditions can prese nt with bleeding manifest ations when the primary problem is vascular integrity or connective tissue disease and laboratory scre ening res ul ts are normal.
Henoch-Schonlein Purpura This condition causes purpura secondary to an immunoglobu lin A (IgA)-mediated vasculitis and has been mistaken for child ab use. J6 ,59 The screening tests of coagulation are nor mal, or the platelet count may even be slightly raised. 16 ,100 The purpura typically begins as an urticatial wheal, which gradually changes to the colour of a bruise (palpable pur pura), occurring on the extensor sutfaces of the body. An important differential feature fro m child abuse is the sym metry of the bruising, particularly when it is extensive (Figs 4.8 and 4.9, pp. 84 and 85).60
Vitamin C Deficiency Vitamin C deficiency (scurvy) is rare but can be manifes t in infan ts betw een the ages of 6 months to 2 years who have a poor dietary intake, especially if milk and fruit juice is boiled. It has also been rep orted in chil dren with very restricted diets. IOI ,102 The bleeding tendency is due to loss of vascular integri ty wi th collagen deficiency, and results
In severa l congenital platelet disorders the platelets are not only low, but also morphologically ab normal. In add itio n, some a re associated with dysfunctional platelets. In practice, as long as a full blo od count an d film are examined, the diagnosis of a haematological abnormality sho uld be easily made. Many of these abnormalities are very rare, how ever, and precise di ag nosis may require fur ther testing including pl atelet function tests and electron microscopy.
WlSKOTT-ALDRICH SYNDROME This co ndition occurs via X-linked recessive inheritance and therefore is seen almost exclusively in boys. It is char acterized by thrombocyto penia with small platelets (Fig. 4. 17) (which a re dysfunctional), eczema an d combined immunodeficiency.104 Bleeding episodes usually present with in the first fe w years of li fe and have a patt ern typical of a primary haemostatic disorder, namely mucocutaneous bleeding and bruising.
BERNARD-SOULIER SYNDROME In this condition there is moderate-to-severe thromb o cytopenia, with associated pla telet dysfunction due to absence of the Gpl b receptor lO5 and giant platelets on examination of the blood film (Fig. 4.18) . The type of bleeding seen is typ ical of tha t of a primary hae mostatic defect and can be severe. The haematological picture can be confused with idiopathic thrombocytopenic purpura (ITP), in which the low platel et co un t may also be associ ated wit h gia nt forms . Howeve r, there is no functional defect in ITP and the bleeding is genera lly less severe.
HERMANSKY-PUDLAK SYNDROME This syndrome is associated with a lack of platelet-dense granules. It is inherited in an autosomal recessive man ner
Abnormalities of platelet number or morphology I
Figure 4.17 Wiskott-Aldrich syndrome, show ing thrombocytopen ia and a sma ll platelet.
95
Figure 4.19 The peripheral blood of a patient with Chediak-Higashi syndrome, show ing a neutrophil with giant abnormally staining granules.
oth er granule-co ntaining cells (Fig. 4.19). There may be neu tropenia and later thrombocytopenia. The pla telets show a storage pool defect on testing, and bleeding is usually mild to moderate and less severe than that seen in Hermansky-Pudlak syndrome. lOS Partial oculocuta neous albinism also occurs. MAY-HEGGLIN ANOMALY
This is a rare, autosomal dominant disorder in which th rom bo cytopeni a and giant platelets are seen. Pale blue-staining inclusion s, similar to Dohle bodies, are also visible in the cytoplasm of neutrophils, eosinophi ls, basophils and mono cytes (Fi g. 4.20) .1 09 Functional abn ormalities of the platelets have been demonstrated. llo Bleeding is seen in nearly one half of the pa tients and is of a primary, haemostatic nature. OTHER MACROCYTOPATHIES
Figure 4.18 Bernard-Sou lier syndrome, show ing thrombocytope nia and giant platelets. Neutrophils and lymphocytes are normal. and is characterized by tyrosinase-positive oculocutaneous albinism and the accum ulation of ceroid in the tissues. \06,107 Bleeding is usually mild to mod erate. CHEDIAK-HIGASHI SYNDROME
The most striking morphological abnormality in this syn drome is the presence of giant organelles in leucocytes and
There are several other condi ti ons in which macrothro mbo cytopenia and leucocyte cytoplasmic inclu sions are a feature and these include Fech tner's syn drome, with associated renal failure, and the Sebastian platelet sy ndrome, witho ut renal failure. 109 Alport's syndrome, in which there is sensorineural deafness, haematuria, cataracts and progressive renal fail ure, can be associated in some cases with macrothrombocyto penia and platelet dysfunction (Epstein's syndrome) III or macrothrombocytopenia with norma l pla tel et func tion (Eckstein's syndrome).lJ 2In these conditions, bleeding can be mild to moderate bu t can be significant after tra uma. The progressive renal failure seen in some of these sy nd romes will also con tribu te to the bleeding di athesis. GREY PLATELET SYNDROME
In this extremely rare recessive condition the platelet 0: granu les are red uced or ab se nt and th e pla telets thus appear agranular on a Wright-Giemsa stain. There is mild thrombocytopenia and occasi onal large forms are seen. !Os Bleed ing is usually mild.
96 I
Haematological abnormalities that can simulate abuse
a platelet count, the diagnosis was initially missed and a mis taken d iagnosis of abuse made.
MYELODYSPLASIA In this condition platelet numbers are often low and may be morphologically abnormal. The condition will be dis cussed later under 'Bone marrow failure syndromes'.
COAGULATION DEFECTS Haemophilia A and B
Figure 4.20
The peripheral blood of a patient with
May-Hegglin anomaly, showing thrombocytopenia, a giant platelet and a small pale blue neutrophil inclusion (Dahle bod y).
Acquired ---~
Acquired defects can be either quantitative or functional, the former being the more common. Acquired thrombocytope nia should be easily identified from a full blood count and film. Further investigations may be appropriate dependent upon initial results and clinical examination.
IM MUN ETH ROM BOCYTOPENIA Idiopathic lor immune) thrombocytopenic purpura (lIP) is often an acute, self-limiting, benign disease in childhood, presenting most commonly with skin purpura and mucocuta neous bleeding. Intracranial haemorrhage is extremely rare and is often associated with additional risk factors. 1lJ-116 The child is usually well or may have an accompanying viral infection and the diagnosis is made on the basis of a normal clinical examination, apart from the bleeding manifestations. The presence of petechial haemorrhages accompanying the bruising is an important clinical finding. A full blood count shows an isolated thrombocytopenia and large platelet forms may be seen on the blood film. Wheeler and Hobbs report three cases of ITP 117 and Harley two cases,B for which, without
Haemophilia A and haemophilia B are the most common of the severe inherited bleeding disorders. They are due to a functional deficiency of coagulation factors vm and IX, respectively, and are clinically indistinguishable from each other. As X-linked conditions, the severe forms occur almost exclusively in males. They are the commonest inherited bleeding disorders to present neonatally (Smith 1990 11B ) and 90 per cent of those with severe disease wiII have presented with bleeding by the age of 1 year. Neonatal presentations include cephalhaematoma, excessive bleeding from a heel stab or venepuncture, or bleeding at the site of intramuscu lar vitamin K (Fig. 4.6, p. 83). Only occasionally does umbil ical stump bleeding occur. Intracranial haemorrhage (ICH) has been reported in neonates with severe haemophilia, especially, but not only, if there has been a long, difficult labour or instrumental deliveryY9-1 2J Diagnosis in these cir cumstances is crucial, as immediate treatment is required. After the immediate neonatal period, the baby is unlikely to bleed unless injury occurs. This may be comparatively triv ial, such as a fall from a sofa onto a padded surface, but the bruising or bleeding that results will be disproportionate to the injury (Fig. 4.21). The child may present with tongue and mouth bleeding as he begins to put objects in to his mouth, especially if he is learning to walk and falls with such an object in his mouth. As he learns to walk haemarthroses will occur, resulting in a limp or reluctance to use a limb. Soft tis sue bleeding and bruising ca n lead t o the mistaken diagno sis of chi ld abuse, particularly as there may be no history of significant injury and bruises of different ages may be pres ent. Although an inherited disease, approximately one-third of cases arise as a result of a new mutation, 56 a positive his tory is only elicited in around 50 per cent of cases. 122 Several reports describe children who have been misdiagnosed as victims of NAI before full evaluation of the possible causes of the bleeding has been undertaken. B,117, 123 In haemophilia A or B, a coagulation screen will show isolated prolongation of the aPTT, which will correct on 50:50 mixing with normal plasma. Specific coagulation assays wi ll identify the defi cient factor, and appropriate treatment and advice can then be instituted. Moderate and mild haemophilia present later. In the case of mild haemophilia, presentation is usually only after trauma . The aPTT assay is not very sensitive to mild, but clinically significant, decreases in factor VIII or IX
The neonate I
97
0.4-0.5 gIL are haemostatic and in prothrombin deficiency lev els need to reach 0.2-0.3 fU/mi. for factor V, factors V + VIII, factor VIl and factor X haemostatic levels are achieved at around 0.10-0.15 fU/mL. This should be borne in mind when attributing bleeding to reduced levels of these factors. 124,125
Deficiencies of Factor XII, High-l\I1olecular Weight Kininogen and Prekallikrein
------
Deficiencies of these proteins result in a prolongation of the aPTI but are not associated with a bleeding tendency. They are important for in vitro haemostasis but not for in lJivo haemostasis. 66 .67
THE NEONATE Figure 4.21 Haematoma and bilateral black eyes in a chi ld with undiagnosed haemophilia after a fall from a height of 40 em on to a eush ion. levels. In a survey as part of the National External Quality Assessment Scheme (NEQAS) in the United Kingdom, test plasma w ith a factor vm level of 0.33 fU/ml was sent for aPlT testing to 572 laboratories. In total, 605 tests were carried out and the overall median aPTI ratio was 1.25, with a median aPTI of 38.1 seconds. 64 Increases in aPTI of only a second or two should not be ignored.
Haemophilia C Factor XI deficiency is found mostly but not exclusively amongst the Ashkenazi Jewish population and is the com monest of the rare coagulation deficiencies after haemophil ias A and B. 58 •67 It is inherited as an autosomal recessive co ndition and there is often no positive family history, unless there is consanguinity within the family. Heterozygotes may have a mild bleeding tendency and have levels of factor XI of between 0.15 and 0.70 fU/mi. Homozygotes have levels below 0.15 IU/mL and generally have more severe bleeding problems although bleeding tendency is not exclusively determined by factor XI levels. 67 Bleeding is usually mucocu taneous in nature with easy bruising, epistaxis and menor rhagia. Haemarthroses are uncommon. Bleeding is manifest after trauma or surgery. A coagulation screen will show an isolated, prolonged aPTI with normal factor VllI and factor IX levels. A factor XI assay will show reduced levels.
Special cons ideration needs to be given to interpretation of haemostatic parameters in the neonate, which differ from those in older children and adults. As the system matures in the growing child, so parameters change towards those of the older child and adult. Age-related normal ranges are therefore necessary and sequential ranges are required in the first few weeks and months of life. There are also some bleeding problems that might manifest themselves in the neonatal period, either because they are specific to that period or because they are congenita l and present after a haemostatic challenge. Most haemorrhagic conditions presenting in the neonatal period are accompanied by abnormal screening tests, the majority of them are acquired. The exceptions to this are severe platelet function defects in the presence of normal platelet numbers such as Glanzmann's thrombasthenia and factor XIII deficiency. Presentation and description of both these conditions have been outlined above. Dysfibrino genaemias rarely present in this period but in general are detectable by abnormal fibrinogen measurements or throm bin time. By contrast, severe hypofibrinogenaemia or afib rinogenaemia frequently presents in the neonatal period with bleeding from the umbilical cord, gastrointesti nal haemorrhage or cerebra l haemorrhage. In afibrinogenaemia both PT and aPTI are markedly long. Von Willebrand dis ease presents, exceptionally, in the neonatal period, as levels of VWF are physiologically increased and high-molecular weight multimeri c forms, which are more active, are prefer entially increased in this period. 126 These physiological changes result in patients with VWD having adequate levels for primary haemostasis. Those with type 3 VWD, however, are at risk of bleeding, as both vv\1f and factor VlII are markedly reduced and screening tests wi ll be abnorma l.
Rare Coagulation Deficiencies These include deficiencies of fibrinogen (factor 1), prothrombin (factor 11), factor V, factors V + vm, factor VIl and factor X. They occur with a frequency of 1 in 500000 to 1 in 2 million. In congenital fibrinogen deficiency, levels of
Thrombocytopenia in the Neonate This will lead to purpura and, in particular, a petechial rash.
Differentiation from NAl can be made from examination of
98 I
Haematological abnormalities that can simulate abuse
the blood count and film. The causes will include the rare inherited thrombocytopenias described above, immune causes such as neonatal alloimmune thrombocytopenia (NAlT) or maternal ITP and non-immune causes. Non immune thrombocytopenia is seen in congenital intra uterine infection such as that caused by cytomegalovirus (CMV) or the human immune deficiency virus (HN), and in neonates who are sick from various causes, including sep sis and neonatal asphyxia, congenital aplastic processes as described above or infiltration of the marrow by leukaemia, for example. History and examination are important in determining the precise cause.
Neonatal Alloimmune Thrombocytopenia Although NAIT classically presents at or soon after birth it should not be mistaken for NAl. The classical presentation is one of bleeding, which includes intracranial haemorrhage in an otherwise well infant. Neonatal alloimmune thrombo cytopenia occurs as a result of transplacental passage of maternal IgG antibody directed against the fetal platelets owing to expression of a paternal antigen that the mother does not possess. The commonest cause of this is antibody directed against the HPA-Ia antigen expressed on the platelets of about 98 per cent of the population. This is significant because the antibody, when bound, blocks the GpIfb-lIIa receptor on the platelet and leads to a platelet function defect as well as moderate to severe thrombocytopenia. Although skin and mucocutaneous bleeding are frequent, the most ser ious complication is that of intracranial haemorrhage, which occurs in up to 20 per cent of neonates with this condition; around one-half of these bleeds occur in utero. The degree of thrombocytopenia is most marked in the first day of li fe and platelet numbers gradually rise over the next 2-4 weeks as the level of the antibody declines in the infant's circulation. It is essential that the condition is correctly identified so that appropriate treatment may be given to the child and advice given to the mother for future pregnancies.
Vitamin K Deficiency Vitamin K is essential for the carboxylation of coagulation factors II, VJI, IX and X. Without this step, these factors are inert and incapable of activation in the coagulation process. Vitamin K deficiency misdiagnosed as child abuse has been described in several reports, for example Wheeler and Hobbs 117; Wetzel et al 127 in the neonatal period (haem orrhagic disease of the newborn, HDN) and by Carpentieri et al 128 and Kaplan 129 later in life. Presentation of HDN can be in the immediate neonatal period and is usually second ary to maternal ingestion of vitamin K antagonists such as warfarin or anticonvulsants. Classical HDN presents at between 2 and 5 days after birth, with purpura and gas trointestinal haemorrhage in an otherwise well infant. Mucocutaneous bleeding can occur, but intracranial
haemorrhage is uncommon. Vitamin K prophylaxis wil l not have been given and the infant is often almost exclu sively breast-fed. Factor VJI has the shortest half-life of the vitamin K-dependent coagulation proteins and therefore a coagulation screen will initially show a prolongation of the PT. As factors II, IX and X then fall, so the aPTT prolongs. A deficient state is demonstrated by correction of the coagula tion times after a 50:50 mix of test with normal plasma. These screening tests are not specific and further investiga tion of individual factors and the carboxylation state of prothrombin will confirm the diagnosis. A therapeutic trial of vitamin K will shorten the PT significantly in this condi tion and may be the most practical approach. Late HDN occurs after the first week of life and is most common between weeks 2 and 8 after birth; however, there have been reports of it occurring as late as 15 weeks after bilih.1JO A significant proportion of neonates (50 per cent) presents with intracranial haemorrhage and its associated morbidity and mortality. A history of exclusive breast feeding and either lack of vitamin K at birth or a single oral dose, is often elucidated. However, it is also seen with other underlying conditions that res ul t in malabsorption of vita min K, such as antibiotic therapy. Laboratory investigation reveals the same results as for classical HDN.
DRUGS ASSOCIATED WITH BLEEDING Warfarin Ingestion of warfarin, whether accidentally or for therapeutic purposes, will lead to a haemorrhagic di athesis, the extent depending upon the amount taken. Bruising is the common est manifestation but mucocutaneous bleeding, including haematuria and gastrointestinal bleeding, can also occur. If investigated, a coagulation screen will show a prolonged PT with normal or slightly prolonged aPTT and normal platelets. When a large amount of warfarin has been ingested, the aPTT will be prolonged, but not to the same extent as the PT. Such an overdose may be accidental, but it must be remembered that a child with fabricated or induced illness (FII) may pre sent with bleeding due to warfarin poisoningYI.132
Heparin Heparin must be given parenterally to exert its antithrombotic effect, which is mediated through antithrombin. Owing to anti-factor Xa and antithrombin effects, the capacity to gen erate thrombin is both decreased and delayed. Heparin can cause excessive bruiSing but also more serious bleeding, including CNS and retroperitoneal bleeding. Heparin can also induce thrombocytopenia. If standard heparin has been given, a coagulation screen will show both prolonged aPTI and TI. The reptilase time will be normal. Low-molecular-weight heparin is most senstively detected using an anti-Xa assay.
Bone marrow failure syndromes I
Platelet Dysfunction Secondary to Drugs
BONE MARROW FAILURE SYNDROMES
Many drugs have been associated with platelet dysfunction in vitro but either do not have a clinically signifi cant effect in vivo or are used under strict medical supervision;1 33 such examples include high-dose penicillin and heparin.
Inherited Bone Marrow Failure Syndromes
Sodium Valproate The anticonvulsant sodium valproate may also cause thrombocytopenia and minor coagulation abnormalities due to an effect on platelet function. Richardson 134 describes a study in which platelet functio n was shown to be abnormal in 6 out of 23 patients on sodium valproate but who had normal bleeding times. A previous report had shown a prolonged bleeding time in four out of five patients, one of whom was thrombocytopenic. A later prospective study showed that one-third of children stud ied had a fall in platelet count, with the lowest reaching 35 X 10 9 /1. 135 Subsequent studies have confirmed that sod ium valp roate-associated thrombocytopenia occurs in 12 per cent to 18 per cent of patients , with th e platelet counts correlating inversely with age of the patient and dosage of the dru g. 136 - 138 Whether the mechanism is immune or due to impaired platelet production has not been clearly e]ucidated . 139 ,\40 There are a few other drugs that have clinically relevant antithrombotic and haemor rhagic effects.
NON-STEROIDAL ANTI-INFLAMMATORY DRUGS Aspirin is rarely given to children because of the risk of Reye's syndrome. There are certain specific indications such as Kawasaki 's disease, for whic h there is a risk of thrombosis and aspirin is given specifically for its ant ipl atelet effects. 141 Accidental aspirin ingestion or poisoning occurs. Inges tion of aspirin is the commonest cause of platelet dysfunction in the population as a whole. Aggregation is impaired due to deficien t thromboxa ne A2 formation by aspirin's inhibition of cyclooxygenase. The defect lasts for the lifespan of the platelet and causes a mild bleeding tendency. It can con tribute significantly to the bleeding seen in those with coag ulation defects such as haemophilia; maternal asp irin within 10 days of delivery may contribute to neonatal bleeding.14 2 O'Hare and Eden 7 describe a case of child abuse in which aspirin resulted in abnormal platelet function tests an d led to severe bleeding from a minor injury. Other anti-inflamma tory drugs such as ibuprofen that might be used in children are rarely associated w ith bleeding problems.
Some drugs a re g ive n specifically for their antithrombotic effects but a re rarely used in ch ildren. They include prosta cyclin and prostacyclin analogues, ticlopidine and dipyri damole.
- -- - - -- -
It is outwith the remit of this chapter to discuss bone mar row failure syndromes in detail, but Freedman and Doyle 143 have reviewed this topic. When there is significant thro mbo cytopenia, presentation with bruising and bleeding is pos sible. Many of these conditi ons have accompanying physical abnorm alities as well as haematological ones. For example, Fanconi's (aplastic) anaemia is often accompa n ied by short stature, pigmentary skin changes and abnor malities of the upper limbs. 144 Othe r conditions in this category includ e Shwach man-Diamond syndrome, dysker atosis congenita and thrombocytopenia with absent ra dii (TAR) . Amegakaryocytic thrombocytopenia presents in infancy with isolated thrombocytopenia due to red uced or absent marrow megakaryocytes. An initial diagnosis of ITP may be made (see below)' but with failure of resoluti on of the thrombocytopenia within 3 weeks a bone marrow asp ir ate is recommended 145 and the diagnosis made. Overall , 45 per cent of patients will go on to develop ap lasti c anaemia.
Aplastic Anaemia In this condition, a child may present with pallor, overt infec tion or bleeding problems typical of a primary haemostatic defec t. Past history is usually unrem arkab le and physical examination normal. A fuJ I blood count will show a combina tion of anaemia, leucopenia and thrombocytopen ia. The diagnosis can be confirmed on bone marrow aspirate and biopsy.
Leukaemia Leukaemia can present primarily with bleeding problems, but the clinician is usually ale11ed to the diagnosis as the child is often unwell and examination may reveal hepatosplenomegaly and lymphadenopathy. A full blood count may show a pancytopenia or a naemia a nd thrombo cytopenia with a normal or high wh ite cell count. Gener ally, blast cells are seen on the blood film and diagnosis ca n be confirmed by bone marrow examin ation. This con dition is rarely misdiagnosed as NAI, although McClain et a!. 146 describe a case of acute lymphoblastic leukaem ia diagnosed at autopsy when abuse was suspected.
Myelodysplasia
ANTI-PLATELET DRUGS
-
99
- - -- - --
Acqu ired storage pool defects can acco mpany the thrombo cyto penia of myelodysplasia and result in significant skin purpura and mucocutaneous bleeding (Fig 4.22).133 Diagnosis is genera lly straightforward, with a full blood
-----
--
Haematological abnormalities that can simulate abuse
100 I
count showing pancytopenia commonly with a red cell macrocytosis and teardrop poikilocytes; neutrophils may show Pelger-Hue! forms. Bone marrow cellularity is normal or increased, often with increased reticulin, and one or more cell lines may be dysmorphic. An iron stain to detect ring sideroblasts and chromosomal analysis is also useful. Early in the disease process, however, mild thrombocytopenia alone may be present, with a disproportionate bleeding ten dency due to the functional defect. Other investigations such as fetal haemoglobin (HbF) level and plasma lactate dehy drogenase measurement may indicate the correct diagnosis. Vitamin B12 or folate deficiency must be excluded.
(al
Bone Marrow Infiltration Malignant conditions with bone marrow infiltration can result in thrombocytop enia and subsequent haemorrhage. Neuroblastoma in particular can present with unilateral or bilateral black eyes due to tumour infiltration (Fig. 4.12, p. 85). Lipid storage diseases, such as Gaucher's disease, can infiltrate the marrow and cause splenomegaly with subse quent hypersplenism. In this condition other coagulation abnormalities have been reported, including factor IX and factor Xl deficiencies and abnormalities of platelet func tion. 147 - 149 Clinical bleeding is generally mild and in most cases is that expected for the degree of thrombocytopenia.
(el
SYSTEMIC DISEASE ASSOCIATED WITH A BLEEDING TEI\IDEI\ICY Although defects of both coagulation and platelet function are described in systemic disorders, the primary condition underlying the defect is nearly always manifest and there fore consideration of the bleeding defect can be taken into account if the question of possible child abuse arises. Johnson 19 points out that children with chronic and stress produ cing diseases may be at increased risk of abuse from their parents or carers.
Renal Failure Bleeding in children with end-stage renal disease is usually from mucosal membranes and is related to the level of uraemia. I SO Qualitative defects of platelet function are well described and are one of the most important contributors to impaired haemostasis. Mild thrombocytopenia can also occur and anaemia is invariabl e; both of these conditions will contribute to the bleeding tendency.
Figure 4.22 Bruising with no history of trauma in a child with mild thrombocytopenia and platelet dysfunction due to a congenital myelodysplasia.
Liver Failure The liver is the main site of synthesis for most haemostatic proteins. An exception to this is factor V1II, which has
Conclusion I
signi ficant extrahepa tic synthesis . Liver failure results in impaired synthesis of coagulation proteins and reduced clearance of activated haemostatic factors. Thrombo cytopenia can occur as well as imp aired platelet function. 49 Owing to the reduction of both procoagulant and anticoag ulant proteins and activation of both the coagulation and fibrinolytic systems, thro mbosis, as well as bleeding, may occur. Factor VII has the shortest half-life of the coag ulation proteins and therefore disappears from the circulation first. The result is an initial prolongation of the PT, foll owed by prolongation of the aPTI, and, lastly, reduced fibrinogen as liver func tion declines. In the absence of DIC, factor VIII is normal owing to its extrahep atic synthesis. Bleeding manifestations include bruising and petechiae, mucosal haemorrhage and gas trointestinal haemorrhage.
Malabsorption Syndromes Malabsorption of fat-soluble vitamins, including vitamin K, can occur in cystic fibrosis ,128 biliary atresia and obstruc tive jaundice. Broad-spectrum antibio tics may alter the normal intestinal flora and result in malabsorption of vitamin K. Bleeding manifestations are those of vitamin K deficiency.
ACTIVATION OF COAGULATION
101
evidence of microangiopathy on the blood film (red cell fragments and reduced platelets) and increased levels of FDPs and/or D-dim ers . A chronic localized form of DIC is seen with giant haem angiomas; this is known as the Kasabach-Merritt syndrome.
Head Injury Head injury in children can be associated with coagulatio n abno rmalities secondary to DIc.152.153 In adults, a coagu lopathy has been associated with delayed brain injury sec ondary to bleeding. 154 ,155 The study by Hymel et al 151 of 265 children with head injury showed that those with parenchymal brain damage were more likely to have a coagulopathy manifested by prolongation of the PT than those without. Of the children who had died from abusive head injury, which had caused parenchymal brain damage, 94 per cent had prolongations of their PT and 63 per cent had evidence of activated coagulation. Hymel and col leagues concluded that it was highly unlikely that the coagulation abnormalities were due to an underlying haemorrhagic diathesis. It is important to recognize th is as a secondary phenomenon , not a primary one, so that inves tigation into possible abuse is not terminated on spurious gro unds.
Cyanotic Congenital Heart Disease
Disseminated Intravascular Coagulation Disseminated intravascular coagulation is due to patho logical activation of the coagulation system and can be precipitated by a variety of disease processes. Both endothelial injury and release of tissue factor can trigger the activation, resulting in both circulating thrombin and plasmin. Throm bin generates fibrin fro m fib rinogen, which is deposi ted in small vessels and causes microvascular thrombosis. Con sumption of coagulation factors occurs during this process, leading to the classical changes in coagulation parameters of a prolonged PT, prolonged aPTI, decreased fibrin ogen and reduced platelets. A bleeding tendency results. Increased circulating plasmin digests both fibrinogen and fibrin, resulting in raised FDPs and D-dimers (see p. 91). These interfere with both fibrin polymerization and platelet function, thus adding to the bleeding tendency. In children, DIC is usually acute and occurs in an ill child; it can be sec ondary to infection, trau ma (such as crush injuries or burns) and malignancy (such as leukaemia and li ver disease). A more chronic, compensated form can occur, with the fol lowing features : (1) less marked derangement of the coagu lation parameters when the platelet count may be only moderately reduced; (2) plasma fibrinogen is often normal or slightly elevated; and (3) the PT and aPTI may be within normal limits. In such patients, the diagnosis is made on
- - - - -
-
~ ~
Abnormalities of coagulation and platelets can occur in cyanotic congenital heart disease. 156 ,157 The exact mech anis m producing the coagulopathy is not known. Hypofib rinogenaemia secondary to poor liver function and reduced synthesis and clearance of clotting factor intermediates can lead to low-grade disseminated intravascular dissemin atio n. Thrombocytopeni a may result from both DIC and shortened platelet survival.
CONCLUSION In a child presenting with haemorrhagic symptoms, a his tory that includes details of ethnic origin, consangu inity and family history, drug history and clinical examination is essential. Initial screening tests should include aPT, aPTI, TI and fibrinogen, a factor VIII and factor IX level and a VWF antigen CVWF:Ag) and activity (RiCoF). If the history, examination and investigations do not account for the bleeding then further investigation might include other specific assays, such as factor XIII or platelet function test ing. Advice on ful1her testing or interp retation of findings should be sought from a haematologist who has experience in this field, especiall y if there is a disparity between the clinical and laborato ry findings. If the case is likely to
102 I
Haematological abnormalities that can simulate abuse
receive a formal legal challenge then even the rarest of causes may require exclusion. Diagnosis of a bleeding diathesis, especially if associated with a mild phenotype, does not exclude NAI and when these are found concur rently the child will be at even greater risk.
REFERENCES Stewart GM, Rosenberg NM. Co nditions mistaken for child abuse: part II. Pa ediatri c Emergen cy Care 1996; 12:2 17-221. 2 Gallenoh H, Oppenheim W. Th e battered child syndrome revisited. Clin OrtllOpaed Related Res 1982; 162:11-19. 3 Wexler R. Wounded il1lIocents: the Real Victims of tile War Against Child Abuse. Buffalo, NY: Pro metheus Books, 1990. 4 Vorenberg E. Di ag nosin g child abuse; th e cost of getting it wrong. Arch Dermatol 1992 ; 128:844-845. 5 Anh NT. 'Pseudo battered child' syndrome. JAMA 1976; 236:2288. 6 Kirschner RH, Stein RJ. Th e mistaken diagnosis of child
abuse: a form of medical ahuse? Am J Dis Child 1985 ;
139:873-87 5.
7 O'Hare AE, Eden OB. Bleeding disorders and non-accidental injury. Arch Dis Ch ild 1984; 59:860-864. 8 Harley JR. Disorders of coagulation misdiagnosed as nona cc id enta l b!1lising. Pedialr Emag Care 1997; 13 :347 -349. 9 Maguire S, Mann MK, Sibert J, Kem p A. Are there patterns of b!1lising in childhood which are diagnostic or suggestive of abuse? A systematic review. Archs Dis Child 2005; 90:182 -186. 10 Johnson CF, Coury DL. Bruising a nd hemophilia: accid ent or
child abuse? Ch ild Abuse and Neglect 1988 ; 12:409-415.
11 Smith. S.M, Hanson. R. 134 batte red children: a med ical and
psychological study. British IVIedical Journal 1974 ; 3:666-670. 12 Knight DJ, Bennet Gc. Non-accidental injury in osteogenesis imperfecta: case report. J Pediatr Orthop 1990 ; 10:542-544. 13 Altieri MF. Child abuse. When to be suspicious and what to do the n. Postgrad Med 1990; 87:153-162. 14 Ladebauche P. Childhood trauma. When to suspect abuse. Reg Nul'S J 1997; 60:38-42. 15 Roberts R. When you suspect abuse: what next? The Practitioner 1997; 241:322-326. 16 Daly KC, Si egel RM. Henoch-Schonlein purpura in a child a t risk o f abuse. Arch Ped iatr Ado/esc IVIed 1998; 152:96-98. 17 Lapo sata ME, Laposata M. Children with signs of abuse: when is it not child abuse? Am J c/in ParI1O/2005; 123(Suppl. I): S 119- I 24. 18 Sibert J. Bruising, coagulation disorder and physical child abuse. Blood Coagulation Fibrinolysis 2004; 15(Supp\. I): S33-39. 19 Johnson CF. Inflicted injury versus accidental injury. Pediatr c/in North Am 1990; 37:791-814. 20 Wilson EF. Estim a tion of the age of cutaneous contusions in chi1.d abuse. Pediatrics 1977; 60:750-75 2. 21 Schwartz AJ, Ricci R. How accu rately can bruis es be aged in abused children? Lite rature re v ie w and synthesis. Pediatrics 1996; 97:254-257. 22 Stephenson T, Bialas Y. Estimation of the age of bruising. Arch Dis Child 1996; 74:53-55. 23 Munan g LA, Leonard PA , Mok JY. Lack of agre e me nt on colour description between clinicians examining childhood bruising. J Clin Forensic Med 2002; 9:171-174. 24 Sugar NF, Taylor JA. Feldm a n KW. BIllis es in infants and toddlers: those who don 't c ruise ra rel y bruise. Puget Sound Pediatric Resea rch Network. Arch Pediatr Adolesc Med 1999; 153 :399-403.
25 Dunstan FD, Guildea ZE, Kontos K et a\. A sco ring system for bruise patterns: a tool for identi fying abuse. A rch Dis Child 2002; 86:330-333. 26 Hamilton M, Jenkin s B, Dunstan F et a\. Prospective analysis of bruising in children with and without an inherited bleeding di so rder. Blood 2006; 108:304a-305 a. 27 Kunicki TJ, Nugent OJ, Staats SJ et a \. The human fibroblast cl ass \[ extracellul ar ma trix receptor mediates platelet adhesion to collagen and is identical to the platelet glycoprotein la-lia compl ex . J BioI Chern 1988; 263 :4516- 4519. 28 Roberts HR, Monroe OM, Oliver JA et a\. New er concepts of blood co agulation. Ha ell10ph ilia 1998; 4:331-334. 29 Bazjar L, Morser J, Neisheim M. TAFI, or plasma procarb oxpeptidas e B, couples the coagulation and fibrinolytic casca des through the thrombin-thrombomodulin complex. J Bioi Chem 1996 ; 271: 16603-1 6608. 30 MacFarlane RG. An enzyme cascade in the bl oo d clotting mechanism and its function as a biochemical amplifier. Nature 1964; 202 :498-499. 31 Davie EW, Ratnoff 00. Waterfall sequ en ce for intrinsic blood clotting. Science 1964: 145:1310-2. 32 Peny OJ. Factor VlI deficiency. Br J Haematol 2002; 118 :689-700. 33 Clauss A. Rapid physiological coagulation methods in determination of fibrinogen. Acta Haemato/ (Basel) 19 57; 17 :237 -246. 34 Mackie IJ, Kitchen S, Machin SJ, Low e GDO. Guidelines on fibrinogen assays. Br J Haemato/2003; 121 :396-404. 35 Medical Dev ices Agency. MDA Evaluation Report : fibrinogen assay reagents and methods. MDA 2000/01 ISBN184J82 139X. Norwich: Her Majesty's Stationery Office. 36 Sutor A. The bleeding time in pediatrics. Semin Thromb Hemostasis 199 8 ; 24:531-543. 37 Valeri CR, MacGregor H, Cassidy G, Tinney R, Pompei F. Effec ts of temperature on bleeding time and clotting time in normal male and female volunteers. Crit Care Med 1995; 23 :698-704. 38 van den Burg PJ, Hospers J .E, van Vliet Met a\. Changes in haemostatic factors and activation produ c ts after exercise in healthy subjects w ith different ages. Thromb Ha emostas is 1995; 74:1457-1464, 39 Lowe GDO. Blood viscosity and cardiovascular disease. Th romb Haemostasis 1992; 67 :494-498. 40 Jakob E. Sutor A. Interaction of platelet aggregation and gra nulocyte function (Abstract). Ann Haemato/ 1991; 62:78. 41 sohngen D, Hattsteill E. Heye ll A et aI. Hematological parameters in fluencing th e Thrombostat 4000. Semin Thromb Hemostasis 1995; 2J(Suppl. 2):20-24. 42 Ivy AC , Shapiro F, Melnick P. Th e bleeding tendency in jaundice. Surg Gynaeco/ Obstet 193 5; 60:781 - 784. 43 Lethagen 5, Kling S. New bleeding time devices with retractable blades evaluated in children, hea lthy volunteers and patients with prolonged bleeding time. Thromb Haemostasis 1993; 70 :595-597. 44 Fe usner JH. No rmal and a bnormal bleed in g times in neo nates and yo ung children utilizing a full standardized template technic. Am J Clin Pa t11011980 ; 74:73-77. 45 Andrew M, Castle V, Mitchell L, Paes B. Modified bleeding time in the infant. Am J Hemato/1989; 30:190-191. 46 Andrew M, Paes B, Bowke r J , Vegh P. Evaluation of an automated bleeding time device in the newborn. Am J Hematol 1990; 35:275-277. 47 Michelson AD. Flow cytometlY: a clinical test of platelet function (Rev iew). Blood 1996; 87 :4925-4936. 48 Michelson AD. Platel et function in the newborn. Semin Thromb Hemostasis 1998; 24:507-512.
References I
49 Kni:il1er R, Weissbach G, Kuhlisch E. Platelet function tests in childhood. Measuring aggregation and release reaction in whole blood. Semin Thromb Hemostasis 1998; 24:513-521. 50 Ha rrison P, Robinson M, Leisner R et al. The PFA-IOO: a potentia l rapid screening tool for the assessment of platelet dysfi.lnction.
Clinical Labomtmy Ha ematology 2002: 24:225-232. 51 Harriso n P. The role of PFA-IOO® testing in th e in vestigation and management of hae mo stati c defects in children and adults. Br] Haematol2005; 130:3-10. 52 Andrew M, Vegh P, J ohnston M et al. Maturation of the h em atopoie tic system during childho od . Blood 1992; 80: 1998-2005. 53 Lane DA, Scully MF, Thomas DP et al. Acquired dysfibrinogenae mia in acute and chronic li ver disease.
Br] Haemato/ 1977; 35 :301-308. 54 Weigart Ai, Schafer AJ. Uremic bleeding: pathogenesis and th era py. Am] !VIed Sci 1998 ; 316 :94- 104. 55 Mannucci PM, Gi angrande PLF. Acquired disorders of coagulation. In Bl oo m AL, Forbes CD, Thomas DP, Tuddenh am EG D (eds) Haemostasis and thrombosis. London: Churchill Livings ton e, 1994, vol. 2, pp. 249-268. 56 Giannelli F, Green PM. The molecular basis o f haemophilia A and B. Bail/ieres Clin HaematoI1996; 9:21/-228. 57 Casaiia P, Martinez F, Haya S, Espinos C, Aznar JA. Significant linkage and non-linkage of type I von Willebrand disease to th e von Willebrand fac to r gene. Br 1 Haematol 2001; 115:692- 700. 58 Bolton-Maggs PHB, Young Wa- Yin, McC raw AH et al. Inh e ritan ce and bl eedin g in factor XI deficiency. Br] Ha ematol 1988; 69 :521-528. 59 Brown J, Melinkovich P. Henoch-Schonlein purpura mi sd iagnosed as suspected child abuse: a case report a nd literature review. ]AtVIA 1986; 256:617-618. 60 Menter A. Symmetrical purpura doesn't mirror ch ild abuse. lAMA 1987; 257:486. 61 Perrot LJ. Masque ecc hymotique. Specific or nonspecific indicator for abuse. Am] Forens ic iVIed 1989; Pathology 10:95-7. 62 Rooms L, Fitzgerald N, McClain KL. Hemophagocytic lymphohistiocytosis masquerading as child abuse ; presentation of three cases and revi ew of cen tral nervous system findin gs in hemophagocytic Iy mphohistiocy tosi s. Pediatrics 2003 ; 111:636-640. 63 Pearson RW, Triplett DA. Factor XI assay results in the CAP su rvey (1981). Am] Clil1 Pathol 198 2; 78(Suppl. 4) :615-620. 64 UK NEQAS for blood coagulation Survey !l8, September 1999. 65 Martin CM, Engstrom PF, Barrett O. Surreptitious self administration of heparin. ]AMA 1970; 212:475-476. 66 Blanchette VS, Dean J, Lillicrap D. Rare co ngen ital hemorrbagic disorders. In Lilleyman J, Hann I, Blan chette V (eds) Pediatric Hematology, London: Churchill Livingstone , 1999,611-628. 67 Seligsohn U. Facto r XI deficien cy. Tinomb I-Iaemostasis 1993;
70: 68-71. 68 Green D, Lechn er K. A survey of 215 non-hemophilic pa tients with inhibitors to fa cto r VlII. Thromb Ha el110stasis 1981; 45 :200- 203. 69 Nakashima K, Miyahara T, Fuj ii S et al. Spontaneously acquired fa cto r VllJ inhibitor in a 7-ye ar- old girl. Acta Haematol 1982; 68:58-62. 70 Reece EA, Clyne lP, Romero R, Hobbins J.e. Spontaneous factor XI inhibitors: seven additional cases and a review of the lite rature. Arch In t Med 1984; 144 :525-9. 71 Beck DW, Strauss RG , Kisker CT, Henriksen RA. An intrinsic coagulation pathway inhibitor in a three year old child.
Am] Ciin Patho11979; 71:470-472.
103
72 Shapiro SS, Martin ez J, Holburn RR. Congenital dysprothrombinaemia: an inherited stru ctural disorder of human prothrombin.] Ciin In vest 1969; 48:2251-2259. 73 Miller CH, Lenzi R, Breen e. Prevalence of von Willcbrand's disease among US adults. Blood 1987; 69:454-459. 74 Werner EJ, Broxson EH, Tucker n, Manaiac LF, Giroux DS, Abshire Te. Prevalence of vo n Willebrand disease in children: a multiethnic study. Blood 1991; 78:68a. 75 Castaman G, Fredericic AB, Rodeghiero F, Mannucci PM. Von Willebrand's Disease in tb e year 2003: towards complete identification of gene defects for th e correct diagnos is and treatment. Haematologica 2003; 88 :94-108. 76 Abildgaard CF, Su zuki Z, Harrison J et al. Serial studies in von Willebrand disease: variability versus 'variants '. Blo od 1980; 56:712-716. 77 Laffan M, Brown SA, Collins PW et al. The dia gnosiS of von Willebrand disease: a guideline from the UK Hae mophilia Centre Doctors' Organisation. Haemophilia 2004; 10: 199-217. 78 Castaman G, Eikenboom JCJ, Bertina RM, Rodeghiero F. In co nsistency of associa tion between type I von Willebrand di sease phenotype and genotype in families id ent ified in an epidemiological investigation. Tinomb Ha emostas is 1999; 82: 1065-1070. 79 Frederici AB, Mannucci PM. Actual management of von Wil!ebrand disease: first report on 880 cases of the Ita lian regis tlY of vWD. Blood 1997; 90:32a. 80 Frederi ci AB. DiagnosiS of vo n Willebrand disease. Semill Thromb Hemostasis 1998; 24:654-660. 81 Silwer J. Von Willebrand 's disease in Sweden. Acta Pa ediatrica Scandanovia 1973 ; 238:1-159. 82 Gill JC, Endres-Broo ks J, Bauer PJet al. The effec ts of ABO blood groups on the diagnosis of vo n Willebrand disease. Blood 1987; 69:1691-1695. 83 Ruggeri ZM, Zimmerman TS. Variant von Willebrand 's disease. Characteri zation of two subtypes by analysis of multimeric composi tion of FVllljvWF in pl asma and platelets. ] Ciil1 fl1/Iest 1980; 65: 1318-1325. 84 Mill er JL, Ca stella A. Pl atelet-typ e von Willebrand disea se: characterization of a new bleeding disorder. Blood 1982;
60:790-794. 85 Rinder MR, Richard RE, Rinder HM. Acquired von Willeb rand's disease: a concise review. Am] Hematol1997; 54:139-145. 86 Sim one N, Co rnet JA. Abildgaard CF. Acquired von Willebrand ' syndrome in systemic lupu s erythe ma tosus. Blood 1964; 31:806-812. 87 Pareti Fl, Mannucci PM, D'Angelo A et aL Congenital deficiency of thromboxane and prostacyclin. La ncet 1980; 1:898-901. 88 Duckert F. Jung E, Shmerling DH. A hitherto undescribed co ngen ital hemorrhagic diathesis probably due to fibrin stabilizing fact or deficien cy. Throl17b Dia thesis Ha em orrl1agic
1960 ; 5:179-186. 89 Sakata Y, Aoki N. Si gnitl ca nc e of cross-linking of a lpha 2-pJasmin inhibitor to tlbrin in inhibition of fibrinolysis and in hemostasis.] C/in Il1vest 1982; 69:536-542. 90 Duckert F. Documentation of the pl as ma fa ctor Xlli deficiency in man. Anl1 New York Acad Sci 1972; 20 2:190-199. 91 Lorand L, Losowsky MS, MiJoszewski KJM. Human factor XIII: fibrin s tabilizin g factor. Progress Hemostas is Tl1romb 1980; 5:245-290. 92 Saito H. Alpha 2-plasm in inhibitor a nd its deficiency states. 1 Lab Ciin Med 1988; 112:671-678. 93 Lind B, Thorse n S. A novel missense mutation in the human plasma inhibitor (alpharantiplasmin) gene associated with a bleeding tendency. Br] Haematol 1999; 107:317-322.
104 I
Haematological abnormalities that can simulate abuse
94 Schleef RR, Higgins DL, Pillemer E, Levitt LJ. Bleeding diathesis due to decreased functional activity of type I plasminogen activator inhibitor.) Clillinvest 1989; 83:1747-175 2. 95 Dieval J, Nguyen G, Gross S et al. A lifelong bleeding disorder associated with a deficiency of plasminogen activator inhibitor type J. Blood 1991; 77:5 28- 53 2. 96 Fay WP, Parker AC, Condrey LR, Shapiro AD. Human pla sminogen activator inhbitor-l (PAl-I) defici e ncy: characterization of a large kindred wi th a null mutation in the PAI-l gene. Blood 1997 ; 90:204-208. 97 Nurden AT, Caen JP. An abnormal platelet glycoprotein
pattern in three cases of Glan z mann's thromba sthenia.
Br) Haematol 1974 ; 28:253-260.
98 Phillips DR, Agin PP. Platelet membrane defects in Gl a nzmann 's thrombasthenia. Evidence for decreased amounts of two maj or glycoproteins.) Gin Invest J977; 60:535-545. 99 Gl a n zman n E. Hereditare hamorrhagische Thrombasthenie.
Ein Betrag zur Pathologie der Blutplattchen. )alzrbuch
Kinderheilkunde 1918 ; 18 : 1-42.
100 Sauslbury FT, Kesler RW. Thrombocytosis in Henoch Schonlein purpura. Clinical Pediatr 1983; 22:185-87. 101 Weinstein M, Babyn P, Zlotkin S. An orange a day keeps the doctor away: scurvy in the year 2000. Pediatrics 2001; 108:55. 102 Chatproedprai S, Wananukul S. Scurvy: a case report. ) Med Assoc Thailand 2001; U4(Suppl. J):S 106-lJO. 103 Stewart GM , Rosenberg NM. Conditions mistaken for child abuse: part I. Paediatr Emerg Care 1996; 12:116-12J. 104 Aldri ch RA. Steinberg AG, Campbell DC. Pedig ree demonstrating a sex-linked recessive condition characterized by draining ears, eczematoid dermatitis and bloody diarrhea. Pediatrics 1954; 13 :1 33-139. 105 Clemetson KH, Clemetson JM. Molecular abnormalities in Glanzman 's thrombasthenia, Bernard-Soulier syndrome and platelet-type von Willebrand's disease. CUTl' Opill Hematol 1994; 1 :388-393. 106 Hermansky F, Pudlak P. Albinism associated with hem orrhagic diathesi s and unusual reticulated pigmented cells in the bone marrow: a report of tVIO cases with histoch emical studies. Blood 1959; 14:162-169. 107 White JG. Structural defects in inherited and giant platelet disorders. Adv HlIm Gen et 1990; 19:133-234. 108 White JG. Inherited abnormalities of the platelet membrane and sec retory granules. Human Pa th ology 1987 ; 1U: 123-139. 109 Greinacher A. Muelle r-Eckh a rdt C. Hereditary types of thrombocytopenia with giant platelets and inclusion bodi es in the leukocytes. BIt//' 1990; 60:53-60. 110 Djaldetti M, Creter 0, Bujanover Y, Elian E. Ultrastructural and functional studies of the platelets in patien ts with May-H egglin anomaly. Haemafologica 1982; 67:530-538. III Epstein CJ, Sahud MA, Piel CF et al. Hereditary macrothrombocytopenia , nephriti s and deafness. Am) Med 1972; 52:299-310. 112 Eckstein JD , Filip OJ , Watts JC. Hereditary thrombocytopenia, deafness and renal disease. Ann Intern Med 197 5 ; 82:639-645. 113 Lilleyman JS. Intracranial haemorrhage in idiopathic thrombocytopenic purpura. Arch Dis Child 1994; 61:251-253. 114 Lilleyman JS. Intracranial haemorrhage in chronic childhood ITP. Pedia tric Hematology and On cology 1997; 14:iii-v. 115 Lilleyman JS. Managem e nt of childhood idiopathic thrombocytopenic purpura. Br) Haematol 1999; 105: 871-875. 115 Bolton-Maggs PHB, Moon I. Assessment of UK practi ce for management of acute childhood idiopathic thrombocytopenic purpura against published guidelines. Lancet 1997; 350:620-623. 117 Wheel er OM, Hobbs CJ. Mistakes in diagnosing non accidental injury: ]0 years' experience. Medical Practice 1988; 296:1 23 3-123 6.
118 Smith PS. Co ngenital coagul a tion protein deficiencies in the perinatal period. Semin Perinatal 1990 ; 14: 384-392. 119 KJetzel M, Miller CH, Becton DL et 31. Post delivery head bleeding in hemophili c neonates. Am ) Dis Clli/d 1989; 143: 1107-1110. 120 Ljung R, Lindgren AC, Petrini P, Tengborn L. Normal vaginal delivery is to be recommended for haemophilia ca rrier gravidae. Acta Paediatrica 1994; 83 :609-61 J. J21 Kadir RA, Economides DL. Obstetric management of carriers of haemophilia A. Haemophilia J997; 3:81-86. 12 2 Conway JH, Hilgarter MW. Initial presentations of pediatric hemophiliacs. Arch Pediatr Adolesc Med 1994; 148:589-594. 123 Schiver W, Brueschke EE, Dent T. Family practice grand rounds: hemophilia.) Fam Pmct 1982; 14:661-662,664, passim. 124 Peyvandi F, Mannucci PM. Rare coagulation disorders. Thromb Haemostasis 1999; 82:1207-1214. 12 5 Bolton-Maggs PH , Peny OJ , Chalmers EA et a!. The rare coagulation disorders : review with guidelines from the UK haemophilia Doctors Organisation. Hael110philia 2004; 10:593-628. 126 Katz JA, Moake JL, McPherson PO et al. Relationship between human devel opment and disappearance of unusually large von Willebrand factor multimers from plasma. Blood J989 ; 73: 1851-1858. 127 Wetzel RC, Slater AJ, Dover GJ. Fatal intramuscular bleeding misdiagnosed as suspected non-accidental injury. Pediatrics 1995; 95:771-773. 128 Carpentieri U, Gustavson LP, Haggard ME. Misdiagnosis of neglect in a child with bleeding disorder and cystic fibro sis. South Med) 1978; 71 :854-855. 129 Kaplan SJ. Pelcovitz 0, Salzinger S et al. Adolescent physical abuse: risk for adolescent psychiatri c disorders. Am J Psychiatry 1998 ; 155: 954-959. 130 Sutor AH , Dagres N, Nied erhoff H. Late form of vitamin K bl eeding in Germany. Klinische Padiatrie 1995; 207:89-97. 131 White ST, Voter K, Perry J. Surreptitious warfarin ingestion. Child Abuse Neglect 1985; 9: 3 49-352. 132 Souid AK, Korins K, Keith 0 et al. Unexplained menorrhagia and he m a turia: a case of Munchausen's syndrome by proxy. Pedin tr Hel17otol Oncol 1993; 10:245-248. 133 Warkentin TE, Kelton JG. Aquired platelet disorders. In Bloom AL, Forbes CD, Thomas DP et al (eds) Haemostasis and Thrombosis. London: Churchill Livingstone, 1994, pp. 767-815. 134 Richardson SG, Fletcher OJ, Jeavons PM, Stuart J. Sodium valproate and platelet function. BM) 1975; 1 :221-2 22 . 135 Barr RD, Copeland SA, Stockwelll\llL et a l. Valproic acid and immune thrombocytopenia. Arch Dis Childh 1982; 57:681-684. 136 Ko CH, Kong CK, Tse Pw. Valproic acid and thrombocytopenia. Hong Kong Med ) 2001; 7: J5-21. 137 Conley EL, Coley KC, Pollock BG et al. Prevalence and risk of thrombocytopenia with valproic acid: experience at a psychiatric teaching hospital. Pharmacoth erapy 2001; 21:1325-1330. 138 De Beradis 0, Campanella 0, Matera V et a l. Thromb ocytopenia during valproic acid treatment in young patients with new-onset bipolar diso rder. ) Ciin
Psychopharma col2003; 23:451-458. 139 Ganick OJ, Sunder T, Finley JL. Severe hematolog ic toxicity of valproic acid: a report of four patients. Am) Pedintr Hematol Oncol 1990; 12:80-85. 140 Gesundheit B, Kirby M, Lau W et al. Thrombocytopenia and megakaryocyte dysplasia: an adverse effect of v alproic acid treatment.) Pediatr Hematol Onco12002; 24:589-590. 141 Durongpisitkul K, Fururaj VJ, Park JM, Martin CF. The prevention of coronary artery anuerysm in Ka wasaki disease: a meta-analysis on the efficacy of aspirin and immun oglobulin treatment. Pediatrics 1995 ; 96: 1057- 106J.
References I
142 Stuart MJ, Gross SJ, Elrad H, Graeber JE . Effects of acetosalicylic- acid ingestion on materna l and neon ata l haemostasis. N Engl J Med 1982 ; 307:909-912. \43 Freedman MH, Doyle JJ. Inh erited bone marrow failure syndro mes. In Lilleyman J, Hann I, Blanchette V (eds) Pediatric Hematology. London: Churchill Livingstone, J 999, 23- 49. 144 Young NS, Alter BP. Clinical features ofFanco ni 's anae mia. In Aplastic Anaemia, Acquired 011d fuherited. Phila delphia, PA: WB Saunders, \ 994 , pp: 275-309. \45 Eden DB, Lilleyman JS. Guidelines for the management of idiopathic thrombocytopenic purpura. Arch Dis Child 1992 ; 67: 1056-1058, 146 McCla in JL, Clark MM, Sandu sky GE, Undiagnosed, untrea ted acute lymphobl astic leukaemi a presenting as susp ected child abuse. J Forensic Sci 1990 ; 35:735-739, 147 Berrebi A, Ma lnick SDH, Vorst fJ , Stein D, Hi gh incidence of factor Xl deficiency in Gau cher's disease. Am J Hema lol 1992; 40:153-161. 148 Humphries JE, Hess CE, Gaucher's diseas e and acquired coagu lopathy. Am J Hemato l 1994; 45:347-353, 149 Kelsey H, Chri stopoulos C, Gray AA. Mach in SI. Acquired pseudo-pseudo Bernard-Soulier syndrome comp licating Gaucher'S disease. J c/in Palh ol 1994; 47:162-165,
1 05
150 Leititis JU, Brandis M, Critical care in uraemic children, Pediatric Nephrology \ 992; 6:88-95, 151 Kelly D, Summerfi eld J, Haemostasis in liver disea se, Semil1 Liver Dis 1987; 7:182-\91. \ 52 Miner ME, Kaufman HH, Graham SH, Haar FH, Gilden berg PL. Disseminated intra vascu lar coagul ation fibrin olytic syndrom e following head injulY in children: frequency and prognostic implications, J Pedialr 1982; 100:687-691. 153 Hymel](P, Abshire Te, Luckey DW, Jenny C. Coa gulopathy in pediatric abusive head trauma. Pe(iialrics 1997; 99:371-375. 154 Stein SC, Young GS, Talucci RC et at. Delayed bra in injury after h ea d trauma: sign ificance of coagulopathy. NeurosurgelY \992 ; 30:160-165. 155 Kaufman HH, Moake JL, Olson JD et al. Delayed and recurrent intracran ial hematolnas related to disseminated intravascular clottin g and fibrinoly sis in head injury, Neurosurgery 1980; 7:445- 449 , 156 Wedemeyer Ai, Ed son JR, Krivit W. Coagulation in cyanotic congenital heart disease. Am J Dis Child 1972; 124:656- 660, 157 Waldman JD, Czapek EE , Paul MH et al. Shorten ed platelet survival in cyanotic heart disease. J Pedialr 1975; 87:77-79.
I
CHAPTERS
I
BIOCHEMICAL INVESTIGATIONS ON POST-MORTEM SPECIMENS Denis R Benjamin
Introduction General evaluation Hypoxia Inflammation Anaphylaxis Infection Dehydration and electrolytes
106 107 109 109 109 110 110
INTRODUCTION
The diagnostic use of biochemical tests on post-mortem specimens dates back to the earliest development of the technology that enabl ed us to measure the blood concen trations of various compounds. I It is safe to assume that almost evelY analyte measured during life has been investigated in post-mortem samples. Unfoliunately, the changes associated with death can profoundly affect the concentration of many intermediary metabolites. As a ll the dynamic energy systems normally maintaining various concentration gradients begin to fail or run out of sub strate, compounds rapidly equi librate across the various body compartments and cell membranes. Such changes occur in the first few minutes to hours after death. Indeed, many such changes may already be under way before a formal declaration of death by the medical a ttendan ts. With time, as cells die, intracellular compounds are rel eased into the surrounding tissue and circulation, com pounding t he difficulties in the interpretation of results. This is well illustrated by the onset of haemolysis in the circulation. The rupture of red cells and t he release of the haemoglobin and al l the intracellular enzymes in to the bloodstream often signa l the end of the useful period for the measurement of many metabolites. Not only does the concentration change in unpredictable ways, but also the ' matrix ' in which a particular ana lyte is normally measured may be sign ificantly a ltered and this can have an important imp act on the analytical method. As the pH of vario us body
Time of death (post-mortem interval) Endocrine disorders Genetic metabolic disorders presenting as sudden unexpected death Technical considerations at the time of autopsy References
111 112 114 117 120
fluids decreases, compounds may dissociate from proteins and other ligands thus changing their measured concentra tion. This is especially important in the interpretation of drug levels. A recent study, which included a large compendium of drugs detected in post-mortem blood, examin ed the pos sible factors affecting their concentrations. 2 Hormones and other compounds may be similarly affected. Some compounds are relatively stable for prolonged periods of time. These are of the most value in post-mortem assessment. Others show relatively predictable changes, increasing or decreasing with some definable relationship to the post-moliem in tervaL Numerous efforts have been made to use these for the assessment of the time since death, or if the time were accurately known, to use some 'correc tion factor ' or equation to calculate what the ante-mortem value would have been. Both uses are fraught with diffi culty as the many factors influencing the rate of change are usually not known, nor can they be readily controlled. Other compounds show erratic and unpredictable behaviour with increasing time after death, which is of almost no ben efit in post-mortem diagnosis. For many analyies, changes in blood concentration are so non-specific that their meas urement offers no signifi cant diagnostic information. Despite the obvious drawbacks of the post-mortem changes, the investigation of sudden, unexpected death in infants a nd children can be substantially aided by the judicious use of post-mortem biochemical testing on a variety of body fluids a nd tissues. The recognition, over the past decade, of a large number of inherited metabolic
General eva lu ation I
defects, which may not be preceded by overt or obvious clillical disease, has greatly ex pand ed the differential diag nosis of such cases. A high index of suspicio n will ensure [ha t these are not missed at the time of the autopsy, as diagnosis has imp0l1ant implications for the family and future pregnancy planning. If ca refully so ught, clues are frequently uncovered in the his tO lY or on the gross and microscopic exam ination alerting the pathologist to a pos sible underl ying disease. Newer technologies, such as ta n dem mass sp ectrometry, now permi t ready scree ning for a wide variety of metabolic defects, an d molecul ar biologica l techniques can be used for specific genetic disord ers. These have considerably exp anded our abi lity to establish diag noses from post-mortem material. It is pruden t to coHect all of the approp ri ate samples, en sure that they are suitably stored an d then decide on how to proceed, once the various lines of evidence unfold. All the caveats pe11aining to post-mortem biochemistry in adults apply in children . Most of the studies relating to the changes in con centration of various metabolites after death have been perfo rmed in ad ults, although a number have been validated in children. Table 5.1 lists the common fac tors influencin g the interpretation of bioch emical resul ts. The literature relating to thana tochemi stry is relatively old. Many of the studies were carried out in the early yea rs, soo n after clinical patho logy techni ques became available. Many have no t been rigo rously repeated with curren t technology and this has created potential problems. For example, flam e photometry, whi ch was the sta ndard for electrolyte measure ments in the 1950s and 1960s, has been largely repla ced with ion-selective electrodes. Each analytic method is sensi tive to its own particular set of interferences, and one cannot necessarily extrap olate results between different methodol o gies. For this reaso n, reference ranges published in the older literature may no longer have currency. For each analyte these variables need to be known and appropriately considered. Because of the large number of possibilities, great precision is frequently not possible. Inter pretation of the test results often requi res considerable judg ment and certainty of diagnosis is seldom achieved , especially with the measurement of traditional intemlediary metabolites and electrolytes Even so, these difficulties should
Table 5.1
Factors affecting the results of biochemical
investigations
Post-mortem interval Temperature at which the body has been maintained Sample type (e.g. blood, vitreous humour, urine, cerebrospinal fluid, etc.) Sample site (e.g . left or right heart, peripheral blood) Ana Iytica I method Interfering substances, e.g. drug s, alcoh ol Resuscitative attempts Mode of death (e.g. rapid, prolonged, presence of hypoxia, etc.)
107
not militate against the thoughtful use of selected biochemi cal tests. On the other hand, the use of molecular biologic tech niques and tandem mass spectrometlY has opened pow erful new avenues for the definitive diagnosis of a host of dis orders in infants and children who die unexpectedly. Failure to appreciate the impact of those factors influ encing results of post-mortem biochemical investigations, or to perform adequate cont ro l studies, has spawned a large literature in which numerous claims have been made for the pathogenesis of certain diseases, only to be refuted by subsequent studies. Nowhere is this more eviden t than in the investigation of sudden infant death sy ndrome (SIDS, or criblcot death). The 'scientific' literature is filled with such poorly designed studies . There has been a characteris tic pattern of publication in this field . The first couple of articles usua lly note an interesting chem ical observation in a small number of infa nts. This is claimed to be sp ecific a nd sensitive. A few articles validatin g the observation may follo w. But soon the naysayers arrive, poin ting out similar findings in a number of other settings. The speci ficity of the original fin ding co mes under se rious scrutin y as more 'controls' are added and the impact of post mortem in terval and other factors are assessed. And not infrequently, the test or obselvation that appea red so hope ful at its inception fades into obscurity. This chapter only covers those analytes or compounds for which interpretation is reasonably possible and there is some demonstrated clinical application . Readers are referred to a number of excellen t general reviews of post-mortem bio chemistlY if they require information about analytes that are not dealt with in this discussion. The following are nat covered in this chapter: lactate, pyruvate, ammonia, serum enzymes, lipids, trace metals, immunoglobulins,) methaemo globin,4 myoglo bin,5.6 thiamine,? and thyroid and parathy roid hormones 8- 14 as well as the other hormones.) It is indeed un fortunate that four of the most important indica tors of metabolic diseases in infants and children, namely, elevated ammonia, low glucose, abnormal pyruvate an d lactate are so unreliable in the post-mortem pelio d. Other fluids, such as pelicardial 15 and synovial16 have had such limited study that fe w ge neralizations about their value in the day-to-day prac tice of pediatlic pathology can be made. In considering the role of biochemical testing in the post-mortem period in pediatric fore nsic cases, tests can be di vid ed into two broad gro ups. Some tests are us eful for the assessment of t he general health of the child, whereas oth ers focus on specific clinical disorders. Table 5.2 lists the role of these anci lla ry pro cedures in pediatric cases.
GENERAL EVALUATION Nutritional Status The nutritional statu s of the deceased is an important con sideratio n in many pa ediatric au topsies. This may have
108 I
Biochemical investigations on post-mortem specimens
Table 5.2
Role of biochemical testing in paediatric forensic cases
General evaluation Nutritional status Hypoxia Inflammation Anaphylaxis Infection Dehydration Time of death (post-mortem interval)
advantages over prealbumin for the assessment of recent nutritional intake. Both these proteins are very sensitive to the presence of the acute inflammatory response. Hepatic synthesis rapidly decreases in the face of any inflammatory stimulus, be it from infection, trauma, neoplasia or autoimmune disease. For this reaso n prealbumin should always be combined with some measure of the acute phase response. At this time C-reactive protein is the recommended test (see below).19
Specific disarders Endocrine disorders Diabetes Adrenal and/or pituitary insufficiency Renal failure Inborn errors of metabolism Fatty acid oxidation defects Other mitochondrial disorders Urea cycle defects Aminoacidopathies Organic acid urias Abnormal carbohydrate metabolism Miscellaneous
implications in regard to parental neglect and/or abuse, or to a chronic, but clinically unsuspected, underlying dis ease. Evaluation of the physical characteristics and stan dard anthropometric measurements should be carefully performed and compared with the appropriate standards for age, sex and ethnic group. These will be abnormal in many chronic diseases, including endocrinolog ica l, meta bolic, immunological and renal disord ers. From the labora tory standpoint there are only a small number of tests that can be utilized to assess both recent and long-term nutri tional status. Obviously, none of the dynamic or functional tests that is helpful during life can be used. RECE NT NUTRITIONAL INTAKE Knowing the recent food intake history of the deceased sometimes has important diagnostic implications. Perhaps the best serum protein for evaluating recent protein and calorie intake is prealbumin. Prealbumin, also known as thyroxin-binding prealbumin or transthyretin, is a transport protein with a very short half-life (2 days). It is synthesized by the liver and has a small pool size. Either protein or calo rie deprivation will result in a decrease in the serum level of prealbumin in 3 to 4 days.17,18 As such it is an indicator of recent nutritional intake rather than an accurate assessment of nut ritional status. Any infant or child who has been recently sick or who may have had significant anorexia could have a depressed prealbumin level. There are no good studies of prealbumin in the immediate post-mortem inter val. Retinal-binding protein, also produced by the liver has an even shorter half-life (approximately 12 hours). It has no
LONG-TERM NUTRITIONAL STATUS The long-term nutritional status is much more problematic to assess from a biochemical standpOint and the laboratory probably offers few significant advantages over anthropo metric studies. When the body is adapting to the decrease in nutrients, including total calories and protein, there is a phase of compensation during which a variety of hormonal adapta tions and changes occur. These are difficult enough to evalu ate under ideal circumstances during life, and become even more so with the post-mortem changes in blood chemistry. Similarly, many tests of immune function become abnormal in the chronically malnourished infant and child but these are almost impossible to evaluate after death. Selected serum proteins have long been used as surro gate measures of nutrition. Albumin is a non-specific and relatively insensitive measure of nutritional status. Its serum concentration is dependent on a variety of factors. Post-mortem levels in the serum are close to ante-mortem values unless there is significant haemolysis. The problem is not with any post-mortem changes, but with t he ma ny factors influencing serum albumin concentration during life . The serum value is the net result of the protein intake and hepatic synthesis, degradation and loss. Each of these is subject to many influences other than protein intake. It should also be remembered that in cases in which caloric insufficiency exceeds the protein deprivation (i.e. maras mus) serum protein levels might be normal. Because of the long half-life of albumin in the circula tion (18-20 days) and its large extravascular pool, changes in serum values are slow to develop. Decrease in serum albumin develops late in the course of malnutrition and only in the more severe cases. In interpreting the albumin level the following factors must be kept in mind: • Serum albumin is dependent on t he hydrational state of the patient, so that dehydration in a malnourished child may result in relatively normal albumin level. • Both synthesis and catabolism are dep endent on hepatic function and the acute infl a mmatory response. The presence of abnormalities of either will significantly change albumin levels. • Renal and gastrointestinal losses of protein can profoundly decrease the albumin levels. If all of these can be excluded then low levels roughly correlate with the severity of chronic malnutrition.
Anaphylaxis I
HYPOXIA
A laboratory test for the detection of ante-mortem hypoxia would be a great boon for the forensic pathologis t. The question arises in a diverse array of circumstances, varyin g from the suspicion of suffoca tion and strangulation to drowning and fire victi ms and to natural causes of death, such as SIDS. Needless to say, although many tests have been investigated, none has proven to be reliable or satis factory. This is hardly a surprise as so many mechanisms of death involve some element of hypoxia as one of the com mon final pathways a nd the rapidity of the post-mortem change on most of the candidate chemical compounds pre cludes their use after more than a few minutes . For exam ple, post-mortem oximetry, with blood being samp led from va rious sites in the circulation, was no t fou nd to be partic ularly helpful or reliable in defining the cause of death in one study of 214 ad ult cases .20 Other compounds have also been investigated from this standpoint,21 such as lactate, which increases in response to anaerobic glyco lysis. How ever, there is such wide variation in the res ul ts that inter pretation in th e indivi dual case is all but impossible. Normally, lactate rises quite promptly in the blood - within a day reaching leve ls of up to 70 times greater than the ante-mortem va lues .22 The rise of lactate in the vi treous humour is not quite so dramatic, but is sti ll substantial. 2J Hypoxa nt hin e is formed from adenosi ne monophosphate in the presence of hypoxia. Its appearance in the vitreous humour has been used as indicatio n of ante-mortem hypoxia. For examp le, some investigators have hypothesized that a SIDS death may be preceded by hypoxic episodes of varying duration. 24 One study investigated the hypoxan thine levels in four g roups of patients including a group of SlDS victims, ch ildren with congenital heart disease, babies who succumbed from respiratory distress and a control group of accident victims in whom there was no evidence of hypoxia. A sizeable fraction of the infants and children in the SIDS, respiratory distress and congenital heart groups had elevated levels of vitreous hypoxanthine.25 This obser vation has been challenged by other stud ies, which have all failed to corroborate the results. 26 - 28 At this time there is conflicting evidence to support a significant difference in the hypoxa nthine concentration between SIDS and other causes of death in infants. 29 ,)O Most recently, the meas urement of vascular endothelial growth factor (VEGF) in cerebrospinal fluid (CSF) has been used to assess the incidence of hypoxia in SlDS victims.)1
INFLAMMATION It is freque ntly useful to know if there is a pre-ex isting infl amm atory state at the time of death. This 'acute phase response' commonly accompanies infections, autoimmune conditions or maligna ncies. The erythrocyte sedimentation rate (ESR) has been the tradition al screening test with the
1.09
appropriate clinica l indications, albeit quite non-spec ific. Performing a satisfactory ESR test after death is not possi ble with all the changes in coagulation and fibrinolysis. In the last two decades, C-reactive protein has found a very useful role as an alternative or a compl ementary indication of inflammation; it can be used as a replacement for the ESR in the majority of clinical situations or in addition to it. C-reactive protein has many advantages over the ESR, as it is not influenced by many of the other proteins that affect the ESR and it is independent of the shape, size and number of the red cells. It is produced by the liver as part of the acute phase response and is a sensitive marker of inflammation . It increases dramatically in the serum, rising to over 10 times the baseline value within 12 hours.)2,)) None of the other proteins that also increase in response to inflammation, such as a -acid glycop rotein (orosmucoid) a nd tra nsferrin , offers any advantages over C-reactive pro tein in terms of rapidity of change, ease of technical meas urement or stability in the serum afte r death. A C-reactive protein level that is greater than 0.8 mg/dL is good indica tion of a pre-existing inflammatory state. The higher the valu e, t he greater the chance of a bacterial infection, alth ough elevations in C-reactive protein should always be considered quite non-specific. A number of studies have confirmed the relative stability of this protein in the imme di ate post-mortem period. )4,)5 A number of recent investi gations have also confirmed its value.)6,)7 A great deal of recent work, from Europe in general and France and Germany in particular, has focused on the use of pro calcitonin as a marker of bacterial infection .)8,)9 There is as yet no consensus of opinion on whether this is more useful t han C-reactive protein. Its behaviour in post mortem specimens has recently been studied,4o and some of the changes associated with tempera ture and storage have also been documented .41
ANAPHYLAXIS An acute anaphylactic reaction is a possible mechanism of death in a small minority of paediatric cases. There is usually a past history of severe hyp ersensitivity to a particul ar all er gen, such as peanuts or shellfish. However, bee venom and other allergens may also be responsible. Generally the clini cal features of overwhelming anaphylaxis are evident, although in some cases they can be so acute or confusing as to go unrecognized. A number of tests have been promoted for the diagnosis of anap hylaxis. Histamine is released from both tissue mast cells and the circulating basophils. It has a very short half-life and so is of no use in the post-mortem diagnosis of anaphylaxis. Of the other tests, serum tryptase, released from degranul ating mast celis, has attracted the most attention. This enzyme is not released from the circulating basophils. A number of studies in living patients, involvi ng cases of witnessed anaphylaxis, have demonstrated raised levels of serum tryptase in association with an increase in
110 I
Biochemical investigations on post-mortem specimens
allergen-specific immunoglobulin E (IgE). In the early studies of post-mortem specimens the resuHs appeared quite encour aging; however, these involved only a small number of cases. 42.43 Subsequent evaluation in a group of 49 patients known to have died from other causes, with no evidence of anaphyla xis whatsoever, showed that many samples had levels of tryptase above the usual detectable levels in the serum.44 Of these, five subjects had levels of tryptase that were greater than 10 ng/mL, which was regarded as a positive result, and one had an extremely high value. The mechanisms for these increases were quite unclear. There was no correla tion with the post-mortem interval. Further studies have demonstrated raised levels in a variety of circumstances. 45 At this time it is advisable to utilize both the senlm tryptase level and an increase in the allergen-specific IgE for the diagnosis of anaphy laxis. A recent study concluded that as many of 13 per cent of deaths in adults may be accompan ied by the activation of mast cells 46 There has been one formal investigation into the lise of tryptase in sudden unexpected deaths in infants. This study suggested that mast cell degranulation was more evident in infants in t he SIDS groups than in the infants in whom a cause of death was found. 47 This was a small study that has not been corroborated in a systematic survey involving a larger number of control subjects, with careful attention to the post-mortem interval and mechanism of death; this issue has yet to be resolved. 48.49
INFECTION If there is any suspicion of infection then the appropriate cul tures should be obtained. The problems of interpreting post mortem bacterial cultures are well known , eve n if standard techniques are followed. Culturing multiple sites may be use ful or separately sampling each ventricle of the heart might provide additional infomlation that assists in the interpreta tion of positive cultures. 5o Viral cultures can be very inform ative and the use of new polymerase chain reaction (PCR) assays for the detection of many human pathogens is becom ing more widespread. These should always be considered in cases of possible myocarditis and meningit is. Over the years we have been struck by the poor correlation between the gross appearance of the heart and the microscopic presence of significant myocarditis. For this reason we tend to err on the side of conservatism and obtain viral studies in most cases of unexpected death. The inherent sensitivity of these molecular diagnostic techniques raises the difficult issue of the clinical significance of a positive finding. We are only just beginning to learn about our 'normal' microbial flora as defined by PCR methodology. Assays for specific bacterial tox ins are described and have been quite useful in selected cases. These are not gen erally available in most clinical laboratories. If a particular infection is suspected from the gross and microscopic find ings then the pathologist should seek out the appropriate
laboratory for assistance in the diagnosis. For example, toxic shock syndrome has been identified as a cause of unexpected death in a child , with the identification of the staphylococcal toxin TTST-1 in brain tissue. 51 The presence of bacterial endotoxin has been sought in a variety of clinical situations. The usual test is based on the limulus lysate assay, of wh ich there are a number of modifications. The level of endotoxin is not significantly affected by the post-moJiem interval in the first 48 hours after death, nor does it correlate with the blood culture results. Detectable endotoxin has been noted in a many dif ferent causes of death, including blunt injury as well as severe infection. However, it is interesting to note that it has not been found in a study of SIDS victims. 52
DEHYDRATION AND ELECTROLYTES The question of possible dehydration is a frequent consider ation , especially in paediatrics when the caregivers may underestimate the extent and severity of vomiting/diarrhoea and there is inadequate fluid replacement. Dehydration develops more rapidly in infants and young children than it does in adults and is often unrecognized by the parents. The chemical diagnosis of dehydration and the accompanying electrolyte abnormalities have been well studied. The ability of cells to maintain their normal concentration gradients of ions is highly energy dependent and this rapid ly decreases after death. Within minutes, intracellular ions, such as potas sium ions, begin to equilibrate across the cell membrane. After an hour the serum potassium may well be six times the ante-mortem level, even in the absence of haemolysis. Once haemolysis develops, the pot assium concentration soars even higher. Therefore, serum potassium should n ever be used to assess the state of dehydra t ion as it increases geometrically after death. The appearance of haemolysis is quite variable, depending on the temperature and the storage of the body. It can take as long as 48 hours for haemolysis to be evident. Serum sodium concentrations' tend to decrease more slowly after death compared with the rapid and very significant changes with potassium. Sodium decreases at a rate of approximately 0.9 mEq/L per hour after death;5J chloride decreases as well , with an average fall of 0.97 mEq/L per hour. However, there is so much vari ation in sodium and chloride ion levels from among subjects that they have not been useful for the assessment of post-mortem interval. In addition, serum is not the ideal body fluid for the assessment of dehydration unless blood can be obtained very soon after death. Sodium and chloride, as well as osmolality, are much more stable in the vitreous humour. Almost all studies of electrolytes have relied on the measurement of their con centrations in v it reous humour,2J, 54,55 which are similar in both adults and children, 56- 58 The sodium and chloride levels in the vitreous humour change in parallel with tbe serum and every study has shown very good correlation
Time of death (post-mortem interval) I
between the known concentration of these during the ter minal hours of life and the post-mortem values in the vitreous humour. The levels of these ions are the most reliable for the assessment of hypernatraemic dehydration. Measured with ion-se lective electrodes, concentrations of sodium of > 165mEq/L and chloride >125mEq/L are excellent indi cations of hypernatraemic dehydration. The blood urea nitrogen (BUN) may also be significantly ele vated. If renal failure has developed, the creatinine will also be increased. Although severe dehydration is the usual cause of hypernatraemia, either as a result of a disease such as gastroenteritis, or neglect by the caregivers, other possi bilities must be kept in mind. Excess loss of water may occur through the skin and lungs in infants and children with fever, or there may be excessive renal excretion, such as occurs in diabetes insipidus (both central and nephro genic), osmotic diuresis (e.g. chronic renal failure, hyper glycaemia) or hypercalcaemia. Both accidental salt poisoning and inte ntional salt poisoning have also been described in chi ldren. 59 Claims that a significant fraction of infants dying of SIDS have significa nt electrolyte distur bance based on vitreous chemistry60 are almost certainly overestimated. In certain cases of dehydration, when fluid witho ut suf ficient electrolytes is given to the patient to drink, such as water or apple juice, hyponatraemic and hypochloraemic dehydration can occur (e.g. excessive vomiting as might accompany pyloric stenosis) .61 The vitreous sodium is generally less than 135 mEq/L and the chloride is less than 95 mEq/L. Hyponatra emic dehydration is much less fre quent than the hypernatraem ic form. Both adrenal insuffi ciency (see below) and salt-losing nephritis are causes of hyponatraemia in the dehydrated patient. Other causes of hyponatraemia and hypochloraemia should also be consid ered, including lethal water intoxication due to intentional, forced water ingestion. In the oedemato us patient, low sodium frequently accompanies congestive cardiac failure, hepatic cirrhosis and the nephrotic syndrome 6 2 However, it must be remembered that not all cases of dehydration are accompanied by significant electrolyte disturbances. Coe J,6J has described four basic patterns of electrolyte abnormalities in the vitreous humour in forensic cases in which they are diagnostically helpful: 1. The hypertonic dehydration pattern, which is the common pattern in most cases of severe dehydration leading to an unexpected death. In this state, both sodium and chloride are elevated, usu ally to a similar degree, and the urea nitrogen may be mildly to moderately increased. 2. The uraemic pattern, in which sodium and chloride are relatively normal, while both the creatinine and the urea nitrogen are elevated (see below). 3. The hypotonic or low-salt pattern, in which both the sodium and chloride are low, as is the potassium concentration; the potassium concentration should be less than 15 mEq/L.
111
4. The decomposition pattern, in which the sodium and chloride are both low but the potassium is markedly elevated (>20 mEq/L). As can be deduced, the hypotonic and decomposition patterns are only distinguished by the potassium concen tration. For this reason it is important to measure the vitre ous potassium level as part of any electrolyte panel in order to separate these t\ovo pathogenetic mechanisms, Vitreous potassium has been well studied, as it shows a linear increase after death and has been utilized for the assess ment of the post-mortem interval 64 (see below). Coe and Apple 65 have studied the influence of the method of measurement on the values of vitreous elec trolytes and showed very sizeable differences between tra ditional flame photometry and ion selective electrodes. It is very important therefore for each laboratory performing these assays to establish their own unique reference ranges, and fo r the pathologist to be knowledgeable about the methodology used for satisfactory interpretation. Studies on the osmolality of vitreous fluid are quite lim ited; in normal subjects it has been repolted to range from 305 mOsm/kg 66 to 346 mOsm/kg.67 No systematic studies of v itreous osmolality in the post-mortem period have been performed in children covering a wide enough spectrum of pathologies. The anion gap, a useful measurement during life for the interpretation of electrolyte and acid-base dis turbances, plays no role in post-mortem chemistry,
TIME OF DEATH (POST-MORTEM INTERVAL) Over the years, in addition to all the gross pathological evi dence and the temperature of various body organs, other methods have been sought to more accurately define the time of death. 68 .69 No other question, apart from the cause of death, has provoked so many studies or spilled so much ink, The answer to this forensic question is often crucial to the entire case. Tests on blood constituents have largely been abandoned in favour of a few ana lytes in vitreo us humour. These have included such compo unds as inorganic phosphorus,7o amino nitrogen, non-protein nitrogen, ammonia and creatinine. 7I Levels of amino acids also rise sharply after death, although there is considerable varia tion between the various amino acids in terms of the rate at which their levels rise. None of these analytes has been proven to be sufficiently rei iabl e or reproducible for the estimation of the post-mortem interval on a routine basis. In the last tvvo decades, potassium and hypo xanthine have gained the most favour for the assessment of the post mortem intervaI. 64 ,72,7] In the uncomplicated adult case, both the potassium and hypoxanthine concentrations in vitreous humour, show a linear increase in the early post-mortem period. 74,75 However, the rate of change after the first 24 hours is different from the initial rate and this is seldom taken into account. It has been suggested that hypoxa nthine shows
112 I
Biochemical investigations on post-mortem specimens
less varia tion in the early post-mortem period, especially in the first 12 hours. 76 A variety of studies have investigated the relationship of these two analytes to the post -mol1em interval and the effect of temperature. Also, equations have been con structed to estimate the time of death, ordinarily based on regress ion analyses from cases in which the time of death was precisely known 77 The most recent example suggests the use of both measurements 78 and the formulae constructed from this study are as follows : post-m0l1em interval (hours) = 4.32 X (potassium [mmol/Lll - 18.35; post-mortem interval (hours) = 0.31 X (hypoxanthine [mmol/Lll + 0.05. The final estimate is based on the mean of these two calculations. Needless to say there are a variety of methodological and conceptual problems with these estimations. Recent refinements in the formulae have not been systematically tested for routine practice in paediatrics. They are far from precise, with considerable variation from case to case. They have not replaced body temperature in the early post mortem period . Their use in adults shou ld be considered once the body reaches ambient temperature. The rate of increase in both ana lytes is not perfectly linear over time and this has been ignored in all the regression equations published to date. Moreover, the mechanism of death and the length and extent of ante-mol1em hypoxi a have not always been well controlled for in the study popu lations. The environmental temperature at wh ich the body remains after death has a major impact on the rate of change of potassium. Finally, the number of studies that have included paediatric cases is very limited. There is evi dence that potassium increases much more rapidly in the vitreous humour of the infant than in that of the adu]t. 3 ,58.79 Some of the problems associated with hypoxanthine are discussed above in the section on hypoxia .
ENDOCRINE DISORDERS Diabetes lVIellitus GLUCOSE Estimation of the glucose level is important for the detec tion of a number of disorders. It is obviously critical for the determination of uncontrolled or untreated di abetes melli tus, during which the severe hyperglycaemia is usually accompanied by significant ketoacidosis. When decreased, glucose is one of the key metabolites indicative of a possi ble metabolic disease, or may be associated with starvation, abuse and neglect. Glucose measurements have been well investigated after death. The specimen type and its proper selection is crucial for evaluation. If blood is taken from the right side of the heal1 or the inferior vena cava, glucose lev els may be very high. This is a result of hepatic glycogenol ysis. With time, the glucose diffuses into the surrounding vessels. On the other hand, very low levels of glucose from
such a sample maybe indicative of hypoglycaemia, depend ing on the length of the post-mortem interval. Note, how ever, that it has been well shown that high levels of glucose may occur in peripheral blood sampled from a distal extremity even in patients who are not diabetic.8o .8 ! This has been observed in a variety of deaths due to different etiolo gies and may be the result of a terminal surge of noradren aline (norepinephrine) and adrenaline (epinephrine) and/or the iatrogenic effects of resuscitation. 82 In interpreting the glucose level, the terminal therapy received by the patient should be carefully scrutinized, including all the adminis tered drugs and the intravenous fiuids. The use of car diopulmonary resuscitation outside the hospital has risen dramatically in many Western countries, often being administered by paramedical personnel, and such interven tion can significantly impact the glucose values. In some communities it is the rare patient who avoids the chemical 'last rites'. Some investigators, such as Coe,] feel very strongly that a diagnosis of diabetes mellitus should never be based exclu sively on a blood glucose measurement. The level of glucose in the CSF (rather than in blood) is more reliable as it is less affected by the terminal conditions that can so dramatically infiuence the serum. On the other hand, the increasing use of vitreous fiuid has largely supplanted CSF as the sample of choice for the diagnosis of hyperglycaemia. In the normal situation, the vitreous humour glucose level falls after death due to continuing glycolysis. In most uncontrolled diabetics the glucose will remain above 200 mg/dL. Ketones invariably accompany the hypergly caemia and can be detected by the standard qualitative lab oratory procedures in the vitreous fiuid (or peripheral blood). Diabetes has been diagnosed in embalmed cadavers.8] Some have recommended the use of both lactate and glucose measurements on the vitreous humour, in addition to the measurement of ketone bodies in the serum, such as acetone, for the diagnosis of diabetes. 82 However, a cl ose examina tion of the data shows no significant benefit of the addi tional lactate measurement. The circumstances and environm enta l tempe rature around the time of death may significantly infiuence the rate of change of the vitreous glu cose. Rapid cooling, as might occur in drowning (cold water immersion) reduces the rate of glycolysis, such that the glu cose may be higher for a given post-mol1em in terval. This may be compounded by the stress response accompanying such deaths. 84 ,8 5 In recent years, glycosylated haemoglobin (haemoglobin A Ic) has been used as a reliable marker of long-stand ing hyperglycaemia. It is stable in the post-mol1em blood and is a refiection of the average blood glucose leve l over the pre ceding 3 months. It is important to recall that there may be other mechanisms for long-standing hyperglycaemia. For exa mple, one study of 28 cadavers demonstrated raised haemoglobin A Ic in all five patients with known diabetes. 86 In addition, seven subjects with malignancies, who had been receiving steroid therapy, also had elevated levels. Other
Endocrine disorders I
studies have confirmed the value of measuring glycated haemoglobin as an assessment of hyperglycemia. B7 - B9 Fructosamine, which is a measure of non-labile glycated serum proteins, has also been used during life to monitor the effectiveness of diabetic therapy. It too is an indication of the mean blood sugar level over the preceding weeks. One of the advantages of fructosamine is that it is influ enced neither by the presence of any abnormal haemoglo bins, nor the dynamics of red cell turnover in patients who may have haematological problems. However, there are only a few post-mortem studies on fructosamine. 90. 91 Although sudden unexpected death in childhood due to undiagnosed diabetes mellitus is very rare, it has been reported. 92 •93 In one particular case an ll-year-old girl died suddenly after a 2-day history of a minor respiratory tract infection. The autopsy was performed 3 hours after death and documented a vitreous humour glucose level of 606 mg/dL, blood acetone level of 24 mg/dL, marked glucosuria (2073 mg/dL) and positive urinary ketones. Her liver con tained microvesicular fat and the pancreas showed typical changes in the islets. 94 The combination of an elevated glu cose in the vitreous humour and acetone in the blood has proven to be quite reliable for the diagnosis of unexpected diabetes mellitus. Sudden death has also been well described in children who have been recently diagnosed and are under going initial therapy. The diagnosis of hypoglycaemia in post-mortem sam ples is much more problematic than the detection of hyper glycaemia. This remains a serious problem in paediatric forensic cases as hypoglycaemia is such an important clue for a variety of metabolic diseases. Ongoing glycolysis generally results in a progressive decrease in glucose level in all body fluids, including CSF and vitreous humour. However, the dynamics of this decrease are quite unpre dictable. Attempts to combine glucose measurements with lactate and other metabolites have not proven to be satis factory for routine diagnosis. At this time there is no acceptable method for the detection of hypoglycaemia.
KETO NES The detection of ketone bodies (acetone, acetoacetic acid and (3-hydroxybutyric acid) is important for the diagnosis of a variety of conditions. Infants and children have a ten dency to generate ketone bodies more readily than adults, especially after a relatively short period of fasting. Children shift their energy production to utilize fat quite promptly, generating the ketone bodies in the process. The dipstick test, which is based on the reaction of acetone and ace toacetic acid with nitroprusside in an alkaline buffer to produce a purple colour, is quite sensitive. detecting as few as 5 mg/dL in the urine. Unfortunately, the detection of ketones in body fluids is rather non-specific and should be interpreted with caution. It accompanies starvation/fasting, vomiting and exercise, as well as any impairment of carbo hydrate metabolism, most notably diabetes mellitus in
----=-- --
-
.'
113
which it is associated with hyperglycaemia. On the other hand, the finding of ketosis with hypoglycaemia is seen in some of the glycogen storage disorders, such as Von Gierke's disease. Perhaps even more important than the finding of ketones in an infant or child with sudden death is the absence of ketosis, especially if there is any evidence of hypoglycaemia. The lack of ketones should trigger the investigations for a possible (3-fatty acid oxidation disor der. Hypoglycemia with negative ketones is also present in nesidioblastosis and other states of hyperinsulinism.
Addison's Disease: Adrenal Insufficiency Sudden death in infants and children as a result of an Addisonian Gisis is well described. 95 - 99 It may be due to congenital adrenal hypoplasia or destruction of the adrenal glands as a result of a number of different pathological processes. These latter cases are more likely to occur in the childhood and teenage years, 100 whereas the congenital forms generally are present in infancy and may mimic SlDS. JOI The gross and microscopic appearance of the adrenal glands normally establishes the diagnosis. The usual meta bolic derangement accompanying an Addisonian crisis is usually reflected by low sodium and chloride concentration in the vitreous humour and very elevated potassium, often in the range of the 'decomposition' pattern described above. As salt wasting and dehydration are frequent fea tures, levels of urea nitrogen, and even creatinine, may be elevated. If urine is available, the elevated sodium level should be evident. This combination of dehydration , hyponatraemia and an elevated level of sodium in the urine is generally indicative of adrenal insufficiency or salt losing nephritis. If both the adrenal glands and kidneys are normal then one of the many causes of inappropriate secretion of antidiuretic hormone should be entertained. Random cortisol values will be markedly decreased in case of adrenal insufficiency, regardless of the time of day. Cor tisol concentration in the blood is relatively stable for up to 18 hours after death, especially if the body is cooled within 4 hours of death.102 Reference values are method, age and sex dependent, so that interpretation depends on compari son with the specific laboratory 's reference values.
Renal Failure Renal failure in childhood may be quite occult, with few clinical indications. Sometimes the family or the attending physicians may not readily detect the growth failure asso ciated with declining renal function. Unexpected death in such situations is quite uncommon but can occur. Gener ally the gross and/or microscopic appearance of the kid neys and the urinary system will establish the diagnosis. The severity of the renal impairment can also be gauged by
114 I
Biochemical investigations on post-mortem specimens
measuring the concentration of urea nitrogen and creati nine in the blood and vitreous humour. Urea nitrogen is remarkably stable in these body fluids 56 ,IOJ-)05 and the diagnosis of renal failure can readily be established. IOG , IU7 Of all analytes, urea nitrogen is perhaps the most stable, showing no diagnostically significant changes over a period of up to 5 days post-mortem. J08 Creatinine measurements in both blood and vitreous humour are also quite stable,109 although some investiga tors have described modest increases in creatinine after 72-96 hours. 106,107 Levels in the vitreous humour are slightly lower than serum values 54 but parallel the changes that occur in renal fai lure. In well-established renal failure, both the urea nitrogen and creatinine levels will be markedly elevated. In states of acute dehydration and pre renal failure, the level of creatinine will be normal and the renal nitrogen will be elevated. The ratio of urea nitrogen to creatinine is a useful estimate of prerenal ver sus renal failure.
GENETIC METABOLIC DISORDERS PRESENTING AS SUDDEN UNEXPECTED DEATH The list of inborn errors of metabolism presenting as sud den unexpected death continues to expand. These can be broadly grouped into two large categories. The first group involves enzyme defects leading to an increase of inter mediary metabolites proximal to the defect. One or more of these compounds becomes toxic as they accumulate, pro ducing the clinical symptoms or unexpected death during an acute metabolic event. Examples in this category include many of the organic acidurias, aminoacidopathies, urea cycle defects and carbohydrate intolerances. The sec ond category involves energy metabolism, based primarily in the mitochondrion. The major organs affected by these defects are the liver, skeletal muscle, heart and central nervous system (CNS). A number of metabolic pathways have been incriminated, including fatty acid oxidation, mitochondrial transport and metabolism , glucose and glycogen metabolism, and the lactic acidaemias. A variety of clues shou ld be sought to help point the way to the most appropriate work-up, as it is seldom possibl e to investigate all the possible alternatives. Table 5.3 is helpful in directing the physician and the pathologists to the appropriate cate gory for further work-up of infants who present with a n acute ill ness in infancy. This is a very heterogeneous group of disorders, so it is to be exp ected that there are a variety of clinical phenotypes and presentations. While this chap ter focuses on the rare event of a sudden unexpected death in an infant, the vast majority of cases do present during life with one of the more frequent clinical manifestations. During life the common laboratory signs suggestive of an acute metabolic decompensation associated with some of the inborn errors of metabolism include hypoglycaemia ,
an increased anion gap (due the accumulation of unmea sured acids), hyperammonaemia, ketonuria, lactic acidosis or a respiratory alkalosis. Unfoliunately, neither ammonia nor lactic acid measurements are reliable after death and cannot be used. They tend to increase in both the serum and the vitreous humour in a fairly lin ear fashion in the first 24 hours, but there are no systematic studies of post mortem ammonia levels in infants and children evaluating their role in diagnosis. Organic acid ana lysis in the urine or v itreous humour is valid after death. Many changes occur in the plasma amino acids (including an elevation of glut amine and alanine) and the neurotransmitters GABA and (1-alanine. This makes post-mortem amino acid measure ments difficult to interpret. Table 5.4 lists a collectio n of metabolic diseases that are associated with 'energy production' and that have reportedly caused sudden unexpected death. This list will undoubtedly grow as defects in each step of the complex pathways leading to the generation of adenosine triphosphate are identified.
Fatty Acid Oxidation Defects Tn the past couple of decades, a large num ber of defects involving the 0 - oxidation of fatty acids have been recog nized. lll ,116-119 Some of these can be entirely asymptomatic and present with unexpected deaths in infancy, usually after a year of age. Some have occurred earlier under clin ical circumstances that closely mimic SIDS. 79 ,120- 132 It is the unusual case that is truly confused with a classic or typical SlDS. In most instances eli nical clues suggest that the infant was not quite normal prior to death. Often the death is later than the usual 2- to 6-month range of most SIDS babies. Clinical indicators include lethargy, hypotonia and an acrid odour to the breath (likened to the pungency of 'smelly feet'). There is enough overlap between classical SIDS and some case of acute metabolic decompensation to justify consideration of this group of disorders always. Certainly, any sudden unexpected death under the age of 5 years should raise the suspicion of a fatty acid oxidation defect. The reported incidence of metabolic disorders contributing to the number of SIDS v ictims or sudden unexpected death after the age of 1 year varies very widely in the available literature. The true incidence is unknown b ecause there have been no satisfactory, systematic and comprehensive studies with a large enough number of cases. Moreover, metabolic defects are continuously being uncovered and few studies have investigated the really wide spectrum of currently known possibilities. Acquisition bias afflicts some of the studies, as does poor case selection and widely valying definitions of what constitutes a diag nosis. Having said all this there is no doubt tbat metabolic diseases can and do result in sudden unexpected deaths in infants and children who were otherwise fe lt to be grow ing, feeding and developing normally.
Genetic metabolic disorders presenting as sudden unexpected death I able 5.3
115
Genetic metabolic disorders which may present acutely during infancy
Acute presentation in infancy
Major disorder
Disease example
Confirming genetic test
MSUD NKH glycinemia Tyrosina em ia-I
Plasma AAs Organic acids Tissue enzymes
NH 3t
GTC Argininosuccinate Citrullinaemia
Plasma/urin e AAs Orotic acid Tissue enzym es
acidosis NH3t (7) urine ketones
Methylmalonic Propio nate/isova leric Glutaric-II
Organic acids Tissue enzym es
'Iytes
Renal loss GI loss
Plasma/urine pH Et HC0 3
Lactic acidosis mitochondria
lactate t
PDH def Mitochondrial S MERRF, MELAS
Pyruvate/lactate Tissue enzymes Mu scle bx, DNA
Glycogen storag e
glucose ...
GSD-I Et II
Glucose/lactate Chol/trig/uric a. Liver bx, DNA
urine ketones ...
MCAD LCAD
Organic acids DNA for MCAD, LCHAD Tissue enzymes
Galactosaemia
RBC gal-I-PUT DNA genotype
Zellweger syndrome Pseudo-Zellweger syndro me
X-rays VLCFA/plasma log ens Liver bx Tissue enzymes
GM-l gangliosidosis MPS disorders
X-rays Tissu e enzymes MPS excretion Oligosaccharides EM of skin
(l ethargy, coma, acidosis, hepatomegaly)
Amino acid disorder
Rule out: • hypog Iycaem ia • sepsis • CNS bleed • congestive heart • toxins
Urea cycle disorder
Organic acid disorder
HC0 3 loss
Physical exam: • norma I phenotype • enlarged liver/spleen • dysmorphic
Fatty acid oxidation
Galactosaemia
Laboratory test: CBC: Urine: 'Iytes pH glucose spec. grav. lactate ketones NH 3
Abnormal clinical test
+ urine-reducing
Peroxi so mal disease
Lysosoma l disease
substance
central nervous system; 'Iytes, electrolytes; MSUD, maple syrup urine disease; NKH, non-ketotic hyperglycinaemia; OTC, ornithine transcarbamilase; GI, gastrointestinal; Glutaric-II, glutaric aciduria type II; PDH def, pyruvate dehydrogenase deficiency; MERRF, myoclonic epilepsy with ragged red fibres; MELAS, mitochondrial encephalomyopathy, lactic acidosis, stroke-like episodes; GSD, glycogen storage disease; MCAD, medium chain acyl-CoA deficiency; LCAD, long-chain acyl-CoA deficiency; LCHAD, long-chain 3-hydroxyacyl-CoA-dehydrogenase; S, syndrome; MPS, mucopolysaccharide; AAs, amino acids; RBC, red blood cell; bx, biopsy; VLCFA. very long cha in fatty acids ; EM, electron microscopy. CI~S,
A minor illness, a short peliod of fas ting, an immuniza or other stress might trigger the acute metabolic decom pensation. There are a number of rather ordinary situations in which physiological fasting might occur and not be obvi ous to the parents. The two classic examples occur at the time of weaning and when an infant begins to sleep through ~he night. It is not unco mmon for in fants to present with sy mptoms at such times and it is useful to specifically ~equest this history from the family. In a recent case of car :1itine transporter deficiency, it was surmised that the fasting stress caused by poor breast-feeding without any supple mentalY formula feedings, and possibly the vegetarian diet ~on
of the mother, might have triggered the metabolic crisis leading to a neonatal death.11 4 The biochemistry of this group of disorders is complex and incompletely understood. It involves the transfer of fatty acids into the mitochondria via the carnitine transport mechanism and their oxidation to produ ce adenosine triphosphate (ATP). The oxidation of fa tty acids is an impor tant energy source in many organs, apart from the brain, but especially the heart and skeletal muscles. In the normal situation, fatty acids enter the cell, are activated to their acyl-eoA esters, linked to carnitine by the action of carni tine palmitoyl transferase 1 and II, and then translocated
116 I
Biochemical investigations on post-mortem specimens Table 5.4
Disorders of eneigy metabolism
Disorder
Reference(s}
Medium-chain acyl-CoAdeficiency Long-chain acyl-CoA deficiency Short-chain acyl-CoA deficiency Long-chain 3-hydroxyacyl-CoA dehydrogenase deficiency Carnitine palmitoyl CoA tranferase II defect Mitochondrial acyl-carnitine/carnitine translocase defect Cellular carnitine uptake defect Trifunctional protein defect Multiple acyl-CoA dehydrogenation defects (glutaric aciduria type II) Carnitine transporter deficiency Electron transport flavoprotein subunit ex defect Electron transport flavoprotein subunit ~ deficiency Electron transport flavoprotein-OQ deficiency Mitochondrial ATPase Cytochrome oxidase deficiency
across the inner mitochondrial membrane into the mito chondrial matrix. The carnitine is removed and the acyl CoA fatty acid then enters the 0-oxidation cycle. Repetitive rounds of dehydrogenation by a number of different dehy drogenases (short, medium and long chain) depending on the length of the fa tty acid, convert them to acetyl-eoA. This then enters the respiratory (electron) chain that gener ates the ATP. The diagnosis of abnormalities in 0-fatty acid oxidation, which may be due to a defect at any of the many points in the process, usually relies on the measurement of organic acids in the urine, carnitine in the serum and the profile of acyl-camitines. The in terp retation of the organic acid pro files on the urine as well as the acyl-carnitine profile requires considerable experience and skill, and this is beyond the scope of this chapter; those interested in pursuing the topic further should consult the suggested references. 13J-lJ5 Depending on the results, more definitive tests involve the measurement of specific enzyme activities or direct detection of a gene defect. The use of the serum measurement of dodecanoic acid was initially promoted as a sensitive and specific screening test for the common fatty oxidation defects. JJ6 Soon after the original publication, issues around its specificity sur faced and it was reported to be much less specific than originally hoped. 137 Until this issue is resolved with addi tional studies and a satisfactory cut-off value estab lished, this method of screening cannot be relied upon. However, it may alert the pathologist to those cases need ing addi tional work-up. One of the most commonly identified disorders in this group is medium chain acyl-CoA dehydrogenase deficiency (MeAD). This is inherited as an au tosomal recessive, and the most frequent mutation accounting for 90 per cent of cases
110-112 111 111 111 111,113
111 111
111 114
111 111 111 115
is the G958 allele. JJ8, 1J9 Medium chain acyl- CoA dehydro genase defiCiency accounts for the majority of cases of sud den death due to a defect in fatty acid oxidation. It usually presents during infancy with lethargy, a Reye-like syn drome, accompanied by hypoketotic hypoglycaemia. As many as 40 per cent of infants are said to die during their first episode. Generally, the liver shows considerable fatty change, although there is great variation from case to case. Moreover, lipid accumulation in the liver is quite non specific. Although much less common, the mutation for long ch ain 3-hydroxacyl-CoA dehydrogenase deficiency is also well describ ed. 140 Tests for the diagnosis of 0-fatty acid oxidation defects include: J. Urine: organic acids by gas chromatography/mass spectroscopy to identify specific glycine conjugates; urine acyl-carnitine profile to identify the specific acyl-carnitine conjugates. If urine is not avai lable then consider collecting either bile or vitreous humour. Both bile l41 and vitreous humorlJl have been used successfully to establish a diagnosis. 2. Liver: enzyme analysis for both medium- and long chain acyl-eoA dehydrogenase activity. 142 Immediate freezing of tiss ues greatly retards the post-mortem decrease in activity. However, the short-long chain ratio remains constant so that the difference between a true enzyme deficiency and post-mortem degradation, in which both values will be decreased, can be distinguished. An alternative approach to enzyme activity is the analysis of various metabolites obtained from the liver after methanol extraction and ultracentrifugation. 14J This has been successfully utilized for the detection of a number of the defects, even in livers that were only obtained 72 hours after
Technical considerations at the time of autopsy death. This technique assays a large number of the farry aci d metabolites, together with glucose and lactate. The pattern of metabolites present can sometimes establish a specific diagnosis. J. D A: medium-chain acyl G985 mutation - a PCR assay is available;144 coenzyme A dehydrogenase de ficiency (MCAD) - the mutation has been detected in D A obtained from a variety of different sources, including frozen blood, llOdried blood spots,145 archived paraffin embedded material, 146 and liver. 147 Detection of the mutation associated with long-chain J-hydroxyacyl CoA dehydrogenase (G 1528C) is also possible. 14o Newborn screening cards are an excellent source for this testing. Many DNA tests can be carried out directly from the newborn screening blood spots, as can acyl-carnitine profiling. ~. Cultu red fibroblasts: globa.l [9,1O- 3Hl myristic acid oxidation assay - detects at least nine different fatty acid oxidation defects;117 specific enzyme analysis.
Other Genetic Metabolic Disorders number of disorders involving amino acid metabolism may cause unexpected death, especially in the first 6 months of life. Some of these are also associated with the ab normal excretion of certa in organic acids. Of these, maple syrup urine disease is an excellent example. 148 Table 5.5 provides a list of potential enzym e deficiencies that might cause sudden death; as will be noted, these involve 'ery disparate metabolic pathways. Some of the disorders may be classifIed as both an organic acidaemia and an a minoacidopathy. It must also be pointed out, however, that this list has been comp iled from a variety of sources, ~e \V of which include the clinical or pathological details of lhe individual conditions or cases.ll8.149-151 Finding the pri mary documentation for all these claims has proved quite difficult. In reality, sudden unexpected death is a very ncommon manifestation of very rare diseases. Almost all cases would have been suspected or diagnosed during the patient's life. :'l,
TECHNICAL CONSIDERATIONS AT THE TIME OF AUTOPSY The following recommendations have been compiled from a 'ariety of sources. 3, 128.1 49.154 The extent and thoroughness of ampling will depend on available resources, the relationship between the forensic pathologist and the reference biochem :cal laboratory and the circumstances surrounding the autopsy. Although some investigators recommend a stan dard protocol for all unexplained infant deaths this has gen erally been limited to research investigations, especially I .. hen there is specific expertise and interest in many of the uncommon inborn errors of metabolism.
I
117
Table 5.5 Other genetic metabolic disorders claimed to be associated with sudden death Carbohydrate metabolism Galacatosaemia Hereditary fructose intolerance Glycogen storage disease, type" Fructose-l ,6-diphosphatase deficiency Urea cycle defects Carbamylphosphate synthetase Ornithine carbamyltransferase Argininosuccinate synthetase Argininosuccinate lyase Amino acids Tyrosinaemia Isovaleric acidaemia Non-ketotic hyperglycinaemia Maple syrup urine disease Organic acidaemias Propionic acidaemia Methylmalonic acidaemia 3-Hydroxy-2-methylbu tyric aciduria Glutaconic acidaemia Isovaleric acidaemia Multiple carboxylase deficiency Glutaric acidu ria type II Miscellaneaus Lysine protein intolerance
Biotinodase deficiency (ref. 152)
Glycerol kinase
Menkes' disease (ref. 153)
Vitreous Humour Vitreous humour is obtained by puncturing the eye at the outer ca nthus, using a 7- to 20-gauge needle and a small syringe. Because of the viscous nature of the fluid, the fluid flows slowly. Gentle suction should be applied, accompanied by a large measure of patience. Vacuum tubes should not be used as they generate too much force. An effort should be made to collect all the avai lable fluid, which may be as little as 0.5 mL in a neonate or as much as 3 mL in the young adult. Fortunately, technology today allows a number of measurements on these small volumes. The vitreous must be centrifuged and the supernatant is then used for the analy sis. The fluid should be perfectly clear, with no evidence of clouding or discolouration. No significant differences have been shown in samples drawn simultaneously from each eye, although this contention has been challenged. IS S Any pre-ex isting eye disease that affects the vitreous humour might confound the results. Fortunately, such diseases are extremely rare in paediatrics. The recommendation to
118 I
Biochemical investigations on post-mortem specimens
remove all the fluid is based on some expelimentaJ data that demonstrated regional differences in the concentration of certain analytes, Because of this, repetitive sampling of the vitreous humour is not possible, Replacing the vitreous humour with a volume of water or saline similar to the amount aspirated can restore the shape of the eyeball by maintaining its turgidity, a courtesy generally appreciated by the mortician. Vitreous humour can be used for the following tests; glu cose, ketones, sodium, urea nitrogen and organic acids. The differences in levels of organic acids in the blood and vitre ous humour have been investigated. 111.156 Amino acid analy sis is also satisfactory if the vitreous humour can be obtained in the first 24 hours after death and is rapidly depro teinized. 139 With time there is a gradual decrease in the sodium and glucose concentration, whereas urea (measured as urea nitrogen) shows a slight rise. The level of magnesium is quite age dependent and also rises with time. The rate of decrease of sodium is approximately 0.5-0.65 mEq/l per hour. Calcium shows no consistent or predictable changes after death. Abnormalities in calcium metabolism have not been successfully diagnosed after death. Glucose declines at a rate of around 0.21-1.6 mg/dl/hour. Urea rises in the vit reous humour at a rate of 0.081-0.62 mg/dL/hour. 56 Bicar bonate values are usually low in the vitreous humour, which may be an artefact due to exposure of the sample to the air with loss of carbon dioxide. Such is apt to occur when a small volume of vitreous humour is placed in a large con tainer and if there is a long delay before the specimen is analysed. Other compounds that have been measured in the vitreous humour include amino acids,157 hormones l50 and lactate. 23
Blood Collection There is considerable variation in the concentration of cer tain analytes between the right and left side of the heart. 159.J 60 Moreover, the influence of post-mortem hepatic metabolism, such as glycogenesis, can affect a number of carbohydrates. As the largest body of clinical biochemical data is based on peripheral blood analysis during life, it makes the most sense to use this source whenever possible. Femoral or subclavian vein puncture will most accurately reflect the ante-mortem values. Blood should be centrifuged as soon as possible and the serum separated and stored at - 20· C or -70· C. The longer the serum is in contact with the formed blood elements, the more difficult interpretation becomes. 16J Haemolysis will affect the measurement of many analytes and should be avoided. If red cells are going to be analysed (e.g. for the presence of a haemoglobinopathy) then some blood should be anticoagulated with ethylenediaminetetraacetic acid (EDTA) in the standard fashion. Collecting blood in EDTA or heparin is always valuable, as the specimen can be immediately centrifuged and the plasma separated. Certain
laboratories prefer plasma for some of the analyses_ It is important that the laboratory's reference ranges clearly specify the type of specimen , as there may be considerable differences between whole blood, serum and plasma, Blood collected in EDTA can be used for the preparation of DNA. Blood collected in heparin or acid-citrate-dextrose (ACD) can be utilized for the preparation of white cells for the analysis of lysosomal enzymes,
Urine Collection The most convenient way to obtain urine is to perform a bladder puncture once the abdomen has been opened. This will ensure that the urine is not contaminated. In the event that the bladder is empty, attempts can be made to collect a small quantity of urine by puncturing the renal pelvises. Other techniques, such as catheterization are more com plex, especially in small infants, as the appropriate size of catheter is often not available. Expressing the urine from a full bladder is also possible, although it is more difficult to collect than a simple bladder tap. Urine should be stored in sterile plastic or clean glass tubes at -70· C, in aliquots of 1-5 ml, until analysis. Urine may not always be available. In one study of infants dying of SIDS a substantial majority (60 per cent) had no urine on opening the bladder. 132 If urine cannot be aspirated from the renal pelvis then Bennett et al 131 recom mend swabbing the surface of the bladder with a cotton swab, storing the swab at -20·C and then submitting it as the specimen for organic acid analysis . However, most laboratories will not have the expertise to process such samples effectively. In the event that the urine cannot be shipped to the ref erence laboratory in the frozen state, a few drops of 6N hydrochloric acid or chloroform can be added as a preser vative in order to prevent the growth of bacteria. However it is preferable to maintain the specimen frozen until the time of analysis. Urine can be utilized for the measurement of ketone and organic acids. If urine is not frozen rapidly then the insta bility of some of the urinary organic acids will result in the disappearance of some of the key groups, such as the oxoacids. There are many potential artefacts and pitfalls in the post-mortem analysis of urinary organic acids. The most common and serious is bacterial contamination. Freezing or inhibiting the growth with acidic compounds such as hydrochloric acid will avoid this. Bacterial over growth can render the entire analysis meaningless. The key for the successful interpretation of urinary organic acids is close communication between the pathologists and the lab oratory worker who is performing the assay. Knowledge of the infant/child's diet and medication history may be cru cial. For example, certain acids such as adipic, furoic and tartaric acid may have a dietary origin. 'Nutramigen', a common supplementary formula, may be associated with
Technical considerations at the time of autopsy I
an increase in 5-oxoproline. Infa nt feeding formulas that a re supplemented with MCT oil will result in elevations of the dicarboxylic acids: adipic, suberic and sebaric. Many pharmaceuticals profoundly affect the analysis, although this can generally be accounted for. Unfortunately, there a re only a velY few organic acid disorders that can be diag nosed from a single urine analysis, even if one pays very close attention to the pattern of additional metabolites.
Fibroblast Culture for Enzyme Analysis If there is any hint of a possible metabolic disease an attempt should be made to establish culture of fibroblasts. This is generally successful with the cell culture media available nowad ays for at least 48 hours after death. The sk in should be cleaned with an alcohol-based disinfectant, and well dried before excising a couple of2- to 4-mm J por lion of skin. These can usuaJJy be taken along one of the sta ndard autopsy incisions. The specimen must contain adequate amounts of dermis. Some investigators have used :;"chilles tendon biopsies for the source of the fibroblasts; this necessitates another incision but may decrease the bacterial overgrowth that is common in skin biopsies, even with apparently adequate disinfection. The transport media should contain antibiotics to suppress the growth of any contaminating bacteria. Generally, the biopsy in the trans port media should be maintained at room temperature prior (Q transfer into the growth media. Once fibroblast culture has been established, the cultures can be frozen and stored :n liquid nitrogen. With such appropriate storage, fibro blast culture can be re-established for many years. Should culture facilities not be immediately available, [he skin biopsy itself should be frozen at -70°C and stored in that manner. It is notable that successful culture has been established from such stored tissue. 162
Tissue for DNA Analysis In any case in which there is a suspicion of an inherited dis ease, tissue should be obtained for potential DNA analysis. If only limited facilities are available, blocks (approximately 5 g) of tissue (such as thymus, lymph nodes, spleen or liver) should be frozen at -70°C. An alternative to the direct stor age of tissue is to extract the DNA at the time of the autopsy and to store the DNA. This probably results in better quality DNA for later analysis. There are numerous standard proto cols for the preparation of DNA from tissue samples.
The 'Acute' Metabolic Autopsy
:c is not uncommon for infants and children to die after a ')fief period in hospital during which the possibility of a
119
metabolic disease is raised by the clinical and associated laboratory findings . In such circumstances, performing the autopsy immedi a tely after death can circumvent the prob lems of post-mortem change in tissue and body fluids. Autopsy consent should be obtained ante-mortem if at all possible, or immediately at the time of death to reduce any delay, and the body should be expeditiously transported to the morgue. Most children 's hospitals offer autopsy serv ices around the clock to expedite the diagnosis of such cases. Tissue and body fluids should be collected on open ing the chest and abdomen and immediately processed to limit any of the post-mortem changes. Tissue should not be allowed to sit at room temperature for any length of time. Three 5-mm cubes of liver, skeletal muscle, healt, CNS and kidney should be snap-frozen in liquid nitrogen and stored at -70°C. Urine and blood should be collected and stored as mentioned above. Very small samples (1 mm maximum) of liver, skin, CNS, healt and other organ should be fixed in glutaraldehyde or other suitable fixative for electron microscopy. Thymus/lymph node or spleen can all be used for DNA studies as indicated above. Once the tissues have been colJected and stored the autopsy can then proceed in a more considered and 'leisurely" fashion . ln many institu tions the formal autopsy may even be delayed to the next working day. However, if it is not feasible to perform an autopsy immediately after death then all is not lost. The largest sac rifices are in enzyme activity and electron microscopy. Although celtain assays, such as enzyme activity determi nations, may become impossible to interpret, considerable information can still be obtained from blood, bile, urine and vitreous humour.1 6J -1 65 As has been demonstrated for many of the fatty acid oxidation defects, analysis of post mortem liver, even as long as J days after death, can still yield helpful and diagnostic information. 14J As DNA tests for more and more of these diseases become available, it is likely tha t t he need for such expeditious sampling may diminish . Tandem mass spectroscopy, using blood spots on newborn screening cards or bile, have also expanded the potential diagnostic repertoire, as these a re quite stable a nd can be mailed to a distant laboratory for analysis. The following list shows the recommended specimens in the majority of cases in which death has occurred many hours previously. Not all of these specimens will be needed. The autopsy findings and any clinical history will s hed light on the possible aetiologies and will guide their subse quent use. • urine (freeze 1-mL aliquots at -80' C); - for organic acid analysis by gas chromatography/mass spectroscopy;
- for amino acid analysis;
- ketones;
- reducing substances;
• blood (peripheral site) ; - anticoagulated with EDTA (for haemoglobins) ; - centri fuged, and the sentm stored at -70°C;
120 I
Biochemical investigations on post-mortem specimens
• vitreous humour; - for electrolytes, osmolality, glucose, BUN,
creatinine;
• bile; - for metabolic studies; • skin or Achill es' tendon biopsy for fibroblast culture; • fro zen sections of liver, muscle, heart and kidney sta ined for fat with Oil Red 0; • spleen, thymus or lymph nod e (5-g portions) snap frozen in liquid nitrogen and stored at -70'C for DNA analysis; • l.iver, heart, skeletal muscle and CNS (5- to lO-g portions for poss ible enzyme or metabolite analysis, frozen at -70'c. 142 ,143
9
99:758- 60, 10
11 J2
14
16
17 18
20
21
22
24
25
26
29: 11 60- 3.
5 Puschel K, Lockemann U, Bartel J. Postmortem investigation of se rum myoglobin leve ls with special reference to electrical fat a liti es. Forensic Sci fnt 1995; 72:171-7. 6 Zhu Bi, Ishida K, Quan L, Tanaguchi M et al. Postmortem urin ary myoglobin leve ls with referen ce to the causes of death. Foren sic Sci Int 2001; 115: 183-8. 7 Wyatt DT, Erickson MM, Hillman RE, Hillman LS. Elevated thiamine levels in SIDS, non-SIDS and adults: postmortem artifact. J Pediatr 1904 ; 104:585-8. 8 Rachut E, Rynbrandt OJ, Doutt TW. Postmortem behavior of serum thyroxine, triiodothyronine and parathormone. J Forens ic Sci 1980; 25:67-71.
Arroyo A, Valero J, Marron T et al. Pericardial fluid postmortem: Comparative study of natura l and violent deaths. Am J Forensic Med Parhol 1998; 19:266-8. Madea B, Kreuser C, Banaschak S. Postmortem biochemica l examination of synovial fluid: a preliminary study. Forensic Sci Int 2001; 118: 29 -35. Haider M, Haider SQ. Assess ment of protein-calorie malnutrition. Ciin Chem 1984 ; 30: 1206-99. Tuten MB, Wogt S, Dasse F, Leid er Z. Utilization of prea lbumin as a nutritional parameter. J Parenteral Enteral NutI' 1985; 9:709-11.
19
23
Evans WED. The Chemistry oj Death. Springfi eld, IL: Charles C Thomas Publishers, 1963 . 2 Druid H, Holmgren P. A co mpilation of fatal and control concentrations of dru gs in postmortem femoral blood. J Forensic Sci 1997 ; 42:79-87. 3 Coe J1. Postmortem chemistry update. Emphasis on foren sic application. Am J Forensic Med Patho11993; 14 :91-117. 4 Reay DT, Insalaco SJ, Eisele JW. PostmOl1em methemogl ob in concentrations and their significance. J Forensic Sci 1984 ;
Schwarz EH, Chaslow FI, Erickson MM et al. Elevation of postmortem triiodothyronine in sudden infant death syndrome and in infants who died of other ca uses: a marker of previous health. J Pediarr 1983 ; 102 :200- 5. Edston E, Druid H, Holm gren P, Ostrom M. Postmortem measureme nts of thyroid hormones in blood and vitreous humor co mbined with histo logy. Am J Forensic Med Pathol 2001; 22:78-83.
15
REFERENCES
Benjamin DR , Beckwith JB . Elevated values of triiodothyro nine in victims of sudden infant death syndrome (Letter). J Pediatr 198 2; 100:841. Ross IS, Moffat MA, Reid IW. Thyroid hormon es in sudden infan t death syndrome. Ciin (him Acta 1983; 129:151-5. Peterson DR, Green Wi, van Belle G. Sudden infant death syndrome and hyperthiodothy roninemia: comparison of neonatal and postmortem measurements. J Pediatr 1983; 102:202 -9.
13
Cerebrospinal Fluid A number of investigators have attempted to use CSF for the measurement of a variety of compounds in an effort to mitigate some of the problems associated with blood analysis, or for a more direct determination of the changes occurring in the CNS around the time of death. The use of this body fluid has not found widespread general applica tion in most forensic laboratories, although there are certainly some clinical situations when this should be con sidered. The change in concentration of many compound s in the CSF with increasing post- mortem interval is similar to that described in the vitreous humour. Enzymes, such as creatinine kinase and lactic dehydrogenase, increase in a linear fashion over time, as do potassium and lactate, whereas glucose, pyruvate, sodium and chloride all decline,I 66
Chacon MA , Tildon JT. Elevated values of triiodothyronine in victims of sudden infant death syndrome. J Pediatr 1981;
27
28
Benjamin DR. Laborato ry tests and nutritional assessm ent. Protein-energy status. Pediarr Clin N Am 1989; 36:139-60. Maeda H, Fukita K, Oritani S et al. Evaluation of postmortem oxymetry with reference to the causes of death. Forensic Sci inr 1997; 83:201. Sturner WQ, Sullivan A, Suzuki K. Lactic acid concentrat ions in vitreous humor: their use in asphyxial deaths in ch ildren. J Forensic Sci 1983; 28:222-30. Jetter WW, McLea n R. Biochemical changes in body fluids after death. Am J Ciil1 Path 1943; 13:178-85. Jaffe FA. Chemical postmortem changes in the intra-ocular fluid. J Forensic Sci 1962; 7:231-7. Rognum TO, Saugstad 00. Hypoxanthine levels in the vitreous humor: evidence of hypoxia in most infants who died of sudden infant death sy ndrome. Pediatrics 1991; 07 :306-10. Poulsen JP, Rognum TO, Hauge S et al. Postmortem co ncentrations of hypoxanthine in the vitreous humor - a co mparison betwee n babies with severe respiratory failure, conge nital abnormalities of the heart, and victims of su dden infant death sy ndrome. J Perinar iVIed 1993 ; 21: 153-63. Madea B, Kaferstein H, Hermann N, Sticht G. Hypoxa nthine in vitreous humor and cerebrospinal fluid: a marker of postmortem interva l and prolon ged (vital) hypoxia? Forensic Sci illt 1994; 65:19 - 31. Belonje PC , Wilson GR, Siraka SA. High postmortem co ncentration of hypoxa nthin e and urate in th e vitreous humor of infants are not confined to case of sudd en infant death syndrome. S AJrican Med J 1996; 86:8 27-8. Carpenter](H, Bonham JR, Worthy E, Variend S. Vitreous humour and cerebrospinal fluid hypoxanthine concentration as a marker ot' ante mortem hypoxia in SIDS. J Clin Parhol 1993; 46:650-3 .
29
Opdal SH, Rogum TO, Vege A, Sa ugsta d 00. Hypoxa nthine levels in vitreous humor: a study of influencing factors in sudden infant death syndrome. Peciiatr Res 1990; 44:192-6.
References I
30 Vege A, Chen Y, Opdal SH et al. Vitreous humo" hy poxanthine levels in SlDA and infectious death. Acta Paediatr 1994; 83 :634-9. 31 Jones KL, Krous HF, Nadeau J et al. Vascular endoth elial growth factor in the cerebrospinal fluid of infa nts who die of sudden infant death syndrome: eviden ce for antecedent hypoxia. Pediatr 2003; 111 :358-63. 32 Putto A, Rllllskanen 0, Meurman 0 et al. C-reactive protein in the evaluation of a febrile illness. Arch Dis Child 1986; 61 :24-9. 33 Peltola H, Homberg D. Rapidity of C-reactive protein in detecting potential septicem ia. Pediatr InJ Dis 1983; 2 :374-6. 34 Benjamin DR , Siebert JR. C-reactive protein and prealbumin in su spected sudden infant death syndrome. Pediatr Parhol 1990; 10:503-7. 35 Uhlin-Hansen L. C-reactive protein (CRP), a comparison of pre a nd post-mortem blood levels. Forensic Sci hrt 2001; 14:32-5. 36 Tsokos M, Reichelt U, Jung R et al. Interleukin-6 and C-reactive protein serum level s in sepsis-related fatalities during the early postmortem period. Forensic Sci Int 200 I; 119:47-56. 37 Fujita MQ, Zhu Bl, Ishid a K et al. Serum C-reactive protein level s in po stmortem blood-an analysis with special reference to the cause of deat h and survival time. Forensic Sci Int 2002; 130:160-6. 38 Chiesa C, Panero A, Rossi N et at. Reliability of procalcitonin concentrations for the diagnosi s of sepsis in critica lly ill neonates. Clin I11J Dis 1998; 26:664-72. 39 Monneret G, Lab a une JM, Issac C et al. Procalcitonin and C-reactive protein leve ls in neo natal infections. Acta Paediatr 1997; 86:209-12. 40 Tsokos M, Reichelt U, Nierhaus A, Puschel K. Serum pro calc itonin (PCT): a valuable biochemical parameter for the postmortem diagnosis of sepsis. Jnt J Legal Med 200 1; 114:237-43. 41 Mei sne r M, Tschaikowsky K, Schanbel Set al. Proca lcito nin influence of temperature, storage, anticoagulation and arterial or venous asservation of blood samples on procalcitonin concentrations. Eur J C/in Chem Clill Biochem 1997 ; 35:597-60J. 42 Yunginger JW, Nelson DR, Squillace Dl e t al. Laboratory investigation of deaths due to anaphylaxis. J Forensic Sci 199 I; 36:857 -60. 43 Ani sa ri MQ, Zamora JL, Lipscomb MF. Po st mortem diagnosis of acute anaphylaxis by serum tryptase levels. Am J c/ill Parh 1993; 99:101-3. 44 Randall B, Butts J. Halsy JF. Elevated pos tmo rtem tryptase levels in th e absence of anaphylaxis. J Forensic Sci 199 5; 40:208- I I. 45 Edsto n E. van Hage-Ha msten M. Beta -tryptase measurements postmortem in anaphylactic deaths and controls. Forens ic Sci lnt 1998; 93:135-42. 46 Schwartz HJ, Yunginger JW. Schwartz LB. Is unrecognized anaphylaxis a cause of sudden unexpected death ? Clin Exp Allergy 1995; 25:866-70. 47 Holgate ST. Walters C, Walls AF et a l. Th e anaphylaxis hypothesis of sudden infant death syndrome (SIDS): mast cell degranulation in cot death revealed by eleva ted concentrations of tryptase in serum. c/in Exp Allergy 1994; 24: 1I 15-2 2. 48 Edston E. Gidlund E, Wickman M et a!. Increased mast cell tryptase in sudden infant death: anaphylaxis, hypoxia or artifact. C/in Exp Allergy 1999; 29: 1648- 54. 49 BuckJey MG, Variend S. Walls AF. Elevated serum concentrations of beta- tryptase, but not alpha-tryptase, in Sudden Infant Death Syndrome (SIOS). A n investigation of anaphylactic mechanisms. C/in Exp Allergy 2001; 31: 1696-704.
121
50 Schonheyder HC, Henriqu es U. Postmortem blood cultures. Eva luation of sepa rate sampling of blood from the right and left ventricle. APMIS 1997; 105:76-8. 51 Bentley AJ. Zorgani AA, Blackwell CC et al. Bacterial toxins and sudden unexpected death in a young child. Forensic Sci lilt 1997; 88:141-6. 52 Platt MS. Elin RJ, Hosseini JM, Smialek JE. Endotoxemia in sudden infant death sy ndrome. Am J Forensic Med Pathol 1994; 15:261-5. 53 Coe JI. Postmortem chemistries on blood: particular reference to urea nitrogen. electrolytes and bilirubin. J Forensic Sci 1974; 19:33-42. 54 Naumann HN. Postmortem chemistry of the vitreous body in man. Arch Ophthalmol 19 59 ; 62: 356-63. 55 Coe JI. Postmortem chemistries on human vitreous humor. Am J C/in Patholl969; 51 :741-50. 56 Swift PGF. Worthy E, Em ery JL Biochemical state of the vitreous humour of infants at necropsy. Arch Dis Child 1974; 49:680-5. 57 Sturner WQ. Dempsey JL. Sudden infant death: chemical a nalysis of the vitreo us fluid. J Forensic Sci 1973; 18:12. 58 Blumenfeld TA, Blanc WA, Devlin J. Postmortem v itreou s humor concentrations of Na, K, CI, Ca and Mg in children. Pediatr Res 1974 ; 8:356. 59 Zumwalt RE. Hirsch CS. Subtle fatal child abuse. Hum Pathol 1980 ; 11:167 - 74. 60 Emery JL, Swift PGF. Worthy E. Hyp ernatremia and uraemia in unex pected death in infancy. Arch Dis Child 1974; 49:686. 61 Coe JI. Use of chemical determinations on vitreous humor in forensic pathology. J Forensic Sci 1972; 17: 54 1-6. 62 Flear CTG, Gill GV Hyp onatremia : mechanisms and management. Lancet 1981; ii: 26 -31. 63 Coe JI. Further thoughts and observations on postmol1em chemistry. Forensic Sci Gazeffe 1975; 5 :2-6. 64 Coe JI. Vitreous potassium as a measure of the postmortem interval: an historical review and clinical evaluation. Forensic Sci Int 1989 ; 42:201-13. 65 Coe JI. Apple FS. Variations in the vitreous humor chemical values as a result of the instrumentation. J Forensic Sci 1985; 30:828-3 5. 66 Sturner WQ. Dowdey ABC, Putnam RS. Dempsey JL. Osmol ality and other chemical determinati on s in postmortem vitreous humor. J Forensic Sci 1972; 17:387-93. 67 Devgun MS, Dunbar JA. Biochemical investigation of vitreous: application in forensic medic ine esp eci ally in relation to alcohol. Forensic Sci Jnt 1986; 31:27-34. 68 Hen ssa ge C. Krompech er T, Madea B et al. The Estimation oj the Time Since Death in the Early Postmortem Period. 2nd edn. London: Hodder Arnold, 2002. 69 Schleyer F. Determinations of the time of death in the early postmortem interval. {VIeth ods oj Forensic Science, vol. 2. Chichester: Inte rscience Publ., John Wiley and Sons, 1963. pp. 253-93. 70 Furuno J . Momura S. Inorga nic phosphorus content in human aqueous humor in rel ation to hours postmortem. Tohoku J Exp Med 1976; 119:293-5. 71 Henry JB. Smith FA. Estimation of the postmortem interval by chemical means. Am J Forensic Med Patho11980; 1:341-7. 72 Stephens RJ, Richards RG. Vitreous humor chemistry: the use of po tassiu m concentration for the prediction of the postmortem interva l. J Forensic Sci 1987; 32:503-9. 73 Adjudanti s G. Co utselinis A. Estimation of the time of death by potass ium levels in the v itreous humor. Forensic Sci 1972 ; 1 :55-60. 74 Madea B. Henssge C. Honig W, Gerbracht A. References for determining the time of death by potassium in vitreous humor. Forensic Sci Jnt 1989; 40:231-43.
122 I
Biochemical investigations on post-mortem specimens
75 Saugsatd OD, Oliasen B. Postmortem hypoxanthine leve ls in the vitreous humour. Foren sic Sci Int 1978; 12:33-6. 76 Rognum TO, Hauge S, Oyasaeter S, Saugstad OD. A new hioch emical method for the determina tion of postmortem time. Forensic Sci Int 1991; 51 :131-46. 77 Madea B, Herm a nn N, Hengbe C. Precision of estimating the time since death by vitreous potassium- co mparison of t'.NO different equations. Forensic Sci Int 1990 ; 46:277-84. 78 James RA, Hoad ley PA, Sampson BG. Determination of postmoltem interval by sampling vitreous humour. Am J Foren sic Med & Patl7 ol1 997; 18:158-62. 79 Mason JK, Harkness RA, Elton RA, Bartholomew S. Cot deaths in Edin burgh: infant feeding and socioeconomic factors. J Epidemiol Commlln Health 1980; 34:35-41. 80 Coe JI. Postm ortem periph era l blood glucos e and cardiopulmonary resuscitation. Forensic Sci Gaz 19 75; 6:1-2. 81 Gormsen H, Lund A. diagnostic value of postmortem blood glucose determinations in cases of diabetes mellitus . Forensic Sci Int 1985 ; 28: 103-7. 82 Peele t C, Picotte P, Jobin F. The use o f vitreous humor levels of glucose, lact ic acid and blood levels of acetone to establ ish ante-mortem hyperglycemia in diab etics. Foren sic Sci Inr 1994; 65:1-6. 83 Coe JI. Comparative postmortem chemistries of vitreous humor befo re and after embalming. J Forensic Sci 1976; 21 :583-6. 84 Bray M. The eye as a chemical indic ator of environme ntal temperature at the tim e of death. J Forensic Sci 1984; 29:396-403. 85 Coe JI. Hypothermia: auto psy findin gs and vitreous gl ucose. J Forensic Sci 1984; 29:395-8. 86 Chen C, Glagov S, Mako M et al. Postmortem glycosylated hemoglobin (HbAlc): evide nce for a history of diabetes mellitu s. Alln Ciin Lab Sci 1983; 13 :407-10. 87 Winecker RE, Hammet-Stabler CA, Chap man J F, Ropero-Miller JD. HbA I c as a postmortem tool to identify glycemic contro l. J Forensi c Sci 2002; 47: 1373-9. 88 Khuu HM , Robinson CA, Brissie RM , Konrad RJ. Postmortem diagnosis of unsuspected diabetes mellitus established by determination of dec ede nt's hemo globin Alc leve l. J Forensi c Sci 1999; 44:643-6. 89 Goulle JP, Lacroix C, Bouige D. Glycated hemog lobi n: a useful post-mortem referen ce marker in determining diabetes. J Forensic Sci 2002; 128 :44-9. 90 Osuna E, Garcia-Villora A, Perez-Carceles!VI et al. Vitreous humor fructosamine concent rations in the autopsy diagnosis of diabetes mellitus. fnt J Legl Med 1999 ; 112:275-9. 91 Osuna E, Garcia-Villora A, Perez-C arcel es M et al. Glucose and lactate in vitreous humor compared with the determination of fructosamine for the post-mortem diagnosis of diabetes mellitus. Am J Forensic Med Pathol 2001; 22:244-9. 92 DiMaio V JM, Sturn er WQ, Coe JI. Sudden and unexpected deaths after the acute onset of diabetes mellitu s. J Foren sic Sci 1977 ; 22: 147-51. 93 Sturner WQ, DiMaio VJM. Sudden and unexpected death following the acute onset of diabetes mellitus (Abstract). Foren sic Sci 1975; 5: 168. 94 Rozin L, Perper JA , J a ffe R, Drash A. sudden unex pected death in childhood due to unexpected diabetes mellitus. Am J Forensic Med Patho11994; 15:251-6. 95 Russel MA, Opitz JM, Visekul C et al. Sudd en infant dea th s due to conge nital ad renal hypopla sia. Arch Pathol Lab Med 1977 ; 101:168-9. 96 O'Don ohoe l\rv, Holland PDJ. Fam i.lial con genita l adrenal hypopl asia. Arch Dis Child 1968; 43:717- 23. 97 Jindrich EJ . Adrenal hypofullction and sudden death. J Forellsic Sci 19134 ; 29:930-3.
98 Sperling !VIA, Wolfson AR, Fisher DA. Congenital adrenal hypoplasia: an isolated defected of organogenesis. J Pedjatr 1973; 82:444-9. 99 Batch JA, Montalto J , Yong ABW et a l. Three cases of congenita l adre nal hypopl asia; a cause of salt wasti ng and mortality in the neonatal period. J Pediatr Child Hea lrh 1991; 27:108-12. 100 Molander N. Sudden natu ral death in later childhood and adolescen ce. Arch Dis Child 1982; 57 :572-6. 101 Favara BE, Franciosi RA, Miles V. Idiopathic adrenal hypoplasi a in children. Am J Clin Pathol 1972; 57:287-96. 102 Finlayson NB. Blood cortisol in infants and adults: A postmortem study. J Pediatr 1965; 67 :248-52. 103 Sa nders FW. The preservatio n of blood for chemical analysis. J Bioi Ch em 1923; 58:1-15. 104 Paul JR. Postmortem blood chemical determinations. Bull Ayer Ciin Lab Pen Hosp 1925; 9:51-62. 105 Leahy MS, Farber ER. Postmortem chemistry of human vitreous humor. J Forensic Sci 1967; 12:214-22. 106 Levonen E, Raekallio J, Saikkonen J. Postm orte m determination of blood creatinine and urea . J Forensic Sci 1963; 10:22-9. 107 Jensen OM. Diagnosis of uremia postmortem. Dan Med Bull 1969; 8(Suppl.):1-97. 108 Coe J1. Postmortem chemistries on human vitreous humor. Am J Clin Pathol 1969 ; 51 :741 - 50. 109 Pol ayes SH, Hersh ey E, Ledered M. Postmortem blood chemistry in renal disease. Arch liltem Med 1930; 46:283-9. 110 Bennett MR, Rinald o P, Millington DS, Tanaka K, Yokota I, Coales PM. Medium-chain acyl-CoA dehydro ge nase deficiency; postmortem diagnosis in a case of sudden infa nt death a nd neonata l diagnosis o f an affected sib ling. Pediatr Patlzol 1991 ; 11 :8139-95. 111 Bennet l\IlJ, Allison F, Pollitt RJ, Variend S. Fatty acid oxidation defects as a cause of unexpected death in infa ncy. Prog Clin Bioi Res 1990; 321 :349-64. 112 Duran M, Hofkamp M, Rhead WJ et al. Sudd en child death and 'healthy' affected family members with medium- cha in acyl- coe nzyme A dehydrogenase deficiency. Pediatrics 1986; 78:105 2-7. 113 Chalmers RA, Stanley CA, English N, Wigglesworth JS. Mitochondrial camitine-acylcarnitine transl ocase deficiency presenting as sudd en neonatal death. J Pediatr 1997; 131:220-5. 114 Rinaldo P, Stanley CA, Hs u BY et al. Sudden neonatal death in carnitine transp orter deficiency. J Pediatl" 1997 ; 131:181-2. 115 Di onis i- Vici C, Seneca S, Zeviani Met al. Fulminant Leigh syndrome and sudde n unexpected death in a family with the T9176C mutation of the mitochondrial ATPase 6 gene. J Inherit Metab Dis 1998; 21:2-8. 116 Ha rpey J-P, Charpentier C, Coude M et al. Sudden infant death sy ndrom e a nd multiple acyl-coenzyme A dehydrogenase deficiency, ethylmaionic-adipic aciduria or system ic carnitin e deficiency. J Pediatr 1987; 110:881-4. 117 Lundemose JB, Kolvraa S, Gregersen N et al. Fatty acid oxidation disorders as primary ca use of sudden and unexpected death in infants and yo un g children: an investigation performed on cu ltured fibroblasts from 79 children who died aged between 0-4 years. J Ciin
Palhol Mol Patho11997; 50:212-17. 1113
119
Emery JL, Va riend S, Howat AJ, Vawter GF. In vestigation of inborn errors of metabolism in unexpected infant death. Lancet 1988; ji:29-31. Hale DE, Benn ett MJ. Fatty acid oxidation disorders: a new class of metabo.lic diseases. J Pediatr 1992; 121: 1-11.
References I
120 Ho wa t AJ, Bennet MJ, Variend S, Shaw L. Medium cha in fatty acyl -coenzyme A dehydrogenase pres en ting as the sudd en infant death. BM] 1984; 288:976. 121 Howat AJ, Bennett MJ, Variend S et al. Defects in the metabo lism of fatty acids in the su dden infant death syndrome. BlVlJ 1985; 290:1771-3. 122 HollOn JB , Allen JT, Green CA et a l. Inherited metabolic disease in sudden infant death sy ndrome. Arch Dis Child 1991; 66:13)5-17. 123 Miller ME, Brooks JG , Forbes N, Insel R. Frequency of medium-cha in acyl-CoA dehydrogenase G-985 mutation in sudden infan t death syndrome. Pediatr Res 1992; 31:305-7. 124 Chinsky J, Tolsm a T, Cowan T, Blitzer M. MCAD deficiency in SIDS. Am] Hum Gen 1991 ; 49(Suppl.):AI83. 125 Arens R, Gozal D, Jain K et al. Prevalence of medium-chain acyl-CoA dehydrogena se deficiency in sudd en infant death syndrome.] Pediatr 1993; 122:71 5-18. 126 Lemieux B, Giguerre R, Cyr D e t al. Sc reenin g urine of 3 week old newb orns: la ck of associ at ion betw ee n sudden infant death synd rome and some metabolic disorders. Pediatrics 1993 ; 85:986-8. ) 27 Bonham JR, Dow ning M. Metabolic deficiencies and SIDS. ] Clin Pathol 1992; 45(Suppl.):33-8. 128 Bennett MJ, Powell S. Metabolic disease and sudden, unexpec ted death in infancy. Hum Pat/1011994; 25:742-6. 129 Roe CR, Millin gto n DS, Maltby KF. Recognition of med ium chain acy l CoA dehydrogenase deficiency in asy mptomatic siblings of children dy ing of s udden infant death or Reye like syn drome.] Pediatr 1986; 108:13-18. 130 Rebuffat E, Sottiaux M, Goyens D et al. Sudd en infant death syndrome, as first expression of a metabolic disorder. In Sch aub J, Van Hoof F, Vis HL (eds) Inborn Errors of Metabolism. New York: Vevey/Raven Press, Nestle Nutri tion Workshop Series 24, 1991, pp. 71-80. 13 1 Bennett MJ, Hale DE, Coates PM, Stan ley CA. Postmortem reco gnition of fatty acid ox idation disorders. Ped iatr Pathol 1991 ; II :365-70. 132 Bennet MJ, Allison F, Pollitt RJ, Variend S. Fatty acid oxid ation defects as causes of unexp ected death in infancy. In Tanaka K, Co ates PM (eds) Fatty Acid Oxidation: Clinical, Biochemical and Molecular Aspects. New York: AJan R Li ss, 1990, pp. 349-64. J 33 Blau N, Duran M, Blaskovics ME (eds) . Physicia n 's Guide to the Laboratory Diagnosis of Metabolic Diseases. London : Chapman and Ha ll Medical. 1996. 134 Scriver CR, Beaudet AL, Sly WS, Valle D (eds). The Metab olic and /VIolecular Bases of Tnh erited Disease. New York: McGraw -Hili, 1995. 135 Applegarth DA, Dimmick JE, Hall JG (eds). Organelle Diseases. London: Chapman and Hall Medical, 1997. 136 Kemp PM, Little BR , Bes t RO, Dawson DB. Whole blood leve ls of dodecanoic acid, a routinely detectable fore nsic marker for a gen etic disea se often misd iagnos ed as Sudden Infant Death Syndrome (SIDS): MCAD deticiency. Am] Forensic NIed Path 1966; 17:79-82. 137 Ross KF, Guileyardo JM, Bennet MJ, Ba rnard JJ. Comment on wh ole blood levels of dod eca noic acid, a routin ely detectabl e foren sic marker fo r a genetic disease ofte n misdiagnosed as Sudden Infant Death Sy ndrom e (SlD S): MCAD deficiency. (Letter). Am ] Forensic Med Path 1966;17:349-50. 138 Blakemore AJF, Singleton H, Politt R et at. The frequency of the G985 MCAD mutation in the gen eral populati on. Lancet 1991; 337:298-9. 139 Coates PM. Histori cal perspecti ve of medium chain acyl-CoA dehydrogena se deficiency. A decade of discovery. Prog Clin BioI Res 1992; 375:40 9-23.
123
140 !jIst L, Ruiter JP, Hoovers JM et al. Common missense mut ation G 1528C in long-chain 3-hydroxyacyl-CoA dehyd roge nas e deficiency. Characterization and expression of the mutant protein, mu tation analysis on geno mic DNA and chro mosomal localization of the mitochondrial trifunct ional protein alph a subunit gene.] Clin Invest 1996; 98: 1028-33 . 141 Rashed MS, Ozan d PT, Bennett MJ et al. Inborn errors of metabolism dia g nosed in sudden infant death case by acylcarnt ine analysis of postmortem bile. Clill Chem 1995; 41:1J09-14. 142 Hale DE, Cornell JE, Bennett MJ. Stability of lon g- chain and short-chain 3-hydroxyacyl -CoA dehydroge nase activity in the postmortem li ver. Clin Chem 1997 ; 43:273 -8. 143 Boles RG, Martin SK , Blitzer MG, Rin a ldo P. Bio chemical diagnos is of fatty acid oxid atio n disord ers by metabolite analysis of postmortem li ver. Hum Pathol 1994; 25:733-4. 144 Gregersen N, Blakemore AIF, Winter V et al. sp eci fic diagnosis of medium-chain acyl-CoA de hyd rogenase (MCAD) deficiency in dried blood spots by a polymerase chain reaction (PCR) assay detecting a point mutat ion (G9 85) in the MCAD gen e. Clin Chim Acta 1991 ; 203 :23-4. 145 Matsubara Y, Na risawa K, Tada K. Medium-chain acy l-CoA dehydrogenas e deficiency. Molecular aspects. Eur] Pedia tr 1992; 151:154-9. 146 Ding J-H , Roe CR , lafolla AK, Chen Y- T. Medium-chain acy l-CoA dehydrogenase deficiency and sudden infant death (Letter). N Engl] Med 1991; 325: 6 1-2. 147 Kelly DP, Hale DE, Rutledge SL et al. Molecular basis of inherited med ium-cha in acyl-CoA dehydrogenase deficiency ca using sudden child death.] Inherit Metob Dis 1992 ; 15:171-80. 148 Hallock J, Morrow G, Karp LA, Barness LA . Postmoliem diagnosiS of metabolic disorders. The findings of maple syrup urine disease in a cas e of sudden and unexp ected dea th in infancy. Am] Dis Child 1969; 118:649- 5 1. 149 Byard RW, Cohle SD. Sudden Death ill Infancy, Childhoo d and Adolescence. Camb ridge: Cambridge University Press, 1994. 150 Clayton PT, Hy land K, Brand M, Leonard N. Mitochondrial phosphoenolpyruvate ca rboxykinase deficiency. Eu r] Pedia tr 1986: 145:46- 50. 151 Norman MG , Taylor GP, Clark e LA. Sudden, unexpected, natural death in childhood. Pediatr Patlwl 1990; 10:7 69 -84. 152 Burton BK, Roach ES, Wolf B, Weissbecker F. Sudd en dearh associated w ith biotinid ase deficiency. Pediatrics 1987; 79:482-3. 153 Danks DM, Campbell PE, Stevens BJ et at. Menke's kinky hair syndrome. An inherited defect in copper metab olism with widespread effects. Ped iatrics 1972; 50: 188-20 t. 154 Forres t AR (ACP Broadsheet no. 137, April 1993). Obtaining samples at postmoliem examin a tion for toxicological and biochemical analyses. ] Clin Pathol 1993 ; 46: 292-6. 155 Pounder DJ , Carson DO, Johnston K, Orihara Y. Electro lyte concentration differenc es between the left and right vitreous humor sampl es. ] Forensic Sci 1998; 43:6 04 -7. 156 Bennett MJ, Ragni MC, Hood 1, Hale DE. Comparison of postmortem vitreous and urinary organic acids. Ann Clill Biochem 1992; 29:541-5. 157 Pallick W JA, Logan RW. Free amino acid content of the vitreous hum or in cot deaths. Arch Dis Chil d 1988; 63:660-2. 158 Chong APY, Aw SE. Postmortem endocrine levels in the vitreo us hum our. Ann Acad Med (Singapore) 1986; 15 :606-9. 159 Hill E. Significance of dex trose and nondextrose reducing substa nces in postmortem blood. Arch Pat/IOI 1941; 32: 4 52-73.
124 I
Biochemica l investigati ons on post-mortem spec ime ns
160 Lythgoe AS. The act ivity of lactic dehydrogenas e in ca daver sera: a comparison of different sampl ing sites. Med Sci Law 1980; 20:48-53. 161 Zhang DJ, Elswic k RK, Mill er WG, Bailey JL. Effect of serum-clot contact time on clinical chemistry laboratory results. Clin Cliem 1998; 44:1 325-33. 162 Fowler KJ . Storage of skin biopsies at -70°C for futu re fibroblast cu lture. J Ciin Patlioil984; 37:1191-3. 163 Bennet MJ , Rinaldo P. The metabo lic autopsy comes of age. Ciin Chem 200 1; 47 :11 45-6. 164 Wilcox RL, Nelso n CC, Stengel P, Stei ner RD. Postmortem screening for fat ty acid oxid atio n disorders by analysis of Guthri e cards with tandem mass spectrometry in sud den unexpected death in infancy. J Pediatr 2002; 141 :833 -6. 165 Chace DH, DiPerna JC, Mitchell BL et al. El ectrosp ray tandem mass spectrometry for analysis of acylca rnitines in dried postmortem blood spec imens collected at autopsy from infants with unexplained cause of death. Clin (hem 2001; 47: 1166- 82. 166 Karke la JT. Critical evaluation of postmortem changes in hum a n autopsy cistern al fluid . Enzymes, elect rolytes, acid base balance, glucose and glycolysis, fre e amino acids and amm onia. Con-elation to total brain isch emia. J Forensic Sci 1993; 38 :603- 16.
Further Reading Applegarth DA, Dimmick JE, Toone JR. Laboratory detectio n of metabolic di sease. Pediatr C/in North Am 1989; 36 :49-66. Applega11h DA, Dimmick JE , Hall JG reds). Organ elle Disms es. Lo nd on: Chapma n and Hall Med ical, 1997.
Beny J, Alibon e E, McKeever P et al. The contribution of ancillary pathology tests to the in vestigation of sudden infant death. In Fleming P, Blair P, Bacon C et al reds} Sudden Unexpected Dea ths ill InJancy. The CESDI SUDI Studies. Norwich: The Stationery Office, 2000, pp. 97 -112. Bla u N, Duran M, Blaskovics ME (eds). Physician's Guide to the Laboratory Diagnosis oJ M etabo lic Diseases. London: Chapman and Hall Med ical, 1996. Brettell TA, Safers tein R. Forensic science. Anal Chem 1995; 67:27 3R- 94R. Byard RW. Sudden Death in inJancy, Childhood and Adolescence, 2nd edn. Cambridge : Cambridge University Press, 2004. Coe J1. Postm ortem chemistries on human vitreous humor. Am J Clin Pathol 1969; 51:741 - 50. Coe J1. Postmorte m chem istlY of blood, cerebrospinal fluid and vitreous humor. In Ted esc hi CG, Eckert WG, Tedesc hi LG (eds) Foren sic Medicine, vol. 2. New York: WB Saunde rs, 1977. Coe J1. Post mortem biochemistry of blood and vitreous hum our in paediatric prac tice. In Mason JK [ed.} Paediatric Forensic 1\1edicine and Pathology. London: Chapman and Hall, 1989, pp. 191- 203. Coe J1. Post mortem chem istJy upd ate. Emph asis on fore nsic appli catio n. Am J Forensic Med Pathol 1993; 14:9 1-117. Scriver CR, Beaudet AL, Sly WS et al (eds). The Metabolic and Molecular Bases oJ Inh erited Disease. New York: McGraw-Hili, 1995.
I
I
CHAPTER 6
OCULAR INVOLVEMENT IN
NON-ACCIDENTAL INJURY
Harry Willshaw
I t roduction of ocular and adnexal injury
~i:ope
125 125
Fundus haemorrhages References
128 134
INTRODUCTION
SCOPE OF OCULAR AND ADNEXAL INJURY
_\11 abused child is most often the victim of damage 'n flicted by a parent or carer. Although both self-mutila TIo n and Munchausen's syndrome by proxy do occur in children, I they are rarities in the setting of child abuse globally. The practice of child abuse was first challenged legally in New York in 1871 2 using legislation written to prevent cruelty to animals. Since that time the incidence of child abuse, or at least its identification, has risen to such an ex te nt that in 1962 it was considered the major cause of child death and maiming in the USA. 3
Fundus changes are often considered the hallmark of ocu lar involvement in NAl, but in fact any ocular or adnexal structure may be damaged.
The frequency of ocular involvement in non-accidental injury (NAI) has been variously reported at between 5 and 6 1 per cent. 4 ,5 Eye damage is usually seen in association ':;ith head and facial injuries. 6 In reported series in which - e injuries resulted in death, the rate of ocular involve ment is high/- 9 suggesting that ocular involvement is a sig nificant marker for severe neurological involvement. Fatal neurological involvement is particularly common in 'ery young victims; all of 13 fatalities attributable to shak in g reported in 1987 were of children aged 24 months or less . 10 A proportion of children will suffer life-long visual andicap as a consequence of NAl. The likelihood of long lerm visual sequelae from a n injury is usually directly re lated to the extent of ocular damage; however, it must be remembered that some children are blinded by their neuro logical injuries with no evidence of persisting ocular pathology. II
Eyelids Swelling, laceration and burning of the eyelids have all been described. 12 Lid bruising and swelling are relatively common events in a young life and should not be overinterpreted, unless they are seen in combination with other injuries that are suggestive of abuse. In children under the age of 12 years, lid injuries are most commonly caused by implements. The characteristic pattern of belt buckles etc. may be obvious in the injured site. Typically, after the age of 12, the injuries are inflicted with fists and feet. 5 Burns to the eyelids are an occasional feature of ab use. Unlike their accidental counterpart, non-accidental facial burns tend to be deep and multiple. At 60°C, heat must be applied to skin for at least 3 seconds to produce even a first-degree burn,13 so that multiple deep cigarette burns, as seen in Figure 6.1, are not compatible with the child having inadveliently brushed against the hand of a smoker. Occasionally more exotic agents of abuse are applied to the lids. Recently, a case has been reported in which super glue was deliberately applied to the eyelids, 14 resulting in a short period of permanent lid closure.
126 l
Ocular involvement in non-accidental injury
Figure 6_1
Cigarette burns on the eyelids - inflic ted injury.
Figure 6.2
Bilateral subco njunctival haemorrhage and
periocular bruising following assault.
External Eye Subconjunctival haemorrhage is ano ther rela tively co mmon finding (Fig. 6.2) associated with facial IDJUlY or Valsalva manoeuvre, and does not in itself indicate deliberately inflicted damage. On the other hand, persisting corneal damage is a great rarity in childhood. In otherwise healthy children with no cause fo r chronic corneal exposure (e.g. previous ptosis surgery), corneal epithelial defects usu ally settle quickly. Persisting epithelial damage may well be caused by chemical in stillat ion into the eye. This is particu larly likely to involve the epithelium of the lower hal f to one-third of the cornea and conjunctival cui de sac. This is because an intact Bell 's phenomenon causes t he eyes to roll upwards as the lids are pulled apart to allow instillation. We have treated a child with unexplained bilateral corneal clouding accompanied by total epithelial loss. The corn ea s healed well when the child was isolated from her parents, and her father subsequently admitted to repeatedly spitting methadone into the child's eyes. Any type of harmful material may be instilled into the eye - no paliicular sub stance is more common than oth ers. In 1987 the app licatio n of hot peppers (capsicum and capsaicin) to the conjunctival fornices was reported as a type of child abuse. 15
Anterior Segment The presence of anterior seg ment ocular injUlY, in the form of iris or lens damage, tends to suggest severe impact trauma and carries a relatively poor visual prognosis (Fig. 6.3). Pupil sphincter rupture (which manifests as slight irregularities in th e pupil and tears at the pupil margin), hyphaema , recession of the drainage angle, lens sublux ation and cataract leading to glaucom a have all been described in a single child. 16 These events usu ally follow direct injury to the gl ob e,17 but can be seen after the rapid deceleration caused by an impact injury. As with all of the ocular features of child abuse, the changes seen are not pathognomonic of NAJ and appropriate
syste mic investigations are required. This is palii cularly so for lens subluxation, in which investigations must exclude Marfa n's syndrome, homocystinuria, sulphite oxi dase defi ciency and Weil-Marchesani syndrome. When lens damage does occur, it may be possible with the slit lam p to detect small tears in the anterior lens cap sule. These tears allow t he ingress of aqueous humour into the protein of the lens cortex leading to delayed cataract formation. The resultant cataract may not become clinic ally app arent for 4 or 5 days after the original injury.
Posterior Segment Features of NAI in the posterior segment are dominated by the finding of fundus haemorrhages. However, retinal detachments can also occur. These may be caused by giant retinal tears 18 or vitreous base avul sion 19 and may be apparent immediately after the injury has occurred. Because the vitreous gel is rel atively firmly adh erent to the retina over the peripheral portion (the ora serrata), violent movements of the vitreous gel exert strong tractional forces on that area of the retina. As a consequence, the retina may tear over more than 90 0 of its periphery, or the vitreous base may tear away. Alternatively, detachments may present later because of traction from the condensation of traumatic vitreous haemor rhage. In one study, post-mortem examination revealed what have been called haemorrhagic retinal detachments in up to 63 per cent of case. 20 In that particular report, they occurred primarily in the retinal periphery and, though they pathologic ally appeared as retinal detachments, were seen ophthalmo scopically as accumulations of sub retinal blood. Peri macular folding of the retina (Fig. 6.4), seen either at the time of firs t assessment or at later fundus ex amination, is extremely sug gestive of NAJ. The folds can result from either direct head trauma or violent shaking, but are almost unique to the eyes of injured children. 21 .22 Late fundus manifestations of abuse include hole for mation in the retina, cysts, gliosis and scarring. 2J Features
Scope of ocular and adnexal injury I
Figure 6.4
127
Retinal haemorrhages and a peri macular fold.
Figure 6.5 Perineural haemorrhages wi thin the optic ne rve (haematoxylin and eosin [HEtEj stain).
such as gliosis, pigmented dema rcation lines and fi xed ret inal folds are strongly suggestive of long-standing damage and, if seen in the company of more acute features , such as nerve fibre layer haemorrhages, indicate injuries of differ ent ages. 24 Many of these posterior segment features carry a poor visual prognosis and, when combined with neuro logical features, can be used to produce a score that indi cates the likelihood of severe neurological or visual morbidity2o - this is discussed in more detail later in this chapter (see section on long-term visual sequelae, p. 134).
Visual Pathways
Figure 6.3 Anterior segment damage. (a) Dislocated lens; (b) tear of peripheral iris (iridodyalysis); and (c) traumatic cataract.
Neurological involvement in NAl is the major ca use of morbidity and mortality and is often closely correlated with the degree of ocular involvement (see section on asso ciated neurological injury). Direct involvement, particu larly of the optic nerve (Fig. 6.5) or visual cortex, may lead to a poor visual outcome in children who show app arently complete ocu lar recovery. In a group of three children with unilateral retinal haemorrhages reported in 1998, two had
128 l
Ocular involvement in non-accidental injury
profound long-term visual handicap because of optic nerve or higher pathway damage,25 even though the fundus involvement was strictly unilatera l and resolved. Bleeding into the optic nerve sheath and peripapillary sclera has been identified in autopsy specimens,26 and complete avulsion of the optic nerve from the sclera may occur, as may optic nerve transectionY Optic nerve damage is often seen with accompanying visual cortex injury. The visual cortex is vulnerable to trauma via a number of mechanisms, including direct con tusion, raised intracranial pressure and the effects of poor perfusion during periods of collapse.
Ocular Motor Control Figure 6.6
Neurological disturbances of gaze have been reported but are relatively uncommon. Abnormalities of the vergence system, regularly seen after accidental head trauma,28 do not feature significantly in the literature of NAJ. More obvious cranial nerve palSi es, gaze palsies and nystagmus can occur 29 .]O but are usuall y seen in the context of other signi ficant neurological injuries. Fin ally, in the most extreme circumstances, the clinician may be faced with an auto-enucleation as a result of either self-mutilation or NAI.'
FUNDUS HAEMORRHAGES Though all ocular and adnexal structures can be affected by child abuse, fundus bleeding is the most typical feature and the one that has probably received most attention in the world literature.
Types The clinical appearance of fundus haemorrhages is depend ent on the site of ex travasation of blood from the retinal blood vessels, and also upon the extent of bleeding. The vessels involved are usually the capillaries or the post capillary venules and, because of the rich circulation at the posterior pole of the eye, haemorrhages ten d to be most plentiful there. However, characteristic bleeding in NAI involves all the layers of the retina and can extend right out to the ora serrata. Fundus examination of an NAI vic tim must therefore involve visualization of the whole retina, usually necessitating pupil dilatation (see section on examining the child with fundus haemorrhages, p. 136). The physical characteristics of retinal haemorrhages are dictated by the layer into which they occur.
NERVE FI BR E Superficial haemorrhage into the nerve fibre layer of the retina tends to be fl ame-shaped or splinter in type, and is
Intraocular haemorrhage: nerve fibre layer and 'blot'
haemorrha ges - some contain white centres (Roth spots).
derived from the superficial capillary bed. The sha pe of the haemorrhage depends on the tracking of blood between the nerve fibres, and for this reason the haemorrhages change shape beyond the posterior pole. The nerve fibre bundles, densely packed at the posterior pole, make a more open network when one examines them more than 7 mm from the optic disc.]' In these more peripheral locations, there fore, the superficial haemorrhages are rounded (Fig. 6.6). Not infrequently they have white centres, and are known clinically as Roth spots. These were first described in 1872 32 and thought at that time to be a pathognomonic feature of subacute bacterial endocarditis. Subsequent clinical reports have identified them in a variety of other disorders, includ ing leukaemia, anaemia, Behc;et's disease and hyperten sion. 3] They consist of a white centre with a round or oval red surround, and it is our experience that they are a com mon form of superficial retinal haemorrh age in NAt
DEEP RETINA The shape of deeper retinal haemorrhages is also dictated by their anatomical relations. Bleeding from deep retinal capil laries is into the veliically orientated neural structures of the outer retina. Physically, the smaller ones are described as dot haemorrhages and may be mistaken for micro aneurysms. Larger areas of bleeding form blot haemorrhages, which extend throughout the thickness of the retina .
SUBRETI NA Bleeding into a subretinal site may take one of two forms. Blood may extravasate from the deep retina and accumu late betwee n the photo receptors and the retinal pigment epithelium layer. In this location the haemorrhages are generally large and tend to be situated at the posterior pole. When they clear, they usually leave a residual visual defect because of damage to both the choriocapillaries and the
Fundus haemorrhages I
129
o "erlying macula photoreceptors. If, on the other hand, the bleeding originates from choroidal vessels, then it lies deep iO the retinal pigment ep ithelial layer, appears slate oloured, has rounded margins and is associated with ele ,;ation of the retinal pigment epithelium (RPE).34 Once bl ood in this site clears, visual recovery tends to be much more complete.
PR E- RETINA Pre-retinal haemorrhages result from breakthrough of blood fro m the retina into either the space between the nerve fibre laye r and the internal limiting membrane35 or the space bervveen the retina and the vitreous gel. The haemorrhage may be extensive, and sedimentation of the solid components of the blood gives rise to a fluid level that can shift under the in fluence of gravity. Such haemorrhages are typically associ ated with significant intracranial bleeding in the form of sub dura l or subarachnoid haemorrhage. 36
Figure 6.7 Retinal haemorrhage stained using haematoxylin and eosin: a photomi crog raph shows recent haemorrhage at several sites. pr, pre-retinal; nf, nerve fibre layer; ir, intraretinal.
POSTERIOR HYALO ID fin ally, blood may break through the posterior hyaloid face and give rise to a localized or diffuse vitreous haemor :hage. Vitreous haemorrh age is thought to happen 2-3 days after the initial injury,37 a feature that may have con siderable forensic significance. The vitreous will usually clear spontaneously, but the process can be slow, taking several months. Since these injuries typi cally affect chil dren in the sensi tive period for visual development,38 the visual outcome following such severe bleeding is often bad not only because of structural damage to the eye, but also because of stimulus deprivation amblyopia. 39 Studies on children with other causes of stimulus deprivation have shown t hat relativel y short periods of asym metric interfer ence with visual input, at the right time in life, can lead to significant amblyopia. 40 Figure 6.7 illus tra tes the distribu tion of haemorrhages within different layers of the retina. No particular pattern of fundus haemorrhage is exclu sive to NAI, but it is frequentl y observed that the presence of bleedi ng at all depths of the retina, throughout 360 0 of the retina and extending from the posterior pole to the ora serrata, is strongly suggestive of bleeding caused by the severe acceleration and deceleration forces seen in NAI. 41 - 43 If such extensive haem orrhages are seen in com bination with peri macular folds 44 or haemorrhagic retinal cysts,45 then anything but the most severe accidental injulY can almost certainly be excluded as the cause.
Mechanisms The mechanism of blood vessel damage during NAI remains debatable, and it may well be that several different mecha nis ms all contribute. The acceleration/deceleration forces gen erated by the vigorous shaking of an unsupported infant head
(whiplash shaken baby) are substantial, but they increase greatly with head impact against a solid surface. These forces then give rise to a number of contributolY events: • Blood vessels within the retina may shear because of the relatively firm adherence between the vitreous and the retina in yo ung children. The violently moving vitreous transmits its mom entum to the layers of the retina, tending to tear the retina and give rise to splits within it (schisis) and blood-filled cystS. 45 Post-mortem examination of 190 eyes 46 demonstrated that bleeding, even when it extended in to all layers of the retina, invariably arose from the vessels in the inner layers . This tends to support the view that tangential forces are responsible for retinal shearing and bleeding. • Raised intraocular venous pressure results from either raised intrathoracic pressure or raised pressure within the optic nerve sheath. In either case, venous stasis results, and subsequently retinal bleeding. There are several potential causes of raised intraocular venous pressure in an abusive situation. A Purtscher type of retinopathy secondary to chest com pression has been postulated as the cause of retinal haemor rhage in child sexual ab use 47 or when an excessively firm grip is taken of the infant chest during shaking. Chest com pression may well be a contlibutory factor in the re tinal haemorrhages commonly found in the neonate after normal vaginal delivery48,49 (see later). Raised intratlloracic pressure during cardiopulmonary resuscitation (CPR) also has been causatively linked with retinal haemorrhage. A number of individual case reports testify to the possibility of fundus haemorrhage following CPR, but large prospective studies have failed to validate that association (see later) .
130 lOcular involvem ent in non-accidental inj ury
Bleedin g into the optic nerve sheath a nd th e perip apil lary sclera is a commonly reported feature of the more seve re cases of NAl. Eleve n of s ixtee n children dy ing from the central nervous syste m (CNS) complications of NAl showed perineural bleeding. The bleeding was into the dis ta l portion of the optic nerve, indicating that it was not simply forward extension of th e accompanying subdural haemorrhage. 20 It is entirely possible that perineural bl eed ing raises the pressure within the optic nelve sheath, thus ten ding to impede ven ous return, wi th fundu s bleeding seco nda ry to the ve nous co ngestion. Simila rly, raised intrac ranial pressure with pap illoedema is a recog ni zed cause of retinal h ae morrhage with a sim ilar mechanism. There is a close correlation between severe CNS injury and prominent fundus bleeding. 50 A third possible source of fundus h aemorrh age, particu larly in the presence of subarachnoid haemorrhage, is Terson's synd rome, 51 in which subhyaloid blood is seen shortly after the development of a subarachnoid haemor rhage. Paton 51 postulated that the subarachnoid blood, under pressure, passes forward through the la mina cribrosa to e nter the eye, but a mechanism rel ated to raised intrasheath pressure is prob ab ly more likely. Examinati on of children w ith intracrani al haemorrhage from a variety of causes other than NAl showed that associated retinal h ae morrhage is extremely rare.52
Forces Required to Cause Bleeding The forces required to tear retinal, optic nerve and intra crani a l blood vessels have proved difficul t to quantify. Biomechanical studies at the Children's hospita l of Phi lade 1 phia were reported in 1987. 53 Using 'j ust bom doUs ', the heads of which were fi lled with soaked cotton of a weight th at simulated th e infa nt brain, the forces generated during both vigo rou s shakin g and during impact with a solid sur face were measured. Mean angular accelerations achieved during shakjng we re calculated at 9.29 G, but leapt to 428.18 G when impact was involved in the injury. On the basis of this work, the authors concluded that severe 'shaken baby syndrome' was not usually caused by shaki ng alon e, and was unlikely to be cau sed by shaking during ordinary pl ay. When they th en went on to examine the heads of 57 children ad mitted with su spected shaking injury, they found evidence of impact, often subtle, in 75 per cent. Othe r rep orts, however, have clearly documented fundus haemorrhages in the absence of any impact inj uryS4 Fu rthe rmore, the forces asso ciated w ith both bungee jumpi ng 55 an d aircraft ejection,56 which carry no impact component, can alo g ive rise to retinal haemorrhages. Despite t his divergence of views, it is agreed that the forces required to gen erate retinal bleeding are cons ider able. These forces ca n almost never be ac hieved during ordinary domestic ac cidents. Three recent case reports 57
have detailed bleeding following accidental trauma. In one case, a 13-month-old boy developed subdural haemo r rhage and intra- and pre- retin al haemorrhage follow ing a fall down 13 concrete steps. The fundus haemorrhages were unilateral and clea red within 3 months. In the second case, a 9-month-old boy fell 1-2 feet, striking his head on the floor, and sustained bilate ral reti nal haemorrhage in associ at ion with a subdural h aemo rrh age. In the third case, a 7-month-old gi rl fell through a stair rail onto a concrete basement flo or and again sustained both subdural and (unilateral) retinal haemorrhage. These repo rts serve to emphasize that no ocular feature can be considered pathognom onic for NAl. None the less, in la rge selies, accidental injury has almost a lways seemed insuffic ient to cause retinal bleeding. For example, of 79 under-3-year-olds admitted to the Sick Children's Hospita l in Toronto with head inj ury, none of the 75 who had experi enced accid ental trauma had fundus haemorrhages S8 Th e three with NAl and the one whose injUlies were of un certain cause all had fundus haemorrhages. Similar reports have been provided from other instituti ons,59,50 indicating that, excluding high-speed road traffi c accidents and falls from extreme heights, accid ental injUly is most unlikely to provoke retinal bleeding. The summary view ex pressed in a recent Roya l College of Ophthalmologists Working Party ana lysis was that 'No absolute values can be given for the a ngula r acceleration forces required to produce retinal bleeding or other injUly, but th ere is good evidence that they must be considerabl e'.51
ASSOCIATED NEUROLOGICAL INJURY The forces that give ris e t o fundus bl eed ing a lso will commonly cause neurological damage. The correlation betwee n retinal hae morrhages , cerebra l bleeding and bo th morbidity and mortality is strong. The full impact of an injury may not be immediately apparent, and it is known that subdural haemorrhage may ap pear 2-3 days afte r the retinal haemorrhages. 37 For this reason, detailed exami natio n a nd documentation will be necessalY for sev eral days following the admission of a child with a suspi cious injury. Attempts to correlate ocular and neurologica l fin din gs have tended to involve pathologi cal exami nation of more severely affected children. Subdural bl eed ing is t he most common CNS finding in a child with traumatic retin al hae morrhages. Of 13 fatally injured children examined in Phil adelphi a , nine had suffe red blunt trauma w h ilst four showed no evidence of blunt trau ma and were co nsidered to be 'shaken bab ies '.62 All the children with subdural h aemorrhage, re ga rdless of the type of injury, showed evi dence of retinal a nd optic nerve haemorrhage. A contro l grou p of six chil dren dy in g from sudden infant death syn d rom e (SlDS) had no evidence of optic nerve bleeding. In the sa me series, subarachnoid haemorrhage was less commonly associ ated with retin al haemorrhages, and only
Fundus haemorrhages - f of the children with subarachn oid h2e morrhage wed fundus bleeding as well. This study also showed :::-.a- me babies who had been severely shaken were more d_ - to show multilayered retinal haemorrhages than were - _0 e who had suffered blunt trauma. One child (subsequently considered to have suffered - m SIDS) who underwent CPR had a few red cell s in - e anrerior optic nerve - a findi ng in sharp contrast to - e extensive bleeding associated with acceleration/ _C{'eleratio n injuries. Whatever the type of injury, cerebral ema was found in all fatally injured children. A similar study from Sheffield 20 examined the eyes and brai ns of 23 fatally injured children; 16 of these children had ead injuries, while the remainder died from suffocation, abd ominal injUlies or other injuries. The children were all Wlder 3 years of age. The authors constructed a 'total eye ore' based on the presence of haemorrhagic retinal detach menr. retinal and optic nerve bleeding. They compared this -',-ith an 'intracranial score' based on the extent of the inn-acrani al bleeding, cerebral lacerations and evidence of pri mary axonal injury. Though cerebral oedema was com mo nly reported from im aging, they found it difficult to iden . pathologically and, therefore, unlike the Philadelphia dy, did not include it in the score. The authors found a close relationship between the -everity of the ocular and t he cerebral injuries, and felt able to deduce a severity sequence for the damage suffered in _-Al. They suggested that with increasing force of trauma :0 the head a ch il d is likely to suffer: • first, subdural haemorrhage, followed by subhyaloid, intraretinal and then perineural optic nerve sheath haemo rrh ages ; • second, haemorrhagic retinal detachment (more accurately subretin al bleeding); • third, choro id al and vitreous haemorrhages, which coincide with subarach noid haemorrhage, intracerebral haemorrhage and cerebral laceration. The authors concluded that the momentum of soft tissue '."ithin a relatively rigid and stab le surround (e.g. the brain :n the skull or the vitreous in the glob e) is responsible for traction on vessels and the subseq uent bleeding. As with other studies, they found that optic nerve bleedi ng tended to be into the anterior portion of the nerve and did not appear to extend forward from the site of intracranial bleeding. They specu late that a sim ilar mechanism could be :nvoked to explai n optic nerve bleeding, i.e. the mobile eye generates axial and rotationa l forces aro und the relatively fixed point that is the origin of the optic nerve from the eye. As noted in other studies, milder eye changes may initially seem to cause no associated CNS damage on imag ing. However, evidence mi ght appea r within a few days, further emphasizing the need for careful review of affected children. Clinically based (rather than patholo gically) scoring systems have also shown a significant correlation between
I
131
the severity of retinal bleeding and the associated neuro logical injuries. 50 However, t hough CNS and ocular haem orrhage commonly coexist, it is quite possible for either to occur independently. A 1991 study 63 fou nd that 39 per cent of subdural haemorrhages had no associated ocular haem orrhage, and that seven per cent of children with retin al bleeding had no evidence of CNS bleeding.
Differential Diagnosis of Fundus Haemorrhages Since no ocular pathology is unique to NAl, it is essential that the clinician excludes other possible causes of the observed clinical signs. This needs to be done as quickly as possible so that appropriate steps can be taken, whatever the outcome of investigations. A comprehensive list of diagnostic differentials has been produced by a Working Party ofthe Royal College of Ophthalmologists G1 ,64 (Table 6.1). Some of these will be obvious following initial assessment whilst others demand particular diligence on the part of the supervising clinici an. Alternative explana tions that need to be consid ered includ e accidental inju ry, CPR, epileptic seizures, arousal shak ing, bleeding disorders, bleeding secondalY to infection, genetic metabolic disorders and birth-associated retina l haemorrhage. ACCIDENTAL INJURY
A hi stO lY of accidenta l injury or injury associated with attempts to aro use or resuscitate is commonly cited to account for neurological and ocular damage. As discussed above, accid ental injury, except the most severe, is almos t never the cause of significant cerebral or ocular bleeding.
Table 6.1
Differential diagnosis of retinal haemorrhages in
children 61 •64
Leukaemia (particularly ac ute lymp hatic) Haemorrha gic disease of the newbo rn Retinopathy of prematurity Sickle cell retinopathy Extraco rporeal membrane oxyge nation Metabolic (galactosaemia, glutaricacid uria) Henoch-Schonle in pu rpura Maternal cocaine ingestion Meningitis Intracranial vascular malformation Optic nerve dru sen, tuberous sc lerosis, X-linked retinoschisis Chronic severe papilloedema Intraocular surgery Severe hypertension Protein C deficiency Von Willebrand's disease Cerebral malaria Vaginal delivery (sponta neous and assisted)
1 32 l
Ocular involvement in non-accidental injury
CARDIO PULM ON ARY RESUSCITATIO N
AROUSA L SHAK ING
As in the case of accidental injury, isolated case reports exist to suggest that vigorous CPR, particularly if per formed by inexperienced practitioners on very young children, may be responsible for retinal haemorrhage. 65 The infant in Kramer's case report was fully documented, with fundus examination prior to the resuscitation attempt showing there had been no bleeding. This, there fore, represents one of the few cases in which the bleeding unequivocally followed the CPR rather than being present prior to instigating CPR. This particular child received 60 minutes of vigorous CPR, after which she was found to have intraretinal and pre-retinal haemorrhage out to the mid-periphery of the retina. Set against such individual reports, prospective series of paediatric patients and experimental studies on piglets have shown retinal haemorrhage secondary to CPR to be extremely unlikely. In a large clinical series, Kanter found only a single small haemorrhage in 40 children requiring CPR. 66 He concluded 'When retinal haemorrhage is detected in the paediatric patient after CPR, prior trauma should be assumed'. More recently, 169 post-mortem ocular examinations were performed in children after failed CPR. 67 Sixty-one of the children had retinal haemorrhage after CPR, of whom 56 had suffered head injuries and four had CNS disease or sepsis; in only one case was cause of death undetermined. No case was found with retinal haemorrhage in which CPR was the only explanation for the bleeding. In a prospective study of 43 children admitted to hospital with non traumatic problems and requiring resuscitation, small punc tuate retinal haemorrhages were seen in a single case. 68
Arousal shaking is always a difficult area, with many practi tioners considering it inconceivable that the forces required to cause intracranial and ocular bleeding could be generated in a reasonable attempt to arouse an unconscious child. Nevertheless, it is a defence used in court with sufficient fre quency for the British Medical Journal to have published an article warning carers of the possible dangers of baby shak ing and an admonishment to exercise caution. 73
Six newborn piglets receiving CPR for up to 50 minutes with monitored intrathoracic and intracranial venous pres sure showed no evidence of retinal bleeding.69
EPI LE PTIC SEIZURES
In our hospital, we examined a series of children admitted with convulsions. 70 The aetiology of the seizure disorder varied but none was traumatic. Even though 40 of the chil dren were less the 2 years of age and, therefore, in the most vulnerable age group, none showed any evidence of fundus bleeding. The children examined were all seen within 24 hours of admission, and examined in detail with an indir ect ophthalmoscope using pupil dilatation in most cases. In a more recent prospective study of 143 children aged between 2 months and 2 years who were seen in an acci dent and emergency department following seizures and examined by an ophthalmologist, unilateral retinal haem orrhages was documented in only one child. 71 As with resuscitation attempts, despite individual case reports72 it should be assumed that, if retinal haemorrhage is found in a child admitted with fits, the likelihood is that injury caused both the fits and the fundus bleeding.
BLEEDING DISOR DE RS
Various types of bleeding disorder have been shown to give rise to fundus haemorrhages. Lymphoblastic leukaemia, in particular, is recognized to cause widespread retinal bleed ing, involving all retinal layers. This occurs particularly when the children are thrombocytopenic. 74 Sickle cell dis ease, aplastic anaemias, Henoch-Sch6nlein purpura and vitamin deficiencies are all capable of producing retinal haemorrhage and must be excluded before pursuing a diagnosis of NAl.
BLEEDING SECONDARY TO INFECTION
Meningococcal meningitis was reported in 1995 as a cause of fundus bleeding?5 In our hospital, we have seen a child with massive, unilateral retinal haemorrhage (associated with fatal meningococcal disease) in whom the provisional diagnosis was NAI and the true diagnosis was revealed only by post-mortem studies.
GENETIC METABOLIC DISOR DERS
Although individually rare, a number of genetic metabolic disorders (GMDs) have been shown to cause retinal bleed ing and must be actively excluded in the evaluation of NAl victims. Galactosaemia may cause widespread bleeding into both retina and vitreous humour, particularly when associated with significant liver dysfunction?6 Less com mon GMDs have attracted attention in recent years, since it became clear that glutaricaciduria could give rise to both retinal and subdural bleeding along with cerebral atrophy. Children presenting with both subdural and intraretinal bleeding associated with glutaricaciduria were reported as long ago as 1987. 77 These reports further emphasize the need to exclude metabolic causes in all suspicious cases. Autosomally recessive protein C deficiency can also be responsible for intracranial (usually subarachnoid) and intraocular bleeding. Vitreous bleeding with severely reduced protein C levels has been reported at birth in a child of 37 weeks' gestation. 78 Other entities, such as hypertension, prematurity and AlDS-related cytomegalovirus (CMY) infection, should be apparent from the history and examination of the child on admission to hospital.
Fundus haemorrhages I
The most common type of retinal haemorrhage in infancy is birth associated. It merits separate consideration.
BIRTH -ASSOCIATED RETINAL HAEMORRHAGE
Retinal haemorrhages are found in the first few days after delivery in up to 59 per cent of children. 4B The haemorrhages clear quickly and, of 1238 children reported in 1970,49 18.9 per cent examined in the first 24 hours showed signs of having had a retinal haemorrhage. When examination was delayed until 3-5 days after birth, only 2.6 per cent showed evidence of fundus bleeding, presumably indicating that most of the haemorrhages clear in the first few days of life. Recent studies found retinal haemorrhages in about one iliird of babies following spontaneous vaginal delivery and in 75 per cent or more following vacuum extraction.79.BG The incidence of ocular bleeding following Caesarean de livery is lower, at around 7 per cent/ 9 ,BO suggesting that the raised intrathoracic and intracrania l pressures associ ated with compression of the baby in the birth canal are not the only aetiologica l factors . It has been specu lated Bl chat hypoxia and hypercapnia are also contributory. The vast majority of neonatal retinal haemorrhages are superficial and clear within 5 days?4 a less frequent but more 1V0rrying injury is a pre-retinal haemorrhage, which clears more slowly. Nevertheless, even pre-retinal haemorrhages associated with delivery should disappear within 6 weeks; fundus bleeding after this stage needs to be fully investigated .
Examining the Child With Fundus Haemorrhages
133
See at earliest opportunity after referral
Document all fin dings (including negati ve fin dings) in notes. In clude time and date of examination. Sign examination record
Do not dilate pupils without discussion with supervising pa edia tri cia n/i nte nsivis t
If dilatation not possible, record this limitation in notes
Notify consultant of findings - immediately if positive
Consultant review as soon as possible if positive findings
Photography of fundus pathology if dilatation possible
Print off copies of fundus photographs
Place in sealed envelope with note - signed by yourself and another witness - indicating the time and date of the photographs
Place sealed envelope in patient's main hospital notes
A second copy - similarly validated - retained in eye department
- - - - -. - - Specify date of review and ensure review undertaken
Though a paediatrician may be the first doctor to suspect ~etina l haemorrhages in a child, it is essential tha t an oph j]almologist is involved at an early stage (Fig. 6.8). An eye assessment shou ld be requested on any child suspected of being a victim of NAI, particularly if the head (including CNS) and neck area are involved. The responsibility of the ophthalmologist is to examine the whole of the retina (where possible), and accurately document the distribution and depth of the retinal haemorrhages along with associ a ted optic nerve and retinal features. The tool of greatest ':alue is an indirect ophtha lmoscope, w hich allows a view of the retina right out to its peripheral extent. Such a view can be achieved only through dilated pupils. Thus, if neuro :ogical monitoring precludes pupillary dilatation at first co ntact, the supervising team must be prepared to notify rbe ophthalmologist when pupil dilatation is permissible. In some instances a short-acting mydriatic, such as tropi (a mide, may be appropriate. However, in other instances, !Ja n icularly in a child with dark-brown irides, a stronger and longer-acting agent, such as cyclopentolate or ~h enylepherine, is likely to be necessary. There is no reason i...n these circumstances to use agents that cause an effect ~-or several days, such as atropine or homatropine.
Figure 6.8
A guide to the examination of children with
suspected non-accidental injury.
Once the appearances have been documented in detail, the child shou ld be reviewed on a regu lar basis, both to observe the changing appearance of the haemorrhages and to determine whether any therapeutic intervention is likely to be needed. It is not possible to date an injury on the basis of retinal haemorrhages (in the way that skin bruising can be used).61 If, however, the appeara nces change with time, knowing the ti mescale of the absorption of the observed haemorrhages may give a guide as to the timing of their original appearance. Dating the time of haemor rhage with post-mortem specimens is more practical and uses the appearance of haemosiderin. The presence of haemosiderin in the eyes of abused children has been con sidered to indicate that an injury is at least 3 days 01d. 82 To further aid the documentation of the retinal haemor rhages, fundus photography may be useful. Traditional fundus photography is unlikely to be possible, but photog raphy with a variety of hand-held cameras certainly will
134 lOcular involvement in non-accidental injury
be. The limitation on such documentation is related to the relatively small angle of view of most cameras and can be overcome using a RetCam device (Massie Research Labora tories, Inc., Dublin CAl. This recently developed digital camera allows a uniquely wide, 1200 angle of view and is combined with a telemedicine facility. 83 It is still being evaluated, but tools of this sort are sure to remove some of the uncertainty surrounding clinical descriptive recording. When the view of the fundus is compromised by vitreous haemorrhage, imaging with ultrasonography can be coupled with electroretinography (ERG) to define the location of the retina and its functional integrity. Fishman and coJJeagues 84 found ERG to be unhelpful in the initial assessment of six children with retinal haemorrhages. Their ERG parameters were not significantly different from those in six age matched control subjects. However, with time, one parameter improved, suggesting that some subtle neurophysiological damage was caused by the injury. They also stressed the importance of ERG in distinguishing between neurological and retinal visual loss at an early stage. In the presence of vitreous haemorrhage, ultrasonographic evidence of retinal detachment and ERG evidence of impaired retinal function should provoke early referral to a vitreo-retinal surgeon. Clearing the vitreous blood and surgical relocation of a detached retina will help long-term visual prognosis. Some authorities have also succeeded in clearing pre-retinal haem orrhage using intravitreal injections.85 All such procedures need to be considered when the child 's long-term visual development is likely to be compromised .
LONG-TERM VISUAL SEQUELAE The majority of victims of NAI who escape major n euro logical injury will have a fuJI visual recovery. A study of 30 child victims of shaking showed that non-reactive pupils at presentation coupled with midline shift on imaging correl ated strongly with a fatal outcome. 85 Of the 22 children who sUlvived, 18 had good vision in at least one eye, with two children having their poor vision on the basis of cor tical visual impairment. A much larger study of ophthalmic morbidity following head injury from any source and at any age 27 revealed that 48 per cent had some long-term morbidity. However, the majority of these patients had ocular motility disturbances, and only a minority had vitreo-retinal, optic nerve or cor tical damage sufficient to cause visual loss. A study of 99 children from Nigeria 87 revealed similar results, with 28 per cent having visual loss in the long term, but most com monly only one eye was involved . Significant visual impairment is difficult to divine from this series, but may have been limited to one child with cortica l visual impairment. It is clear from these studies that, despite the potential for visual loss secondary to macular folds, retinal detach ment, persistent vitreous haemorrhage and other structura l injuries to the eye, such visual loss is in fact relatively
uncommon, except in the child who has also sustained serious neurological injury.
ADDITIONAL OBSERVATIONS It is a surprising observation that, despite the nature of the injury, strictly unilateral retinal haemorrhages may result from severe shaking. In 1997 we reported three consecutive cases in which the bleeding was unilateral,25 and high lighted the fact that such a presentation should not cast doubt on the diagnosis of NAl. Two of those three children suffered severe long-term visual loss, despite the unilateral nature of their ocular injury. A review of the literature revealed strictly unilateral bleeding referred to in other series, and in those cases, too the associated neurological injury was often severe. 52 .G3 The presence of unilateral eye signs does not diminish the likelihood that the injuries are a consequence of abuse; nor do they predict a more minor level of eye and neurological damage. Non-accidental injury is a harrowing topic for all clin icians. We owe it to the children in our care to consider it as a diagnosis to avoid the child being exposed to further injury. However, we also owe it to the family to ensure that the vista of NAI is only raised in appropriate circumstances. Detailed examination of the eye and modern documen tation wiJJ contribute to the diagnosis, but none of the ocu lar features can, in isolation, establish a diagnosis of NAI . Despite recent reports, domestic accidents rarely, if ever, cause the level of ocular damage that is commonly associ ated with severe shaking or impact injuries.
REFERENCES 1 Voutilain en R, Tuppurain en K. Ocular Munchausen syndrome induced by incest. Acta Ophthalmol (Copen) 19 89; 67:3 J9-21.
2 RadbiJi SX. A his tory of child abuse and infanticid e. In Helfer RE, Kempe CH (eds) The Battered Child. Chicago, IL: University of Chicago Press, 1968. 3 Kempe CH , Silverman WA, Steele BF et al. Th e battered child syndrom e. JAMA 1962; 181:17-24. 4 Cam eron JI\1 , Johnson HR, Camps FE. The battered child syndrome. Med Sci Law 1966 ; 6:2-7. 5 Johnson CF, Showers 1. Injury variabl es in child abuse. Child Ab1lse Negl 1985; 9:207-11. 6 Hendeles S, Barber K, Willshaw HE. The ri sk of ocula r involvement in non- accidental il\iuly. Child Care Health De)) 1985; 11 :345-8.
7 Riffenburgh RS, Sathyavagiswaran L. The eyes of child abuse
victims: autopsy findings. J Forensic Sci 1991 ; 36:741-7.
8 Rao N, Smith RE , Choi JH et al. Autopsy findings in fourt ee n
fatally abused children. Forensic Sci 1I1t 1988; 39:293 - 9. 9 Hadley MN, Sonntag VKH, Rekate HZ, Murphy A. The infant whiplash-sh ake injury syndrome: a clinical and pathological study. NeurosLiTgelY 1989; 24:536-40. 10 Duhaime A- C, Gennarelli TA, Thibault LE et al. The sh aken baby sy ndrome. A clinica l, pathological and biomechanical study. J Neurosurg 1987; 66:409-15.
References I
11 Jacobi G. Damage patterns in severe child ab use with and
without fatal seq uelae. Monatssch r Kinderl1eilkd 1986; 13 4: 307-15. 12 Harcourt B, Hopkins D. Ophthalmic man ifestations of the batte red-baby sy ndrome. BM} 1971; 3 :390 . 13 Feldman KW, Schaller RT, Feldma n JA. Tap water scald burns in children. Pediatrics 1978 ; 62: 1-9. 14 Landers A, Belfer KF, Jenki ns AD. Supergill e Jids: possibly non acc idental and a medicol ega l problem. Eye 1996; 10:402 -3. 15 Tom inack RL, Spy ker DA. Capsicum and caps aicin - a review:
case rep ort of the lise o f hot peppers in child abuse. } Toxicol Ciin Toxicol1987; 25:591-60 1. 16 Tse ng SS, Keys MP. Battered child syndrome simul ating congenital glaucoma. Arch Ophthalmol 1976; 94:8 39-40. 17 Fri en dl y DS. Ocu lar manifestations of phys ical child ahuse. Tran s Am Acad Ophthalmol OtolarYllgol 19 71 ; 7 5:318 - 3 1. 18 Weidenthal DT, Levin DB. Retinal detachment in a battered infant. Am J Ophthalmol 1976; 81:725-7. 19 Gonzales CA . Bilateral rhegmatogenous detachments with unilateral vitreo us base avu lsion as the presenting sig ns of child abuse. Am} Op hthalmo/ 1999 ; 127:475-7. 20 Green MA, Lieberman G, Milroy CM, Parso ns MA. Ocu lar and ce rebral trauma in non- accidental injury in infan cy : underlying mecha nisms and implications for paediatric practice. Br} Oph tl1OImoi 1996 ; 80:282-7. 21 Massic ote SJ, Folberg R, Torc zynski E et al. Vitreoretinal traction and peri macular retinal fold s in the eyes of deliberately tra umatized children. Ophthalmology 199 1; 98:1124-7. 22 Munger CE, Peiffer RL, Bouldin TW et al. Ocular and
associated neuropathological observa tio ns in suspected whiplash shaken infan t syndrome. A retrospective stu dy of 12 cases. Am} Forensic Med Pathol 1993; 14:193-200. 23 Giangiacomo J, Barkett KJ. Ophth alm oscopic findin gs in occ ult child ab use.} Pediatr Ophth almol Stra bismus 1985; 22:234-7 . 24 Levin AV. Oc ular mani festa tions of ch ild abus e. Ophthalmol Clin No rth Am 1990 ; 3 :249 -64. 25 Tyagi AK, Willsha w HE, Ai nswo rth JR. Unilateral retinal haemorrh ages in non- acc idental injury. Lancet 1997; 349:1 224. 26 Lin KC, Glasgow BJ. Bilateral peri op tico intrascleral
27 28 29
30 31 32 33
34
hemorrhages associated with traumatic child abuse. Am} Oph tlw lmol 1999; 12 7 :473- 5. Ktitzinger EE. Pers ona l com mun ication. Kowal L. Ophthalmic manifestati ons of head injury. Aust NZ } Oph thalmol 199 2; 20:35-40. Frank y, Zimmerman R, Leeds NM . Neurologic manifestation in abused children who have bee n shaken. Dev IVIed Child Neurol 1985 ; 27:31 2-1 6. Jensen AD, Smith RE, Olson Ml. Ocular clues to child abuse. } Pediao' Ophthalmol Strabismus 1971; 8:270- 3. Ball anty ne AJ. Th e nerve fibre pattern of th e huma n retin a. Trail S Ophtha lm ol Soc UK 1946; 66: 179-91. Roth M. Uber Netzhau ta ffection en bein Wund feibe rn. Deutsch Z ChiI' 1872 ; 1:471-00. Falcone PM, Larrison WI. Roth spots see n on ophthalmoscopy: diseases with whi ch they may be associated. Conn Med 1995; 59:271-3. Kaur B, Taylor D. Fundus haem orrhages in in fancy. SurIJey Ophthalmo/ 1992 ; 37 :1- 17. Fred HL, Chandler FW. Traumatic asphyxia. Am} IVIed 1960;
135
38 Va ega n, Taylor D. Critical peri od for deprivation amblyopi a in ch ild ren. Trall s Ophthalmology Soc UK 1979; 99 :432-9. 39 Mathews GP, Das A. Dense vitreo us hem orrh ages predict poor
visual and neurologica l prognosis in infants with shaken baby syndrome.} Pedia tr Ophthalmol Strabismus 1996; 33:2 60 - 5. 40 von Noorden GK. New clini ca l aspe cts of stimulus deprivat ion a mblyo pia. Am} Opllthalmol1981; 92:416-21. 41 Ober RR. Hem orrhagic retin opat hy in infancy: A clinico patho log ical report.} Pedia tr Ophthalmol Strab ismus 1980; 17:17-20. 42 Lambert SR , Johnson TE, Hoyt CS. Optic nerve sheath
hemorrh ages assoc iated with shaken baby sy ndrome. Arch Ophthalmol 1986; 104: 1509. 43 Mun ger CE, Peiffer RL, Bouldin TW et al. Ocular and associated neuropat hologic observ ations in suspected w hipl as h baby sy ndrome: a retros pective study of 12 cases. Am} Forellsic IVIed Pathol 1993; 14: 193- 200. 44 Mass icote SJ, Folberg R, Torczynski E et al. Vitreoretinal traction and perimacular retin al folds in the eye of deliberately traumatized children. Oph thilimology 1991 ; 98: 1124-7. 45 Greenwald MJ, Weiss A, Oesterle (S, Friendly DS. Traumatic retinoschisis in battered babies. Ophthalmology 198 6; 93:618-25. 46 Riffe nbllrgh RS, Laks hmanan S. Ocular findings at autopsy of child abuse victi ms. Ophthalmology 199/ ; 98: 1519-24. 47 Tomasi LG , Rosman P. Purtscher retinopat hy in the battered child syndrome. A m) Dis Child 1986; 93:1435-9. 48 Ba um JD, Bulpi tt CJ. Retina l and conjun ct ival haem orrhage in the newborn. Arch Dis Child 1970; 45:344-9. 49 Sezen F. Retinal haemorrhages in newborn infants. Br} Opllthalmol 1970; 55: 248-53 . 50 Wilkinson WS, Han DP, Rappley MD, Owings CL. Retin al
hemorrhage pred icts neurolog ic injuty in the shaken baby sy ndrome. Arch Ophtlwll1lol 1989; 107:147 2-4. 51 Paton L. V 11 diseases of the nervo us system: ocular sy mptoms in subarachn oid haemorrhage. Trans Ophthalmol Soc UK 192 4; 44: 110-26. 52 Schlo ff S, Mullaney PB, Armstrong DC et al. Retinal findings
in children with intra crania l haemorrhage. OphtllGlmol 2002; 109: 1472-6.
Duhaime A-C, Gennarelli TA, Thibaul t LE et al. Th e shake n ba by syndrome - a clini cal, pathological a nd biomecha nical study.} Neurosu rg 1987 ; 66:409-1 5. 54 Gilliland MG, Fo lberg R. Shaken babies: so me ha ve no impact injuries.} Fore nsic Sci 1996; 41:/14-1 6. 55 David DB, Mears T, Quinlan MP. Ocul ar co mplications of bllngeejumping. Br} Op/lthalmol 1994; 70:234-5. 56 Lyle DJ, Sta pp JP, Button RR. Ophthalmologic hydrostatic press ure syndrom e. Am J Opllthalmo/1 95 7; 44: 652 -7. 57 Christian CW, Taylor AA, Hertle RW, Duhaime A- C. Retinal haemorrhages caused by acc idental household traum a. } Pediatrics 1999; 135:1 25 -7. 58 Buys YM, Levin AV, Enzenauer RW et al. Retinal findin gs after head tra uma in in fants and young children. Ophthalmology 1992; 99: 1718-23. 59 Billmire ME, Myers PA. Serio us head injury in infants: accident or ab use. Pediatrics 1985; 75 :340- 2. 60 Elder JE, Taylor RG, Klug GL. Retina l haemorrhage in accidental hea d traum a in childhood. } Pediatr Child ffealth 53
29:508-17. 36 Holl en horst RW, Stein HA. Ocular signs and prognosis in
1991 ; 27:286. 6 1 The Ophth almology Chi ld Abuse Wo rkin g Patty. Child abuse and th e eye. Eye 1999; 13 :3-10. 62 Budenz DL, Fa rber MG, Mirchandani HG et al. Ocular and
subd ural and sub arach noid bl eeding in young children. Arch Ophthalmol1958; 60:187-92. 37 Mushin AS. Ocular damage in the battered baby sy ndro me. BM} 1971; 3:40 2.
optic nerve hemorrhages in abused in fa nts with intracranial injuries. Ophthalm ology 1994; 101 :559-64. 63 Riffenburgh RS, Lakshm anan S. Ocular findin gs at autopsy of ch ild abuse victims. Ophthalmology 1991; 98:15/9-24.
35
136 l
Ocular involvement in non-accidental injury
64 Adams G, Ainsworth J, Butler Let al. Update from the child abuse working party: Royal Coll ege of Ophthalmologists. Eye 2004; 18:795-8. 65 Kramer K. Retinal hemorrhages following cardiopulmonary resuscitation. (lill Perliatr 1993 ; 13 :366-8. 66 Kanter RK. Retinal hemorrhages after cardiopulmonary resu scitation or child abuse. ) Pediatr 1986 ; 108:430-2. 67 Gilliland MG, Luckenbach MW. Are retinal hemorrhages found after resusc itation attempts? A study of the eyes of J 69 children. Am) Forensic Med Pathol 1993; 14:187-92. 68 Odom A, Christ E, Kerr N et al. Prevalence of retinal haemorrhages in pediatric patients after in-hospital cardiopulmonary resusc itation: a prospective study. Pediatrics 1997; 99:E3. 69 Fackler JC, Berkowitz !D, Green R. Retinal hemorrhages in newborn piglets foll ow ing cardiopulmonalY resusc itation. Am) Dis Child 199 2; 146:1294-6. 70 Tyagi AK, Scotcher S, Kozeis N, Willshaw HE. Can convulsions alone cause retinal haemorrh ages in infants? Br) Ophtha/mol 1998; 82:659-60. 71 Mei-Zahav M, Uziel y, Ra z J et al. Convulsions and retinal haemorrhage: s hould we look fUJther. BM) 2002; 86:334-5. 72 Feyi-Waboso AC, Bec k L. Minerva. BM) J997; 314:688. 73 Markovitch H. Opening eyes to child abuse. BM) 1999 ; 318:950. 74 Guyer DR, Sch achat AP, Vitale S et a1. Relationship between fundus lesion s and haematological parameters at diagnosis. Ophthalmology 1989; 96:860-4. 75 Frase r SG. Retinal haemorrhage in meningitis. Eye 1995; 9:6 59 -60. 76 Levy HL, Brown AE, Williams SE, de Juan E. Vitreous hemorrhage as an ophthalmic complication of galactosemia. ) Perliatr 1996 ; 129:922 -5.
77 Hoffman GF, Nau ghten ER. Abuse or metabolic disorder. Arch Dis Child 1998; 78:399. 78 Pulido JS, Lingua RW, Cristol 5, Byrne SF. Protein C deficiency associated with vitreous hemorrh age in a neonate. Am) Ophthalmol J987; 104:546-7. 79 Emerson MY, Pieramici OJ, Stoessi K]\il et al. Incid ence and rate of disa ppearance of retinal haemorrhages in newborns. Ophthalmology 200J; 108:36-9. 80 Hughes LA, May K, Talbot JF, Parsons MA. Incidence, distribution and duration of birth-related retinal haemorrh ages : a prospective study. ) AAPOS 2006 ; 10: 102-6. 8l Von Barsewisch B. Perilla tal Retinal Haemorrhages: iVlorplw/ogy, Aetiology and Significance. Berlin: Springer Verlag, 1979. 82 Gilliland MG, Luckenbach MW, Massicote SJ, Folberg R. The medicolegal implications of detecting hemosiderin in the eyes of children who are suspected of being abused. Arch
Opi1thaimoI1991; 109:321- 2. 83 Schwartz S. Telemedicine and ophthalmology Jules Stei n Eye IJlst: Clin Update 1998; 7:3-4. 84 Fishman CD, Dasher WB, La mbert SR. Electroretinographic findings in infants with shaken baby syn drome. ) Pediatr Ophthalmol Strabismus 1998; 35:22-6. 85 Conway MO , Peyman GA , Recasens M. Intravitrea l tPA and SF6 promote clearing of premacular subhyaloid hemorrhages in s haken and battered baby syndrome. Ophthalmic Surg Lasers 1999; 30:435-41. 86 McCabe CF, Donahue SP. Prognostic indicators for vision and mortality in sh aken baby syndrome. Arch Ophthalmol 2000; 118 :3 73-7. 87 Shokunbi T, Agbeja A. Ocular co mplications of head injuly in children. Childs NeTlJ System 1991; 7:147-9.
I
CHAPTER 7
I
THE DEATH SCENE FOLLOWING THE SUDDEN DEATH OF A CHILD Anthony Busuttil
Introduction Scene management The crime scene manager Sequence of events at the death scene Unclothing the body A good look around Sudden infant death syndrome or non-sudden infant death syndrome External petechiae
137 137 138 139 139 139 140 140
INTRODUCTION
The proper, co mprehensive, planned and professional investigation of a suspicious death of a child requires a carefully managed, orde rly and thorough inspection of the scene where the death has occurred, wit h its full documen ta tion. J This should take place whether the bo dy of the child is still in situ or whether it has already been removed. Th e scene investigation forms as much an essential part of ~he death investigation as does the autopsy and an access ~ o all the medical records of the deceden t. The patho logist investigating the death should always be given app ropriate access to the scene, as this will assist fur iher in the reconstruction of the events that led to the death of the ch ild. 2 ,3 Access to the scene may be required again a fter the au topsy has been completed to attempt to identify -le vario Lls sites at which blunt trauma may have been SLlS (ained. If a cou nter- argument of an accidental infliction of Jl e injuries is being pursued, for example rolling over from a table or couch or a fall down stairs, then it is essential that " is hypothesis be tested at the scene in the exact location -,';here it is suggested to have occurred. This reconstruction should a lso involve the forensic pathologist. On occasions, carers of a baby or child who has died - nexpectedly give a complex narrative of the events that
Bruising Abandoned neonates Deaths from trauma Dyadic and multiple deaths Sudden deaths of older children Sensitivity and stress of the investigation Inquests and inquiries References
140 140 141 141 142 142 143 143
they feel have led to the injuries an d the demise of the child. The police have often found it useful to encourage these wit nesses to re-enact quite carefull y and meticulously at the actual scene what t hey are suggesting has taken place, using dolls as substitutes for the child or infant. These re-enactments, which ha ve been resorted to more frequen tly given the con troversies that may arise about the manner of whi ch head and neck injuries have been sustained, are videoed by the police. Forensic pathologists should view these and comment thoughtfully on the plausibility or otherw ise of what is being suggested. It has also been known for legal defence personnel to do this; if it is known that this tactic will be employed, it is essential that the pathologist sees such reconstructions in good time before the case comes to court.
SCENE MANAGEMENT It has to be decided at a very early phase, after tile deatll has been brought to the attention of the police, whetller it is thought that the death under investigation is suspicious or not. It is always the counsel of perfection to be safe rather than sorry, and to upgrade - at least initially - the investigation of a scene to suspicious status when one is unsure. This decision may require an early input from a forenSically trained medical
138 I
The death scene following the sudden death of a child
practitioner, though not necessarily a pathologist, and the viewing and careful inspection of the body. In situations that speak for themselves, in particular when evidence of violence is present, this, of course, can be dispensed with. The fact that the body of the child may already have been taken to hospital for attempted resuscitation does not dispense with the scene examination, and a close retrospective inspec tion of the scene where the death has occurred is called for at the earliest possible opportunity. Once the attention of the police has been drawn to the occurrence of a suspicious, vio lent or unexpected death, every effoli should be made to return to the scene where the death took place. In these circumstances, it is essential that hospital staff and emergency personnel are interviewed soon after the event to obtain from them informa tion regarding: the findings on the child when brought into hospital , dead or moribund; the results of any emergency bio chemical , haematological and radiological investigations car ried out; and the various aspects of the resuscitative process that have taken place, including drugs administered to the child. It is also useful to obtain access at an early stage to - and indeed to secure seizure of, against appropliate receipting any samples collected from the now-deceased child prior to any medication or intravenous infusions having been admin istered. These so-called 'pre-transfusion specimens' may prove extremely useful for further analysis. Similarly, if clinical photographs have been taken of the injuries of the child then police should obtain access to these. Any clothing removed from the now-deceased child, including nappies, should be handed over to the police. Slightly more controversial decisions may have to be taken, as to when and whether information should go to the police, in cases in which a child is admitted with severe injuries and is likely to die as a consequence. If injuries sus tained are serious and, perhaps, 'are likely to prove' fatal, a decision to inform the police or not has to be taken at the earliest possible time by the clinical staff. There should be no dubiety in the minds of the doctors in accepting that this would not constitute a breach of medical confidentiality due to the patient giving their oveniding obligations as cit izens of a country to ensure that crime is investigated and its perpetrators detected. A senior member of the medical team should make this disclosure to the police, speaking to a senior police officer. Hospitals should produce written protocols about the procedures that should be followed in such circumstances. The child's carers must be told that the police have been informed.
THE CRIIVIE SCENE MANAGER The scene will often be placed under the control of a spe cially trained scientist or police officer designated as the crime scene manager (CSM). The CSM's g'uiding principle is that enunciated by Edmond Locard over a hundred years ago - namely that every contact between two persons or a person and a scene will leave a trace of this contact, which
if looked for and found, thereby proves that such a co ntact had indeed occurred. 4 The agenda set before the CSM comprises the following: 1. From the time the police first attend the crime scene, an accurate and detailed record must be maintained of the comings and goings and of who did what and when. At an early phase, thought should be given to a retrospective timetabled account of events occurring prior to the arrival of the police. 2. That the scene itself be documented fully, both photographically and by video recording, and, if required, by sketches to sca le also. It is an essential good practice for individuals attending the scene to make their own notes and to make their own rough sketches as aide-memoirs. Such documentation will enable 'best evidence' to be produced in front of the courts. All this documentary material will form part of documentary disclosure when a criminal procedure is in force. The photographs and videos of the scene will also be available later to refresh the minds of those working on the case, which allows better briefings of further additional personnel joining the investigative team. J. All trace evidence has to be collected painstakingly from the scene. A systematic, planned and thorough search has to be made of the entire scene and, initially, everything that may have some relevance has to be co llected. To be exclusive and selective in evidential collection at an early phase of an enquiry may be shown later to have been counterproductive and inefficient. The approach to the collection of evidence should be that each item collected is logged and a 'continuity of the chain of evidence' is established for every item. This refers to the progression of each item from the scene to the forensic laboratory and eventually to couli; each step of this procedure has to be accounted for fully and not found to be wanting. Every person handling the item of evidence signs the original label attached to it at the scene. Strict attention has to be given to ensure that appropriate containers, methods of preservation, labelling and packaging are used. The collection, labelling and logging of the items collected is delegated to a police officer designated as the exhibits' (or productions', in Scotland) officer. 4. All relevant investigative personnel should be given access, in turn, to the scene, ensuring that they are all wearing appropriate protective clothing in the course of their attendance to decrease scene contaminati on . They will be briefed to work in unison and in mutual collaboration; altho ugh, in order to avoid congestion and overcrowding at the scene, each team may have to work separately, with the crime scene manager present at all times. The pathologist's role at the scene should principally concern the body. However, even if the body has left the scene, there is often still scope for the pathologist to inspect the scene of death, particularly if it is being treated as a suspicious death.
A good look around I
5. It is essential that there is adequate lighting at the scene. It is foolhardy to attempt to carry out decent and proper investigations without appropriate lighting. The crime scene manager should ensure that portable generators or other sources of light are made available. The commencement of the investigation may have to await the arrival of such equipment and if possible daylight. 6. Information is sought from the scene that will assist with an estimation of the time of death; this requires a recording of the ambient temperature. If death occurred indoors and sequential changes to room temperature over time are known to occur (e.g. central heating cycle), these should be appropriately logged. If the death occurred out-of-doors, meteorological reports should be sought for the area to assist with such estimates. These will have to include details about humidity, ground temperature, wind-chill factor and rainfall. 7. Exposure to carbon monoxide should be considered in all domestic childhood and infant non-violent deaths, and any potential sources of this gas, for example a blocked chimney, faulty kerosene heaters, blocked gas flue and coke burners, should be carefully looked for and expert advice sought about any possible malfunction. 8. If possible, the scene should be secured after all the searches have been carried out, as it may become necessary to return to the scene at a later time to check over certain facts or to look for other evidence; new additional members of an investigative team also may wish to see for themselves the scene of the fatal incident.
SEQUENCE OF EVENTS AT THE DEATH SCENE Once the crime scene manager has been designated, the sequence of events at the death scene often follows the fol lowing pattern of activity:5,6 1. If the body is still in situ, establish a path to the body that will be adopted by all those wanting to inspect the body, this is referred to as th e Co 111 1110 17 Approach Pathway (CAP). 2. The CAP may have to be demarcated and preserved
by elevated platforms or stepping plates that have
bee n thoroughly cleaned previously.
3. If in the open, construct protection (e.g. tent) for the body and its immediate surroundings. 4. Set up a written log of named comings and goings to the scene. 5. (A medical practitione r or a suitably trained
paramedic formally pronounces life extinct).
6. Provide adequate lighting or wait till daylight. 7. Photographs a nd video team document the scene
fully.
8. Appoint an exh ibits' officer.
139
9. Specially trained 'scenes of crime' officers inspect the scene and collect trace evidence. 10. Forensic scientists inspect the scene and collect evidence. 11. Pathologist is given access to the body. 12. Trace evidence from the body, in terms of tapings from the exposed surfaces of the body (for fibre evidence), is collected, and swabs are taken from accessible orifices. 13. The body is transferred on to a clean cotton or polythene sheet and tra nsferred to the mortuary for more detailed inspection. 14. If required, finger print experts cover the scene. 15. The scene is secured. 16. Further examinations as required - all fully logged.
UNCLOTHING THE BODY This should be avoided at the scene as it may lead to loss of evidence, to potential contamination and perhaps also cross-transference of evidence to it. It is a better policy to postpone the unclothing till later, and for this to take place under controlled conditions, in a good light, in a mortuary. This would enable each item of clothing to be removed sequentiaJJy and to be carefully preserved. Contents of pockets can be removed, identified and considered, and areas of damage to items of clothing can be identified and considered, as each layer of clothing is removed in sequence. In babies, the clothing is of particular importance in attempting to assess whether the baby has been subjected to excessive heat;l headgear, in particular, retains a signif icant amount of heat. If the clothing is soaked in sweat it is a very good indication of excessive heat exposure (or of a recent convulsion or hypoglycaemia). Inadequate, flimsy or wet clothing worn by the child in a cold bedroom may be similarly of importance if a death from hypothermia is being considered.
A GOOD LOOK AROUND From the room in which the body is found, the search has to proceed to every other room in the house in a systematic fashion. In addition to looking for other evidence, a good look around the house will give a better appreciation of the lifestyle of the parents and the other inhabitants of the house, and also the prese nce of pets. The police will be alert to signs of chronic controlled substance misuse and alcohol misuse or overuse. This would also mean some degree of attention to sites of waste disposal and garbage containers, to gardens and outhouses. Special care must be taken in examining any bathrooms and kitchens. In the latter, any partially consumed food or made-up bottles of formula milk should be retained. These
140 I
The death scene following the sudden death of a child
should be refrigerated to prevent fermentation, which leads to spurious results of ethanol estimations. They can also be tested for medical and illicit drugs as well as alcohol. Any medication found within the household, in particu lar antipsychotics, antidepressants, benzodiazepines, anti convulsants and controlled drugs, should all be noted and removed.
SUDDEN INFANT DEATH SYNDROME OR NON-SUDDEN INFANT DEATH SYNDROME The most common cause of post-perinatal motility in the western hemisphere is still sudden infant death syndrome (SIDS). However, as the incidence of SIDS has declined worldwide,S the number of deaths related to abuse and neglect has increased proportionally, and the task of distin guishing a natural fr om a suspicious dea th is becoming increasingly frequent. 9,1O,1I In such instances, a thorough scene investigation assists greatly in reconstruction of the events that preceded the death. 12. IJ ,14 The contribution that death-scene examinations make in relation to the definition of such infant deaths has been a source of some controversy. In the 1969 defini tion of SIDS produced by Beckwith,15 there is no mention of the scene inspection. However, in 1976 there was a recommendation by an American Working Party in their Special Report that scenes of death should be visited. 16 In his second attempt at more elaborately defining SIDS, during the second SIDS International Conference held in Sydney in 1992, Beckwith suggested the introduction of a scene investigation into a three-tiered definition; this also failed to meet with general approva l. During the second SIDS Glob al Strategy Meeting held in Stavanger, a further attempt at making a satisfac tory scene investigation an essential criterion for the diag nosis of SIDS also failed.17 However the National Institute of Child Health and Development's (NICHD's) definition of SIDS has, since 1989, required that an examination of the death scene be performed before a diagnosis of SIDS can be made. IS The definition of the scene has been widened 3 to incor porate not only the location where the infant was discov ered dead or unresponsive, but also the no rmal home environment of the child and any other locations that may have been visited by the child immediately prior to the death. Protocols have been published from the United States outlining the scientific and medical input in such scene investigations. 19 If the baby has been found dead in a cot, crib, Moses bas ket, or something similar, then it is important to discover the position in which the baby was found prior to the atten da nce of ambu lance crews and paramedics. Life-sized flex ible dolls have been made use of, in this respect, to enable the carers of the child to recollect the initial discovery of the child. Well over 50 per cent of babies dying of SIDS are found in the prone sleeping position. s This reconstruction is
even more important, if the baby has been co-sleeping, in order to assess the possibility of overlaying. 2I ,22 It is important to retrieve the bedding in order to dis cover whether this was stained wi th blood or saliva or was covered in vomit, and also to assess whether it could have made the baby's environment too warm or, ind eed, inap propriately cold. When various theories related to the possible production of inespirable and toxic gases from cot mattresses were being investigated, it was considered important to retain the mat tress for appropriate testing.2J,24 This is no longer necessary. Great caution must be exercised, if on investigation it is discovered that the 'cot death' under investigation is the second, or even more concerning, the third sudden infant death in the same family. Greater suspicion and , therefore, vigilance should be entertained in these circumstances. 25 ,26
EXTERNAL PETECHIAE Although there is some dispute as to the percentage inci dence of the presence of petechiae on the serous membranes of children dying of SIDS, there is no doubt that th e presence of extern al petechiae in the face, behind the ears, in con junct ival membranes and on t he mucous membranes of the lips and gums of a baby dying of SIDS is exception al. 27 ,2B It is a simple enough test to carry out in a good light with the baby in situ. If petechiae are found localized to the upper t runk and face, in the absence of exposure to meningitis or known coagul ation problems, then that par ticular baby's death has to be treated as suspicious; it should be considered a result of mechanica l asphyxiation, an accident or a criminal act, until proven otherwise by the investigation. Accidental mechanical asphyxial deaths do occur in babies and young children and only an examination of the scene of death would enable the mechanisms of death in such instances to be elucidated. 29
BRUISING The presence of bruises on children, and particularly on babies, is always a cause for concern and suspicion, particularly in children who have not started to toddle and if bruises are of different ages . Bruising in unusual sites, such as the back, arms, neck, chest, thighs and genitals, is also particularly worrying in infants.
ABANDONED NEONATES The abandoned body of a neonate is still a matter that greatly exercises the sensitivity of the public and the interest of the media. On few occasions, it is the consequence of homicide. Nowadays, the emphasis is clearly biased towards
Dyadic and multiple deaths I
attempting to trace and then assist the mother of the aban doned baby, addressing any medical, social welfare or finan cial problems being experienced by the mother, rather than in seeking punishment. The potential for a criminal tlial, however, still exists, particularly if the body shows features of injury and the mother is eventually traced. Therefore, the police have to investigate the matter and carefully search the scene where the body is discovered. The incidence of neonate abandonment varies between countries but even in Western countries such instances are not exceptionaI. 2o ,21,JO,JI It is often the case that the baby has been concealed prior to its disposal. Although putre factive changes are delayed in newborn infants, owing to the sterile environment from which they have been deliv ered and due to the absence of bacteria in their intestines , some decomposition, albeit incomplete, may be present. Over time, there also may be partial mummification of the body in dlY and well-ventilated surroundings. Often the body of the child has a lready been attacked by animals and bears the marks of this. These have to be distinguished fioman~-mortemtrauma.
In all these instances it is important to ensure that a sea rch of the scene is made for the placenta, and for any other items that may have been discarded together with the baby, for example distinctive baby clothes, newspapers and wrappings; this may give a clue to the locus at which the baby was delivered. Occasionally, notes may have been pinned to the body and sometimes items of jewellelY left with it; forensic examination of these may ass ist with trac ing the mother. This may be indispensable in the identifi cation of the abandoned body, which is often a major problem and often rema ins unsolved.J2.J 3 Infanticide has been committed in a number of ways, such as intentional neglect, exposure, killing with a weapon (e.g. stabbing or by the head being struck against a solid surface). suffocation (by the insertion of foreign material into the mouth or with a soft pillow) and by drowning (see Chapter 18). In all instances, it may be use ful to attempt to distinguish whether trauma at the time of birth was brought about by inexpeJiise in delivery or by precipitate labour from intentional injury infliction. 34 Occasionally, abortuses and products of conception are found and the police may have to investigate. It is essential, in the first instance, to ensure that one is indeed dealing with human remains. Products of conception of non-human origin may closely reassemble human material, for example rabbits and other small mammals. Some years ago, in many co untries, toy manufactures produced android toys, which had the same size and general appearance of a fetus, and many a n investigation had to be abruptly termi nated with a ceJiain amount of 'red face' all around when it was discovered that it was not a human fetus which had been discarded but one of these toys, sometimes as a prac tical joke, in poor taste, or simply accidentally. Another common finding is the presence of bones of babies that appear to have been lying undisturbed for
141
decades and which come to light with renovations and dem olition of old buildings (see Fig. 10.2, p. 183), a remnant of the days where the birth of an illegitimate child was a mat ter of public shame and dishonour and contraception was primitive. These cases are velY difficult to investigate, not least because with the passage of time the cause of death and indeed the potential perpetrator cannot be identified. J5 An attempt will usually be made to assess the age of the skele ton to determine whether it is pre- or post-viable.J 6,37
DEATHS FROM TRAUMA A careful review of the scene is a sine qua non in traumatic deaths, particularly when different scenarios have been alluded to in statements given by those who were there when the injUlY took place. The potential veracity of one or other scenario has to be investigated by a careful inspec tion of the scene, with measurements being taken as appro priate22.J8,39 A debate still rages over whether a head impact against a solid unyielding surface, causing acute deceleration, is essential to produce the pathological changes that a re ascribed to the 'shaken baby' syndrome.4o ,41 Evidence for such an impact may be scanty, such as small indents in fur niture or the deposit of a sma ll amount of hair and skin from the baby, but shou ld be carefully sought.
Deaths in Fires In Britain, particularly in Scotland,42 there are a number of deaths every year in which children are killed by household fires . This incidence has hardly changed over the years despite major educational attempts by fire brigades and cen tral governments to ensure that houses are fitted with work ing smoke alarms, and discouraging the cultures of smoking in bed or cooking while under the influence of alcoholY,44 The deaths may be accompanied by adult deaths,45,46 and it is frequently the case that smoke inhalation is the cause of the death. The children may be too young to be able to find their way to safety or may become disorientated and terrified at an early phase and succumb to smoke inhalation before reaching safety. The possibility of drug or alcohol admin istration to children in this circumstance should not be forgotten. The dynamics of the commencement and spread of the fire and the use of accelerants will be reported on by the special ized fire officers and the forensic team that attends the fire. In all such cases, the possibility of arson should be kept in mind.
DYADIC AND MULTIPLE DEATHS A dyadic death is one in which the perpetrator of a homi cide later commits suicide,47 often at the same site where
142 I
The death scene following the sudden death of a child
the homicide occurred, with multiple bodies found in the same scene. Many of these instances of homicid e cum sui cide are th e consequence of severe, occasionally previously undiagnosed, mental disorde rs, such as severe dep ression or paranoid schizophrenia. In many instances, there is an acute, emotional trigger for such occurrences, for example bankruptcy, prosecution of the fath er for some alleged offence, the discovelY of a n extramarita l affair, an impend ing threatened sepa ration or divorce. Men and women can both be involved in such deaths. Often, an entire family can be eliminated in the same event. Mechanical asphyxial methods, fire arms and hefty sharp weapons, suc h as axes or machetes, are made use of. The wea r and tear of modern living appears to have taken its toll, resulting in an increasing incidence of such deaths. These scenes are particularly harrowi ng for the investigators, but in all these instances it is essential to ensure that there need not be a search for another person. Even more harrowing and soul-destroying are scenes in which many children have met their death together in a violent manner. In major incidents of natural or unnatu ra l causation, children are often among the fat a lities. It is, however, the exception al and unusua l situation in which all the victims in a majo r incident are children. 48 The noto rious exception in this co unt ry was the incident in Dun blane primary school; othe r countries, such as the USA have had to deal with several simil ar incid ents . These inci dents have to be dealt with as if one is dealing wi th multi ple single homicid es, each decedent being treated individually as a vic tim of homicide. The scene in its entirety, no matter how large, becomes a scene of crime.
SUDDEN DEATHS OF OLDER CHILDREN PsychiatIists are beginning to build up a better profile of child sex abuse rs and killers. Most child murders are carried out by fami ly members - often mentally ill. In other cases the perpetrators have been abused as a child themselves. However, there are also a small number of paedophilic crim inals responsible for child homicides. In such cases, the chil dren involved are older children and the motive for their murder is sex ual g rati fication of the perpetrator. These clim inals will do their best to conceal their crime and it is often the case that the body of the deceased child will not be fo und until an exhausti ve search has been carned out; when the body is eventua lly found, it is often in a state of decomposi tion. The exa mination of such scenes is even more fraught and requires a meticulous attention to detai l. In order to glean as much information as is possible fro m the sites where such bodies are deposited, in an effort to solve the crime, it is often the case th at the expeltise of a number of other experts may have to be called into play. These include, for example, arc haeologists a nd anthropol ogists, soil experts, experts in pollen and entomologists. The final resting place of the body may well differ from the
site where the murder has been perpetrated or whe re the body has been stored prior to its disposal. In such instances, more than one scene will have to be carefully examined once these have been identified, each being organized and supervised by a different CSM. To enable a ll the various experts to congrega te at the scene, it may be necess ary to retain the body in situ for a lengthy period. occasionally of several hours' duration. The public media may suggest that this is a callous and insen sitive way in which the police de al with such gri m discov eries ; from personal experience, there is no substitute to ensuring that such investigations are thorough, compre hensive and as all-encompassing as is possible. Delay in carrying out the autopsy is a sm all penalty to pay for this. Id entification of the victi m becomes paramount in such situations, and the assistance of forensic odontologists and deoxyribonucleic acid (DNA) exp erts are invaluable in suc h instances.
SENSITIVITY AND STRESS OF THE INVESTIGATION Deaths of babies and children are becoming increasingly uncom mon and thus when such a death occurs, particu larly if it is sudden or violent, it wi ll cause major emotional upheaval and grief in the immediate family. If there is an index of suspicion that has to be investigated and excluded then great care must be taken to ensure that the fami li es are treated with resp ect and sensitively at all times, even when there is high index of suspicion re lating to their direct involvement in the death . The police have, over recent years, trained officers as 'family li aison officers' who can speak to and communicate with the families, keeping them informed of the progress of the investigation. There wilJ be a certain amo unt of antag onism and animosity shown towards the police and what they are perceived to represent. It may be n ecessa ry to liaise with primary medical care rs and the social wel far e department to ensure that the family gets best treatment. If it is thought that other children in the same family may be at risk of physical injury or neglect, the appropriate social welfare legislation must be utilized to safeguard other siblings. The family may wish to view the body of the deceased baby: they have a right to do so, a right enshrined in the European Convention of Hum an Rights. Unless there are very good reasons to the contrary, their wishes should be accommodated as soon as possible. The dead baby should be presented in surroundings that make this harrowing experi ence less stark. The infant ca n be held and touched and mementos, such as a hair lock or an imprint of the hands or feet in ink, wax or plaster of Paris be provided. Matters relating to post-traumatic stress cannot be ignored or side-stepped for any reason , not least the possi bility of issues of compensation. The pathologist at the scene may feel that the officer in charge of the scene, the senior
Referen ces I
investigating officer (SIO), should set in motion appropriate provisions, which his police force will p ut in place. In cases involving mutilation of the body, whatever the cause, this attains an even more acute importance.
INQUESTS AND INQUIRIES All that transpires at the scene will be scrutinized in court. This may be a coroner's court where a n inquisitori al scheme of thin gs is in operation, or in subseq uen t criminal cases in which an adversarial and confro ntatio nal system is at wo rk, an d all aspects of the investigatio ns may be a nd a re likely to be scrutinized through the fine sieve of cross examinatio n. Certa in cases may also come to the attention of Tribunals, Public Inquiries or other judicial investigations in which dif ferent nuances of procedure may hold, but evidence will stil l have to be printed under oath or affirmation.
REFERENCES Wagner GN. Crime scene investigation in child abus e cases. Am J Forensic Med Pathol 1986; 7:94-9. 2 Usher A. Th e role of the pathologist at the scene of crim e. J Fo rensic Sci Soc 1970; 10:213-8. 3 Bass M, Kravath RE, Class l. Death scene investigation in sudden infant death. N Engl J Med 1986; 315:100- 5. 4 Locard E. The analysis of dust traces, Parts I-Ill . Am J Police Sci 1930; 1: 276- 298,405-406, 496-514. 5 Vanezis P, Busuttil A. Suspicious Death Scene investigation. London: Hodder Arnold, 1996. 6 Wh ite P. Crime Scene to Court. The Essentia ls oj Forensic Science. Lond on: The Roya l Soc iety of Chemistry, 1998. 7 Stanton AN. Sudden infa nt death. Ove rheat in g and cot death. Lancet. 1984; 2: 11 99-201. o Byard RW, Krous HF. Sudden Injant Death Syndrome. Lond on : Arnold, 200 I. 9 American Academy of Ped iatrics. Distinguishing sudden in fant death syndrome from child abuse fata lities. Pediatrics 1994; 94( I): I 24 - 6. 10 Christoffel KK, Zieserl EJ, Chiaramonte J. Should child abuse and neglect be considered when a child dies unexpectedly? Am J Dis Child 1985; 139:876-80. II EmelY JL. Chi ld abuse, sudden in fant death syndrome, and
unexp ected infant death. AnI J Dis Child 1993;
1476( 10) : 1097 - 100.
12 Bya rd RW, Carm ichae l E, Bea l S. How useful is the post
mortem examination in sudden infant death syndrome?
Pediatr Patho1 1994; 14:817- 22. IJ Bya rd RW, Becker LE, Berry 1. Formal retrospective case
review and sudden infant death. Acta Paediatr 1997;
86:10 11-1 2.
14 Newlands J, Emery JS. Child abuse and cot deaths. Child
Abuse Negl 1991; 15(3):275- 8.
15 Beckwith JB. Discussion of terminology and defin iti on of th e sudde n in fant death syndrom e. In Bergma n AB, Beck with JB , Ray CD (eds) Sudden Infant Death Syndrome. Proceedings oj the Second international Conjerence on Causes oj Sudden Death in Injants. Seattle: Uni versity of Washi ngto n Press, 1970, pp. 14- 22.
143
16 Jones AM, Weston JT. Examination of the sudden infant death syndrome: investiga tive and autopsy protocols. J Forensic Sci 1976; 21 :833-41. 17 Hilton J, Berry JP. Pathol ogy. In Fitzgerald K (ed.) Second S/DS Global Strategy Meeting. Norway: Stavenger, August 5-6, 1994, p. 334. In Rognum TO (ed.) Sudden Infant Death Syndrome. New Trends in the Nineties. Oslo: Scandinavian University Press, 1995. 18 Willinger M, James S, Catz C. Definin g the SIDS: deliberat ions of an expert panel conceived by the National Inst itu te of Child Health and Human Development. Pedilltr Pathol J 994 ; II :677-84. 19 Center for Disease Control (CDC). Guidelines for dea th scene in vest igatio n of sudden , unexplained infant dea th s: reco mmendations of the Inter Age ncy Panel on Sudden Infant Death Sy nd ro me. MMWR Rec0111111 Rep 1996; 45:RR-JO. 20 Iyasu S, Ha nzlick RE, Rowley D, Wi llin ger M. Pro ceedings of Workshop on Guidelines for Scene Investigation of Sudden Unexp lained Infant Deaths. J Forens ic Sci 1994; 39: 1126- 39. 21 Dragon DA, Lan denberg Ai. Infant mechanical suffocation deaths in the United States, 1980-1977. Paediatrics. 199 9; 103 :59. 22 Nakamura AS, Wind M, Danello MD. A review of hazards associated with children placed in adult beds. Arch Pediatr Adolesc Med 1999;153:1018-23. 23 Fact Sheet on the Tox ic Gas Hypothesis of Cot Death, www.sids.org.uk/ed itpics/364-I.pdf 24 Coghlan A. .Mattress bugs may link to cot deaths. New Scieutist 2002; 10:41. 25 Meadow R. Recurrent cot death and suffocation. Arch Dis Ch ild 1989; 64(1): 179-80. 26 Emery JL, Gi lbeli EF, Zugu ibe F. Three crib deaths. A babyminder a nd probab le infa nti cide. Med Sci La w 1988 ; 28(3) :205 -11. 27 Byard R, Petechia l hemorrhages and unexpected infa nt deat h. Legal Medicine 2003; 1(4): 193-7. 28 Betz P, Hausmann R, Eisen menger W. A contribution to a possible differe ntiation betwee n SIDS a nd asp hyxiatio n. Forensic Sci Int 1998; 9 1: 147 -52. 29 Shepherd RT. Accidenta l se lf-s trangula tion in a yo un g chi ld. Med Sci Law 1990; 30: 11 9-23. 30 Mend lowicz MV, Jean -Lo ui s G, Gekker M, Rapaport MH. Neonaticide in the city of Ri o ele Janeiro: forensic and psycholegal perspectives. J Forensic Sci 1999; 44:741 - 5. 31 Cheung PI. Materna l filicide in Hong Kong, 1971-85. jVled Sci Law 1986; 26: 185 - 92 . 32 Lee ACW, Li CH, Kwong NS, So KI. Neonat icide, newborn abandonm ent, and denial of pregnancy - newborn victimisation associated with unwanted moth erhood. Hong Kong jVled J2006; 12: 6 1-4. 33 Herman -Gidd ens ME, Smit h JB , Mittal fvJ et al. New born killed or left to die by a pa rent: a population-based study. JAMA 2003; 209:1425-9. 34 Mitchell LB, Dav is JH. Spontaneous births ill to ilets. J Forensic Sci 1984; 29(2):591-6. 35 Kerley ER. Forensic Anth ropology and crime involving ch il dren. J Forensic Sci 1976; 21[2):33 3- 9. 36 Hargrave DR. A study of chil d hom icide over two decades. Med Sci Law 1992; 32[3): 196- 245. 37 Kerley ER. The ide ntificatio n of battered- infant skeleto ns. J Forensic Sci 1978; 23(1): 133- 8. 38 Kasim MS, Cheah I, Shafie HM. Childhood deaths from physical ab use. Child Abuse Negl 1995 ; 19(7):84 7- 54. 39 Kellet RJ. Infanticide and ch ild des U'llCtion - the historica l, legal and pa thological aspects. Forensic Sci Int 1992: 53 (10):1-28.
144 I
The death scene following the sudden death of a child
40 Alexander R, Sato Y, Smith W, Bennett T. Incidence of
impact trauma with intracranial injuries ascribed to shaking. Am] Dis Child J990; 144 :557-63 . 4J Duhaime AC, Christian CW, Rorke LB, Zimmerman RA. Non-accidental head injury in infants - the 'shaken-baby' syndrome. N Eng] Med 1998 ; 330:1822-7. 42 Squires T, Busuttil A. Child fatalities in Scottisb house fires: a case of child neglect? Child Abuse Neg l J995; 19:6 5-73. 43 Roberts I. Deaths of children in house fires. EM] 199 5; 311: 1381-2. 44 Squires T, Busuttil A. Alcohol and house fire fat aliti es in Scotland. Med Sci Law 1997; 37:325-40.
45 Runyan CW, Bangdiwala IS, Lin zer MA et a1. Fatal house tires: who dies and who su rvive? ]AMA 1998; 279: 1633- 7. 46 Istre GR, McCoy BS, Barnard JJ, Bolton A. Deaths and injuries from house fires. N Eng] Med 2001; 344(25): 1911-16. 47 Buteau J, Thompson B, Sexton D. Homicide followed by suicide: a Quebec case-series, 1988 -90. Cal1 ] Psychiatry 1993; 38(8):552-6. 48 Busuttil A Lockerb ie Ct Dunblane. Disasters and dilemmas. Med Leg] 1998 ; 66(4):34-42.
I
CHAPTER 8
I
POST-MORTEM EXAMINATION IN BABIES AND CHILDREN Jean W Keeling
Introduction Death scene investigation Rectal temperature Medical and family history Other important information Radiological examination Photography Microbiological samples Toxicological investigations Biochemical and metabolic investigations
145 145 146 146 146 146 147 148 149 149
Weights and measurements External examination Estimating blood loss Dissection (infants and older children) Examination of the brain, spinal cord and eye Exam ination of the newly born Histological samples Retention of organs Exchange of information and multidisciplinary review References
150 150 152 152 156 158 162 163 163 164
INTRODUCTION
DEATH SCENE INVESTIGATION
This chapter considers the investigation of deaths in the perinatal period, infancy and early childhood. The investi gation of deat hs in all th ree groups follows a similar gen eral line, but there are some diffe rences in techn iqu e, emphasis and even in interpretation of the same patho logical findings, for example the presence of facial and upper trunk petechiae, among the gro ups. Such differ ences wi ll be emphasized at appropriate points within the chapter. Sudden death in early life may be the outcome of a var iety of natural diseases (see Chapters )0-12) as well as acc idents and homicide. Sometimes, a pre-existing con genital or acquired disorder may predispose a child to acci dents and, possibly, to deliberate assault. In an y sudden death , whether or not a cause of death is precisely defined an d particularly when some necropsy findings , such as petechial haemorrhages, can be the res ul t of natu ral disease, it is most important that a full range of inves tigations is carried out and that their results are clearly presented. 1,2 Failure to do so weakens any assert ion of non-natural death and enhances the scope of cross examination.
Thorough appraisal of the locus is as relevant to deaths occurring in the home as it is to motor vehicle collisio ns and other incidents away from the domestic environment. Bass et al3 and Sturner4 stress the importance of detailed eval ua tion of the death scene in unexpected infant deaths. This includes the physical state of the infant's place of sleep and the precise position of the in fa nt when found. The latter is particularly important when positio nal asphyxia is a possi bility and when co-sleeping is practised. The use of a life sized doll may assist the recall of witnesses and improve the precision of information about position when fou nd. 4 Evaluation of an inappropriate sleeping environment in respect of temperat ure, humidity and the presence of noxious gases can only be made by prompt evaluation of the scene. 3 ,5 Seizure of feeding bottles and domestic items, such as jugs and spoons used in preparation of feeds, for toxico logical analysis is important. The contribution of drugs and alcohol to sudden deaths in babies and children is probably un de resti mated. A visual record of the death scene can aid the interpret ations of necropsy findings and high-quality photographs
146 I
Post-mortem examination in babies and children
or video recording are often extremely useful. Death scene investigation is discussed in detail in Chapter 7.
RECTAL TEMPERATURE Measurement of rectal temperature at the death scene or on arrival at hospital may be a useful corroborative investiga tion. If elevated, it raises the possibility of pyrexial illness or an inappropriately warm sleeping environment. There are no tables or formulae relating rectal temperature and likely time of death that are appropriate for use follow ing the death of babies or children. Surface area in children is relatively greater and body mass less than that of adults, so that cooling is likely to be more rapid and probably more influenced by the amount of clothing and other coverings. 6
MEDICAL AND FAMILY HISTORY Information about the individual's medical history and family background should be available to the pathologist prior to necropsy examination. Details of pregnancy, birth and progress in the neonatal period are essential in infant deaths, together with a detailed account of recent illness and of events in the 48 hours prior to death. In older children, a brief history of birth and early life may be sufficient; however, information about any ill nesses and accidents, particularly those resulting in hos pital admission, are important. Information about members of the immediate family should include major problems of a social, legal or finan cial nature, recent illnesses and documentation of any familial disorders. Information about sudden death or col lapse in the wider family, particularly in early life, should be specifically sought. A history of many consultations with the general practitioner or hospital, either by the deceased or siblings, is important information for the pathologist. Multiple hospital attendances can be difficult to ascertain, as children may have been taken to a variety of different institutions. The 'At Risk' register, or its equivalent, should be checked carefully, not only for the deceased individual, but also for siblings and half-siblings, ta king cognisance of the multiplicity of surnames used in some of today's compli cated family units. Information about prescribed medication for the decedent and any other family member is required, togetl1er willi information about the availability of recreational drugs and alcohol in the decedent's immediate environment.
OTHER IMPORTANT INFORMATION Information about the deceased's mobility and physical capabilities are important to the interpretation of any
injuries present, particularly in the infant period when there is much individual variation. Non-mobile infants should have few, if any, bruises. 7 ,B Awareness of supervi SOlY arrangements around the time of death is also impor tan t. An inexperienced carer may unwittingly permit hazardous activities or inappropriate freedom and be less aware than parents of recent recommendations about best childcare practice. 9 Carers who are distracted by illness, depression or other adverse events, such as marriage break down and bereavement, are likely to be Jess vigilant in respect of both symptoms of illness and environmental hazard.
RADIOLOGICAL EXAMINATION A fuJJ radiological examination is an essential part of the evaluation of any sudden death in infancy and childhood, and this is further discussed in Chapter 3. Necropsy exam ination should not be contemplated in establishments where radiography is not available. It is better to accom modate a small delay to necropsy so that radiological evalu ation can be carried out rather than having a major disruption to dissection or an incomplete examination after the post-mortem examination has been completed. A single-film 'Babygram' is not sufficient; it is esse11tial to have a complete set of appropriately penetrated regional views. The radiographs should be taken by an appropri ately qualified and experienced radiographer. This individ ual is important for his/her expertise in producing the appropriate views to encompass classical injUly sites and ensuring technical excellence. An experienced radiog rapher may suggest additional films. Take the advice! Recent rib and skull fractures cause most problems for the radiologist; however, a pathologist is unlikely to miss them as they can be observed directly and should always be carefully sought. Small fractures in long bones in the vicinity of joints (see Chapter 3) are best demonstrated radiographically but require histological confirmation. Histological examination of any area when there is dubiety about the nature of a radiological abnormality is essential. Sections through identified fractures can contribute to assessment of the age of the fracture. In any suspicious death, it is important that the radio graphs are reviewed prior to necropsy examination by an experienced radiologist, preferably one with paediatric experience. A verbal report will suffice. It is often useful to X -ray individual bones or regions (i.e. rib cage) following removal during necropsy. In partic ular, it ensures better views of posterior rib fractures (Fig. 8.1), which, being close to the costovertebral articulations, are often obscured by thoracic viscera. Localized abnor malities of uncertain aetiology may be positively identified using this procedure. Re-X-ray of a skull vault with wormian bones ensures an optimal record and avoids later argument.
Photography I
147
Radiological examination adequately documents gas or air within chest, abdomen and pericardium and within large vessel s (e.g. pulmonary arteries. portal venous system and cerebral sinuses). Cranial axial X-ray, following instill atio n of contrast medium into the superior sagittal sinus. has been used suc cessfully to demonstrate torn bridging veins in infants. 10, II Other modalities. such as magnetic resonance imaging. may be useful in particular cases but are usually more dif ficult to arran ge (see Chapter J).
PHOTOGRAPHY High-quality photographs are an important part of the record of any necropsy on infants or children and are par ticularly important when investigating sudden deaths. Whole-body views give a useful visual reminder of body proportions and sta te of nutrition (Fig. 8.2). and assist in the local ization and quantification of cutaneous injuries such as burns or sca lds. Close-up photographs of injuri es, with and without a scale, and sometimes from different angles. permit assess ment of injuries much more effectively than pages of
Figure 8.1 Radiographic examination of an infant admitted with subdural haemorrhage. (a) Portable radiograph taken on the intensive care unit; rib fractures are difficult to see. (b) Anteroposterior radiograph following evisceration; posterior rib fractures are readily visible. (c) Portion of rib cage X- ra yed after removal from the body; fine detail of fracture and callus aids dating of injury.
Figure 8.2 Posterior view of whole body prior to necropsy. There is little subcutaneous fat, muscle bulk is reduced, and napkin dermatitis and hypothermic injury are present - asa resu lt of inadequate care.
148 I
Post·mortem examination in babies and children
(a)
(b) ... f
i" ) .L./~""t)'% ),.\ ..:,: -)..., ~'
.
Figure 8.3 (a) Bruises photographed on admission to an intensive care unit. (b) Comparable photograph taken 36 hours later, prior to necropsy. detail ed written description and measurements ever can. Good photographs of injuries mean that referral to experts is more likely to produce a useful opinion. They are most helpful for a defence pathologi st and are particularly use ful when cases are re-opened or charges brought some years after the death occurred. Additionally they are a helpful aide-memoire before court appearance. It is important that th e pathologist has an 0ppOltunity to rev iew all of the photographs taken following a violent death, and not merely those sel ected for possible court use. Comp arison of photographs taken during life, for ex ample on the intensive care unit, with post-mortem photographs may enable a better estimate of the age of bruises than either set ex amined in isolation (Fig. 8.3). Sturner4 emphasizes the importance of a photographic record of petechiae and injuries on admission to hospital as changes take place quickly. A further external examination and re-photography of cutaneous injuries on the day following necropsy often yields useful information (Fig. 8.4), as may photographs taken using ultraviolet light.
IVIICROBIOLOGICAL SAMPLES Microbiological investigations are an important palt of the investigation of sudden death in both infants and children. 12
~ ,. ~ '\ > . ~'
,
1
"
-,
.r
Figure 8.4 (a) Bruises photographed prior to necropsy. (b) Image (a) re-photographed 24 hours later.
It is good practice to obtain a core set of samples as early in the examination as possible (Table 8.1) to minimize contam ination. In some institutions, these samples are taken in the acciden t and emergency department (emergency room) according to a protocol agreed between pathologist an d clin ical coll eagues, with prior authorization of the legally responsible department. IJ Additional sa mples are sometimes indicated by necropsy findin gs. The expansion of molecular methods in microbiology has resulted in quicker results and
Biochemical and metabolic investigations I
Table 8.1
Microbiological samples usefully token in sudden death investigation in early life
Bacteriological samples
Nose/throat swabs Cerebrospinal fluid Blood culture Bronchial swab Lung Large bowel or rectal swab
149
Table 8.2
Useful toxicological samples in the investigation of sudden death in early life
Sample
Blood Urine/bladder washout
Virological samples
Toxological agent
Alcohol Com mon sedatives Common analgesics Recreational drugs Carers' prescribed medication Recreational drugs Volatiles
Nose swabs/nasopharyngeal aspirate Tracheal ring Lung Heart Small intestine
Lung (whole) Hair - cut Earlier/long-term drug ingestion Hair - plucked Stomach contents
has extended the period over which samples may be use ful, 14, 15 but the significance of detection of viral RNA must be considered in association with relevant histopathological changes in appropriate organs. Although microbiological investigations are positive in only a small proportion of cases, negative results are partic ularly important in the sudden unexpected deaths in infancy (SUDl)/sudden infant death syndrome (SlOS) context l6 and it is important that these investigations are carried out, even when death seems likely to be the result of trauma; failure to do so can make for vexatious cross-examination. It is worth remembering that some microbiological causes of sudden death in early life, such as group B streptococcal and some viral infections (such as echovirus), elicit no spe cific findings at either gross or microscopic level and so that positive cultures are essential for diagnosis.
possibility in mobile infants and small children when pre scribed medication or recreational drugs are present in their environment. Alcohol ingestion can produce rapid and marked metabolic dis turbance in the young. Rapid onset of drowsiness/incoordination and metabolic disturbance, such as hypoglycaemia, are indications for full toxicological investigation. Paracetamol (acetaminopohen) is recommended only on presCliption for infants < 3 months of age. 19 A single dose may be given following immunization. Detection of non-prescribed paracetamol in young infants requires a full explanation. Suitable samples for toxicological investigation are tabu lated (Table 8.2).
TOXICOLOGICAL INVESTIGATIONS
Post-mortem biochemical investigations on blood samples are unsatisfactory on the whole, with the exception of toxicological investigations. Vitreous humour, urine and cerebrospinal fluid (CSF) are much more useful, although normal ranges for CSF values for the paediatric age group are often not available in individual laboratories. Some groups regard biochemical investigations as part of the basic SUDI protocol, 13,20 whereas others only perform them when there are historical features or abnormal pathological findings. 17 Sampling the vitreous must be omitted if it is likely that histological inv estigation of the eye is important. Whilst a range of investigations is possible, the size of the sample is usu ally the limiting factor. This is a particu lar problem with urine, as the bladder is empty in very many SUD!. A bl adder washout with - 10 mL of normal saline can be used to investigate the possibility of organic acid abnormalities. It can also detect low levels of com monly used analgesics when the blood levels are within the therapeutic range (personal observation) .
Toxicological investigations are not undertaken routinely in many jurisdictions as part of the investigation of SUD! or sudden death in older children. There may be cost con straints or perceptions of a low return for effort. This is par ticularly so in fetal deaths and amongst those infants found dead in their cots on household awakening.17 Sims and Collins 18 found evidence of drugs, particularly recreational drugs, in 21 per cent of fetal deaths that were investigated in a forensic pathology depa11ment. Such investigations are more likely to be done when concerns are expressed in a police rep0l1. Pathologists should be ready to recommend that, as a minimum, screening procedures for some types of medication and recreational drugs are carried out and should ensure that appropriate samples are retained for more detailed investigation if indicated. Drugs with hyp notic or sedative effects are sometimes given inappropli ately to keep an infant quiet. It should not be forgotten that infants are inquisitive and casual ingestion is always a
Common sedative/analgesics
BIOCHEMICAL AND METABOLIC INVESTIGATIONS
150 I
Post-mortem examination in babies and children
Prioritization of investigations is usually necessary and is based on a combination of history and pathological findings. Sodium, urea and osmolality are useful baseline investi gations JJ and levels are generally stable in vitreous humour for about 72 hours. Sodium and osmolality pro vide corroborative evidence of hypernatraemia or dehydra tion. An elevated level of urea is not usually found until 2-3 days into an illness and provides a general indicator of subacute illness. Glucose levels in vitreous humour are not so stable but are useful for up to 12 hours after death,13 which makes their usefulness at time of necropsy questionable. The reliability and usefulness of post-mortem investiga tions is discussed further in Chapter 5. In recent years, molecular methods have replaced lengthy biochemical investigations, with somewhat unceliain results, in the identification or exclusion of a number of genetic metabolic disorders (GMDs) that might underlie some SUD!. That these investigations can be carried out on frozen tissue or even thick sections from paraffin embedded tissue (especially spleen) is a particular advantage, as fluid samples can then be used for other tests. Molecular methods on tissue samples can also detect the molecular rearrangements of some forms of long QT syndrome. 21 As the potential for molecular investigation of more and diverse genetic disorders continues it becomes essential to reserve both frozen and formalin-fixed wax-embedded material indefinitely when sudden death in infants and children is initially unexplained or incompletely explained . Another sample that has proved useful in the detection or exclusion of some GMD is the Guthrie card blood spot. Wilcox et al 22 undertook tandem mass spectrometry for acyl-carnitines. They identified three GMDs amongst 247 SUm.
WEIGHTS AND MEASUREMENTS The value of recording body weight and external measure ments is much greater in the post-moliem evaluation of the fetus, infant and child than it is in adults. For full assessment of the fetus and infant, body weight, crown-rump (sitting height), crown-heel (standing height) and foot length, together with the occipitofrontal head cir cumference and biparietal diameter, are impoltant. These are compared with normal values 23 - 25 (see Appendix B, p. 471) and provide information about gestational age in the fetus and the appropriateness of postnatal development and nutritional status in babies and infants. In older children, body weight, standing height and head circumference give information about nutritional status and appropliateness of development. Normal values by sex are readily available in chart form from paediatric wards or
outpatient departments. The most widely used charts in the UK are based on the work of Freeman et al 26 (Appendix B). Weight and length are important when estimates of blood volume and calculation of administered drug doses are required. Caution should be exercised when comparisons are made between body weights recorded during life and post mortem weight. Weight in life may include clothing, a number of different weighing scales may be used and any may be inaccurate. (When were your mortuary scales last inspected and calibrated?) Weights may not have been recorded accurately - check them yourself if in doubt, especially when using an unfamiliar mortuary or when examination of babies is carried out infrequently. Has there been a transcribing error? Go back to the original record and compare with that in the report. Accurate organ weights in babies and children can pro vide pointers towards natural disorders and may be useful for corroboration when investigating postoperative deaths. An electric balance calibrated in O.l-g increments is appro priate and inexpensive. Gestation-related normal organ weights for comparison from 12 to 42 weeks' gestation are those of Singer et al,27 Hansen et al 28 and Maroun and Graem 25 (Appendix B) . Age-related organ weights in infants have been collated by Thompson and Cohle 29 (Appendix B). Organ weights in older children are more difficult to access. The most recent papers, 30, 31 and Haddad et ai, are statistical analyses of older data. Age-related weights of major organs are found in Altman and Dillmer32 (Appendix B). The brain should be weighed before and after fixation, being aware of a potential confounding factor when different balances are used.
EXTERNAL EXAMINATION A detailed external examination by region is essential, par ticularly when there are injuries or external changes of uncertain aetiology present. Outline figures with infant body proportions 33 are invaluable in the recording of focal lesions (Fig. 8.5). Tbe general description includes racial origin, clothing, jewellery, skin colour, regions affected by livor mortis (including areas of sparing). extent of rigor mortis, cleanliness and state of nutrition . Signs of medical intervention are listed separately from other external features. Kaplan and Fossum 34 describe resuscitation injuries of the fa ce and neck, and relate them to different types of apparatus used in resuscitation . A careful search is made for petechial haemorrhages, including eversion of both eyelids to examine tarsal plates, examination of the oral cavity and behind the ears. Petechiae are unusual in healthy infants, 35 and are unusual in SUD! that remain unexplained. J6 Careful note should be taken of any secretions emanating from the mouth or nose. Blood staining or streaking should
External examination I
Name ____________________________
Race _ ________
Sex ________
Autopsy number ________________________________________ _ _________
Date ________
Figure 8.5
Age __________
151
Body surfaces with infant proportions aid accurate recording of injuries (after ref. 33).
be described carefully and photographed, with the appear ance (e.g. fro thy, fluid, dried) noted. When secretions are bloody, Krous et ae 7 urge careful examination of the mucous membra nes of the nose and mouth with an a uroscope.
Any tears or bruising of the frenula of the tongue and lips are recorded (Fig. 8.6), as well as scarring suggestive of old injUly. The impression of teeth (even in an edentulous infant) and patterned bruises of the inner aspects of the lips
152 I
Post-mortem examination in babies and children
(b)
Figure 8.7
Mongolian blue spot above the gluteal fold in a baby
of Indian parentage.
p. 37. If assault is possible, swabs from the vagina and rectum are obtained for DNA analysis of secretions and microscopy for spermatozoa, foreign material and microbiology.
ESTIMATING BLOOD LOSS
Figure 8.6
Injury to the upper frenulum (a), accompanied by
excoriation of the lower lip, with bruising of the adjacent alveolar margin (b).
suggest peJioral pressure. Examination of the pharynx for injury, secretion or foreign bodies can be carried out easily using an infant laryngoscope. Remember that a foreign body may have been removed in accident and emergency but associated abrasions may be present. The epiglottis and vocal folds can be visualized as well, giving a better impression of upper airways' obstruction as a result of tis sue swelling from oedema or inflammation rather than an examination after removal when swelling can diminish rapidly as tissue fluid escapes. Each injury is described and measured with reference to a fixed bony point. Direct dictation of this part of the examination while it is carried out is most accurate and can easily be done while photography is under way. Care should be taken that natural phenomena, such as Mongo lian blue spot (Fig. 8.7), haemangiomata and pigmented naevi 38 are not inappropriately interpreted as injuries. The genitalia are examined carefully. There is a wide variation of appearance of the genitalia (see Table 2.1, p. 30, and Table 2.2, p. 31), which should not be construed as injury. Signs of abuse are listed in Table 2.4 (p. 39). Post mortem anal dilatation, a natural phenomenon ,39 should not be confused with sexual assault (see Table 2.5, p. 40); other perianal findings are also tabulated - see Table 2.3,
Body cavities should be opened with care and any blood or fluid measured. In infants, this is probably achieved most easily using a large syringe without a needle. The extent of blood loss should be related to the esti mated circulating blood volume, calculated with reference to body weight (Table 8.3). Blood loss may be directly rele vant to cause of death, but its significance can be missed unless calculated as a propoliion of normal blood volume. The extent of intracranial blood loss is difficult to calcu late. There may be a haematoma, which can be measured by weight or volume, but if it is attached to the dura or arach noid membranes it should not be disturbed for volumetric measurement as histological assessment for injury to death interval takes prioJity (see Table 8.4) A thin film of haemor rhage over one or both cerebral convexities is particularly difficult to measure and one is left with only residual blood in the cranial fossae after removal of the brain. Blood loss into pericranial tissues and into muscles attached to the skull can be considerable and is also diffi cult to measure, but an assessment should be attempted and its contribution to overall blood loss stressed. See sec tion on 'Birth injury' later in this chapter.
DISSECTION (INFANTS AND OLDER CHILDREN) This section relates to the examination of infants beyond the neonatal period and older children. Methodology appropriate to the examination of the newly born is found later in this chapter (p. 158). Protocols for the examination
Dissection (infants and older children) I
Table 8.3 Age
Newborn 1 month 6 month 12 month 2 years 3 years 4 years
5 years 10 years 15 years
153
Blood volume/oge for infants and children
Volume (mL/kg)*
80-85 80-85 75-80 75-80 75-80 70-75 70-75 70-75 70-75 70-75
Boys
Girls
Average weight (kg)+
Average blood volume (mL)
Average weight (kg)'''
Average blood volume (mL)
3.6 4.5 8.2 10.2 12.5 14.7 16.5 18.6 31.5 55.5
288-306 364-383 615-656 765-816 938-1000 1029-1103 1155-1238 1302-1395 2205-2363 3885-4163
3.4 4.2 7.5 9.4 12.0 14.1 16.2 18.2 32.2 53.3
272-289 336-357 563-600 705-752 900-960 987-1058 1134-1215 1274-1365 2254-241 5 3731-3998
'Ref. 86.
+Boys' growth chart (birth-18 years) 1996/1 Child Growth Foundation.
'Girls' growth chart (birth- 18 years) 1996/1 Child Growth Founda tion.
Table 8.4
Histological appearance of subdural haematomata of different durations (after ref. 46)
Time after injury
Haematoma
Dural aspect
Arachnoidal aspect
Fibrin Fibrin
Fibrin Fibrin
Fibroblasts present at dural junction Fibroblast layer 2-5 cells layers thick
Fibrin
Histological features of subdural haematamata
To 24 hours 24-48 hours
2-5 days 4-5 days
1 weeks 2 weeks 3 weeks 4 weeks
1-3 months 3-12 months >1 year
Fresh red blood cells Fresh red blood cells, polymorphon uclear leukocytes and fibrin Macrophages replace polymorphonuclear leukocytes Mixture of intact and lysed erythrocytes, siderophages present Red blood cells lysed; early angiofibroblastic proliferation Haematoma liquefies, sinusoids, 'giant capillaries' appear Vascular sinusoids are well developed Haematoma is liquefied
Fibroblast layer up to 12 cell la yers thick Fibrobla st layer is about half of the thickness of dura
Fibrin
Single layer of fibroblast cells present Fibroblast layer is a few cells thick, with an occasional capillary
Fibroblast layer is as thick as Fibrous membrane and a few capillaries see n the dura, siderophages present Fibroblast layers are hyalinized and form membranes on both surfaces, large (giant) capillaries appear early, secondary haemorrhages often develop The neomembranes fuse and contain mature fibrous tissue and scattered siderophages, after 3 months it is no longer possible to accurately date (age) the haematoma The neomembrane is a distinct fibrous connective tissue layer that closely resembles the adjacent dura mater, occasional ca lcification and/or ossification
of sum cases are detailed by Valdes Oapena et al40 and KroUS. 41 Comment will be directed paJiicularly at those organs or regions for which a different approach or empha sis is appropriate in the older paediatric age group . The task is easier when the following equipment is available: shOji
(10- to 12-cml, non-toothed forceps to mJlllmlZe tissue damage; a range of tapered, round-ended scissors (Mayos); and a variety of scalpels, rather than an autopsy knife. Reference to the need for a digital balance was made in an earlier section .
154 I
Post-mortem examination in babies and children
Forma l neck dissection to detect injUly is carried out in a sim il ar fas hion to that in adults and invo lves layer-by-Iayer examination of the anterior neck muscles. In the infant, it can be achieved following a vertical submental-symphysis pubis incision as tissues stretc h readily but is probably eas ier with a shoulder-to-shoulder appro ach. Posterior neck muscles are conveniently inspected by ensuring that th e posterior scalp refection extends to the cervica l sp ines. The thorax is best approached by serial division of costal cartilages, avoiding damage to the osteochondral junctions, some of which will be submitted for histological examina tion, a nd has the advantage of leaving the ribs intact. Fresh rib fractures may be apparent following reOection of skin and soft tissue (Fig. 8.8). Resuscitation Jib fractures are un com mon, and are usually situated at the anterior ends of lower ribs; accompanying haemorrhage is minor and there is no vital reaction on histological examination. Rib frac tures are discussed further in Chapter 11, pp. 213- 215. Relationships of orga ns are observed before disse ction commences, some abnormal relati onsh ips may suggest a syndrome diagnosis. The thymus is a prominent organ in early life, weigh ing 10 ::!: 4 g at term. It is of normal size (often described as large) in sudden death. A reduction in thymic weight below th e normal ran ge for age is a non-specific indicator of ante mortem stress, probably in excess of 12 hours' duration 42 Petechial haemorrhages v isible through the thymic capsule, particularly on its posterior aspect, a re present in about 80 per cent of SIDS 43 (see Chapter 11, p. 212). Larger, blotchy haemorrhages may be present in asphyxial death. The appearance of the lungs, noting the degree of expan sion or collapse, is recorded. The lu ngs usua lly fill the chest in SJDS but are often collapsed following attempted resuscitation
Figure 8.8
Fresh rib fractures accompanied by bruising, visible
after reflection of thoracic skin fl aps.
in hospital with a high concentration of oxygen. Fluid in the pleural and pericardial cavities is observed, measured and described. A sample of pericardium can be obtained at this stage using sterile inst ruments and placed in tissue culture Ouid for immed iate culture, if appropriate, or stored at - 280°C for future availability for investigation of putative genetic abnormality. This is followed by sampling of heart blood, lung and myocardium for microbiological purposes. External examination of the heart is conveniently canied out at this point by fully opening the pericardial sac. Cardiac abnormalities, both congenital and acqu ired, are common causes of explained SUD! (see Chapter 11, pp. 208-210) and older children (see Chapter 12, pp. 226-235). Cardiac situs is dependent on atrial morphology. The comparative size an d relationsh ip of the ascending aorta and pulmonary trunk are noted - they should be equal in size. The course ofthe an te rior descending branch of the left coronalY artery should be observed and the connection of t he vena cavae a nd pul monary veins noted. If all are normal then congenital heart disease is excluded, with the exception of septal defects and anomalous pu.lmonary venous return to the coronary sinus. The liver is relatively large in infants and protrudes below the costal margin. Colour and consistency are noted . Fatty infiltration and evidence of trauma are sought. The former requires frozen section and histochemica l staining for confirmation. Preservation of samples to investigate the possibility of GMD and full toxicological exami nation should be considered. The stomach is often distended, either because of attempted 'bag and mask' ventilation or microbial fermentat ion. The mesentery is inspected for tears and contusio ns (Fig. 8.9). Completeness of intestinal rotat ion is observed, along with the presence of volvulus, intussusception and herniae - all of wh ich are potential causes of intestinal obstruction, shock and sudden death - are sought.
Figure 8.9
Contusion in the mesentery close to the
duodenojejunal junction.
Dissection (infants and older children) I Organs should be removed in blocks and not piecemeal. In particular, the heart should remain attached to the lungs until completely dissected. Th e Rokitansky technique is suitab le but the following proced ure is easier: removal of (he intestines from the duodenojejunal fl ex ure to the rec tum (having checked the mesentery for contusions) (Fig. 8.9), then rem oval of the thoracic and upper abdominal viscera en bloc thus opening the inferior vena cava and abdominal aorta, followed by removal of genitourinary tract. When sexual assault is suspected, a cuff of perineal skin is removed in continuity with the lower genital tract and anus, usually including the coccyx, the pubic area and the med ial aspect of the thighs. Swabs for forensic exam ination should be taken before dissection is begun . Following removal of the viscera, the thoracic and abdominal cavities should be cleared of blood and fluid and their walls examined. The ribs should be carefully inspected for fractures. Rib fractures are often undisplaced in babies and in fants. Attention will be drawn to fresh frac tures by subperiosteal an d intramuscular haemorrhage. Older fracture sites will be supported by callus and are usu ally stable (Fig. 8. 10), although re-fracture can occur. Cal lus is particularly marked on the internal aspect of the rib and is readily palp able. Fractures at the posterior ends of ribs are easily missed and should be specifically sought, with the remova l of pali of the rib cage fo r radiographic and histological examin ation wh en there is suspicion of injury. Previous avulsion of the osteochondral junction of
155
the lower ribs (6-10) is common when there are fractures elsewhere and is seen as expansion of the internal aspect of the bone adj acent to the cartilage (see Fig. 8.20, p. 160). Stripping of the pleura to look for rib fractures in infants is not recommended. The stripping process results in tissue dam age, including any subperiosteal reaction and may render histological evaluation probl ematic. Th e upper airway is examined carefully and may be fixed before opening when trauma to the neck has occurred. Laryngeal fracture is unlikely but small haemor rhages may be present in the intrinsic lary ngeal muscles. FUliher microbiological samples can be taken from the lower trachea/main bronchi. A convenient meth od of opening the heart is shown in Figure 8.11. It is opened before separation from the lungs. Should a cardiomyo pathy be suspected, a horizontal slice through both ventricles, midway between base and apex, demonstrates it well (Fig. 8.12) without interfering with dissection; chamber connections and morphology, appea r ance of valves, septal defects and the origin and course of coronary arteries are examined. The appearance and thick ness of ve ntricular myoca rdium is noted. Any subendocar dial fibroelastic thickening or haemorrhage is noted. Heart and lungs are separated and weighed. Pulmonary arteries are examined for evidence of thromboembolus and hypeliension - both causes of sudden death in early life. The former is most likely to be found in the poorly mobile, postoperatively and follo wing the use of intravenous devices.
... ...
...
\ \ \ \ \
-
Figure 8.10 Thorax after evisceration; old posterior rib fractures with ca llus are readily visible.
\4 \ \. \ ...
\
"
\
...
...
...
"
"-,,- 2 " .... ....
\
...
\
...
\
\
\
... ...
\ \
.......
.........
...
...
...
"
Figure 8.11 Opening the heart, cutting lines are numbered sequentially (from Fetal and Neonatal Pathology, 4th edn, 2007, p. 39, figure 2.20, with kind permission of Springer Science and Business Media).
1 56 I
Post-mortem examination in babies and children
Figure 8.12 Transverse slice through the ventricular myocardium half way between base and apex. There is marked left ventricular hypertrophy in this ll-month-old infant with hypertrophic cardiomyopathy.
Subpleural petechiae are a common post-mortem finding in infants. Larger, blotchy or confluent haemorrhage is less common and may indicate an asphyxial episode (see Chapter 11, p. 218). Lung scarring may be present in babies who were born preterm and who have received intensive care. The stomach is opened and its contents noted. A sample may be retained for toxicological examination. Foreign material, such as paper or non-food items, should evoke a high suspicion of non-natural death. The nature of food residues should be related to the description of ante mortem events. The amount of food in the stomach is a most unreliable guide to length of survival after the last meal. One should remember that gastric aspiration may have been carried out in the accident and emergency department. When both stomach and intestines are found to be empty, careful note of the history and nutritional state is required. Microbiological samples from the intes tine are mandatory in these circumstances and detailed questioning of carers may be appropriate. Haemorrhagic infarction of the intestines, myocardium and sometimes brain is commonly seen in infants who have been subjected to intensive care procedures for periods of > 12 hours prior to death. Care should be taken not to over interpret these changes (see Chapter 11, p. 215). Pulmonary haemorrhage is common after attempted resuscitation involving ventilation but siderophages, seen on microscopic examination, require at least 2 days to develop.44
EXAMINATION OF THE BRAIN, SPINAL CORD AND EYE Before the cranial cavity is opened, the scalp and calvarium are examined for evidence of injury; if injuries are found then these are recorded as described previously. In the
Figure 8.13 Cranium opened by the modified Beneke technique. Bone flaps are reflected following incision of suture lines. There is a brown membrane attached to the parietal dura, due to organization of a subdural haemorrhage that occurred several weeks before death. newborn and in early infancy, the calvarium is opened using the modified Beneke technique. 45 It allows visualiza tion of bridging veins before the bony flaps are completely reflected and of ' membranous' adhesions that may be pre sent between dural and arachnoid membranes during organ ization of old subdural haematoma (Fig. 8.13). Any tearing of bridging areas is recorded and a search made for a related focal subarachnoid haemorrhage (Fig. 8.14) and thrombosis of torn veins (Fig. 8.15). The presence of extradural and subdural haematoma is noted. The former are unusual in early life because of the firm adherence of the dural membranes to the inner table of the calvarium, particularly along the suture lines, a phe nomenon that may introduce difficulty in removal of the calvarium following a circumferential saw-cut. The infant brain swells rapidly following hypoxia or trauma, largely obliterating the subdural space so that sub dural haemorrhage may be reduced to a confluent film over the convexities, running in the falcine fissure and with small collections in the cranial fossae. Any abnormality of the convexities is recorded, the brain is removed, weighed and fixed in 10 per cent buffered formalin before examination. It is not advisable to attempt photography of the brain prior to fixation, other than pic tures taken before removal of the brain from the skull base. The infant brain is particularly soft following oedema and
Examination of the brain, spina l cord and eye
I
157
Figure 8.16 Fresh infant brain with flattening and posterior separation of the hemispheres. There is stretching and tearing of the corpus callosum.
Figure 8.14 Superior convexity surface; focal subarachnoid haemorrhage may indicate the site of torn bridging veins.
Figure 8.17 Slice of the cerebral hemispheres after fixation; death followed a motor vehicle collision. There is haemorrhage in the right putamen/internal capsule and focal haemorrhages in the corpus callosum.
Fi gure 8.15 Surface of the brain after fixation from a baby with oJbdural haematoma and haemorrhage. Thrombosis of a torn ridging vein is present. . :"Poxic insult. Artefactual tearing of vulnerable structures, ch as the corpus callosum, can result from manipulation of the unsupported brai n (Fig. 8.16). When the brain is very - ft, weighing in the fresh state can be omitted. Following the removaJ of the brain, standard samples for rristological examination are obtained from the parietal dura w d the dural folds. Any dura with attached clot, organizing ~em brane or brown staining should be removed, additionally, - r histological examination. Hardman 46 describes the histo .ogical appearances of subdural haematoma with approximate _uration prior to death (Table 8.4, p. 153).
In babies, infants and young children, the spinal cord is most conveniently removed from the front. An intervertebral disc is incised in the lower lumbar region and the proximate vertebral body grasped and elevated; the pedicles can then be divided with bone forceps on either side up to the upper cer vical region. The cord is examined through its intact dural covering and any extraduraJ haemorrhage is noted. The nerve roots are incised on either side as far laterally as possible to preserve dorsal root ganglia and the cord is removed after incising the dura at Cllevel. The cord is fixed flat for optimal examination. This is easily achieved by carefully laying it on a strip of card and permitting adherence for 2 or 3 minutes before immersion in fixative. Examination of the fixed brain and cord (preferably by a neuropathologist) is carried out systematically. The brain is photographed prior to slicing and any abnormalities identified subsequently are also photographed (Fig. 8.17). Samples for microscopy should follow appropriate fetal and infant blocking schedule (Fig. 8.18). additionally, sam ples of any abnormalities are obtained.
158 I
Post-mortem examination in babies and children
Rand L frontal parasagittal
Rand L parietal parasagittal Rand L parietal convexity Rand L basal ganglia at level of mamillary bodies
affords a view of the total orbital contents so that any extraglobal lesion can be sampled. 47 It is important that vitreous sampling is omitted when detailed examination is considered appropriate. The eyes are fixed separately in labelled containers. Exam ination of the fixed eye is described by Lee.48 Photographs of any external lesions should be made together with a photo graphic record of the interior after opening before sampling for microscopy. Processing of the eye requires care to min imize artefactual distortion. Perls' Prussian blue reaction (PBR)-stained sections are mandatory when assault is likely and immunohistochemical examination against beta amyloid precursor protein of the optic nerve is more effective than haematoxylin and eosin staining for the demonstration of axonal swelling when optic nerve trauma is suspected. 49 In some of these cases, referral of the whole eye for specialist examination may be the safest option.
Rand L thalamus Rand L hippocampus at level of lateral geniculate body
Rand L occipital
Midbrain Pons Medulla Rand L cerebellar hemispheres, inCluding dentate nucleus vermis
Figure 8.18
EXAMINATION OF THE NEWLY BORN Detailed techniques of necropsy examination appropriate for the fetus and neonate are described by Wiggleswolth 50 and Keeling.4 5 The following observations are relevant to the examination of the newly born for medicolegal purposes. This may be required because the body was concealed, abandoned , born in the absence of appropri ately qualified attendants or was unexpectedly still born. Other situations when a medicolegal examination is required are when attending clinicians cannot complete a death certificate or if there are complaints by parents or others and following maternal trauma or death (see Chap ter 10). Important considerations during the examination of perinatal deaths are those of live birth and separate exis tence. These are discussed in Chapter 10, pp. 187-192.
Diagram for sampling of the infant brain for
histological examination. Any lesions are sampled in addition
Radiography
(from Fetal and Neonatal Pathology, 4th edn, 2007, p. 43, figure 2.24, with kind permission of Springer Science and Business Media, courtesy of Professor JE Bell, Edinburgh).
The spinal dura is incised anteriorly and posteriorly, and the cord surfaces photographed. The presence of blood or clot is noted. The cord is sectioned transversely from the upper cervi cal region. This can be done while it remains attached to the dura. Any lesions within the cord are photographed. Sam ples for microscopic examination are taken at identified levels, and the cord and dura are preserved in continuity so that other identifiable blocks can be obtained if required. The eye can be removed from the front (anterior approach) following incision of the conjunctival reflections and division of the intrinsic ocular muscles in turn. The eye is prolapsed and the optic nerve divided. In the young, it is easier to approach the eye from the floor of the anterior cranial fossa (posterior approach), which is thin. This
A single whole-body radiograph is not sufficient in susp i cious fetal or neonatal deaths; imaging should follow the procedures outlined on p. 147 earlier in this chapter. As well as providing a record of any fractures, radiographs in the perinatal period provide useful information about ges tation (and, therefore, viability) ;51. 52 they also provide good proof of skeletal malformations, which can contribute to a syndrome diagnosis. A radiograph may demonstrate gas in the stomach and intestines in live birth.
External Findings Weights and measurements mLlst be carefully recorded. A careful external examination is most important. Any changes of maceration are carefully recorded as they provide useful information about the fetal death to delivery interval. Serial maceration changes are tabulated (Table 8.5);
Examination of the newly born I Table 8.5
External changes of maceration by death to delivery
interval (after ref. BB)
Observed feature
No maceration No discolouration of cord insertion Desquamation ?! 1cm Cord discolouration, brown/red Desquamation face/back/abdomen Desquamation ?! 5 per cent of body surface Desquamation in two or more zones' Skin brown/tan in colour Moderate/severe desquamation Mummification (dehydration :t fetal compression)
Death to delivery interval
< 6 hours < 6 hours ?! 6hours ?! 6 hours ?! 12 hours ?! 18 hours ?! 18 hours ?! 24 hours ?! 24 hours ?! 2 weeks
'Sody zones: scalp, face, neck, chest, abdomen. back. arms, hand, leg, foot,
scrotum.
Ch anges of maceration are accelerated by fetal hydrops and retarded in
the presence of fetal growth restriction.
early changes of maceration are ill ustrated in Figs 10.4- 10.6 p. 184. The presence of excessive meconium staining indicates hypoxic fetal stress in the mature baby. If there is much meconium, blood or vernix caseosa on the skin it is advisable to remove it after initial photography so that the skin can be examined in more detail. Cutaneous pallor may be evident following significant fetomaterna l haemorrhage. 53 The presence and nature of any cutaneous injury is recorded and photographed. It is not unusual to find cutaneous petechial haemorrhages over the presenting part in fresh still birth, particularly when there is placental abruption. They are commonly found over the face, head and neck and, occasionally, on the chest wall in a mature fetus when vertex presentation is usual. They are often particularly prominent following placental abruption (see Fig. 10.10, p. 188) and should not be interpreted as evi dence of stra ngu lation or deliberate airways obstruction without corroborative evidence. Cutaneous petechiae, and sometimes more extensive bru ising, are seen on the legs following breech delivery or the arm if one has prolapsed through a partly dilated celvix. These findings are more usual in the premature fetus, for which breech presentation is more common. 54 At 32 weeks' gestation approximately 16 per cent of babies present by the breech, compared with around 5 per cent at term. The presence and nature of any dysmOlvhic features are noted. These may contlibute to a syndrome diagnosis, directly related to death. The presence oftvvo or more dysmorphic fea tures is an indication for chromosome examination. The umbilical cord and its insertion are examined and any discolouration or dehiscence is noted. The free end of the cord may indicate the method of division. Differences in appearance of the cord ends are described 55 - traction,
159
picking apart, and cutting with scissors or bl un t/sharp knives - and assessment is aided by dissecting microscopic examination. It may correspond with the free end of the cord attached to a placenta discovered elsewhere. The pres ence or absence of a clip or tie is noted and described.
Examination of The Scalp and Cranium EXTERNAL
Examination of the head and cranial contents is undertaken systematically. The head circumference should be approxi mately equal to the crown-rump length in the second half of pregnancy. A larger measurement raises the possibility of hydrocephaly, whereas a smaller one may result from crania l distortion or, in its absence, chronic in trauterine brain pathol ogy. An increase in the occipitomenta l diameter (see Fig. 10.9, p. 188) is a useful observation. Some moulding of the cranium is nOlmal but excessive moulding of the cranium is an indica tion of long labour. In vertex presentation there is often an area of localized oedema, usually with marked congestion and sometimes with frank haemorrhage, over the presenting part. INTERNAL
The scalp is incised from behind the ear over the posterior fontanelle to the opposite ear and reflected forwards and backwards, sufficiently far to completely expose the bones of the cranial vau lt. The presence of haemorrhage within the scalp is noted and a check is made for any corres ponding external injury. Marked congestion of the scalp is a very frequent finding in antepartum still births after vertex presentation. It should not be confused with tra uma. Histological examination shows no vital reaction. The con tour of the occipital bone is examined. Occipital osteodias tasis , in which the inferior pali of the OCCipital bone is displaced inwards resulting in damage to the brain or venous sinuses, may occur after breech delivery or, occa siona lly, following forceps delivery;56 it is eas il y missed if the sca lp is not ful ly reflected posteriorly. The posterior neck muscles are conven iently ex amined at this point and any contusion is photographed and sam pled for histological examination. The atlanto-occipital membrane sho uld then be incised under direct vision. Gentle pressure on the cranium wiiJ result in a flow of CSF, which is likely to be heavily blood stained when there is selious intracranial injury. The fontanelles and suture lines are obselved. Congestion or haemon'hage is sometimes seen at the margins of sku ll bones and probably indicates the stress of forcible displacement.
Cranial Cavity The cranial cavity is opened using the modifi ed Beneke technique (Fig. 8.19), which allows examination of the
160 I
Post-mortem examination in babies and children
Figure 8.20 Birth injury. Tear at the junction of the falx and tentorium follo wing instrumental delivery, displayed by removal of the cerebral hemisphere on the affected side.
Figure 8.19 Neonatal cranium opened using the Beneke technique; there is confluent subarachnoid haemorrhage, of hypoxic origin, over the right parietal region; this is not accompanied by significant subdural bleeding. superior surface of the brain while preserving the falx and tentorium for subsequent examination. Extradural haemorrhage is uncommon amongst peri-natal deaths because of strong adherence of the dural membrane to skull bones. Localized, non-space-occupying haemorrhage can accompany skull fracture. The presence and position of haemorrhage in the subdural and subarachnoid space is noted. The form er is the result of trauma; the latter may be traumatic when observed in the vicinity of torn bridging veins or when, as a result of hypoxia, the presence of multi ple petechial haemorrhages is likely, although confluent sub arachnoid haemorrhage of hypoxic oligin is recognized in the perinatal period, often over the temporal poles. Observation of the cortical gyral pattern provides useful information about fetal maturity. The gestation-related appearance is remarkably constant and is illustrated by Dorovini-Zis and Dolman 57 and Feess-Higgins and Lar roche 58 (see ref. 45). The falx and tentorium should be examined with care. This is best done by tipping the head forward then gently eleva ti ng each occipital pole in turn so that the posterior part of the falx and the tentorium can be viewed. The usual site of significant injury is from the free edge of the tentorium running into straight sinus (Fig. 8.20). Sheari ng tears in the free edge of the tentorium , not involving venous sinuses, are an indication of rapid moulding of the head and are not infrequent following spontaneous delivery, particularly in still birth. They do not, in themselves, contribute to death. Tears in the falx are less usual and it is imp ortant to be aware that lacunae, smooth-edged holes, can occur naturally within the falx during development.
In view of the incompleteness of myelination of the fetal brain, fixation of the whole brain before examination is essential otherwise important information may be lost.
Spinal Cord When intrapartum death occurs following in stru mental delivery and no intracranial patho logy is app arent, the cer vica l spinal cord should be removed within its bony cover ing using Yates' method. 59
Body Cavities Examination of thorax and abdomen follows a method appropriate to this age group.45.50 Petechial haemorrhages are numerous in the thoracic viscera follo wing acute hypoxic stress, and may be particularly prominent around the ductus arteriosus. They are particularly promin ent and numerous after placental abruption. Potentially fatal congenital anom alies seen in externally normal infants are desCIibed in Chapter 10, p. 193 . In the macrosomic fetus (>4000g) it is important to look for evidence of birth injury.
Placenta EvelY effort should be made to examine the placenta. It may contribute important information to case evalu atio n in most circumstances when there is a potential legal interest in a perinatal death .6o Examination of the placenta is detailed by Wigglesworth 5o and Keeling 45 . Cord length, weight and measurements should be recorded as they contribute to the overall assessment of the case. Gestation-related placenta weights for twins and singletons are given by Pinar et al 61 and cord lengths by Naeye 62 (see Appendix BJ. Histological examination of the placenta may contribute to the assess ment of the death to delivelY interval in stillbirth. 63 Help with interpretation of placental abnormality can be found in Benirshke et al,64 Fox and Sebire 65 and Khong 66 .
Examination of the newly born
I
161
.-\ltshuler67 and Macpherson 68 discuss the relevance and limitations of placental data and pathological abnormalities in relation to alleged obstetric malpractice.
Recognized Birth Injuries Birth injury is more common fo llowing instrumental, and sometimes operative, delivery. However, there are occasional case reports of birth inju ries following spontaneous delivery. Birth injuries are considered in more detail by Wig glesworth 69 and Keeling.70 After delivery in hospital it is important to look for signs of birth trauma such as a chinon following Ventouse delivery, bruising and abrasions in the shape of a forceps bl ade, and perhaps a linear skin incision following Cae sarean section. Scalp injury is a recognized complication of vacuum extraction. 7] Bruising of the legs is common fol lowing breech delivery, especially in preterms. In term breech delivery, it is important to look for predisposing fac tors, such as neuromuscular pathology or renal abnormal ity (because of oligohydramnios), which might be the underlying cause of death. Subaponeurotic (subgaleal) haemorrhage is an infre quent but serious complication of forceps or Ventouse delivery, or a combination of both methods (Fig. 8.21). 72 A large volume of blood can be lost into the su baponeu rotic space. Robinson and Rossiter 7J calculated blood losses exceeding 100 mL using a formula based on cranial diam eter and scalp thickness. Subperiosteal haemorrhage, particularly over the par ietal bones, is common (Fig. 8.22) and of little consequence. It may be found in live borns following spontaneous, easy vaginal delivery with vertex presentation, as well as fol lowin g instrum ental delivery. The amount of bleeding is limited by periosteal attachment at the margins of the bone and by the limited distensibility of the membrane. Although it may be found following difficult delivery, with other pericranial lesions, as an isolated finding, it makes little contribution to the overall assessment of the case. Subdural74 and cerebral haemorrhage7 5 and very occa sionally, extradural haemorrhage 76 or posterior fossa haem orrhage 77 can complicate vacuum extraction as well as forceps delivery. Subdural haemorrhage is almost always due to trauma. When present over the cerebral convexities it is the result of tearing of bridging veins; trauma here includes excessive moulding of the skull from obstructed labour or rapid moulding of the cranium in precipitate deliv ery. Minor, focal subdural haemorrhage is not uncommon following normal vaginal delivery.78 Evidence of these minor haemorrhages in the form of focal brown staining of the dura is not unusual in sudden infant death . Any sugges tion that major subdural haematoma is th e result of unas sisted delivery should be addressed very critically.79 Tears of the dural folds that extend into cerebral sinuses are rare in non-instrum ental delivery but cause catastrophic fatal haemorrhage. Subdural haemorrhage solely at the base of
Figure 8 .21 Birth injury. Subgaleal haemorrhage, massive blood loss into soft tissues. Instrumental delivery.
Figure 8 .22 Birth injury. Subperiosteal haemorrhage following spontaneous delivery; blood loss is trivial. the brain is unlikely to be traumatic and a local cause should be carefully sought. Fra ctures of the skull vault should be sought with care; th ey are not always apparent on radiographs. A single fracture in the parietal bone from the sagittal suture line running down towa rds the ear is seen occasionally follow ing forceps delivery or vacuum extraction. An undisplaced Iinear fracture is not of great significance in respect of cause of death, although it is an indication of difficult delivery. Of more significance is an accompanying sub aponeurotic or intracranial haemorrhage, or cerebral hypoxic injury. Clavicular fracture is the commonest birth injury. There is usually a history of difficulty in delivering the shoulders in a large baby. Bruisin g of the neck muscles (Fig. 8.23) and fracture of the humerus can occur in similar circumstances. It is essential to correlate any injuries found with events in labour. Babies in the extended breech position may suffer fracture of a femur (or occasionally both) when the legs are brought down. Rib fractures are exceedingly rare birth injuries. Careful attention should be given to delivery events, birth weight and events in the neonatal period
162 I
Post-mortem examination in babies and children Table 8.6
Samples for histological examination in perinatal and infant forensic deaths20,45
Figure 8.23 Birth injury. Bruising of anterior neck mu sc les. Instrumental delivery. befo re accepting birth injury as a cause for rib fract ures in an infant. Rib fractures are discussed further in Chapter 11. The commonest visceral injury is to the liver, when rup ture of a subcaps ular haematoma can result in catastrophic haemo rrhage. The baby is usually mac rosomic. A simil ar injulY is seen in preterm infants, particularly following breech delivelY, as the liver is poorly protected by the rib cage at this time. Splenic rupture is a rare birth injury.
HISTOLOGICAL SAMPLES Thorough sampling for histological examination is an essential part of perinatal, infant an d childhood post mortem examination. In ea rly life, much diagnostic infor mation is derived from microscopi c examination when macroscopi c ap pearan ces may be normal, non-specific or not understood by the inexperienced. Sampling should fol Iow a standard protocol (Table 8.6). It is the in vestigation most likely to make a positive contribution in SUDl. 2o Norma l his to logical appearances of organs in the infant period are illustrated in Valdes-Dapena et al. 40 In addition to convent ional sectio ns of fo rmalin-fixed para ffin-emb edded samples, it is important to reserve sm all fresh samples of heart, liver, kidney and adrenal gland for frozen sections in order to look for lipid infiltration ; when small in amount, lipid infiltration is probably st ress-related, but massive lipid infiltration is su ggestive of genetic meta bolic disease (see Chapters 11 and 12). These sa mples can be kept frozen, perhaps wit h a sample of spleen, for molecular studies should the need arise. If inhalation is suspected, a froze n sample of lung for lipid staining is also useful. In SUDI, it is advisable to examine lung sections stained fo r iron (PBR) to assess the possibility of previous pulmonalY haemorrhage. It may be useful in other circum sta nces, Wh en death occurs suddenly in the neonata l period, staining for reticulin is very usefu l in the evalu-ation of microscopic lung anatomy: it permits exclusion of congenital alveolar dysplasia and better assessment of the appropriateness of
la rynx Trachea Lung - five lobes, PBR reaction on all Heart - right and left free ventri cu lar wall, interventricular septum Kidney - two Liver - right and left lobes Small and large intestine Pituitary Thyroid Pan crea s Adrenal - two Gonad Diaphragm Costochondra l junction Any macroscopic lesions Dural fold Dura - parietal In SUOI (including sudden neonatal death)
Frozen section for oil red 0 staining from: He art Liver Ki dney Adrenal ~L u ng
In perinatal death
Pla centa X 2 Umbilica l cord Extra placental membranes lung maturity. An elastic stain, (Elastic van Gieson is my preferred technique) allows assessment of postnatal adaptation of pulmon ary vasculature, ventilation-induced damage and other residua of neonata l intensive care. Exa mination of the ca rdiac conduction system is desir ab le following instantaneo us death or observed rapid col lapse that is unexplained after conventional investigation. It should be considered when sudd en death follows cardiac surgelY and when a second sudden death occurs within the family. In these circumstances the heart should be retained complete with this in mind. Conventional sampling is desc ribed and illustrated by Davies et al. BO Michaud et alBI and Ashworth B2 describe a simplified method for analysis of the conduction system. It is imp ortant no t to over interpret minor differences in microanatomy, either in distributio n of conduction tis sue B3 or of the process of so -called 'resorptive degeneration' of the atrioventricular nodal tissue and the His bundl e. 83 ,84 The so-call ed marke rs of SIDS, pulmonary haemosiderin and the significance of minor degrees of infla mmation in SUD! are discussed in Chapter 11, pp, 209-212, 215-2 16 an d p. 210, respectively.
Exchange of information and multidisciplinary review I
Figure 8.24 Posterior end of rib trimmed for histological examination after fi xa tion and decalcification. A fracture with callus is present.
163
Figure 8.25 Whole section of the anterior end of a lower rib. There is expansion of the interna l aspect of the rib and irregularity of the osteochondral junction as a result of prior trauma.
Demonstration of Bony Injury Histological exa mination of bones following radiographic :dentification of fracture or unidentified localized abnor mality is essential. The bone(s) in question can be excised completely or a segment excised with a clear margin on all aspects of the area to be investigated. It is useful to re-X-ray the excised bone, either before or after fixation. These films are useful when sampling for microscopy. Sampling should be done following decalcifi cation, otherwise further damage to the bone is likely and histological assessment rendered more difficult. Blocks for microscopic examination are taken at right angles to the fracture line (Fig. 8.24) or along the long axis of the lesion (Fig. 13.25). Blocks should include cortical bone, periosteum and adjacent soft tissues. Histological examinatio n of frac ture contributes to assess ment of timing of injury, particu larly those that have occurred within 1 week of death, when radiological abnormality is minimal. The significance of fractures in perinatal and infant deaths are discussed further in Chapter 3 (pp. 51-581, Chapter 10 (p. J92) and Chapter 11 (p. 218).
RETENTION OF ORGANS In the course of the medicolegal post-mortem examination in babies and children, careful thought should be given to the need to retain organs. Many coroners and procurators fiscal are unwilling to sanction organ retention, particu larly the brain. It is importa nt that pathologists ensure that legal personnel are appropriately informed of the benefits of so doing in any particular case. It is difficult, and in some cases impossible, to examine the brain of a still born or young infant in the fresh state. The information obtained from an appropriately thorough examin ation of
the central nervous system is of utmost importan ce, in the context of both assessing injury when inflicted injury is suspected and in respect of cause of death in SUDI and pelinatal deaths. Local expelience of the contribution of formal neuropathological examination in deaths investi gated by a procurator fiscal is greater than that cited by the CESDI SUDI studies. 8s Brief (overnight) fixation of other organs, particul arly the heart, makes sampling for microscopic examination easie r and more accurate. In cases of inflicted injury, organs may require fixation so that they are in an appropriate condition for examina tion by another pathologist. In these cases, if there is doubt concerning the identity of lesions then the organs or tissues should be retained.
EXCHANGE OF INFORMATION AND MULTIDISCIPLINARY REVIEW In complex cases, it is important that there is regular exchange of information between all the professionals involved so that appropriate interpretation of findings ensues. Th e pathologist should ensure that the results of any ancillary investigations are available and that conclusions are based on the appropriate interpretation of all findings. Multidisciplinary reviews are conducted on all sudden and unexpected death in a number of areas within the UK, con vened either through hospital paediatric units, community paediatricians or via child protection agencies. Following the publication of an intercollegiate report, such reviews are likely to become standard practice throughout the UK. 86 The empha sis of such reviews is likely to vary depending on the lead agency and the composition of the review gro ups. Although these activities are time-consuming, they are productive both
164 I
Post-mortem examination in babies and children
in respect of child protection a nd in our understanding of causes and contributing factors in sum, putting in place app ropriate support for famili es and in the appropriate plan ning of selvices for chiJdren.o 1
21
22
23
REFERENCES
2
3 4 5
6
7 8 9 10
11
12 13
14
15 16
17
18 19
20
Vujanic GM Cartlidge PHT, Stewart JH. Improving the quality of pe rinatal and in fant necropsy examinations: a follo w up study.] Clin Patho/1998; 51:850-3. Sheehan KM, McDonnell M, Doyle EM et a!. The qua li ty and val ue of sud den in fan t death necropsy reporting in Irelan d. ] Ciin Path 01 2003; 56:753 - 7. Bass M, Kravath RE, Glass L. Death-scene investiga tion in sudden infant death . N Engl] Med 198 6; 315:100-5. St urner WQ. Commo n errors in forensic pediatric pathology. Am] Forel1 sic Med Pat110/1998; 19:317 -20. Bass M. The falla cy of the sim ultaneous sudden infant
dea th syndrome in rwi ns. Am] Forensic Med Pathol 1989;
10:200-5.
Sa ukko P, Knight B. The pathophysiology of death. In Kn ight's Forel1s ic Pathology, 4th edn. London: Arnold , 2004, pp. 52-97. Ca rpenter RF. The prevalence and distribution 0 f bruising in babies. Arch Dis Cilild 1999 ; 80:363 - 6. Sugar NF, Taylor JA, Feldman KIN et a1. Bruises in infa nts and toddlers. A rch Pediatr Adoles c Med 1999; 153:3 99- 403 . Moon RY, Patel KM, Schaefer SJ. Sudden Infant Death Syndrome in child care settings. Pediatrics 2000; 106:295-300. Maxeiner H. Demo nstrati on and interpreta tion of bridg ing vein rup tures in cases of infantile subdura l bleedings. ] Foren sic Sci 2001; 46:85-93. Erlich E, Maxeiner H, Lan ge J. Postm ortem radiological in vestigation of torn bridging vein ruptures. Legal Med (Tokyo) 2003; 5lSupp1. 1) : 225-7 . Sadler DW. The value of a thorough pro tocol in the in vestigation of sudden infant dea ths.] Ciin Pathol 1998; 51 :589-94. Wigglesworth JS , Keelin g JW, Rushton D, Berty PJ. Tech nica l method. Pathological inves tigations in cases of sudd en infan t death.] Ciin Pat110/198 7 ; 40:1481-3. Dettm eyer R, Baas ner A, Schlamann M et a!. Coxsa ckie BJ myocarditis in 4 cases of suspected sudden infant death syndrome: diagnos is by immunohistochemical and molecular pa thologic invest igation. Patlwl Res PraU 2002; 198 :689-96. Bajanowski T, Rolf B, Jorch G et al. Detection of RNA viruses in sudden infant death (SID), Int] Legal Med 2003; 117 :237- 40. Carmichael EM, Gold wate r PN, Byard RW. Routin e microbiological testing in sudden and un expected infant death. ] Pediatr Child Health 1996; 32:412-15. Mitchell E, Krous HF, Donald T, Byard RW. An ana lysis of the usefuln ess of spec ific stages in the pathol ogic investigation of sudd en infant death . Am ] Forensic Med Pat flol 2000; 21 :395-400. Sims MA, Collins KA. Fetal death: a 10-year retrospective study. Am] Forensic Med Pathol 2001; 22:251 - 5. British Med ica l Associatio n. British Natiol1al Formulary, Paediatric. Londo n: Roya l Pharm aceutical Society of Great Britain, 2005, p. 225. Berty J, Allibone E, McKeever P et al. The pathol ogy study: The contribution of ancillary pathology tests to the investiga tion of unexpected infant death . In Fleming P, Bacon C, Blair P et al (eds) Sudden Unexpected Deaths in Infancy the CESDI SUDI Stu dies 1993-1996. London: The Statio nery Office, 2000, pp. 97-11 2.
24
25
26
27
28
29
30
31
32
33
34 35 35 37
38 39 40
41
Ackerman MJ, Tester DJ , DriscoU DJ. Molecular autopsy of sudd en unexplained death in the you ng. Am] Forensic Med Patho1 2001; 22:105-11. Wilc ox RL, Nels on CC, Stenzel P, Steiner RD. Postmol1em screening for fatty ac id oxidati on di sord ers by analysis of Guthrie ca rds with tand em mass spec trom etry in sudden unexpected death in infancy.] Pediatr 2002; 141 :833-6. Yudkin PL, Aboua lfa M, Ey re JA et al. The influence of elective preterm delivery on birth- weight and head ci rcu mference sta ndard s. A rch Dis Child 198 7; 62: 24-9. Chambers HM, Knowles S, Staples A et al. Anthropo metric measurements in the seco nd tri mester fetu s. Early Human Dev 1993; 33:45-59. Maroun LL, Graem N. Autopsy standards of body pa rameters and orga n weights in non-macerated an d macerated human fetuses. Pedia tr Deuel Patho/2005; 8:204-17. Freeman N , Co le TJ, Chinn S, Jones PR M, White EM, Preece MA. 1995 Cross-s ecti ona l stature and weight refere nce CUlves for the UK, 1990. Arch Dis Child 73:17- 24. Singer D, Sung CoR, Wiggleswonh JS. Feta l growth and maturation: with stand ards for body and organ development. In Wigglesworth JS , Singe r D (eds) Textbook of Fetal and Perinatal Pathology, 2nd edn. Ma lden, MA: Blackwell Sc ience, 1998, pp. 8-40. Hansen K, Sung CJ , Huang C et al. Referen ce values for second trimes ter fetal and neonata l organ weights a nd meas urements. Pediatr Develop PatI101 2003; 5:160-7. Thompson WS, Cohl e SD . Fifteen-year retrospective stud y of infant organ weights a nd revisio n of sta ndard weight tabl es. ] Forensic Sci 2004; 49:1-11 . Shan kle WR, Landing BH, Gregg J. Normal organ weights of infa nts and children: Graphs of va lue by age, with confide nce intervals. Pediatric Patho l 1983; 1:399-408. Haddad S, Res tieri C, Krishn an K. Ch aracte rization of age related cha nges in body weight and orga n weights from birth to ado lescence in humans. ] Toxic ol Enviroll Health 2001 ; 54:453-64. Altma n PL, Dittmer DS. Growth: In cluding Reproduction alld Morp hological De velop men t. Washington DC: Federation of Ame rican Socie ties for Exp erime ntal Biology, 1961. Handy TC, Nichol s GR , Buchin o JJ. A pediatric fo rensic medicine prog ram. In Dimniick JW, Singer DB (eds) Forens ic A spects ill Pediatric Pathology. Perspectives in Pediatri c Pathology. Basel: Ka rger, 1995, pp. 87-95. Kaplan JA, Fossum RM. Patterns of facial resuscitation injury
in in fan cy. Am ] ForellS ic Med Path o/199 4; 15:187-91.
Downes AJ, Crossland DS, Mellon AF. Prevalence and distribution
of petechiae in well babies. Arch Dis Child 2002; 86:291-2.
Byard RW, Krous HF. Petechial hemorrhages and unexpec ted in fant death. Legal Med (Tokyo) 1999; 1:193-7. Krous HF, Nad eau JM, Byard RW, Blackbourn e BD . Oronasal blood in sud den infant death. Am ] Fore nsic Med Path ol 200 1; 23 :346-51. Mayes C, Macleod C. When 'N AJ' means not act ually injured. BM] 1998; 318:1127 - 8. McCann J, Reay D, Siebert J et al. Postmortem perianal findings in children. Am] Foren sic Nled Patho/2006; 17:289-98. Valdes-D ape na M, McFeel ey PA, Damus KH et al. Histopa thology Atlas fo r the Sudden Infant Dea th Syndrome: Fi ndings derived from the Natio l1 al/llstitute of Child Health and Hum an Develop ment Cooperati1!e Epidemiological Study of Sudden Infant Death Syndrome (SIDS) Risk Factors. Washington : Armed Forces Institute of Pathology, 1993. Kro us H. The interna tional sta ndardize d autopsy protocol for sudden unexpected infant death. In Rognum TO (ed.) Sudden Infant Death Syndrome. New Trends in the Nineties. Os lo: Sca ndin avian Uni ve rsity Press, 1995, pp. 81-95.
References I
';1
";3
~4
-lS
46
47 48
49
50
51 52
53
54
55 56
57 58
59 60
61
62 63
64 65 66
van Baarlen J, Schuurman H-J, Huber J. Acute thymus involution in infancy and childhood: a reliable marker for duration of acute illn ess. Hum Pathol 1988 ; 19: 1155-60. Beckwith JB. Intrathoracic petechial hemorrhages; a clue to the mechanism of death in sudd en infa nt death sy ndrome? Ann Acad NY Sci \988; 533:37-47. Muir R, Niven JSF. The local formation of blo od pigments. J Pathol Ba cteriol 1935; 41 :183-97. Keeling JW. The perinatal n ecropsy. In Keeling JW, Khon g TY (eds) Feta l and Neona ta l Pathology, 4th edn. London: Sprin ger, 2007, pp. 20-53. Hardman JM. Cerebrospinal trauma . In Davis RL, Ro bertson DM (eds) Textbook of Neuropathology, 3rd edn. Baltimore, MD: Williams Et Wilkins, 1997. Parsons MA, Smart RD. Necropsy techniques in ophthalmic pa thology. Best Practice 164 J Clin Path 01 2001; 54:417-27. Lee WR (ed.). Examination of the globe. Techni cal aspects. [n Oplithalmic Histopathology, 2nd edn. London: Springer Verlag, 200 2, pp. 1-33. Reichard RR, White CL Ill , Hogan RN et at. Beta-amyloid precursor protei n immunohistoch emistry in the evaluation of pediatric traumatic optic nerve inju ry. Ophthalmology 2004; 111 :822-7. Wigglesworth JS. Performa nce of the perinatal autopsy. In Wiggleswowrth J S (ed.) Peri/lOtal Pathology, 2nd edn. Phil adelphia , PA: Sau nders, 19 86, pp. 23-40. Russell JGB. Radiolog ical assessment of fetal maturi ty. J Obstet Gynae Brit Cwlth 1969; 76:208-19. Ste mpfle N, Huten Y, Fondacci C et al. Fetal bone age rev isited: proposal of a new radiographic score. Pediatr Radiol 1995 ; 25:551-5. Biankin SA , Arbu ckle SM, Graf NS. Autopsy findings in a series of five cases of feto maternal haemorrh age. Pathology 2003; 35:319-24. Ha rlow FHD, Spencer JAD. Obstetrics for the neonatologist. In Rennie JM, Roberton NRC (eds) TeXTbook of Neonatology. London: Churchill Livin gston e, 1999, pp. 157-77. Morris JF, Hunt AC. Breaking streng th o f the umbilical cord. J Forensic Sci 1966; 11:43-9. Wigglesworth JS, Husemeyer RP. Intracra nial birth trauma in vaginal breech delivery: the continued importa nce of injury to the occipital bone. Br J Obstet Gynecol 1977; 84:684-91. Dorovini-Zis K, Dolman CL. Gestational develo pment of brain. Arch Pathol Lab Med 1977 ; 101 :192-5. Feess-Higgins A, Lan'oche J-c. De1Jelopmel1 t of the Human Foetal Brain. An Anatomical Atlas. Pa ris: Inserm CNR S Masson, 1987. Ya tes PO. Birth trauma to the vertebral arteries. A rch Dis Child 1959; 34:436-41. Marchet ti D, Belviso M, Marino M, Gaudio R. Evalu ation of the pl acen ta in stillborn fetuses to estimate the time of death. A m J Forensic Med Pat/wi 2007; 28:38-43. Pinar H, Sung CJ , Oyer CE, Si nger DB. Reference values for singleton and twin placental weights. PeliiMr Patlwl Lab Med 1996 ; 16 :901-7. Naeye RD. Umbilical cord length: clini cal significance. J Pedi atr 1985; 107:278-81. Genest DR. Estimating th e tim e of death in stillborn fetuses: II. Histologi c evaluation of the pla centa; a study of 71 stillborns. Obstet Gyneco11992; 80:585-92 . Benirschke K, Kaufmann P, Bae rgen R. Pathology of the Human Place nta , 5th edn. New York: Springer, 2006. Fox H, Sebire N. Pathology of the Placen ta, Major Problems in Pat/wlogy, 3rd edn. London: Saunders, 2007. Khong TY. The placenta and umbilical co rel. In Keeling JW, Khong TY (eds) Fetal and Neonatal Pathology, 4th edn. London: Springe r, 2007 , pp. 54-89.
67 68 69
70
71 72
73 74 75
76
165
Altshuler G. Placenta w ithin the medicolegal imperative. Arch Pathol Lab Med 199 1; 11 5:688 -95. Macphe rson 1. Fact a nd fancy. What can we really tell from the placenta? Arch Path Lab Med /991; 115: 672 -81. Wigglesworth JS. Pathology of intrapartum and early neon atal death in the normally form ed infa nt. In Wigglesworth JS, Singer DB (eds) Textbook of Fetal and Perinatal Pathology, 2nd edn . Ma lden , MA: Blackwe ll Science, pp . 251-68. Keeling JW. Intrapartum problems. In Keelin g JW, Khong TY (eds) Fetal and Neonatal Pathology, 4th edn. London: Springer, 2007, pp. 273-96. Teng FY, Sayre JW, Vacuum extraction: does duration predict scalp if\iury? Obstet Gynecol 1997; 89 :281-5. Amar AP, Aryan HE, Meltzer HS, Levy ML. Neonatal subgaleal hematoma causing brain compression: results of two cases and review of the literature. Neurosurgery 2003; 52: 1470-4. Robinson RJ , Rossiter MA. Massive subaponeurotic hae monhage
in babies of African origin. Arch Dis Child 1968 ; 4 3 :684-7.
Doward W, Sgouros S. Acute su bdural haematom as follo wing
Ventouse-a ssisted delivery. Pediatr Neu rosurg 2001; 35:336, Towne r D, Castro MA, Eby-Wilkens E e t al. Effect of mode of delivery in nulliparous women on neonatal intra cranial injury. N Engl J Med 1999; 341: 1709-14. Vachharajani A, Mathur A. Ultrasound-guid ed needle aspiration of cranial epidura l hema toma in a neonate: trea ting a ra re complication of vacu um extraction. A m J
Perinatoi2002; 19:401-4. 77 Perri n RG, Rutka JT Drake JM et al. Ma nagement and outcomes of posterior fo ssa subdural hematomas in neonates . Neu rosu rgery 1997; 40:1190-9; discussion 1199- 2000. 78 Whitby EH, Griffiths PO, Rutter S et al. Frequency and na tural histo ry of subdural haemorrhages in ba bi es a nd relation to obstetric factors . Lancet 2004; 363 :846-51. 79 Salman M, Crouch man M. What ca n caus e subdura l haemorrhage in a term neo nate? Paediatr Today 1997 ; 5:42-5. 80 Davies MJ , Pomerance A, Lamb D. Tec hniques in exa mination and a natom y of the heart. In Dav ies MJ , Pomerance A (eds) The Path ology of the Hea rt. Oxford: Blackwell Sc ientific Pu blica tions, 1975, pp. 26-31. 81 Michaud K, Romain N, Taroni F et al. Evaluation ofa simplified method of th e conduction system a nalysis in 100 forensi c cases . Forensic Sci Int 2002; 130:13-24. 8 2 Ashworth MA. The cardiovascular system. In Keeling JW, Khong TY (eds) Fetal and Neonatal Path ology, 4th edn. London: Springer, 2007 , p. 613. 83 Ho SY, Anderson RH. Conduction tissue and SIDS. Am NY Acad Sci 1988; 533:176-90. 84 Ma tturri L, Otta viani G, Ramos SG, Rossi L. Sudden infant death syndrome (SID S): A study of cardiac conduction system . Cardiovasc Pat1lO1 2000; 9:137-45. 85 Keeling JW, McKenzie KJ, Ironside J et al. Does detailed neuropathological ex amina tion o f the fi xed brain in sudden, un expected dea th in infa ncy contribute to the evaluation of the death? J Pat/wi 2002 ; I 98(Suppl.) :2 4A.
86 Report of a Working Group . Sudden Unexpected Death in Infancy. London: Royal College of Path ologists and Royal College of Paedi atrics a nd Child Hea lth, 2004. 87 Fleming P J, Blair PS, Sid ebotham P, Hay ler T. Investiga ting sudden unexpected deaths in infancy and childh ood and ca ring for bereaved families: an integra ted multiagency appro ach. BMJ 2004; 328:331-4. 88 Genest DR , Singer DB. Es timatin g the time o f death in stillborn fetuses: JII. Extern al fetal examination; a study of 86 s tillborns. Obstel Gynecol 1992 ; 80:593-600.
I
CHAPTER 9
I
PATHOLOGY OF NEUROLOGICAL ABNORMALITY IN EARLY LIFE Waney Squier
Introduction
166
cerebral palsy Timing of injuries by histology Acquired intra-uterine damage
Birth-related injury Stroke in the developing brain
Clinical manifestations of early brain damage:
167 167 169
INTRODUCTION
Examination of the nervous system forms a critical part of the autopsy exami nation of any infant who dies, whether suddenly and unexpectedly or following recognized illness. The cause of sudden infant death is not always proximate insult or disease; pre-existing diseases of, or damage to, the nervous system may cause sudden death. Babies with neuro logical disease, for example, are more pron e to aspiration of feeds and to epilepsy. Even quite extensive congenital brain damage acquired in utero or at the time of bilih may not present clinically until months or years later, while still hav ing the potential to cause sudden death . It may be that a 'secon d insult' is required, for example velY mild brain mal formation (microdysgenesis) is seen in patients with epilepsy but may not cause seizures until there is another insult such as trauma. Similarly, babies with metabolic or mitochondrial disease may become profoun dly unwell a nd die when chal lenged by a respiratory infect ion. In these babies careful exa mination of skeletal muscle with histochemistlY and review of pre-mortem biochemistry is necessary if an accu rate diagnosis is to be made. Chronic subdural haemorrhage may go quite unrecognized for weeks or months until dis covered by routine head circumference measurement or even by parental observation of increasing head size. The pathologist making an autopsy exa mination in sud den infant death is faced with an awesome task. Th ere will often be great pressure from those who have cared for the child in life to find a pa rticular cause of death . Opinions concerning the possibility of natura.1 or unnatural causes
Metabolic disorders Infections References
173 176 177 177 178
may be very strongly held. The findings in the brain are of paramount importance in such cases but may also be very subtle. If trauma is suspected the possible mechanisms must be fully considered and due attention paid to exami na tion of the stru ctures of the neck and spinal cord as well as the brain. Meticulous neuropathological examination can identify no t only the cause of neurological disease in childhood but, in the case of acquired damage, can also assist in evaluating when the damage has occurred. How ever, the effects of the terminal condition of the baby must be taken into consideration. Hypoxia and resuscitation, as well as a period of ven tilation , will lead to brain swelling and ge neral organ failure, disturbance of blood clotting, leaking from blood vessels and slowing or cessation of cel lular reactive processes, serious ly hampeling the ab ili ty to time the injUIY. Careful correlation with brain scans take n closer to the time of coll apse will assist in distinguishing primary injUly from these secondary changes. Neurological abnormalities may resu lt from genetic or meta bolic disease, diseases acquired in utero, during deliv ery, in the pelinatal period or in early life. Intra-uterine damage may not be symptomatic until weeks or months after birth, for example cerebral palsy is not usually definitively diagnosed until 5 years of age. Con versely, babi es damaged immediately before or during bilih tend to exhibit signs or symptoms such as floppiness, depression and asphyxia in the first days of life. Once the brain has been damaged there will be tissue loss followed by atrophy. Tissue regeneration and plasticity due to compensatory hyperplasia has been demonstrated in
Timing of injuries by histology
I
167
the immature human brain.l Following even static and non-progressive insult, clinical signs may evolve due to atrophy of associated and connected brain areas.
fragments of cerebellum in the su barachnoid space at mul tiple levels of the spinal cord. Histologically, oedema is better seen in densely packed white matter tracts than in grey matter.
CLINICAL MANIFESTATIONS OF EARLY BRAIN DAMAGE: CEREBRAL PALSY
Cell Death
Cerebral palsy remains the mo st common form of chronic motor disability in children (1-2 cases per 1000 live births). The full effects of early damage may not be apparent until several years of age. Signs include abnormal control of movement or posture, cognitive impairment, seizu res and blindness. Sudden death may be due to seizures or swallow ing disorders and aspiration. Timin g the damage is crucial not only to the under standing the possible aetiology, but also it will be critical in any potential litigation. The cause of cerebral palsy remains unknown in the majority of cases. In term babies the most impoltant causes are stroke (17 per cent), maternal infection (12 per cent), mul tiple pregnancy (10 per cent) and birth asphyxia (6 per cent). Birth injury, genetic and metabolic diseases, fetal infections and toxins account for a small percentage of cases. 2
TIMING OF INJURIES BY HISTOLOGY Histological timing cannot be regarded as precise and must always be interpreted in the light of all the other aspects of the case, partiCUlarly the clinical history. Much of the data below has been previously presented with source refer ences.] The earliest reactions in the infan t brain are oedema, neuronal death and cellular reactive changes (Table 9.1). Timings must be interpreted with caution, palticularly if the baby was venti lated prior to death. During ventilation the brain may become very swollen and the blood supply compromised (respirator brain). In these circumstances the reactive processes may be slowed or otherwise modified. The dura retains a better blood sup ply via the emissary veins of the skul.! bones and in the case of subdural haemorrhage dural histology may prove to be a better indicator of timing.
Cerebral Oedema Brain swelling may begin within minutes of injUlY, reach in g a maximum after 1 week. The speed of swelling is sub ject to huge individual variation. It causes narrowing of the sulci and compression of the ventricles. On slicing the brain, the cortical ribbon may appear prominent, with soft ening and grey discolouration of the underlying white matter. Flattening of the gyri is uncommon as the unfused infant skull can accommodate considerable brain swelling without compression. However, it is not uncommon to find cerebellar tonsillar herniation and necrosis with displaced
Th e two best-described patterns of cell death are necrosis and apoptosis (Fig. 9.1). Necrotic neurones develop intense cytoplasmic eosino philia, the nuclear membrane lyses and the chromatin dis perses into a fine web. Cytoplasmic eosinophili a alone may be a reversible or artefactual change, often seen in surgical material. Nucleolysis is part of the irreversible process. These changes take less than 12 hours, in the adu lt brain they are thought to take 5-6 hours to develop but timing in the infant brain has not been accurately documented. Apoptotic cells undergo pyknosis, when the nucleus becomes shrunken, rounded and intensely basophilic. Later the nucleus forms multiple rounded densely staining masses (kalyorrhexis). Apoptotic cells are readily phagocytosed and cause no inflammatory response. This process probably takes 12 hours to complete but, again, accurate fig ures are not available for the infant. Apoptosis is a more frequent in the immature brain than in the adult brain.
Macrophage Response During the first 24 hours after hypoxic-ischaemic injury (HIl) , microglial cells (the intrinsic phagocyte population of
Table 9.1
Early responses to injury within the brain
Cerebral oedema Cell death Necrosis Apoptosis Macrophages Microgl ia Phagocytosis Gliosis Capillary reaction End othelial swelling Capillary proliferation Haemosiderin Mineralization Axonal damage" Beta amyloid precursor protein expression Silver staining bulbs Haematoxyli n and eosin staining bulbs "Timing from ad ult tissue studies.
Minutes to 1 week 5-12 hours <12 hours <24 hours 3 days-yea rs 12 hours-6 days 1-3 days 5-7 days 48 hours-years 8 days-years 0- 3 hours 12-18 hours 15-24 hours
168 I
Pathology of neurological abnormality in early life
Figure 9.2 Acute severe hypoxic-ischaemic injury. Reactive capillaries in the white matter showing branching and proliferation. The endothelial cells are plump, several layers thick and many are apoptotic (haematoxylin and eosin).
Figure 9.1 Acute severe hypoxic-ischaemic injury. Section of cerebellum. Asingle remaining Purkinje cell shows intense cytoplasmic eosinophilia and nucleolysis - typical features of necrosis. In contrast, the granule cells are pyknotic with rounded, densely basophilic nuclei. Many contain rounded apoptotic bodies (haematoxylin and eosin). the brain) are prominent and are recognized by their long, rectangular nuclei. These are replaced by macrophages in up to 2 days after injury and by 3 days these cells may contain phagocytosed debris or red cells within their cytoplasm. Damaged tissue is removed by these phagocytes, which frequ ently cluster around blood vessels where they may remain for months or years . If there has been haemorrhage then red cells are seen in the cytoplasm of macrophages in the early stages; they are later converted to haemosiderin, which is brown in haematoxylin and eosin (HE) stain but after about 48 hours may be detected with Peds' stain.
Gliosis
injury (Fig. 9.2). Endothelial markers CD31 and CD34 assist in identifying reactive vessels. Capillary proliferation takes 5-7 days and is seen at the periphery of necrotic areas of brain. It is often particularly marked in the ischaemic cortex, particularly in the depths of the sulci, and is associated with the magnetic resonance imaging (MRI) appearance of 'cortical highlighting'.
Haemosiderin This is stainable in Perls' reaction 48-72 hours after haemor rhage. In late stages, iron is seen in astrocytes or free within tissues as well as in macro phages. It is common to find specks of haemosiderin within the dural layers in infants.
Cyst Formation Tissue breakdown and removal by macro phages causes cysts to develop, usually \0 days to several weeks after injury.
Mineral ization
Swollen astrocytes are seen from 12 hours after insult. Fib rillary gliosis takes about 6 days to occur. On the basis of radiological observations, it has been assumed that gliosis does not occur in the first half of gestation; however, fetal injury during amniocentesis at 17 weeks elicits a brisk gli otic response. 4
Mineral is readily deposited in damaged areas of the fetal brain. It begins as early as 8 days after injury and persists for years (Fig. 9.3). The pattern of mineralization may be a helpful diagnostic guide. Widespread irregular mineraliza tion is seen after infections such as herpes; periventricular mineralization typically occurs in cytomegalovirus (CMY). Patches of mineral in areas vulnerable to ischaemic injury suggest this cause.
Capillary Reaction
Axonal Damage
Endothelial cell swelling is a helpful and sensitive marker of tissue damage, becoming prominent 1-3 days after
Damage to the axonal membrane causes interruption of the intra-axonal transport of proteins, leading to irregular
Acquired intra-uterine damage I
Figure 9_3
1 69
Periventricular leucomalacia. The edge of the lesion
consists of a band of mineralized cell bodies and axonal fragments. Within this area, the scarred tissue contains glial cells
(b)
and many macrophages with pigmented cytoplasm (haematoxylin and eosin).
swellings or varicosities of the axon; if sufficiently severe, rounded swellings or axonal spheroids develop. These have been traditionally demonstrated with HE staining or with silver stains, but immunohistochemistry using anti bodies to the protein 'beta amyloid precursor protein' (0APP) has proved a highly sensitive and specific marker of axonal damage (Fig. 9.4) . Axons express 0APP after a vari ety of insults, including HII and metabolic damage as well as trauma;5 however, this stain cannot distinguish the causes of injury - this depends upon the specific pattern and distribution of expression, which, again, alone, is not diagnostic. Studies in the adult brain have shown that 0APP may be expressed from less than 30 minutes after injury6 but it is not a reliable indicator of timi ng. Silver staining takes some 12-18 hours to develop and axonal retraction balls are not seen with HE staining until about 15-24 hours after injuryJ Beta amyloid precursor protein staining persists for about 2 weeks in spheroids but may persist in irregular granular deposits for several months . Microglial cells accu mulate in areas of axonal damage within 4-5 days.
ACQUIRED INTRA- UTERINE DAMAGE Early
,
.-.
.
Figure 9.4
Damage in the first 20 weeks of gestation causes malfor mation. Specific malformations may indicate the timing of the causative insult as structures are malformed only dur ing th eir period of formation . Once formed, structures may show evidence of destruction. An example is seen in the corpus callosum, which develops from front to back
Periventricular leucomalacia. Small focus of
periventricular leucomalacia in an infant who died 11 days after severe hypoxic-ischaemic injury. There is a small collection of glial cells and macrophages, among which are proliferating capillaries and axonal spheroids. (a) HEtE; (b) C034, showing the newly formed capillaries; (c) ~ APP stain, showing the axonal swellings in the infarcted area.
170 I
Pathology of neurological abnormality in early life
between 11 and 17 weeks of gestat ion . If development is interrupted the posterior part will be deficient, whereas lesions in an terior or mid-p arts with an intact posterior part suggest later insults. AGENESIS OF THE CORPUS CALLOSUM
The corpus callosum may fail to form at all. In some cases this is due to severe und er-development of the cerebral cortex, wit h insufficient fibres generated to cross the mid line. In others, a bundle of fibres forms but fails to cross the midline and mns instead in an anterior-p osterior direction . This is known as a bundle of Probst and causes distortion of the ventricles, which take on the typica l 'Viking's horn' appeara nce that is recognized on scans (Fi g. 9.5). Agenesis of the co rpus callosum may be an isolated findin g, but is more frequently associated with other mal formations such as cerebella r hypoplasia. Its presence should prompt a search for these.
NEURONAL MIGRATION DISORDERS
One of the better-understood developmen tal processes is neuro nal migration. All neurones migrate from the place of their genesis in the germinal matrix to their final posi tion in the mature brain. Neuronal migration to the cere bral cOltex takes pl ace in the human brain by radial an d tangential pathways at between 6 and 23 weeks of gesta tion. Interruption with this process leads to cOltical dyspla sia and neuronal ectopias. These malformations may be identified on MR scans of the brain and are a frequent find ing in patients with intractable epilepsy. They should be sought in any child who dies unexpectedly as they may themselves be responsible for seizures 8 Neuron al migra tion disorders may be genetic in origin, but many are the result of early acquired damage occurring before migratio n is complete.
PORENCEPHALY
Porencephaly means a hole in the brain. The term is applied to any parenchymal defect, resultin g, for example, from old trauma, haemorrhage or infa rction. The term is also applied to asymmetric outpouching of the ventricle due to deep white matter hae morrhagic infarction. Heterotopic grey matter in the walls of a porencephalic cyst ind icates an origin before the end of neuronal migration.
SCHIZENCEPHALY/POLYMICROGYRIA
Schizencephaly and polymicrogyri a are sometimes grouped together due to their frequent coexistence. 9 Schizencephaly is a defect invo lving the whole thick ness of the brain wa ll with communication between the ventricle and the brain surface. The cleft is frequently lined
Figure 9.5 (a) Age nesis of the co rp us callosum. Coronal slice of fetal brain at 22 weeks' gestati on. A large bundle of fibres is running parallel to the midline on each side and displacing the ventricles laterally. There is no corpus cal losum; instead, the med ial borders of the hem ispheres curl in wa rds and a bu ndle of Probst (P) is seen in the medial wa ll of the lateral ventric le. (b) Section of the same bra in stained with haematoxylin and eosin. Note that there are small bilateral germinal matrix haemorrhages. P, bundle of Probst. (c) Hi gher power ima ge shows the latera l ventri cle and germ inal matrix. In the right wall of the ve ntricle, fibres of the bundle of Probst (P) have been transacted (haematoxylin and eosin) . GM, germinal matrix.
Acquired intra-uterine damage I
by ectopic neurones, and the adjacent cortex may be dysplastic. Schizencephalic clefts are often bilateral, and most commonly found in the region of the Sylvian fIs sure. Although originally considered to be a malformation due to a focal defect in the germinal matrix,1O schizen cephaly is now generally considered to be a destructive lesionY Rare familial cases are described. Mutations in the EMX2 gene were implicated but this association has not been confIrmed. 9 Polymicrogyria is a cOliical malformation resulting from damage prior to 28 weeks of gestation. iJ.12 Macro scopically, gyri are irregular and small , but the diagnosis can be confirmed only by histology. Two histological pat terns are described: unlayered and four-layered, but their distinction has no aetiological significance and they com monly coexist. A common and determining feature of polymicrogyria is abnormal fusion of adjacent gyri, lead ing to continuity of the superficial cortical surface while the neuronal laminae form festoons below. Large blood vessels are trapped in the seams of fused gyri. It is not a uniform entity; several specific patterns of polymicrogyria are described, both by histology and macroscopic distribu tion. Acquired polymicrogyria occurs randomly throughout the cortex or on the borders of destructive lesions, whereas symmetrical and bilateral patterns may correlate with spe cific genetic disorders. I] Polymicrogyria and schizen cephaly are both described after ischaemic events and infections such as CMY. Classification systems of cerebral malformations based on MRI studies are not always accurate, owing to the limi tations of MRI resolution. What may appear to be specific malformations on imaging may be both structurally and aetiologically heterogeneous.
171
It has long been accepted that white matter disease results from ischaemia of the developing white matter. However, in recent years intra-uterine infection and inflammation have been implicated in its pathogenesis. Epidemiological studies have demonstrated an association between placen tal and amniotic infection and maternal pyrexia and neonatal neurological abnormality. IS Histological evidence of inflammation in the placenta is associated with brain lesions and later cerebral palsy. J9 Three pattems of white matter damage are described. Com mon to all is gliosis and capillary proliferation. When seen alone this is described as telencephalic leucoencephalopathy.2o The most common form, pellventllcttiar leucomalacia, is recognized by multiple infarcts in the deep periventricular white matter that can be seen by the naked eye as tiny cysts, often rimmed by a yellow mineralized zone (Fig. 9.6). The infarcts tend to occur in the deep frontal white matter and around the trigone. Histologically, they appear as focal glial scars - centrally cystic and surrounded by macro phages and damaged axons, which are frequently calcifIed (see Fig. 9.3, p. 169). Recent lesions are accompanied by capillary prolif eration (see Fig. 9.4). The most dramatic manifestation of white matter dam age is multicystic leucoencephalopathy (Fig. 9.7), in which large areas of damaged white matter are replaced by cysts,
Late WHITE MAnER DAMAGE
This is the commonest cause of cerebral palsy. It typically occurs before the brain is mature, usually between 24 and 36 weeks of gestation, but this is not exclusive and some white matter damage almost invariably accompanies severe hypoxic-ischaemic injury at term. I Studies of infants undergoing cardiac surgelY have shown white matter dam age to develop after term. Magnetic resonance imaging studies at day 5 have confirmed development of white mat ter injury, 14 and it has been determined by neuropathological examination that it can develop up to 11 months postnatally. IS White matter damage causes diplegic cerebral palsy by involvement of the motor tracts as they pass through the deep white matter of the cerebral hemispheres. This form of cerebral palsy is typically seen in very-Iow-bilih-weight and premature infants. There is little doubt that grey matter damage, particularly the deep nuclei, accompanies most white matter damage. 16 • 17
Figure 9.6 Periventricular leucomalacia. Fixed brain slice with tiny areas of periventricular leucomalacia in the deep white matter, close to the ventricular wall; although some have yellow flecks of mineralization, others are more subtle and show as dusky patches adjacent to vessels.
172 I
Pathology of neurologica l abnormality in early life
Figure 9.7 Multicystic leucoencephalopathy. Almost the entire white matter is replaced by large cysts, some is traversed by glial strands. The overlying cortex is thin and gliotic. Microscopical examination showed extensive gliosis through grey and white matter.
particularly in the subcortical areas. They stretch the over lying cortex, which becomes gliotic with loss of neurones and there is often associated gliosis in deep grey nuclei. Infections such as herpes, varicella and listeria may pro duce similar white matter damage. 21 ,22
Figure 9.8 'Cardiac arrest encephalopathy'. Section of medulla showing well-defined symmetrical areas of infarction in the dorsal medulla, close to the floor of the fourth ventricle (haematoxylin and eosin).
of asphyxia in utero. 26 In the human infant, almost all occur after labour has begun, suggesting that these insults also represent acute on damage on a brain that has already been subjected to episodes of relatively mild hypoxia.
HYPOXlC-ISCHAEMIC INJURY AT TERM
There are three basic patterns of term hypoxic-ischaemic injury, which depend on the nature, intensity and duration of injury. Evidence largely depends on animal models, as well-documented data from human series is limited. Predominantly Brainstem Damage (Cardiac Arrest Encepha lopathy) This pattern results from an acute or total interruption of cerebral blood supply, probably for a perio d of 10-25 min utes, such as may occur in a cardiac arrest. Owing to the speed of interruption of circulation there is no time for adaptive redistribution of blood supply, and damage occurs in the most highly metabolically active brain areas - the nuclei of the brainstem and thalamus (Fig. 9.8). Babies so affected either die or show absent gag and corneal reflexes, and swallowing and breathing difficuities. 2J Basal Ganglia{Tha lamu s Severe acute, profound asphyxia causes damage predomi nantly to the deep grey nuclei, thalamus, internal capsule, hippocampus and central cortex 24 ,25 (Fig. 9.9). The mecha nisms include sudden severe episodes of asphyx ia : for example cord prolapse, uterine rupture or placental separa tion, acute or chronic hypoxia, or multiple severe episodes
Cortical and Watershed Da mage Chronic or repeated asphyxia or hypotension of 30 minutes to several hours' duration allows adaptive changes to 25 OCCUr. Blood is shunted form the forebrain to the vital and highly metabol ically active brainstem areas. Although this preserves the deep grey nuclei, the cerebral cortex becomes vulnerable, particularly in the distal, watershed territories. The most frequently involved is the parasagittal cortex. Gyral crests are relatively spared, as well as the more vulnerable depths of the sulci atrophy. The gyri become mushroom-like in appearance - 'ulegyria' (Figs 9.10 and 9.11). GERMINAL MATRIX HAEMORRHAGE
This is most common before 36 weeks of gestation while the germinal matrix persists in the ven tricular wall. The thin-wa lled capillaries of the germinal matrix may rupture after hypoxic damage and surges in cerebral blood pres sure. The most common time for haemorrhage to occur is within the first 48 hours of postnatal life. If the bleed remains confined within the matrix then the prognosis is good (Fig. 9.5a, p. 170). However, if blood ruptures into the ventricles it causes shedding of the ependymal lining, reac tive gliosis and ependymal proliferation. If this occurs in the narrow aqueduct or drainage channels of the fourth
-
-
.-
-
~
Birth·related injury I
Figure 9.10
1 73
Old hypoxic-ischaemic injury. There is loss of
cortex from the depths of parasagittal sulci, leading to focal ulegyria (arrowheads).
•
most commonly frontal, distinguishes this lesion from leucomalacia, which is much more commonly bilateral. Infants who suffer venous parenchymal infarction are at risk from later hemiplegia due to destruction of motor fibres as they pass through the deep white matterY
BIRTH-RELATED INJURY Figure 9.9 Acute profound hypoxic-ischaemic injury. Basal ganglia damage. (a) Fresh brain slice showing dusky discolouration of the deep grey nuclei, particularly the lentiform nuclei (arrows). The cortex appears as a wh ite ribbon and the ventricles are narrow, owing to oedema. (b) High-power image of the thalamus, many months after hypoxic-ischaemic injury, showing neuronal loss and dense gliosis. Many remainin g neurones are mineralized (haematoxylin and eosin).
ventricle it will cause obstruction of cerebrospinal fluid (CSF) flow, resulting in ventliculomegaly or hydrocephalus. The histological sequel is formation of ependymal rosettes, gliosis and Perls' positive material, which may persist for months or years. PARENCHYMAL VENOUS INFARCTI ON About 15 per cent of germinal matrix haemorrhage leads to wedge-shaped infarction of the adjacent deep white matter, fanning out from the haemorrhage. The draining veins are congested and macrophages containing red cells pack the white matter. Resorption of the damaged tissue leads to porencephalic cyst formation in the deep white matter or unilateral dilatation of the ventricle. The unilateral damage,
Brain injury acquired at birth is usually due either to trauma or to asphyxia. Today, birth trauma is relatively uncommon owing to improved obstetric care and high rates of Cae sarean section. Despite a fivefold increase in Caesarean sec tions over the last 30 years, the rates of cerebral palsy have remained constant in industrialized countries? with many of those affected being low-birth-weight or premature babies who survive with white matter damage.
Birth Trauma One of the largest pathological studies of bilth il1iury is that of Towbin, published in 1970, in which he studied 600 still births and neonatal deaths. 28 In this study, trauma accounted for some 1-2 per cent of newborn deaths, but many more infants suffered non-lethal injuries. Birth trauma is associated with both precipitous and prolonged labour, with large infant size, fetal malpresentation and difficult instrumental extraction. 28
Scalp and Skull Scalp and, in particular, subgaleal haemolThage may follow instrumental delivery, particularly vento use extraction.
174 I
Pathology of neurological abnormality in early life
Figure 9.12 Intradural haemo rr hage in 2-day-old infant fo llowing birth asphyxia. There is extensive fresh haemorrhage into the falx adjacent to the sagittal sinus.
Figure 9.11 Old hypoxic-ischaemic injury. Cerebellar cortex showing old hypoxic-ischaemic injury with large areas of neuronal loss and gliosis. (a) The damage is most severe in the depths of folia. (b) At higher powe r only one Purkinje cell remains. The granule cell layer is replaced with glial cell s.
Subgaleal haemorrhages may be very large and cause cra nial compression as well as significant blood loss from the circulation. 29 Skull fracture is seen occasionally after normal delivery but, more usu a lly, this follows instrumental deliv ery.30 Extracranial birth injuries are considered in Chapter 8.
Intracranial Haemorrhage The commonest form of birth injury is intracrani al haemor rhage, of which subdural haemorrhage is the most
frequent. Among 41 consecutive cranial bilih injuries in term infants, Pollina et al31 found that 73 per cent had subdural haem OiThage, 20 per cent had subarachnoid haemorrhage and 20 per cent had parenchymal haemorrhage. Overall, 24 per cent had cephalhaematomas, 5 per cent had intraventlic ular haemorrhage and 5 per cent had skull fracture. The fre qu ency of birth injury in the institution was not stated. Method of delivelY is important. Towner et al 32 found an increased incidence of intracranial haemorrhage after instru men tal delivery with ventouse or forceps and emergency Caesarean section. Incidence was lower after Caesarean sec tion before labour. Pollina et al found a higher incidence of intracranial injury after forceps and vacuum deliveries, whereas Chamnanvanakij et al 33 found subdural haematoma on computerized tomography (CT) in 26 symptomatic term infants, half of which followed instrumental delivery and half after s ponta neous delivery. In a UK-wide study, O'Mahoney et al 34 reported a lower incidence of cranial birth trauma. Amongst 873 intrapartum deat hs of mature babies, they found only 37 cases out of 181 in which cranial trauma a ppeared likely from the history. Instrumental delivery was clearly associated with trauma; multiple pulls and use of multiple instruments were particular risks.
INTRADURAL HAEMORRHAGE When the dura is examined, intradura l haemorrhage is fre quently observed - particularly in the falx and tentorium. It is seen in the n eonate in many circumstances, often when there is evidence of hypoxic-ischaemic injury. It is also described in spontan eous abOliions and intra-uterine deaths when the fetus is likely to have experienced hypox ia. 35 It is particularly obvious in the dural folds and may be confined to the dura itself or leak on to th e subdural surface, although this is not common (Fig. 9.12).
Birth-related injury I
175
SUBDURAL HAEMORRHAGE
Subdural haemorrhage is the most common form of birth related intracranial haemorrhage. Recent studies using MR scans in asymptomatic neonates have shown a surprisingly high incidence of subdural haemorrhage. Looney et al 36 ;o und small subdural haemorrhages in 26 per cent of vagi na l deliveries and Whitby et al 37 in 9 per cent, the differ ence probably due to use of scanners of different resolution . Infants with congenital heart disease had an incidence of asymptomatic intracranial haem orrhage of 62 per cent after normal delivery.38
Method of Delivery Most subdural haemorrhages follow normal delivery, but are proportionally more frequent after forceps and ve ntouse deliveries,37 although, in the above studies, not after Cae sarean section.
Location The source of bleeding in bilth-related subdural haemor rhage remains uncertain. It is striking that the great maj ority of birth-related subdural haemorrhage is found in the poste rior fossa or around the occipital poles, close to the tento rium. 36 .37 The preferential location at the back of the brain is likely to relate to the origin of bleeding in the dural folds. The dural sinuses and their tributaries have traditionally been considered the origin for bilth-related subdural haem orrhage,J9 even in the absence offrank tentori al tearing. The falx has particularly prominent venous sinuses within it 40 and bleeding here is associated with asphyxia. 41 lntratentor ial bleeding in asphyxiated infants has been demonstrated by CT scan and autopsy correlative studies. 42 Another consideration in the location of these bleeds is gravitational redistribution when babies are nursed, and indeed scanned, on their backs.43 Blood also tracks into the subdural space around the spinal cord and is frequently identified at autopsy as a crescentic collection over the pos terior cord at lower spinal levels. Subdural haemorrhage is occasionally found overlying an area of cerebral infarction in the term infant, usually in the middle cerebral artery territory.44 The pathogenesis is unknown, but Steinbok et al 44 considered the most likely sequence to be cerebral infarction, with secondary subdural haemorrhage resulting from rupture outwards of a haemor rhagic infarct. Animal studies have suggested an alternative explanation: that subdural blood damages the underlying brain. 45 Other rare causes of subdural haemorrhage include vas cular malform ations, meningitis and after neurosurgical intervention. Very rarely, subdural haemorrhage compli cates metabolic diseases associated with brain atrophy, such as Menkes disease and glutaric aciduria in older infants. 46
Figure 9.13 Chronic subdural haemorrhage. (a) Very vascular membrane (M) adherent to deep surface of dura (D) . Note the large numbers of capillaries at the junction wi th the dura and at the free edge (CD34 immunocytochemistry). (b) Subdural haemorrhage 4 months after head injury. There are collections of fresh red cells and macrophages with pigmented cytoplasm. Peris' positive material is seen within the macrophage cytoplasm and free in the tissue, deposited on fibrous strands. These features indicate bleeding of several different ages (Perls' stain).
Natural History of Infant Subdural Haemorrhage The natural histOIY of subdural haemorrhage is that it resolves by forming a granulating memb ra ne (see Fig. 9.13). This membrane contains wide capillaries that may rebleed. The characteristic histology of a healing subdural membrane is a layer of fibroblasts containing fresh and old red cells, pigmented macrophages and haemosiderin consistent with repeated episodes of rebleeding (the chronic active subdural haemorrhage) (Fig. 9.13). Birth-related subdural bleeds were seen to resolve on a single MR scan at 4 weeks in 9 cases. J7 However the membranes may persist, Rogers et al 47 found evidence of chronic subdural membranes in up to 31 per cent of sudden infant death syndrome (SmS) case
176 I
Pathology of neurological abnormality In early life
•
subjects not suspected of having chronic subdural haemor rhage in life. Some infant subdural haemorrhages develop into chronic fluid collections th at are visible on brain scans. The mechanism is unknown but may be related to immaturity of the arachnoid villi or rupture of the arachnoid mem brane, allowing leakage of CSF into the subdural space,43 osmotic accumulation of fluid or repeated small bleeds. 48
.
EXTRADURAL HAEMORRHAGE
Extradural haemorrhage can be seen without skull frac ture. It may result form 'springing' of a suture, causing vascular tearing without fracture. Away from sutures, it probably results from impact related deformat ion of the thin infant skull an d tearing of the dura from the skull without frac ture [ping pong fracture). Extradural and periosteal bl eed ing is sometimes seen on the edges of bones th at are adjacent to widely sepa rated sutures where there has been acute brain swelling and raised intracranial pressure. SUBARACHNOID HAEMORRHAGE
Subarachnoid haemorrhage over the brain surface is very common following hypoxic-ischaemic injury, particularly in preterm infants. More extensive subarachnoid haemor rhage is seen at the base of the brain after intraventricular haemorrhage, when blood tracks down from the ven tri cu lar system and leaks out through the exit foramina of the fourth ventricle. Considerable amounts of blood may fill the spaces beneath the brainstem in the cisterna magna.
I
.'
Figure 9.14 Axonal injury. Cervical nerve root showing many axonal swellings stained with beta amyloid precursor protein. This is a very sensitive method of identifying axonal damage in this site (beta amyloid precursor protein immunocytochemistry).
As noted above, the majority of birth-related subdural haemorrhage is found, at least in part, in the posterior fossa. Another rare birth-related injury to the posterior fossa is occipital osteodiastasis, characterized by displacement of the squamous part of the occipital bone during delivery. It tends to occur in breech, forceps or prolonged and difficult vagi nal deliveries.49 The compliant skull of the premature infant allows displacement of the occipital bon es forwards and upwards when under pressure during breech or instrumental delivery. This can lea d to bleeding by compression of the venous sinuses causing venous congestion and leakage or by tearing of the superficial vessels of the cerebellum. Further more, direct pressure from the displaced bones wiH compress and damage the cerebellar tissue. 50
root damage in one-third of these. He ascribed the damage in most cases to vertebral artery injury. Spinal nerve root tears and haemorrhages and spinal cord injury were seen in 10 per cent of infants, due to stretching or tearing during delivery. Cument incidence is unknown as the cord is not frequently removed and examined. Infants with very severe brain swelling and tonsillar herniation through the foram en magnum will have considerable dis ruption of the cervica l cord and nerve roots, which may, itself, cause local nerve root damage. Immunohistochemistry with antibodies to 0APP is extremely helpful in demonstrating nelve root injury (Fig. 9.14) but is not sp ec ific for the cause of injury. Distinction of hypoxic-ischaemic damage from tru e axonal tearing may be difficult, particularly as the two almost invariably coexist. Towbin 28 noted epidural haemorrhage as the most fre quent injury in his series of neonatal post-moliems. However, small amounts of epidural bleeding are a common finding in infants who have died from non-traumatic causes, suggest ing that in some cases they may represent post-mOIiem artefact - perhaps related to method of cord remova1. 52
Spinal Cord and Nerve Roots
STROKE IN THE DEVELOPING BRAIN
Early studies have noted spinal cord and nerve root injury after birth.28 ,5 1 Yates 51 described damage to the cervical spine in almost one-half of neonatal autopsies, with nerve
Unilateral cerebral artery infarction is common, occurring in 1 in 4000 term infants. The middle cerebral artery is most commonly involved. The damage is due to failure of
Posterior Fossa Damage
Infections I
blood supply through a m ajo r v essel, which may be due to its blockage by a b loo d clot or tissue fragment, for exam ple fronl the placenta, or by externa l compression. Ex tern al co mp ression is rare. The cause of vascular obstruction is not often demonstrated at autopsy. Stroke may occur in the an ten atal or perinatal periods period or at the time of birth, and the timing of the damage may be reflected in the clinical presentation. Ba bies injured in the perin atal perio d tend to present early with neo natal seizures, whereas infants damaged in utero present later and develop hemiplegia and intractable epilepsy after sev eral months of pos tna tal life. 53 However, many strokes are clinica lly sile nt at the time w hen they occur. The aetiology is obscure and prob ably multifactoriaP4.55 Overall, 30 per cent of neonatal stroke is due to v eno us sinus thrombosis. 56 Factors associated with neonatal stroke include: blood clo tting disord ers in fec tions, placental dys fun ction, growth retardation, pre-eclampsia, tw in pregnancy, peripartum asphyxia and maternal thrombophilia with thrombotic lesions in the placen ta. 54.57 ,58 Hypoxic-ischaemic injlllY is the most common association.59
Cerebral Sinovenous Thrombosis Venous thrombosis in sudden deat h is probably under reported and not always sought. Ante-mortem clo ts are distinguis hed by their dry or lamina ted appearance, an d are often adherent at some point. There is a high incidence of sinovenous thromb osis in the first month of life. Dehy dration, sepsis, polycythaemia a nd protein C deficiency have been implicated as causes but are no t always pres ent. 5O ,61 In to ta l, 84 per cent of neon ates with sinove no us thromb osis have perinatal complications, of w hi ch hypoxic encephalopathy is the most common. 56 ,62
METABOLIC DISORDERS Hypoglycaernia Cortical an d subcortical damage in the parieto-o ccipital region has been described by MRl studies of infants who suf fered hypoglycaemia in the neonatal period ,6),64 but studies of the neuropa tho logy of neonatal hypo gJycaemi a are few. 65 There is little informa tio n re garding the effects of maternal diabetes on fetal brain development, although white matter damage is described and considered to result from delayed brain matura tion 66
Mitochondrial Diseases Impairment of mitochondrial fun ction can lead to sudd en death. Babies with abn ormalities of mitochon drial DNA
177
usually h ave raised lactate in their blood and CSF, which may be exacerbated on exercise or when stressed by inter curre nt infections. Clin ical presentation includes poor feeding, vomiting and exercis e-induced lactic acidosis, res piratory deficiency and lethal ap noea. 67 Autopsy examination of skeletal muscle may s how typ ical ragged red fibres with Gomori's tr ichrome stain, abno r mally increased succinate dehydrogenase ac tivity and cytochrome oxidase deficiency in many fibres. However, th ese changes a re less frequ ent in infant than in ad ult cases and diagnosis depends on mitoch ondrial DNA an alys is. Brain examination may show a number of abnorma lities. In Leigh's syndrome there are typ icall y symmetrical lesions of tissue rarefaction, gliosis and capiUary proliferation in the basal ganglia and brainstem. In other cases, for example pyruvate dehydrogenase deficiency, there may be migration disturbances, congen ital cerebral at rophy an d ventricular dilatation. 67
Lipid Storage Diseases These diseases are rarely a cause of sudden death. Presenta tion is usually between 6 and 24 months, and includes vo m iting or coma triggered by fasting. The most commo n lethal fo rm is long-chain methyl CoA deficiency, bu t medium- a nd mul tiple-ch ain CoA deficiencies also occur. Skeletal muscle and li ver sho uld be sampled and snap-frozen for histochem ical study. In froze n sections of skel eta l muscle, abnormally stored fat forms multiple large globules within muscle fibres; this is in contrast w ith normal neonatal muscle, which usu ally contains very li ttle fa t. Fat may also be excessive in the liver and other organs.
INFECTIONS Bacterial infections of the fe ta l brain are extremely rare. The most common organisms to damage the fetal brain by direct invasion are toxoplasmosis, rubella, CMV, herpes simpl ex and li sterios is. Human imm uno defic ien cy v irus (HN) an d herpes simplex v irus (HSV) may be tra nsmi tted to the infa nt during delively. Viral infections may result in calcification and cystic damage . Listeria causes granulo mata aro und blood vessels in t he meninges and ch oroid plexus. CMV may be asso ciated with polymicrogyria (Fig. 9.1 5) , possibly due to its a ffini ty for the end oth elial cell s, causin g interferen ce w ith vascular supply to the cortex in its later stages of development. Bilateral, multifocal cysti c lesions an d calcifica tion , haemorrhage and cerebral oedema have been described on MR and CT sca ns of babies with v iral encephalitis.22 ,27 Extracranial manifestations, for example limb hypoplasia, cataracts, micro-ophthalmia, choreoretinitis, organ calcifi cation o r skin scars supp ort the diagnosis of a congeni tal viral infection.
178 I
Pathology of neurological abnormality in early life
Figure 9.15 Polymicrogyria. Fixed brain of fetus with cytomegalovirus. There are areas of polymicrogyria in the middle regions bilaterally, an appearance which is simi lar to perisylvian polymicrogyria of genetic or ischaem ic origin. The distribution can be misleading.
REFERENCES Rutherford MA, Pennock JM, Cowan FM et at. Does the brain regenerate after perinatal infarction? Europ J Paediotr Neural 1997; 1:13-17. 2 Nelson KB. Can we prevent cerebral palsy? N El1gl .J Med 2003; 349:1 765- 9. 3 Keeling JW. Feta l lind Neonatal Pathology, 2nd edn. London: Springer-Verlag, 1993. 4 Squier M, Chamberlain P, Zaiwalla Z et al. Fi ve cases of bra in injury follo win g amniocentesis in mid-term pregnancy. Deu Med Child Neuro/2000; 42:554-60. 5 Geddes JF. What's new in the dia gnosis of head inj ury? J Ciill Pothol 1997; 50: 271 -4. 6 Hortobagyi T, Wise S, Hunt N, Cary N et al. Traumatic axonal damage in the brain can be detected using beta-APP immun ohistoc hemistry within 35 min after hea d injUly to hum an adu lts. Nellropati1ol Appl Neurobi ol2007; 33:226-37. 7 Oehmichen M, Meissner C, Schmidt Vet al. Axo nal injury - a diagnostic too l in forensic neuropathology? A review. Forensic Sci Inr 1998; 95:67-83. 8 Hannan AJ, Servotte S, Katsnelson A et al. Characterization of nodul ar neuronal heterotopia in ch ildren. Brai l7 1999; 122:21.9- 33. 9 Barkovich AJ, Kuzniecky RI, Ja ckso n GD et al. Classification system for malformations of cOl1ical develop ment: update 2001. Neurology 2001; 57: 2168-70. 10 Yakovlev PI, Wadsworth RC. Schizencep halies: a study of the congenita l clefts in the cerebral mantle. l. Clefts with fused lips. J Neuropathol Erp Neurol 1946; 5: 116-30. 11 Barkovich AJ, Kjos BO. Schizencephaly: correlation of clinical findings with MR characteristics. AJNR Am J Neuroradiol '1992; 13:85-94. 12 Bar1h PG. Migrational disord ers of the brain. Curl' Opin Neuro! Neurosurg 1992; 5:339-4 3.
13 Ja nsen A, Andermann E. Genet ics of the polymicrogy ria syndro mes. 4. J Med Genet 2005; 42 :369-78 . 14 McQuillen PS, Barko vich AJ, Hamrick SE et a l. Temporal and anatomic risk profile of brain injury with neo natal repair of congenital heart defects. Stroke 2007; 38 (Suppl.):736-41. 15 Kinney HC , Panigrahy A, New burger JW et al. Hypoxic-ischemic brain il'\iury in infants with congenital heart disease dying after cardiac surgery. Acta Neuropathol (Berl) 2005; 110:563-78. 16 Krageloh-lVlann I, Toft P, Lunding J, Andresen Jet al. Brain lesions in preterms: origin, consequenc es and compensation. Acta Paediatr 1999; 88 :897 - 908. 17 Krageloh-Mann I, Helber A, Mad er I et al. Bilateral lesions of thalamus and basal gangl ia: origin and outcome. Del! Med Child Neural 2002; 44:477-84. 18 Peebles DM, Wyatt JS. Sy nergy between antenatal exposure to infection and intrapal1um events in ca usation of perinatal brain injury at term. BJOG 2002; 109:737-9. 19 Redline RW. Severe fetal placental vascular lesions in term infants with neurologic imp airm ent. I. Am J Obstet Gynecol 2005; 192:45 2-7. 20 Gilles FH, Leviton A, Dooling ED . The Delle/oping Human Brain: Growth and Epidemiolog ic Neuropathology. Boston: John Wright, 198 3. 21 de Vries LS, Gunardi H, Barth PG et al. The spectrum of cranial ultra so und and magnetic resonance imaging abnormalities in congenita l cytomegalovinls infection. Neuropediarrics 2004; 35:113-19. 22 Kurtz J, Anslo w P. Infantil e herpes simplex encephalitis: diagnostic featu res and differentiation from non- accidental injury. J !njcct2003 ; 46:12-16. 23 Pasternak JF. Hypox ic-ischemic brain dama ge in the lerm in fa nt. Lesso ns from th e laboratory. Pediatr Ciin North Am 1993; 40:1061-72 . 24 Sie LT, van der Knaap MS, Oosting J et al. MR pattern s of hypoxic-i schem ic brain damage after prenatal , perinatal or postnatal asphyxia. Neuropediatrics 2000; 31: 128-36. 25 Naeye RL, Lin HM. Determination of the timing of fetal brain damage from hypoxemia-ischemia. Am J Obstet Gyneeol 2001 ; 184:217-24. 26 Myers RE. Two patterns of perinatal brain damag e and their conditio ns of occurrence. Am J Obstet GYllecol 1972; 112:246-7 6. 27 Rutherford M. Magnetic resonance imaging of injury to the imm atu re brain. In Squier W (eeJ.) Acquired Damage to the Developing Brain: Timing and Causation. London: Arnold , 2002, pp. 166-92. 28 Towbin A. Central nervous system damage in the human fetus and newborn infant. fvlechanical and hypoxic injury incurred in the fetal-neonatal period. Am J Dis Child 1970; 119 :529-42. 29 Amar AP, Aryan HE, Meltzer HS, Levy ML. Neonatal subgaleal hematoma causing brain compress ion : repol1 of two cases and review of the literature. Neurosurgery 2003; 52:1470-4. 30 Dupuis 0 , Silveira R, Dup ont C et at. Comparison of 'instrument-associated' and 'spo nta neous' obstetric depressed sku ll frac tures in a cohort of 68 neo nates. Am J Obstet Gyneeol 2005 ; 192: 165-70. 31 Poll ina J, Dias MS, Li V, Kachurek D, Arbesman M. Cranial
birt h injuri es in term newborn infants. Pediat/' Neurosu rg
2001; 35:1 13-19.
32 Towner D, Castro MA, Eby-Wilkens E, Gilbert VVM. Effect of mode of delivery in nulliparous women on neonatal intracranial injury. N Engl J Med 1999; 341:1709-14. 33 Chamnanvanakij S, Rollins N, Perlman JM. Subdural hematoma in term infants. Pediotr Neural 2002; 26:301-4.
References I
34 O'Mahoney F, Settatree R, Platt C, Johnson R. Review of singleton fetal and neonatal deaths associated with cranial trauma and cephalic delivery during a national intrapartum related confidential enquiry. BJOG 2005; 112:619-26. 35 Geddes JF, Tasker RC, Hackshaw AK et a\. Dural haemorrhage in non-traumatic infant deaths: does it explain the bleeding in 'shaken baby syndrome'? Neuropathol Appl Neurobiol 2003; 29:14-22. 36 Looney CB, Smith JK , Merck LH et a\. Intracranial hemorrhage in asymptomatic neonates: prevalence on MR images and relationship to obstetric and neonatal risk factors. Radiology 2007; 242:535-41. 37 Whitby EH, Griffiths PD, Rutter S et a\. Frequency and natural history of subdural haemorrhages in babies and relation to obstetric factors. Lancet 2004; 363 :846- 51. 38 Tavani F, Zimmerman RA, Clancy RR et al. Incidental intracranial hemorrhage after uncomplicated biI1h: MRI before and after neonatal heart surgety. Neuroradiology 2003; 45:253-8. 39 Volpe JJ. Intracranial haemorrhage. In Neurology of the Newborn, 3rd edn. Philadelphia, PA: WE Saunders, 1995, pp. 377-8. 40 Tubbs RS, Louis RG , Acakpo-Satchivi L, Salter EG. Anatomy of the falcine venous plexus. J Neurosurg 2007; 107:155-7. 41 Friede RL. Hemorrhages in asphyxiated premature infants. In Friede R (ed.) Developmental Neuropathology. Gottingen: Springer-Verlag, 1989, pp. 44-58. 42 Kibayashi K, Shojo H, Sumida T. Dural hemorrhage of the tentorium on postmortem cranial computed tomographic scans in children . Forensic Sci lilt 2005; 154:206-9. 43 Vinchon M, Noizet 0 , Defoort-Dhellemmes S et al. Infantile subdural hematomas due to traffic accidents. Pediatr Neurosurg 2002; 37:245-53. 44 Steinbok P, Haw CS, Cochrane DD, Kestle JR. Acute subdural hematoma associated with cerebral infarction in the full-term neonate. Pediatr Neurosurg 1995; 23 :206-15. 45 Durham SR, Duhaime AC. Basic science; maturation dependent response of the immature brain to experimental subdural hematoma 1. J Neurotrauma 2007; 24:5-14. 46 Kemp AM. Investigating subdural haemorrhage in infants. Arch Dis Child 2002; 86:98-102. 47 Rogers CB, Itabashi HH, Tomiyasu U, Heuser ET. Subdural neomembranes and sudden infant death syndrome. J Forensic Sci 1998; 43:375-6. 48 Markwalder TM. Chronic subdural hematomas: a review. J Neurosurg 1981; 54:637-45. 49 Currarino G. Occipital osteodiastasis: presentation of four cases and review of the literature. Pediatr Radio12000; 30:823-9 . 50 Volpe JJ. Intracerebellar haemorrhage. In Neurology of the Newborn, 3rd edn. Philadelphia, PA: WE Saunders, 1995, pp. 384-9.
51 52
53
54
55 56 57
58 59
60 61
62 63
64
65
66
67
179
Yates PO. Birth trauma to the vertebral at1eries. Arch Dis Child 1959; 34:436-41. Rutty GN, Squier WM, Padfield CJ. Epidural haemorrhage of the cervical spinal cord: a post-mortem artefact? Neuropathol Appl Neurobiol 2005; 31 :247-57. Squier W, Salisbury H, Sisodiya S. Stroke in the developing brain and intractable epilepsy: effect of timing on hippocampal sclerosis. Dev Med Child Neurol 2003; 45:580-5. Arias F, Romero R, Joist H, Kraus FT. Thrombophilia: a mechanism of disease in women with adverse pregnancy outcome and thrombotic lesions in the placenta. J Maternal Fetal Med 1998; 7:277-86. Mercuri E. Timing and aetiology of neonatal cerebral infarction (Letter). Paediatrics 2000; 106 :615-16. DeVeber G, Andrew M, Adams C et al. Cerebral sinovenous thrombosis in children. N Engl J Med 2001; 345:417-23. Mercuri E, Cowan F, Gupte G et a\. Prothrombotic disorders and abnormal neurodevelopmental outcome in infants with neonatal cerebral infarction. Pediatrics 2001; 107: 1400-4. Marrett S, Lardennois C, Mercier A et a\. Fetal and neonatal cerebral infarcts. Bioi Neonate 2001; 79:236-40. Evans D, Levene M. MRI Imaging of injury to the immature brain. In Squier W (ed.) Acquired Damage to the Developing Brain; Timing and Causation. London: Arnold, 2002. Shevell MI, Silver K, O'Gorman AM et a\. Neonatal dural sinus thrombosis. Pediatr Neurol 1989; 5:161-5. Rivkin MJ, Anderson ML, Kaye EM. Neonatal idiopathic cerebral venous thrombosis: an unrecognized cause of transient seizures or lethargy. Ann Neurol 1992; 32:51-6. Fitzgerald KC, Williams LS, Garg BP et a\. Cerebral sinovenous thrombosis in the neonate. Arch Neurol 2006; 63 :405-9. Traill Z, Squier M, Anslow P. Brain imaging in neonatal hypoglycaemia. Arch Dis Child Fetal Neonatal Ed 1998; 79:F145-7. Murakami Y, Yamashita Y, Matsuishi T et at. Cranial MRI of neurologically impaired children suffering from neonatal hypoglycaemia. Pediatr Radiol 1999; 29:23-7. Rossiter JP, Anderson LL, Yang F, Cole GM. Caspase-3 activation and caspase-like proteolytic activity in human perinatal hypoxic-ischemic brain injury. Acta Neuropathol (Berl) 2002; 103:66-73 . Lynch N, Roland EH, Poskitt K et a\. Gestational diabetes mellitus: risk of newborn brain injury. Dev Med Child Neurol 2007; 49(Supp\. 108):7. Brown GK, Squier MV. Neuropathology and pathogenesis of mitochondrial diseases. J Inherit Metab Dis 1996; 19:553-72.
I
CHAPTER 10
I
FETAL AND PERINATAL DEATH
Jean W Keeling
Introduction Defi n itions The law Background information Concealed pregnancy Unattended delivery Was the baby born alive? Is the baby of sufficient maturity to survive?
180 180 181 182 182 182 183 187
INTRODUCTION
There are several circumstances in relation to fetal deaths, still births and deaths in early life when an investigation is likely to faU within the rem it of a coroner, a procurator fis calor a medical examiner. These include concealment of pregnancy; unattended delivery, with or without abandon ment of the newborn infant; and suspected infanticide. The size of the problem is difficult to assess, although the num ber of infanticides (neonaticides) has fallen in some jurisdic tions, as has the number of maternal fatalities attributable to criminal abortion, probably for similar reasons.i Neverthe less, abandoned fetuses and infants are still discovered in a variety of circumstances. Fetal death following maternal injury, sudden fetal death and those fetal deaths in which the attendant midwife or obstetrician is unable to complete a still birth certificate also require formal investigation in some jurisdictions. 2 A further circumstance in which it has become increas ingly likely that the pathologist's report will acq uire medicolegal significance is following intrapartum still bilth or early neonatal death (perinatal deaths) after hospital delivery. The perinatal mortality rate has fallen to lo w levels in many countries, particularly Western Europe, North America, and Australia and New Zealand; it was less than 8 in 1000 total births in the UK in 2002 with an intra partum death rate of 0.62/1000 total births 3 This low risk of labour-related mortality has, in the public's perception,
Is there evidence of prolonged or difficult labour? Are there any significant injuries? Fetal death following maternal injury Is there a natural cause for death? Can I give a cause of death? Should the intrapartum still birth be a medicolegal autopsy? References
187 188 190 193 193 194 195
been translated as 'no risk'. Thus, many consider that there must be fault attached to any death occurring during or shortly after labo ur.
DEFINITIONS Definitions relating to death in early life differ in different countries and in different jurisdictions. This section lists some important and relevant definitions and draws attention to different usages of the same term. These are related to both offences that may have been committed and charges arising therefrom. It is important to be aware of them when comparing statistics or papers from different countries. • Fetal death is defined as death before 22 completed weeks of gestation (World Health Organiza tion [WHO]) or before 24 completed weeks of gestation (UK), when the conceptus exhibits no sign of life after complete separation from the mother. Fonnal registration of feta l death is not required, although many maternity units issue a certificate of fetal death, both as an acknowledgement of existence4 and to facilitate funeral arra ngeme nts in accordance with parental wishes. Concealment of the death of a fetus is not an offence (in England). • Still birth is death after 22 completed weeks of gestation (WHO) or after 24 completed weeks of gestation (UK) of a conceptus who exhibits no signs of
....
-
---------------~
The law I
•
•
•
•
•
•
•
•
•
life after complete separation from the mother. A certificate of still birth must be delivered to the appropriate registration authority. It can be completed by either a midwife or a medical practitioner who was present at the birth, or who examined the body after delivery, and is able to certify that the child was not born alive. If there is any doubt, the procurator fiscal, coroner or medical examiner should be informed. Neonatal death is death following live birth, i.e. exhibition of signs of life after complete separation from the mother, irrespective of the length of gestation, in the first 28 days (UK) or 30 days (USA) of life. Both birth and death certificates are required. Signs of life, from a medical standpoint, are heartbeat, spontaneous respiration and voluntary movement. Only one of these is necessary to fulfil the definition of live birth. Complete separation from the mother is defined as complete delivery of all parts of the baby (fetus) from the body of the mother. Severance of the umbilical cord or delivery of the placenta has no part in definitions relating to completeness of separation from the mother or to live or still birth. The concept of what constitutes a separate existence is both important and difficult; cardiac function and the ability to breathe have been variously considered the gold standard. 5 However, it is not only the initiation of respiration, but the ability to sustain respiration, which is essential for survival,6 that matters. Abortion is the termination of pregnancy before the legal limit of viability. Unqualified, the term encompasses natural pregnancy loss as well as del iberate termination of pregnancy. Induced abortion is the deliberate termination of a pregnancy before the legal limit of viability whether by medical or surgical means. Child destruction is the destruction of the life of a child who is capable of being born alive, i.e. after the legal limit of viability. Infanticide (England and Wales) occurs when a mother (acting alone) causes the of death of a child who is less than 12 completed months of age by a wilful act or omission. Infanticide (USA and elsewhere) is used more generally to mean causing the death of a child in the first week of life. Neonaticide has no legal definition but is used variously to mean the unnatural death of a liveborn child within 24 hours of birth 7 or within the neonatal period.
THE LAW The law, in relation to the unnatural death of the fetus and newly born infant, is considered in detail by Bowen,8
181
Kellett 5 and Mason. 6 They review the effects of sequential changes in legislation in this area in terms of the definition of particular offences and restrictions on medical practice. They set out differences in legislation in Scotland com pared with England and Wales. The law will be considered only briefly here.
The Offence of Procurement of Abortion This is defined in the Offences Against the Person Act 1861, Sections 58 and 59. This Act is still extant. The Abortion Act 1967 serves mainly to define the term 'unlawfully' in the 1861 Act.
Child Destruction Child destruction is an offence under the Infant Life (Preservation) Act J 929, except when performed in pur suance of 'preserving the mother's life'. It applies to the killing of a viable fetus both before and during labour, thus offering protection to doctors faced with performing cran iotomy during labour for cephalo-pelvic disproportion. This procedure is rarely performed now in developed coun tries because of the infrequency of contracted pelvises and the availability of prenatal diagnosis and Caesarean sec tion. The Infant Life (Preservation) Act of ]929 does not apply in Scotland.
Concealment of Birth Concealment of birth, originally a criminal offence, is leg islated in the Offences Against the Person Act 1861, Section 60. It encompasses both live and still births. It is not neces sary to show how death occurred, or even to produce the infant's body, provided that the mother can be shown to have delivered a baby in the recent past. It thus dispenses with the thorny problem of demonstration of a separate existence and is a less serious charge than infanticide. In Scotland, the Concealment of Birth (Scotland) Act 1809 additionally requires that the mother conceals both pregnancy and delivery, and that the child was viable. As in England and Wales, it is not necessary to produce a body for successful prosecution.
Infanticide This offence was introduced in England and Wales by the Infanticide Act 1922 as an alternative to a charge of mur der or manslaughter. It related solely to death of an infant caused by the mother acting alone, because of disturbance of her mind following birth. It applied originally only to the death of a newly born child. This was amended by the
182 I
Fetal and perinatal death
Infanticide Act 1938 which specifIes 'any wilful act or omission , ca uses the death of her child being less than 12 months of age' and 'the balance of her mind is disturbed ' 'not havin g fully re covered from the effects of giving birth' or 'lactation'. This is, in effect, a charge of manslaughte r. In Scotland, where there is no similar act, a mother who kills her infant under similar circumstances is charged with cul pable homicide.
BACKGROUND INFORMATION Before commencing post-monem examination, It IS most important to be fully informed about the circumstances of the case. A detailed police report, including how the baby was found and by whom, together with details of the JOCLIS, is essential for bilihs and deaths occurring outside hospital. In some circumstances photographs of the scene may provide information to suggest modification of necropsy techniques. However, important information may be missed if a patholo gist does not attend the locus, and so such a vis it should be encouraged. When the mother is known, as much informa tion as possible about the pregnancy should be obtained from her genera l practitioner and hospital records.
CONCEALED PREGNANCY There are a number of important areas that must be co nsid ered in relation to concealed pregnancy, unattended delivery and abandonment which are peliinent to charges that may be brought. It is important, then, that these are addressed as precisely as possib le. These include fetal maturity, time of death in relation to delivery, whether the baby was livebom, cause of death and evidence of trauma. Women who conceal their pregnancy have many fea tures in common. 9 In the UK and USA, they are frequently, but not invariab ly, experiencing a first pregnancy. Th e mother may never be identifi ed, although some seek medical assistance because of retained placenta, postpartum haemor rhage or infection. They are usually young, of low educa tional attainment and g ive birth unaided, or at best with inexpert assistance. The neonatal monality follo wing these pregnancies is more than twice that of hospital deliveries;lO the still birth rate is likely to be much higher. Labour may be protracted and there is usually denial of live birth . Efforts are made to concea l the body (Figs 10.1 and 10.2). It is, perhaps, not surprising that concealment of preg nancy and delivery continue in conservative societies where sex ed uca tion is su ppressed . II It was not unco mmon in UK rural communities within the last two generations 12 and is still occasionally encountered in Europ e following the chance finding of fetal remains. 13 - IS Serial concealment of pregnancy and infanticide is not uncommon in Japan, despite the ready availability of abor tion . The motive seems to be largely economic. 16 ,17 The
Figure 10.1
Conce aled pregnancy and delive ry. Mature baby
wrapped in bedding and discovered in a cupboard. (a) The skin is covered in blood and mecon ium. (b) The umbilica l cord is t orn close to its fetal insertion.
practice is not confined to single women alth ough perpe trators are frequently single. 18 Usuall y, husbands of mar ried perpetrators deny knowledge of the pregnancies. 19
UNATIENDED DELIVERY Legislation in the UK restricts thos e persons who may attend a woman in childbirth to practising midwives and medical practitioners. 2o These indiv iduals must be regis tered with their appropriate professional body and be able to demonstrate both that they have underta ken appropriate training and that they are fit to practice. In general, unattended birth can occur in two different circumsta nces and the findin gs on examin ation of the baby mi ght be expected to be different in each case. Either birth may follow a concealed pregnancy, or emergency professional help is unavailable because labour is either short or has not been perceived in time and delivery is
Was the baby born alive? I
183
(a)
Figure 10.2 Mummified remains of a baby found behind panelling during renovations. Body length and amount of subcutaneous fat indicate that the baby was mature.
rapid and occurs before appropriate help is availab le (Fig. 10.3). In the case of mature baby, the mother is often of high parity and has usually, but not invariably, registered for antenatal care and planned confinement. In unattended delivery in general, perinata l mortality is higher than in the case of local hospital delivery or planned home birth .21
Delivery into Toilets AJleged spontaneous delivery of a mature baby into a lava LOry pan shou ld be viewed with circumspection. 22 Although perinea l sensations preceding delivery may be confusing, only the rapid delivery of previable fetuses in this way is redible. The subsequent actions of the mother and other persons present in the domicile are important. Prompt and eta iled documentation of these is required. Disposal of unwanted, liveborn infants in toilets is not uncommon in so me parts of Japan. 17 - 19
WAS THE BABY BORN ALIVE? t may not be possible to distinguish with absolute <:enainty between live birth and still birth from pathological
Figure 10.3 Precipitate delivery onto the bathroom floor. (a) A baby of about 26 weeks' gestation. The skin is extensively reddened as a resu lt of hypothermia. (b) A section of skin showing extensive haemorrhage in the deep dermis. findings alone, but there are some findings that can answer this question absolutely. When there is doubt, this must be clearly expressed and the likelihood of still birth promoted.
184 I
Fetal and perinatal death
Figure 10.4
The earliest sign of maceration, skin slippage, over
an extremity.
Dark-red discolouration of the fetal insertion of the umbilical cord is a sign of maceration which takes about six hours
Figure 10.6
to develop.
Figure 10.7 A 4-day-old neonate. Desiccation of the umbilical cord is an indication of live birth with several days' survival.
(Courtesy of Professor N Mcintosh, Edinburgh.)
Early maceration with focal epidermal loss over bony prominences. Drying has resulted in dark-red discolouration, which might be confused with inflicted injury.
area is involved, we can estimate that death occurred about 18 hours before delivery.23 These times are only approximate. Estimates of the time of death may be modified by the appearance of the umbilical cord and internal organs. It is important to remember that cutaneous maceration is retarded in the presence of intra-uterine growth retardation (fUGR) and accelerated in the presence of oedema (fetal hydrops).
MACERATION
UMBILICAL CORD INSERTION
Changes in the skin due to maceration preclude live birth. Sequential changes owing to maceration are set out in Table 8.5, p. 159. Cutaneous maceration is unlikely to be present until about six hours after death, when focal loss of superfi cial epidermis is seen over extremities and bony prominences (Figs 10.4 and 10.5). This loss of epidermis (skin slippage) becomes more widespread with the increasing length of time that the dead fetus spends in utero. Maceration of face, back or trunk develops in about 12 hours. When more than one
Deep-red discolouration of the umbilical cord at its fetal insertion is an indication that fetal death has taken place at least six hours before delivery (Fig. 10.6). Loss of Wharton's jelly from this part of the cord, often accompanied by marked twisting of the cord, takes longer to develop and is also a marker of still birth. Desiccation or separation of the umbilical cord stump indicates live birth with survival for a few days. The cord stump progressively desiccates (Fig. 10.7) over 3 or 4 days and then separates from the
Figure 10.5
Was the baby born alive? I
umbilicus. A vital reaction is apparent w ithin the cord stump w ithin 2- 3 hours of birth on microscopic exam ination; this indicates live birth with, at least, brief survival.
STOMACH Mil k curd (or other foreign ma terial) in the sto mac h indi cates live birth with survival. Care must be taken no t to confuse mucoid secretions, which can be turb id in the matu re fetus, particularly in the presence of ascending infection w hen the fet us has swallowed infected liquor am ni i, w ith milk, which is usually flo ccu lent. Smears of gastri c co ntents stained w ith haematoxylin an d eosin, Gram 's method for bacteria l organisms and oil red a for fat will often clarify any uncertainty. Rad iographic examination may demonst rate gas in th e stomach and intestines in liveborn infants. This, however, is not abso lute proof of live birth as contamination with gas- form in g organisms by swallowing infected liquor am nii ca n result in a gast ric gas bubble. Meconium may be present in the stomach when there has been acute hypo xic stress to the fetus either just befo re or during labour. Its presence or absence does not help to distinguish between live birth an d still birth.
185
examp le tissues, cotton w ool or fabric, is occasionally fo und in the mouth or pha ry nx in deliberate a ttempted a ir ways obstruction. Abso rbent material is so metimes found in or over the mouth, presumably to take up unwanted fluid secretions, when other findings suggest still birth. The presence of frothy fluid in the upp er a irways, which has been fo und to withstand freezing of the infant's body, 28 is a lso an indica tion of li ve birth.
LUNG Gross Appearance When the lun g has been fully expanded by establishment of respiration in a mature baby, it w ill be pink and crepita nt. It is virtually impossible to expel a ll of the air by external pressure. The lungs are unexpanded in still birth. They are smaller, darker in colour and occupy the posterior, paraver tebral parts of the thoracic cavity. Knight 29 found the naked eye appearance and feel of lu ngs a better dis criminant between live a nd still birth than the lung flotati on test. Petech ial hae morrhages are often visible throu gh the pleural surface, but they may be confined to the pleural fissures. They are considered a sign of acu te hypoxia . Petechial h aemo rrhages are p art icularly numerous when death is du e to placental abruption.3o They lose definitio n and then fade afte r a fe w days, both in the fetus reta ined in utero and in the delivered baby.
STOOLS Before birth, the large intestine contai ns meconium, a mix ture of amniotic squames, intestinal secretions, desqu a ma ted intestinal epithelium and bile acids. 24 Many babies defecate in the first few minutes after birth ,25 although many mature ba bies who are subjected to hy pox ic stress before or during labour w ill often defecate before birth. Meconium is dark green and sticky, but sterile. Over the first few days after birth, a mixture of meconium and normal stool is passed. Breast-fed babies have very soft, acidic, mustard-yellow stool; bottle fed babies have a firmer, paler stool, w hich is less acid ic. 25 Colonization of the intestine wit h bacterial organisms takes place rapid ly. There is geographic variation . More than one-hal f of vaginally delivered infants in Pakistan have positive cu ltures for faec al organisms at 1 day of age,26 but the proportio n in Swedish infan ts was found to be much lower. Stool bacteria l flora was well estab lished at 4 days of age in babies born in the UKY The stool of milk fed newborns wi ll cont a in abundant fat globules; bacterial organisms will be seen diffusely through the stool sample. There are di fferences in the bacterial flora of breast-fed, compared with form ula-fed, bab ies. 26.27
Flotation Test The use of the property of lu ngs to float in water (or buffered formalin) as a determinant of live birth is fra ught w ith diffi cul ty. 29 It is un w ise to rely on it as the only deter minant of live birth even when so me or any of the pub lished modifications, which all egedly impro ve reliability, are introdu ced. It may be falsely positive because of putre faction, even to a minor degree. When putrefactio n has occurred, gas may be in terstitial or both in terst itial and within terminal ai r spaces (see Fig. IO .Sc and d). The value of the flotatio n test is negated by mouth-to-mouth or other p ositive-press ure ventilation J The flotation test may be falsely nega tive in the pres ence of well-developed hyaline membrane disease, a fre quent find in g in the livebo rn preterm baby as well as liveb orn babies born at term who have experienced severe hypoxic stress, or follo w ing ventilation with 100 per cent oxygen. Despite these dra w backs, Moar 31 eva lu ated lung flotation along with liver flotation, lun g compression, the gross appea rance of the lungs and histol ogica l examina tion. He fo und it of valu e in carefully controll ed co nditions but stressed the need for corroborative tests.
UPPER AIRWAYS
Histological Appearance
Plugging of the upper airways by thick meconium pre cludes estab lishment of respirati on. Foreign material, for
Both the presence of large amou nts of squamous debris and mu cus fro m the amnioti c fluid and the presence of meconium
186 I
Fetal and perinatal death
.':.... Figure 10.8 Patterns of lung distension. (a) Antepartum still birth at term: there is partial expansion of terminal air spaces up to the pleural surface as a result of hypoxia-induced inspiratory efforts. There is meconium aspiration. (b) Delivery after motor vehicle collision (MVC) at 34 weeks' gestation, survival for 13 hours. There is overdistension of term inal bronchioles and alveolar ducts with collapse of intervening lung parenchyma; these changes are seen in early respiratory distress syndrome. (c) Irregu lar, focally marked, distension of air spaces and interstit ial tissue as a result of putrefaction, concealed delivery. (d) Putrefaction has resulted in overdi stension of terminal air spaces and accumulation of gas in pleural connective tissue.
plugging of large intrapulmonalY airways and groups of ter minal air spaces indicate severe hypoxic stress before birth. The former are readily demonstrated using Alcian blue/phloxine staining, promoted by Attwood 32 to demon strate amniotic fluid embolism in the maternal pulmonary vasculature, or an immunohistological marker against high molecular-weight cytokeratin, such as CK1. 33 When the fetus has made strong and repeated inspiratory efforts before birth, partial expansion of terminal air spaces is apparent and the alveolar walls create a saw-tooth pattern (Fig. 1O.8a). The presence of hyaline membranes within the lung on histological exa mination is a marker of live birth . They take around 6 hours to develop and up to 12 hours to become widespread within the lung. Before that time, a char acteristic pattern of overdistension of terminal bronchioles and alveo lar ducts is usually apparent (Fig. 1O.8b). Although
hyaline membranes are classically associated with the immature lungs of a preterm infant, hyaline membranes may be seen in the mature fetus following hypoxic injury to the type II pneumocytes. When respiration has become fully established the lung will be uniformly expanded (Fig. 10.S). This can be demon strated by histological examination. The terminal air spaces will not be completely expanded after death, but some expansion should be present right out to the pleura. Pul monary interstitial emphysema has been found to be a use ful additional marker of live birth. 34 Septal distension due to interstitial air trapping and distension of lymphatics can be distinguished using an endothelial marker. It is also impor tant to consider that this might be part of the putrefaction process and indicate the presence of gas-folming organisms (Fig. 1O.8c and d).
Is there evidence of prolonged or difficult labour? I
IS THE BABY OF SUFFICIENT MATURITY TO SURVIVE? In the UK, the legal definition of viability was, in 1992, revised downwards to 24 completed weeks of gestation from the previous level of 28 completed weeks of gestation. 35 Whilst some babies born at low gestations survive in a hospi tal setting where modem facilities and highly skilled carers are available, should they be born outside hospital they are unlikely to survive for very long in a domestic environment. The hazards of unassisted labour and delivery are particularly great for the immature baby. Subsequently, hypothermia (see Fig. 10.3, p. 183) and respiratory problems compromise infant well-being. Opinions about the chances of survival of a par ticular baby should take into account both the level of care required, above that likely to be available at the time of birth, and the proportion of babies who survive at that gestation when given appropriate care. This is between 17 and 49 per cent at 24 weeks' gestation and between 70 and 74 per cent at 27 weeks' gestation, taking into account fully informative studies performed in the 1990s. These are reviewed by Evans and Levine. 36 Likelihood of survival should take into account the standards of care current at the time of death, when fetal or infant deaths are investigated or re-investigated several years after death occurred. 3? Assessment of fetal maturity is best made using a com bination of different factors. These are external measure ments, fetal body weight, organ weights (see Appendix 1), the complexity of the cerebral gyral pattern and histologi cal assessment of some organs. Basic measurements of crown-rump, crown-heel and foot length, together with head circumference and bipari etal diameter, are compared with normal standards and will provide a reasonable working assessment of maturity;38,39 body weight can be assessed for normality using the measurement-derived gestational estimation. This is rele vant to the likelihood of survival, as both growth-restricted and macrosomic fetuses are less likely to survive than appropriately grown infants. The assessment of appropri ateness of growth is relevant to the overall interpretation of the case. Anthropomorphic standards are also available for comparison with skeletonized fetal remains. 4o The appearance of the cerebral gyral pattern is particularly useful in the second half of pregnancy, when the complexity of pattern develops in a regular fashion. It is particularly use ful when there is severe growth restriction. Body lengths and, in very severe cases, head circumference may be reduced below gestation-related norms. It can be assessed against illustrations of gestation-related normal appearance. 39 ,41.42 The histological appearance of the renal cortex is a useful aid to assessment of gestation from 18 to 36 weeks of gestation using a combination of the number of generations of nephrons and the appearance of the nephrogenic zone. 39 A whole-body radiograph will provide additional informa tion about fetal maturity based on the presence of ossification centres and length of long bones. 43 ,44
187
LUNG MATURITY A key factor in the newborn's ability to achieve an inde pendent existence is lung maturity. If the lungs are imma ture or abnormally developed, then the baby may not survive despite being of such dimensions that survival would be expected. Lung weight, despite confounding fac tors such as haemorrhage, oedema and inflammation, is a good, simple indicator of lung growth. The appropriateness of lung size can be conveniently assessed by suspending the thoracic viscera by the larynx. The lower borders of the lungs should be level with the apex of the cardiac ventri cles. 45 Histological examination provides further informa tion. It is not usually necessary to resort to morphometric assessment, but sometimes this may be thought desirable in a particular case (see Baak and Oort 46 for methods). DID THE BABY HAVE A SEPARATE EXISTENCE? To establish live birth, it is not sufficient merely to rule out still birth - the baby must have exhibited signs of life, i.e. breathing, a heart beat or spontaneous movement, when completely outside its mother's body. Neither division of the umbilical cord nor expulsion of the placenta enters into the definition of live birth. Of these signs the pathologist can comment only on the likelihood of established respira tion. Even this must be done with caution as some respira tory effort can occur during the birth process, before separation from the mother is complete. The skin of a term baby will usually be covered with vernix caseosa, often mixed with blood. When no vernix is present at all, not even in flexures, deliberate action to clean the baby's skin has occurred. As set out earlier, desiccation or separation of the umbilical cord, a vital reduction in the cord stump, the presence of milk curd in the stomach, faeces (not meconium) in the terminal colon or the presence of hya line membranes on histological examination of the lungs indicates not just live birth but survival for some hours or days. Thus, they are clear markers of a separate existence.
IS THERE EVIDENCE OF PROLONGED OR DIFFICULT LABOUR? Concealed pregnancies are often first pregnancies. Without appropriate assistance, labour may be particularly long and delivery itself achieved with difficulty. The stresses experi enced by the fetus during labour may be compounded by prolonged pregnancy. It is well established that perinatal mOliality is lowest between 39 and 41 weeks of gestation (term) and is higher outside those limits.4? When subjected to severe hypoxic stress, a term baby often passes meconium into the amniotic fluid. This will adhere to the skin and will also stain the chorionic plate of the placenta a greenish brown colour (see Fig. 10.16, p. 194). If the skin has been cleaned, meconium staining can still
1 88 I
Fetal and perinatal death
Figure 10.9 Elong at ion of the verticomental diameter of the cranium can be an indication of long labour.
Figure 10.10 Blotchy facial petechiae in a non-macera ted antepartum still birth as a result of retrop lacental haemorrhage.
usually be found beneath fingernails , in skin creases and in and behind the ears. In cephalic fetal presentation, prolonged labour is likely to produce excessive moulding of the head. When the ver tex is the presenting palt, this resu lts in elongation of the verticomental diameter of the cranium (Fig. 10.9). With vertex presentation there is often an area of localized oedema, usually with marked congestion, but sometimes with frank haemorrhage, over the posterior fontanelle or adjacent parietal or occipital bones. In antepartum fetal death, very marked congestion of the whole of the scalp can occur, simulating subgaleal haemorrhage which may complicate instrumental or, very occasionally, spontan eous delivery4B (see Chapter 8, p. 160). Examination of the head and face for localized conges tion and oedema may suggest an unfavourable presenta tion, which might be expected to add to the difficulty of delivery. It is not unusual to find cutaneous petechial haem orrhages over the presen ting part. They are particularly likely when placental abruption has occurred (Fig. 10.10). Another cause is difficulty in delivery of the shoulders. In this situation, there is increased venous pressure in the delivered head and neck whilst the thorax is still constricted by the birth canal. Petechial haemorrhages are commonly found over the face, h ead and n eck and occasionally on the chest wall. Such haemorrhages should not be interpreted as evidence of strangulation or deliberate airways obstruction without corroborative evidence. Shealing tears in the tentorium, unassociated w ith dam age to cerebral sinuses, indicate excessive and/or rapid dis tortion of the cranium. A baby presenting in the breech position is at particular lisk during unattended delivery. The
-
--
-
breeched baby can pass through an incompletely dilated cervix, which then impedes passage of the head. Breech pres entation is suggested by the presence of cutaneous petechial haemorrhages over the lower trunk and legs. Sometimes, haemolThage into underlying muscle is extensive. Unskilled effOlts to complete delivery by traction on the baby's trunk can compound the probl em causing injury to the liver, which will be tense because of congestion. A baby's liver is poorly protected by the rib cage, particularly when delivered preterm. When a term baby appears to have presented in the breech position, it is important to look for factors that predis pose to breech presentation, such as neuromuscular problems or renal agenesis (because of oligohydramnios), as they may make a major contribution to failure to survive; death is then attributable to natural disease. The umbilical cord shlmp should be examined carefully and any discolouration or dehiscence noted (see ab ove). The appearance of the free end of the cord may indicate the method of its division (see Chapter 8, p. 160). It may corre spond with the free en d of the cord attached to a placenta located elsewhere (Fig. 10.]]). The presence or absence of a clip or tie on the cord should be noted and described.
ARE THERE ANY SIGNIFICANT INJURIES? Before ascribing significance to any injury observed, it is important to identify commonly described bilth injuries and to be aware of the circumstances in which they occur. They are discussed in Chapter 8 (p. 161).
.a_
~_~
Are there any significant injuries? I
Figure 10.11
189
Torn umbilical cord, concealed delivery.
It is clearly important, when examining a baby found dead after concealed pregnancy or unattended delivery, to look carefully for evidence of inflicted injury. Not every injury found will be deliberately inf1icted. It is impOitant to identify injuries that might occur as a result of maternal efforts to self-deliver as well as any which might arise during the course of prolonged labour. Other injuries may occur after death, during or followin g efforts to conceal the body. Evi dence of post-mortem animal-inf1icted injury is sometimes seen in babies who have been aba ndoned out of doors. The presence and nature of any irUury is recorded and photographed. Injuries are infrequent, death is more usually due to lack of cares Drowning and hypothermia from expo sure are other causes of death in these circumstances. 49 However, in one study from South America,50 violent injury was fou nd in a large proportion of in fant icide cases.
EXTERNAL FEATURES
Linear or slightly curved, parallel, abrasions on the neck are sometimes seen. Examination of skin tags will show them to run from the shoulder to t he head. These can be infli cted by the mother trying to grasp th e chin/occiput in an effort to expedite delivelY of the shoulders. Early skin slippage, which has dri ed out, co mmonly seen over bony prominences such as the brow, cheek bone or chi n, can be mistaken on firs t glance for abrasion due to deliberate injury (see Fig. 10.5, p. 184). Other changes of maceration should be carefully sought. Cutaneous bruises are unusual bilth injuries. Any found should be sampled for histological examinatio n to distin guish t hem from injuries caused post-m oltem. 'Bru ising' of buttocks and legs may be see n in infants presenting in the preterm breech position as a result of hypoxic/hypostatic haemorrhage into underlying muscles. A Mongolian blue spo t (see Fig. 8.7, p. 152), usually pres ent at the apex of the gluteal fold, can sometimes be very extensive and should not be interp reted as trauma. Althou gh more common in Afro-Caribb ean infa nts and
Figure 10.12 Complicated skull fra ct ure of the parietal bone following maternal abdominal trauma. (Courtesy of Dr KJ McKenzie, Ed inbu rg h.) those whose parents are from the Indian su bcontinent, it is somet imes seen in babies of European parentage. FRACTURES
Skull fractures are unusu al foll owing non-instrumental delivery, so any fracture should be documented both photo grap hically and radiologically, and adjacent tissues exam ined with care. As in older individuals, fractures which extend across suture lines result from impact trauma (Fig. 10.12). Fractures resulting from the process of birth are dis cussed in Chapter 8 (pp. 161-162). INTRACRANIAL HAEMORRHAGE
Intracran ial haemorrhage should be carefully documented and photographed. Much information about causatio n can be addu ced by careful observation and recording of the site of haemorrhage 48 (Fig. 10.1 J). Extradural hae mato ma is excessively rare in newborns, apart from a trace of blood in the vicinity of a fracture. Subdural haemorrhage is almost always due to trauma. Heavy blood sta ining of the cerebrospinal fluid (C SF). or haemorrhage in the posterior fossa may be the result of sub dural haemonhage over the convexities but, particularly in preterm infants, may result from intraventricul ar haemo r rhage, when it emerges into the subarachnoid space through the fourth ventricular fora mina. It may sometimes have a local cause. Shearing tears of the falx or, more usu ally, the tentorium indicate excess ive moulding of the head.
190 I
Fetal and perinatal death Trauma
Asphyxia
Cutaneous petechiae
~--------------------
Haemorrhagic infarction
Caput succedaneum
Subaponeurotic ---"-.A:---'I------------------ (galeal) haemorrhage
""""""""'c-------------
Subarachnoid haemorrhage
Cepha Ihaem atom a
,,-'Ic-'\--\-\--\\----------- Ex trad ura I
haemorrhage
Falcine haemorrhage
Intraventricu la r haemorrhage
Germinal matrix (subependymal) haemorrhage
Brain
rf
r
----+\-\-\--+-\-\-------- Su bd ura I
haemorrhage
Lep
J IIII
Bone
Figure 10.13 Sites of pericranial and intracranial haemorrhage in the newborn by causation. (From edn, 2007, p. 287, Fig. 13.16, with permission, Springer Science and Business Media.)
Subarachnoid haemorrhage is usually focal but can be confluent, usually over the temporal poles (but see Fig. 8.19, p. 160). It is usually hypoxic in origin. Focal subarachnoid haemorrhage is often found at the base of a torn bridging vein when the vessel has retracted beneath the membrane. Cerebral haemorrhage is rarely traumatic. It is usually related to bleeding into an area of prior hypoxic injury. The same is true of intraventricular haemorrhage (NH), which arises from hypoxic injury to small vessels in the periventricular germinal matrix or choroid plexus. Although NH most commonly occurs in the preterm baby postnatally, there are well-documented examples of NH occurring in utero. 51 Therefore, its presence in a perinatal death should not be interpreted as a marker of live birth.
FETAL DEATH FOLLOWING MATERNAL INJURY Fetal deaths following maternal injury are not common, although trauma complicates 6-7 per cent of pregnancies. 52 The extent of legally directed investigation of the baby in these circumstances depends on a number of factors: the causation of the injury, whether the baby was liveborn or still born and the jurisdiction under which it falls. The procu rator fiscal (Scotland) is more likely to instruct an investiga tion into a fetal death or still birth than is the coroner (England) if the status of the fetus as a person seems some what ambiguous. Fetal maturity is an important factor and is most impOliant when prosecution of any sort is being con sidered?8 Studies of the outcome of minor trauma in preg nancy are few. Fort and Harlin5J identified women from the
:::0'"'"'
(9;''''1
Periosteum
Fetal and Neanatal Pathalagy,
4th
American Collaborative Study who had suffered non-cata strophic trauma during pregnancy and looked at the out comes. Traumatic episodes comprising blows to the abdomen, falls and minor road traffic accidents complicated 210 (5.7 per cent) pregnancies, half occurring in the third trimester. No increase in pregnancy loss was found when compared with non-traumatized control subjects. Fetal loss is more likely following major maternal trauma (61 per cent) than following minor trauma (27 per cem)54 and when the maternal injury severity score is high 55 57 and the mother is in shock. When trauma is non catastrophic, complications are unusual. Fetal loss occurred in about nine per cent of cases in one study.58 The reason is most commonly premature labour, but placental abrup tion sometimes occurs. Some advocate careful observation, even after minor trauma, for this reason. 59 However, a prospective study of nearly 50000 cases found pregnancy related complications in only 0.001 per cent.50
BLUNT ABDOMINAL TRAUMA
The most common cause of trauma to the abdomen of a preg nant woman is motor vehicle collision (MYC).51 In one study, maternal injury following MYC accounted for 80 per cent of cases of abdominal trauma to pregnant women, of whom 10 per cent were pedestrians. 55 Other injuries, falls, assaults and self-harm, are much less common. One institution found that serious assault accounted for 31.5 per cent of trauma to preg nant women over a 9-year period 52 and for 22 per cent of injury in another study over a similar period. 52
Fetal death following maternal injury I
With expanding vehicle ownership, pregnant women are increasingly involved in MVCs whether as a driver, passen ger or pedestrian. One study found the highest incidence of fetal death following MVC in teenage mothers. 63 Occasion ally pregnant women are involved in accidents as cyclists. In my experience, fal ls from bicycles can provoke premature onset of labour and fetal death as a result of prematurity or placental abruption. Although accidents can occur at any stage of pregnancy, deaths occurring in the last trimester of pregnancy (after 28 weeks' gestation) are the most likely to be subjected to medicolegal investigation, although the deaths of less mature fetuses are sometimes investigated. When assessing the likelihood of feta l death being the result of any accident, an acquaintance with the incidence of spon taneous pregnancy loss at that particular time of gestation is useful. Spontaneous fetal loss is less common as pregnancy advances. In the absence of ma terna l predisposing factors, some of which may be revealed on exa mination of the pla centa, co incidental fetal demise or spontaneous onset of labour is unlikely after 32-34 weeks of gestatio n. The prime concerns after materna I involvement in an MVC are for maternal and fetal well-being, the former bear ing critically on the latter. 64 Observation of mother and fetal monitoring are necessary for several hours, even when maternal injury is slight. 61 Any maternal injuries present should be carefu lly recorded . A careful examination of the abdomina l wa ll for bruising is essential. When the accident involves the mother as a pedestrian or cyclist, the full cir cumstances and extent of injury to the mother are recorded and are usefully made available to the pathologist before fetal examination commen ces. When the mother was the driver or passenger in a vehi cle, information about the type of restrain ts and whether they were used is important. Fetal mortality and morbidity has been associated with lap strap use. Early reports of uter ine rupture 65 and fetal inju ry66 have deterred some women from using restraints. One study found that fetal mortality increased from 14.4 per cent to 16.7 per cent when this type of restraint was used, a small and comparatively non significant increase, but maternal mortality was reduced by half67 Despite more recent advice that the risks to the fetus are reduced with appropriate use of three point restraints,68 the message is difficult to get across to some women. 55 Impact severity is the best prediction of adverse fetal out come, but appropriate seatbelt use (three-point restraint) improved fetal outcome following impacts of mild or mod era te severity.69 The position of the restraining straps across the woman's body is criti ca l to fetal well-being. Defo rma tion of the uterus from forcible contact with the restraint is the cause of retropla cental haemorrhage producing a shear strain across the uteroplacenta l interface. There may be additiona l stresses from tensi le failure during rapid deceler ation of the uterus following impact. 69 Restraints shou ld be positioned above and below the gravid uterus. 6S Premature onset of labour is a common complication (28 per cent) when trauma occurs between 25 and 37
191
weeks' gestation 70 but is not usually a cause of fetal loss. Uterine rupture is uncommon/! but when it occurs fetal mortality is close to 100 per cent. Amniotic fluid embolism is an occasional compJication n Other causes of blunt abdominal trauma are falls and assault, such as ki cking or heavy blows to the maternal abdomen. Whilst the velocity component of trauma enco untered in MVC is missing, nevertheless these insults can result in pregnancy loss and fetal injury even when the maternal injury severity score is zero. 62 The in cid ence of assault during pregnan cy is difficult to assess and likely to be under-estimated. Systematic questioning of mothers attending antenatal clin ic found that an assault had occurred at some time during pregnancy in 14.1 per cent of 384 women. 73 In another study, physical violence occurred in 11.1 per cent of 6143 women in the year prior to deliv ery, with an increase in materna l morbidity including preterm delivery, but no other adverse fetal outcome was noted l4 A study from Sa udi Arabia reported physical vio lence in pregn ancy in 21 per cent of over 7000 records;75 there was an increased rate of placental abruption and pre mature delivery. Confidential enquiries into maternal deaths in the UK found that more than one-third of domes tic abuse cases start during pregnancy and the risk of fetal loss or still birth is doubled. 76 The subsequen t triennial report cites continuing vio lence to pregnant women. Eleven deaths in that triennium were the result of murder by the woman's partner, higher than the number of deaths due to MVC or amniotic fluid embolism and equal to deaths from sepsis. 77 Fatalities are the tip of the iceberg of domestic viol ence in pregnancy. Adverse feta l outcome fol lowing domestic violence, including premature delivery and a 3.5-fold in crea se in neonatal death, is described in a North American study.78 A literature review of 30 informa tive studies of intimate partner v iol ence 9 found an increase in fetal and maternal mortality and morbidity, with a feta l death rate around 16 in 1000 in affected pregnancies. Chronic subdural haematomas have been demonstrated in the neonatal period in offspring of women subjected to physical abuse, which included kicking and blows to the abdomen dUling pregnancy.80,8! In the Stephens et aisl study, cerebral atrophy and haematomata of different ages were apparent on imag ing shortly after birth. Repeated assault in the context of domestic violence increases the risk of peri partl.lm complications. B2 Subdural haematoma have been desClibed in 47 offspring of immigrant Pacific Islanders in New Zealand. s3 Forty-four of these babies were still born; some had intraventricular or intracerebral haemorrhage as well as a subdural haematom a. Some had sca lp or significant subgalea l haematoma. The haemon'hages were not accompanied by skull fra ctures or tears of the dural folds. Coagulation studies were normal in all mothers tested; a further eight had normal platelet counts. The ca use of the intracranial and pericranial haem orrhages in these babies was thought to be particularly
192 I
Fetal and perinatal death
forceful abdominal massage, resulting in slow head com pression. It is probably significant that more than half of the injured babies were presenting in the breech position.
ELECTRICAL INJURY TO THE PREGNANT WOMAN The literature pertaining to electrical injury, including lig ht ning strikes, is small, with a heavy emphasis on single case reports. The outcome is very variable. Lieberman et al 84 report six cases of electrical injury related to domestic acci dents occurring between 20 weeks' gestation and term. In two cases in which there was an immediate reduction in fetal movements, fetal death occurred within one week of the accident. A furth er fetus died 12 weeks after an insult at 2 1 weeks' gestation and was growth restlicted at birth. A causal relationship here is tenuous. The other three were born at full term and survived. The same groups report abortion after an electrical accident. 8s Jaffe et al 86 report fetal death at 14 weeks occurring within hours of a domestic mishap. Reports of fetal death from the late nineteenth and early twentieth century (cited by Mehl 87 ) describe immediate cessation of fetal movements following the incident. Interval deaths and chronic problems such as oJigo hydramnios 84 are more difficult to evalu ate from the point of view of causation . The frequency of pregnancy loss from other (natural) causes at that particular stage of pregnancy must be taken into account. Yoong88 describes massive placental abruption and fetal death 24 hours after electri cal injury. Fetal mortality following maternal electrical injury, when based on accumulated reported cases, is high. Fatovich 89 found a fetal mortality of 73 per cent among 15 reported cases. However, reporting bias is likely w hen the subjects are single case reports. The only prospective study of pregnancy outcome following maternal electrical injury followed 31 women and fo und no adverse outcome related to the insult. 9o Fetal outcome following matern al lightning strike is variable. Of the 12 reported cases 91 - 95 half of the fetuses died. Most mothers sustained burns. Infant survival bore no relationship to maternal loss of consciousness or being thrown to the ground . Loss of fetal movements was observed soon after the incident when the fetus died. There is a Single case report of taser injury in custody 87 resulting in miscarriage 7 days after the incident. There were other inauspicious features in that case.
PENETRATING TRAUMA TO THE MATERNAL ABDOMEN The frequency of penetrating trauma to the gravid uterus has wide geograp hic variation. It is, not surprising ly, enco un tered more frequently in war zones 96 and those countries where gun ownership is common. 97 Fetal well-being is com promised by penetrating injUly to the uterus, but not all cases are fatal, even when there is some bleeding into the amnio tic cavity or minor fetal injury is sustained. 98 ,99
Maternal death is infrequent. The gravid uterus provides protection to other maternal organs. As the uterus increases in size, organs, particularly the in testines, are displaced upwards and may be spared from direct injury. In the case of gunshot wounds, the uterus, being of dense texture, absorbs most of the energy, reducing the risk of damage to other maternal abdom in al viscera. loo Intra-uterine fetal death, premature onset of labour and fetal injury are more likely following penetrating abdominal trauma than is the case following blunt trauma. S5 Fetal injury is seen in 60-90 per cent of cases, about half being serious, including penetrating wounds of thoracic and abdominal viscera. Injury to the placenta and umbilical cord is also described. IOI Pelinatal mortality of 47-71 per cent is likely.loo Among Awwad et al's96 cases, half of the fetuses died because of maternal hypotension or direct fetal injury. Sel f-directed penetrating trauma has, bizarrely, been used in attempts to terminate pregnancy. Both self-inflicted gunshot wounding 102 and vo luntary repeated stabbi ng l03 are recorded .
Fetal Injuries and Cause of Fetal Death The usual cause of fetal death followin g blunt trauma to the maternal abdomen is retroplacental haemorrhage. It usually occurs within hours of the injU1Y and can be predicted by increased uteline contractions. 61 Delayed (several days) placen tal abruption has been recorded. Fetomatemal haemonhage is more common in pregnancies complicated by maternal trauma. 104 It is more likely when the placenta is anterior61 and can result in fetal dysrhythmias, anaemia and fetal death from exsanguination. Placental abruption follo wi ng MVC is signifi cantly associated with higher maternal injulY severity score a nd higher vehicle speed. 105 Still birth occurred in more than 50 per cent of cases in which abruption followed the collision, with no fetal deaths in the absence of abruption. Rogers et al 106 also fo und an association between placental abruption and high maternal severity score. They found that fetal cardio graphic monitoring could reduce fetal mortality. Any injulY resulting in maternal hypotension puts the fetus at risk. If prolonged or severe, it will result in fetal death from hypoxic/hypotensive injury.107 These changes are likely to be seen in the brain and kidney and occasion ally the heart, liver and intestines. Pathol ogical findings are of extensive haemorrhagic infarction. 108 When the effect on the maternal circulation is less severe, the fetus may survive, but wit h brain injury. Rapid resuscitation of the mother gives the fetus the best chance of survival. 59 Blunt injury to the maternal abdomen can result in direct injury to the fetus, altho ugh it is not common. 58 This is most likely to occur when maternal injury occurs late in pregnancy. The fetal head is afforded protection by the maternal pelvis but is vulnerable when the pelvis is frac tured. Sometimes the fetal skull is fractured when maternal injury is trivial. Skull bones are the most common sites of fractures susta ined by the fetus (Fig. 10.12, p. 189) and may
Can I give a cause of death? I
be multiple. 109 Not all are fatal. 110 Skull fractures may be complicated by intracranial haemorrhage, perhaps the more significant injury. 54 Intracranial haemorrhage and tentorial tears are also described. 66 Subdural and pericranial haemorrhage in t he absence of skull fracture and tears of dural folds can cause death within hours, although some fetuses apparent ly survive for several days.B3 Not all antepartum subdural haematomas are fatal. Some babies are liveborn with significant morbidity owing to cerebral atrophy, hydrocephaly or porencephalyBI .83. 111 Whilst the vast majority of in utero subd ural haemorrhage can be related to tra umatic insult, there are occasional reports of this type of haemorrhage being found in situations of reduced blood coagulability. Bilateral fetal subdural haemorrhage in late pregnancy has been described as a com plication of maternal warfarin therapy, 11 2 although no infor mation about precise timing or repetition of haemonha ge was ava il a ble. It has been described occasionally as a feta l complication of inherited clotting disord ers, such as factor X deficiency.113 Usua lly, however, fatal thrombocytopenia and clotting factor deficiencies give rise to intracranial haemor rhage in other sites, predominantly intracerebral and in tra ventricular. 114 See Chapter 4 (pp. 94- 97). Cerv ical injury is rare, bu t lower cervical injury (C5) is describ ed follo wi ng ]'vTVC when the mother was using a seatbelt. 115 Some authors suggest that protection from direct feta l injury in MVC is afforded by an anterior pla centa, but the placenta itself is vulnerable, resulting in retropl acental haematoma. The posterior situation of the placen ta was thought to be significant in one infant who sustained lace ration of the liver, ha emoperiton eum and subdural and subp leural haemo rrhage. 116 Splenic injury with haemoperitoneum is also described.54.117 Fetal chest injuries with h ae mothora x and pulmonary contusion can follo w MCv. 11 8 Scalp bruising was presen t in one infant foll owing MVC, although the cause of death was rupture of the placen tao U9
193
Malformations which should be carefully so ug ht when resuscitation is unsuccessful include laryngeal atresia (Fig. 10. 14) or severe laryngea l stenosis, laryngeal cleft (Fig. 10.15), diaphra gmatic hernia and rena l age nes is. Cardiac disorders do not usually present as fai lure to respo nd to resuscitation. Those cardiac anomalies responsi ble for peJinatal death are usually ductus-d ependent lesions. Death usually occurs after sudden deterioratio n at 1 or 2 days of age. Both the cardiac anatomical anoma lies and some degree of cl osure of the ductus arteriosus w ill be readily apparent on exami natio n of th e thorax. Causes of su dd en de ath in the neon ate are considered furt her in Chapters l1 (p. 209) and 12 (pp. 248-2 50). Placental findin gs asso ciated with sudden death are vela mentous cord insertion with tearing of major vessels, tight cord knots, meconium staining of the fetal surface (Fig. 10.16) and retroplacental haemorrhage (Fig. 10.17).
CAN I GIVE A CAUSE OF DEATH? Whilst the answer to t his question before post- mortem should, in many cases of antepartum still birth, be 'no', few of these deaths are, in fact, reported to a coroner or procu rator fiscal - the ass umption being that examination of the placenta, even if fetal exami nation is denied, will provid e the answer, or at least substantial clues, to t he underl y in g problems.
IS THERE A NATURAL CAUSE FOR DEATH? When investigating apparent in t rapartum-related deaths it is important to be awa re of congenital or acquired diseases which might cause intrapartum death or failure to respond to resuscitation. It is important, too, that appropriate techniques are used to best demons trate any birth trauma a nd that microbiological samples are taken as unsusp ected asce nding infection can cause intrapartum death. These samples are most important when group B streptococcal infection is invo lved as fetal death is due to toxaemia, a nd histological evidence of response to infection may be slig ht or absent. The presence and nature of any dysmorphic features are noted. These may contribute to a syndrome diagnosis, which may be related to death. Remember that the presence of two or more dysmo rp hic features is an indication for chromosome examination .
Figure 10.14
Failed resuscitation of a mature, non-dysmorphic
baby. Laryngeal atresia: the upper larynx appears normal; complete airway obstruction by a thick bar of cartilage in the lower larynx.
194 I
Fetal and perinatal death
Figure 10.17 A placenta after concea led delivery. There is sign ificant blood clot attached to the maternal surface of the placenta.
causes of death, more likely to be able to recognize any iatrogenic contribution to death and less likely to compound the problem by wrongly interpreting trivial findings.
Figure 10.15 Posterior laryngeal cleft can facilitate massive aspiration or, as in this case, repeated displacement of an endotracheal tube, result ing in hypoxic brain damage. (From Fetal and Neonatal Pathology, 4th edn, 2007, p. 538, Fig. 20.7, with permission, Springer Science and Business Media.)
Figure 10.16 Meconium staining of the fetal surface of the placenta is an indication of fetal distress in mature babies; it persists for several days. In the face of an unexpected intrapartum (fresh) still birth, both obstetrician and paediatrician are more likely to report the death to the lo cal medicolegal officer, particularly if there is any expression of dissatisfaction from the family in respect of the care given during labour. Whatever the likely cause of death in this situation, it is important that assistance is sought from a perinatal pathologist. He or she will be more conversant with the range of likely natural
SHOULD THE INTRAPARTUM STILL BIRTH BE A MEDICOLEGAL AUTOPSY? Post-mortem examination following intrapartum fetal death should be done wi th particular care. A forensic pathologist presented with this problem is advised to seek advice of, and perhaps even collaborate with, an experi en ced perinatal pathologist before embarking on the necropsy. A full perinatal necropsy is needed with photo graphy and radiological examination, bacteriological sam ples fro m baby and placenta, complete external examination, full dissection of organs and fixation of the brain prior to detailed examin ation. It is important to examine the placenta, umbilical cord and membranes and to take routine samples from fatal organs and gestation sac for his tological examination. 39 The clinical history may indicate the need for Yates' dis section to remove the cervical spinal cord en bloc within the vertebrae. 120 If not, then in the absence of intracranial haem orrhage remova l of the spinal cord is recommended. Full sampling of organs for histological examination is important. Examination of the umbilical cord, placenta and membranes is essential in every case as it may demonstrate a natural cause of death.1 21 Even when the investigation has not been directed by the procurator fiscal/coroner/medic al examiner's departmen t, it has become increasingly likely that the pathol ogist's report will be scmtinized by lawyers and medical expelis on behalf of the bereaved parents. Bea ring this in mind, it is important to check through the report ca refully for factual or transcribing errors (particularly numbers and deci mal points). It is equally important not to promote the per ception of wrongdoing by the inappropriate interpretation of pathological findings, such as ascribing cephalhaematoma to trauma when this lesion can occur during an easy, normal vaginal delivery. Similarly, shearing tears of the tentorium are a marker of defo rmation of the skull but not a cause of
References I
death. They should also selve as a reminder for detailed neuropathological examination, which is only possible after fixation of the brain, even in mature babies. It is current practice for obstetricians or neonatologists to give a copy of the post-mortem report to bereaved parents. For this reason, it is our departmental practice not to issue to clinicians the clinical summary made by the pathologist for departmental use in case it is based on incomplete, or occa Sionally erroneous, information. Such errors can aggravate parental concerns about the possibility of substa ndard care. It is sensible to avo id the use of potentially emotive telms such as asphyxia or a noxia, which might exacerbate perceptions of malpractice. Hypoxia is a more appropriate and accurate term. Another potential aggravation is the overinterpretation of amniotic fluid debris, which is frequently present in the intrapulmonary airways, as meconium, which may be present in air spaces in abundance fol1owing hypoxic stress in utero in a mature baby.
REFERENCES Adelson L. Pedicide revisited . Am J ForeJ1 sic Med Pathol 1991 ; 12: 16-26.
2 Sims MA, Collins KA. Fetal death. A lO-year retrospective
review. Am J Forensic IvIed Patlwl 2001; 22:261-65.
3 Confidential Enquiry into Maternal and Child Health. Still
Birth, Neonata l and PostlleoJ1atal Mortality 2000-2002.
London: RCOG Press, 2005. 4 Schott J, Henley A, Kolmer N. Pregnancy and the Death oj a Baby: Guidelines Jor ProJessionals, 3rd edn. London: SANDS (St illbirth and Neonatal Death Society). 2007, pp. 171-2. 5 Kellett RJ. Infanticide and child destruction - the historical , lega l and pathologica l aspects. Forensic Sci Inter 1992; 5 3: 1- 28. 6 Mason JK. Unnatural death of the fetus and the newborn: abortion, child destru ction and infanticide or child murd er. In Maso n JK (ed.) ForeJ1 sic Medicine Jor Lawyel·s. Butterworths: London, 2001, pp. 254-68. 7 Di Maio YJ, Di Maio D. ForeJ1sic Pathology, 2nd edn. Boca
Raton: CRC Press, 200 1, pp. 336-7.
8 Bowen DA. Concealment o f birth. child destruction and in fant icide. In Mason .JK (ed.) Paediatric Foren sic MediciJ1e ft Pathology. London: Chapman and Hall Medical, 1989, pp. 178-90. 9 Ophoven JJ. Pediatric forensic pathology. In Gilbert-B arness E (ed.) Potter's Pathology oj the Fetus, InJant alld Child, 2nd edn. Mosby: Elsevier Inc. , 2007, pp. 816-20. 10 Goldenberg RL, Hale CB, Houde J et al. Neonatal deaths in Alabama. Ill. Out-o f-h os pital births. 1940- 1980. Am J Obstet Gynecol 1983; 147:687-93. 11 Ong BB, Green M. Infanticide in Malaysia: two case reports and a review of litera ture. Am J Fo rensic Med Pathol 2003; 24:64-9. 12 Hamilton A, English S. Baby bones uncover an island murder mystery. The Times, January 18,2002 ; p. 3. 13 English S. Woman charged after children's bodies found. The Times, April 18, 2001; p. 1. 14 Nurse hid her dead babies in A ttic. The Herald, August 27 2005, p. 6. 15 Boyes R. Mother tells of her dead babies' secret. Th e Times, August 3 2005, p. 32.
195
16 Funaya ma M, Sagisaka K. Consecutive infanticid es in Japan. Am J Forensic iVIed Pathol 1988 ; 9:9-11. 17 Yamauchi M. Usami S, Ikeda R et a l. Medico-legal stud ies on infanticid e: statistics and a case of repeated neon at icide. Forensic Sci Inter 2000; 11 3 :205-8. 18 Shiono H, Maya A. Tabata N et al. Medicolegal aspects of
infanticid e in Hokkaido District, Japan. Am J Forensic Med
Pathol 1986; 7: 104-6.
19 Funay ama M, Ikeda T, Tab ata N et al. Case repon: repeated neonaticides in Hokkaido. Forensic Sci Int 1994; 64:147- 50. 20 Davies RJ. The midwife's role in the managem ent of normal labour. In Cha mberlain G (ed.) Turnbull's Obstetrics. Edinburg h : Churchill Li v ingsto ne, 1995, pp. 581-90. 21 North ern Region Perinatal Mortality Survey Coordinating Group. Co ll aborative survey of perinatal Joss in planned and unplanned home births. Br Med J 1996; 313:1 306-19. 22 Mitchell EK, Davis JH. Spon taneo us bil1hs into toilets. J Foren sic Sci 1984; 29: 591 -6. 23 Genest DR, Singer DB. Estimating the time of death in still born fetuses: Ill. Ex ternal fetal examination; a study of 86 sti ll borns. Obstet Gynecol 1992; 80:593-600. 24 Harries JT. Meconium in health and disease. Br Med Bull 1978 ; 34:75-8. 25 Rob erton NRC. Care of the normal term newborn baby. In Rennie JM, Roberton NRC (eds) Textbook oj Neonatology. London : Churchill Livingsto ne, 1999, pp. 373 -87. 26 Adlerberth J, Carlsson B, Man de Pet al. Intest inal colonization with enterobacteriaceae in Pak istan and Swedish bospita l-delivered infants . Acta Paediatr SCGnd 1991; 80:602- 10. 27 Balmer SE, Wharton BA. Diet and faecal fl ora in tbe newborn: breast milk and infant formula. Arch Dis Child 1989 ; 64:1 672- 77. 28 Tabata N, Morita M. Azumi J. A frozen newborn infant: froth in the air-passage after thawing. Forensic Sci Int 2000; 108 :67-74. 29 Knight B (ed.). Infanticide and still birth. In Forensic Pathology. London: Ed wa rd Arnold, 199 1, pp. 402-JJ. 30 Wi gglesworth JS. Pathology of Intrapartum and early neonatal death in the no rmally form ed infant. In Wigglesworth JS, Singer DB (eds) Textb ook oj Fetal and Perinatal Pathology, 2nd edn. Malden, MA: Blackwell Sci ence, 1998, pp. 251-68. 31 Moar JJ. The hyd rostat ic test - a valid method of determining li ve birth? Am J Forellsic IvIed Pathol 1997; 18: 109-10. 32 Attwood HD. The histological diagnosis of amniotic fluid em bolism. J Pathol Bacteriol 1958; 76: 211-15. 33 Konstanidou AE, Agapitos E, Korkolopoulou P, Davari s P. Tracheoesophageal malformation: pathogenetic evidence provided by two cases. Teratology 2001; 63:lJ-14. 34 Lavezzi WA, Keough KJV! , Der'Oha nnesia n PJD et al. The use of pulmonary interstitial emphyse ma as an indicator of li ve birth. Am J Forensic Med and Pathol2003; 24:87- 91. 35 St ill birth (Definition) Act 199 2. Eliz II, 1992, c29. 36 Eva ns DJ, Levene MJ. Evidence of selection bias in preterm surv ival studies: a syste matic review. Arch Dis Child Feta l 2001; 84:F79 - 84. 37 Huxley AK, Froede RC, Birkby WH. Strangulation of pregnant woman leads to one first-degree murder indictment for the death of the mother: A medicol ega l reconsideration of maternal/fetal/homicide. Am J Forens ic IvIed Pathol 2001; 22:51-4. 38 Singer DB , Sung C-R, Wigglesworth JS. Fetal growth an d maturation ; with stan dards for body and organ development. In Wigglesworth JS, Singer DB (eds) Textbook oj Fetal and
196 I
Fetal and perinatal death
Perinatal Patho logy, 2nd edn. Malden Ma: Blackwell Science, 39
40
41 42
43
44
45
46
47
48
49 50
51
52
1998, pp. 8-40. Keeling JW. The perinatal necropsy. In Keeling JW, Khong TY (eds) Fetal and Neonaral Pathology, 4th edn. London: Springer, 2007, pp. 20-53. Huxley AK, Kosa F. Estimation of lunar age from diaphyseal length from the fetal materials in the fetal collections at Albert Szent-Gyorgyi Medical Uni ve rsity, Hun gary (abst). Proc Am Acad Forensic Sci 1999 ; 5:222-3. Dorovini-Zis K, Dolman CL. Gestational developm ent of brain. Arch Pathol Lab Med 1977; 101:192-5. Feess-Higgins A, Lan'oche J-C. Del!eiopmel1t oj rhe Humlln Fetal Brain: An Anatomical Atlas. Paris: INSERM CNRS, Masson, 198 7. Foote GA, Wilson AJ, Stewart JH. Perinatal post-mortem radiography - experience with 2500 cases. Br] Radiol 1978; 51 :351-6. Russ ell JGB. Radiological assessment of age, retardation and death. In Barson AJ (ed.) Laboratory In/iestigation oj Feral Disease. Bristol: Wright, 1981, pp. 3-16. Gould SJ. The respiratory system. In Keeling JW, Khong TY (eds) Feral and Neonatal Pathology, 4th edn. London : Springer, 2007, pp. 531-70. Baak JPA, Oort J. A Manual oj Morphometry in Diagnosric Pathology. Berlin: Springer-Verlag; Heidelberg, 1983, pp. 131-6. Hilder L, Costeloe K, Thilaganathan B. Prolonged pregnancy: evaluating gestation -speci fi c risks of fetal and infant mortality. Br] Obstet GYl1aeco/1998; 105:169-73. Keeling JW. Intrapartum problems. In Keeling JW, Khong TY (eds) Fetal and Neonatal Patho logy, 4th edn. London: Springer, 2007, pp. 273-96. Saunders E. Neo nat icid es following 'secret' pregnancies: seven case reports. Public Health Rep 1989; 104:368-72. Mendlowicz MV, Jean-Louis G, Gekker M, Rapaport MH. Neonaticide in the city of Rio de Janeiro: forensic and psycholegal perspective.] Forensic Sci 1999; 44:741-5. de Vries LS, Rennie JM. Preterm brain injury. In Rennie JM, Roberton NRC (eds) Textbook oj Neonatology, 3 rd edn. London: Churchill Li vingsto ne, 1999, pp. 1252-7 \. Connolly AM, Katz VL, Bash KL et al. Trauma and pregnancy.
Am] Perillatol1997; 14: 33 1-6. 53 Fort AT, Harlin RS. Pregnancy o utco me after noncatastrophic maternal trauma during pregnancy. Obstet GYl1ecol 1970; 35:912-15. 54 Rothenberger D, Quattlebaum FW, Perry JF et al. Blunt maternal trauma: a revi ew of 103 cases.] Trauma 1978; 18:173-8. 55 Shah KH, Simons RK, Holbrook T et al. Trauma in pregnancy: maternal and fetal outcomes.] Trauma 1998; 45:83-6. 56 Curet MJ, Schermer CR, Demarest GB et al. Predictors of outcome in trauma during pregnancy: identification of patients who can be monitored for less than 6 hours. ] Trauma 2000; 49: 18-24. 57 Theodorou DA, Velmahos GC, Souter I et al. Fetal death after trauma in pregnancy. Am Surg 2000; 66:809- 12. 58 Good w in TM, Breen MT. Pregnancy outcome a nd fetomaternal haemorrhage after non catastroph ic trauma . Am ] Obster Gynrcol 1990 ; 162 :665-71. 59 Siebenga J, van del' Schoot JT, Keeman IN. Trauma during pregnancy. Ned Tijdschr Geneesk.d 1999; 143:1144-8 . 60 Runnebaum IB, Holcberg G, Katz M. Pregnancy outcome after repeated blunt abdominal trauma. Eur] Obstet GYllecol Reprod Bioi 1998; 80:85-6. 61 Pearlman MD, Fintinalli JE, Lorenz RP. Blunt trauma during pregnancy. N Eng/ ] Med 1990; 323: 1609 - 13.
62 Poole GV, Martin IN Jr, Perry KG Jr, Griswold JA, Lambert CJ, Rhodes RS (1996) Trauma in pregnancy: the role of interpersonal violence. Am] Obstet Gynecol 174:1873-7. 63 Weiss HB. Songer TJ, Fabio A. Fetal deaths related to maternal injury. ]ANlA 2001: 286: 1863-8. 64 Sugrue M, Kolkman KA. Trauma during pregnancy. Aust ] Rural Health 1999; 7:82-4. 65 Rubovits FE. Traumatic rupture of the pregnam merus from 'sea t belt' injury. Am ] Obstet Gynecol 1964; 90:828-9. 66 Raney EH. Fetal death secondary to nonpenetrating trauma to the gravid uterus. Am] Obstet GynecolJ970; 106:313-14. 67 Crosby WM, Costiloe JP. Safety of lap-belt restraint for pregnant victims of automobile collisions. N Engl ] Mer! 1971; 284:632-6. 68 Pearce M. Seatbelts in pregnancy: above and below the bump, not over it. BM] 1992; 304:586-7. 69 Pearlman MD, Klinich KD, Schneider LW et al. A comprehensive program to improve safety for pregnant women and fetuses in motor vehicle crashes: a preliminary report. Am] Obstet Gynecol2000; 182:1554-64. 70 Williams JK, McClain L, Rosemurgy AS, Co lorado NM. Evaluation of blunt abdominal trauma in th e third trimester of pregnancy: m a ternal and feta l considerat ion s. Obstet
Gyneco/1990; 75:33-7. 71
72
73
74
75
76
77 78
79
80
81 82
Matthews CD. Incorrectly used seatbelt associa ted w ith uterine rupture follo wi ng ve hicular co lli s ion. Am ] Obstet Gynecol 1975 ; 121:1115-\6. Jud ich A, Kuriansky J, Engel.berg J et aJ. Amniotic fluid embolism following blunt abdominal trauma in pregnancy. Injury 199 8; 29:475-7. Covington DL, Diehl SJ. Wright BD, Piner M. Assessing for violence during pregnancy using a systematic approach. Matern Child Health] 1997; 1:129-33. Cokkinides VI, Coker AL, Sanderson M et al. Physical violence during pregnancy: maternal comp li cations and birth outcomes. Obstet GYl1eco/ 1999; 93:661-6. Rachana C. Suralya K, Hisham AS et al. Prevalence and comp li cations of phys ica l vio lence during pregnancy. Eur] Obstrt Gynecol Reprod Bioi 2002; 103:26- 9. Depal1ment of Health. Confidential en quiri es of maternal deaths in the UK. London: Depal1ment of Hea lth , 2001 (http//www.doh.gov.uk/cmo/mdeaths.htm). Drjfe J. Why mothers die. ] Roy Coil Physicians Edin 2005; 35:332 - 6. Lipsky S, Holt VL, Easterling TR, Critchlow Cw. Impact of police-reported intimate partner violence during pregnancy on birth outcome. Obstet Gynrcol2003; 102:557-64. Boy A, Salihu HM. Intimate pal1ner violence a nd birth outcomes: a systematic review. Int ] Ferti! Wom ens Mer! 2004; 49: 159-64. Soka l MM, Katz M, Lell ME, Fox A. Neonatal sLl lvival after traumatic fetal subdural hematoma.] Reprod Med 1980; 24:131-3. Stephens RP, Richardson AC, Lew in JS. Bilatera l subdu ra l hematomas in a newborn infant. Pediatrics 1997; 99:619-21. Pak LL, Reece EA, Chan L. Is ad verse preg nancy outcome predictable after blum abdominal trauma 7 Am ] Obstet
Gyl/eco/1998; 179:1 140-4. 83 Becroft DMO, Gunn TR. Prenatal cranial haemorrh ages in 47 Pacific Islander infants: is traditional massage the cause? NZ iVIed] 1989; 102:207-10. 84 Lieberman JR, Mazor M, Molcho J et al. Electri ca l acc iden ts during pregnancy. Obstet Gynecol 1986; 67:861-3. 85 Ma zor M, Lieberma n JR. Abortion caused by electric curren t. Arch GYllecol Obstet 1987; 241 :71-2. 86 Jaffe R, Fejgin M, Ben Aderet N. Fetal death in early pregnancy due to electric current. Acta Obstet GYll ecol SClmd 1986; 65:283.
References I
87 88 89 90
91 92 93 94 95
96
97
98
99
100
101 102
103
104 105
Mehl LE. Electrica l inju ry from tasering and miscarriage. Acta Obstet Gynecol Swnd 1992; 71 : 118-23. Yoo ng AF. Electrical shock susta ined in pregna ncy fo llowed by placenta abruption. Postgrad Med ) 1990; 66:563 - 4. Fatovich DM. Electric shock in pregnancy. ) Emerg Med 1993; 11:175-7. Einarson A, Bailey B, ln oce ncion G et al. Accid enta l electric shock in pregnancy: a prospective cohan study. Am) Obstet Gyn ecoi l997; 176:678-81. Rees WD. Preg nant woman struck by lightning. BM) 1965; 1: 103-4. Chan Y-F, Sivasamboo R. Lightn ing accidents in pregnancy. ) Obstet Gynaecol Br Commonw 19 72; 79:761 -2. Guha-Ray DK. Feta l dea th at term due to li ghtnin g. Am) Obstet Gynecol 1979; 134: 103-5. Fl a nnery DB, Wiles H. Follow- up of a survivo r of intra -uterine light ning exposure. Am) Obstet GYl1eco11982 ; 142:238 - 9. Pierce MR , He nde rson RA , Mitchell JM. Cardiopulmonary arrest secondary to lightning inj ury in a pregnant woman. Ann Emerg Med 1986; 15: 597-9. Awwad IT, Azar GB , Seoud MA et a1. High-velocity penetrating wounds of the gravid uterus: review of 16 yea rs of civil war. Obstet GYl1eco l 1994; 83 :259-64. O'Shaughn essy MJ. Conservat ive obstetric management of a gunshot wound to the seco nd-trimester gravid uterus: a case repon. J Rep rod Med 1997; 42:606-8. Piers on R, Miha lovits H, Thomas L, Beatty R. Penet rating abdominal wou nd s in pregnancy. Ann Ell1erg iVIed 1986; 15: 1232 - 4. Grubb OK. Nonsurgica l management of pe ne trating ute rine trauma in preg nancy: a case report. Am ) Obstet Gyn ecol 199 2; 166 :583-4. Franger AL, Buchsbaum HJ, Peaceman AM. Abdomina l gunshot wo und s in pregnancy. Am ) Obstet Gynecoi 1989; 160: 1124-8. Lavin JP, Polsky SS. Abdom in al tra um a during pregnancy. Ciin Perinotol1983; 10: 42 3- 38 . Buchsbaum HJ, Staples PJ. Self-infli cted gunshot wound to the pregnancy uteru s: report of two cases. Obstet Gyneco/ 1985; 65:32S. Sakala EP, Kort DD. Management of stab wounds to the preg nant uterus: a case report and a review of the literature. Obstet Gynecol Sum 1988; 43:319-24. Rose PG , Strohn PL, Zus pan FP. Feto maternal haemorrhage follo wing trauma. Am) Obsrer Gynecol 1985; 153:844- 7. Reis PM, Sander CM, Pearlman MD. Ab rup tio place ntae after auto accidents. A case-control stu dy. ) Reprod Med 2000; 45:6-10.
197
106 Rogers FB, Rozycki GS, Osier TM et al. A multi-in st itutio nal study of facto rs associated with feta l death in injured pregnant pati ents. Arch Surg 1999; 134:1274-7. 107 Drost TF, Ros emurgy AS, Sherman HF et al. Major trauma in pregnant women: materna l/fetal outcome. J Trauma 1990; 30:574-8. 108 Fries MH, Hankin s GOY. Motor vehicle accident associa ted with minimal maternal tra uma but subsequent fetal demise. AIIII Em Med 1989: 18:30 /- 4. 109 Evrard JR, Stmner WQ , Murray FJ. Fetal skull fracture from an automobile accident. Am ) Forensic Med Patho l 1989; 10:232-4. 110 Hartl R, Ko K. In utero skull fra cture: a case report. ) Trauma 1996 ; 41 :549- 52. III GUIlIl TR , Mo ra JD , Becroft DM. Co nge nital hydrocephalus seco ndary to pre nata l intracrania l hae morrhage. Aust NZ ) Obstet Gyl/aecol 1988; 28: 197 - 200. 11 2 Robinson MJ, Cameron MD, Smith MF, Ayres AB. Fetal subdural haemorrh ages presenting as hydrocephalus. Br lVIed J 1980; 281:35. ]] 3 De Sousa CD, Clark T, Bradshaw A. Ante natally diagnosed subdural hemorrhage in congen ita l factor- X deficiency. Arch Dis Child 1988; 63:1168 - 70. 114 Shere r DM, Anyaegbunam A, Onyeije C. Antepartum fetal intracranial hemorrhage, predisposing factors and prenatal sonography: a review. Am ) Perinatol 1998; 15:431-41. 11 5 Weinberg L, Wyatt P, Busuttil A. Tra umatic intrauterine fetal sp in al fracture follo wing seatbelt use: a case repon. J Trauma 200 1; 51 : 1195- 16. 1/ 6 Fakhoury GW, Gibson JRM. Seatbelt ha za rd s in pregnancy: a case report. Br J Obs tet Gynaeco/ 1986; 93:395-6. 11 7 Siddall-Allum IN , Hughes JH, Kal er S, Reginald pw. Splenic rupture ill utero following a ro ad traffic acc ident. A case report. Br ) Obstet Gynaecol 199 1; 98:3 18-1 9. 11 8 Litmanovitz I, Dolfin T, Arnon S et al. Feta l intrathoracic injuri es following mild maternal motor vehicle acc id ent. J Perill at ivIed 2000; 28: 158 - 60. 119 Griffiths M, Hillman G, Usherwood MM. Seat belt injUly in pregna ncy resulting in fetal death. A need for ed ucation? Case reports. Br J Obstet Gynaeco/ 199 1; 98:320- 4. 120 Yates PO. Birth tra uma to the verte bra l arte ri es. Arch Dis Child 1959; 34 :43 6-41. 121 Ito Y, Ts ud a R, Kimura H. Diagnostic va lue of the placenta in med ico-legal practice. Forensic Sci lilt 1989; 40:79 - 84.
I
CHAPTER 11
I
SUDDEN UNEXPECTED DEATH IN INFANCY: SUDDEN INFANT DEATH SYNDROME OR SOMETHING ELSE? Jean W Keeling
Introduction The definition of SIDS Epidemiology Sleeping environment
198 199 201 203
INTRODUCTION
Sudden, unexpected deaths are a large and important group of deaths in the post-neonatal infant age group (1 month to 1 year). These deaths require a consistent high standard of investigation, which goes well beyond anatomic dissection and microscopic examination of tissue samples from a few major organs. Among the deaths pre senting to the pathologist as sudden and unexpected will be those that are fully explained by a recognized disease process (see Chapter 12), some that are non-natural, some explained by minor or long-standing disease and others that occur in the absence of any pathological abnormality. Cases in the last three groups are often problematic and require the fullest, multidisciplinary investigation. The term 'sudden infant death syndrome' (SIDS) was devised to be used in death certificates in cases of (sudden) infant death in which no disease was identified at post mortem examination and death was deemed natural. It was convenient for pathologists - no need to ascribe the death to tenuous 'respiratory tract infection' or 'gastroenteritis' - and it avoided the use of terms such as 'unasceI1ained', which necessitated fulwer inquiries and an inquest or, worse, 'aspir ation of vomitus', which not only provoked an inquest (non-natural death) but implied poor infant care practices by parents or carers.
Pathological findings in SUDI Death certification References
205 218 219
In the UK, SIDS became an option as a registrable cause of death in 1971, and the number of registrations increased over the following decade as the option was taken up by more pathologists. The International Classification of Dis ease separated 'SIDS' (789.0) from 'Sudden Death - cause unknown' (795.0) in 1979; the latter term was included for the first time in the previous revision in 1968. Whilst the term 'cot death' has been used since the early 1950s and was refined in 1965 to encompass only unex plained deaths, I it was not until the mid-1970s that the general acceptance of Beckwith's definition 'The sudden death of any infant or young child, which is unexpected by history and in which a thorough post-mortem examination fails to demonstrate an adequate cause of death',2 enabled meaningful comparisons between different studies. Clearly, there are ambiguities that beset this definition: How sud den? Unexpected by whom? Which historical events exclude deaths from the SIDS definition? What constitutes a 'thorough' necropsy examination? What type or extent of pathological abnormality comprises an adequate cause of death? Despite these reservations, this definition remains the focus of our approach to the problem. However, the importance of the immediate circumstances of death has become increasingly appreciated. J Berry et al 4 found a detailed history, particularly of the precise circumstances of death, to be a most imp0!1ant pal1 of the evaluation of
The definition of SIDS I
sudden unexpected death in infancy (SUD!). The formal consideration of the circumstances of death as part of the definition of SIDS has been proposed,5 but not widely implemented. The greatest benefit of the SIDS designation has been that it has permitted much better estimates to be made of the size of the problem than has hitherto been pos sible. The number of cases has been very much greater than had been supposed initially and constitutes an important public health problem. There are, however, disadvantages in the use of the term SIDS. One has been the development of an assumption that this 'syndrome' has a single aetiology - 'died of SIDS' is an unfortunate phrase seen in many publications. Another has been a tendency among some in child care to stop thinking about the problem because it is an accepted, registrable designation. The term has been accepted uncritically and in any particular case, until recently,6 few attempts were made to unravel any predisposing factors. A fUl1her prob lem is the unwillingness of some pathologists to investigate cases fully 'because there is nothing to find'! This is a most unfortunate view as some of the specific causes of sudden death, which may be missed by slipshod investigation, are relevant to the medical management of other family mem bers or of siblings yet unborn. Of equal concern is that some of these poorly investigated deaths may be the result of homicide. Bajanowski et al 7 found 5 per cent of non natural deaths among 339 externally normal victims of sudden infant death. Non-natural deaths accounted for 7.9 per cent of all deaths in the Confidential Enquiry into Still births and Deaths in Infancy Sudden Unexpected Deaths in Infancy (CESDI SUDI) studies. 6 Gilbert-Barness and Barness 8 suggested that the use of SIDS be repJaced by 'sudden infant death' with case-specific qualification. Green 9 has argued for the abandonment of the terms 'cot death' and 'SIDS', and for wider use of 'not ascer tained' on the death certificate. He points out that with the considerable reduction in infant deaths following the suc cessful 'Back to Sleep' campaigns, the proportion of non natural deaths is likely to be higher among sudden infant deaths at the end of the 1990s than it was 10 years earlier. Meadow lO supports this view and considers that unexpected infant deaths require a multidisciplinary investigation of high standard and that the term SlDS should be more dosely defined or else abandoned. The opposite view is put by Limerick II who argues that the number of non-natural deaths is small, and that there is a need for a dear category of sudden, unexplained natural deaths. Her statement that a case review following post-mortem examination would avoid any need for the pathologist to use the term 'cause unascertained' is, in my view, incorrect. SUDI can fall outside the SlDS definition for reasons related to the baby's family history, past medical history, circumstances around the time of death, or because of the results of post-mortem investigations carried OUt. 12 Some of the inconsistency of terminology can be related to the provision of inadequate information to the pathologist prior
199
to the commencement of post-mortem examination,13 although evaluation of the effect of bed sharing poses prob lems for individual pathologists. Fleming et al 14 emphasize the role of the multidisciplinary team in the overall evalu ation of any SUDI, using a non-hierarchical grid. Krous et al 15 have proposed a plan of subdivision of SUDI!SlDS with precise definitions. They acknowledge the problems of evaluation of cases after prolonged resuscita tion or incompleteness of investigations, both of which introduce serious problems for reviewers. This chapter will consider those questions arising out of Beckwith's definition of SIDS. It will explore family back ground and discuss why prior events and some patho-logical abnormalities require a different, sometimes more specific, designation. It will also describe pathological criteria which supp0l1 the SIDS categorization and discuss concerning fea tures in presentation or history and asses the appropriateness and significance of pathological findings.
THE DEFINITION OF SIDS How Sudden? A very tiny minority of SUDI occur whilst the infant is being observed. it is, however, an important group, as a high proportion will be explained. This important fact must be conveyed to the pathologist as additional investigations are necessary. The three likely groups of underlying path ology which must be carefully considered are trauma, acci dental and non-accidental, cardiac pathology and upper respiratory tract obstruction. A larger, but still relatively small, group of babies are those who collapse or die suddenly during the course of a recognized acute illness, but before a specific diagnosis has been reached. In this group, an explanation for death is likely to be found if an adequate examination, induding microbiology and investigation for possible genetic meta bolic disease, is carried out. The majority of infants dying unexpectedly are found dead in their usual place of sleep when the rest of the household wakes. The CESDI SUDI studies 6 found that 64 per cent of deaths were discovered betvveen 05.30 and 11.30 hours. A further 17 per cent died during a daytime sleep. Few deaths in that study occurred between 18.00 and 24.00 hours. Time of death should not be an excuse for dis missing the death as natural and for incomplete investiga tion of the death. Important, treatable pathology is found in some infants in this group, and it is important that both the parents and family practitioners are made aware of such findings. A group of deaths that do not fit easily within the defini tion of sudden are those in which the infants, found col lapsed, are maintained on life support systems for more than 12 hours. These should, in my view, be investigated with the same vigour. Clinical management can result in findings at
200 I
Sudden unexpected death in infancy
necropsy that are difficult to interpret and a full interven tional history is essential. Discussion with the medical staff involved may assist evaluation of findings (see Chapter 19). Clearly, there is no necessity to repeat those investigations done in hospital, but care should be taken that the full range of investigations is completed.
How Old? National statistics define the infant period as from birth to 1 year, but deaths in the first week of life are generally birth related and not sudden. SUDI studies, anxious not to miss cases, may set age limits widely, 1 week to 2 years in a UK multicentre study.IG However, most SUD! occur within a more restricted age range. Many consider 1 month to 1 year to be appropriate limits. Most deaths occur prior to the eighth month. Unexplained sudden deaths before 1 month of age are unusual. They are discussed fuliher on p. 209. Sud den neonatal (0-4 weeks of age) deaths accounted for only three per cent of deaths in a Scottish study. 17 Unexpected deaths over I year of age are also unusual (three per cent). Most sudden deaths among infants over 6 months of age are likely to be explained. Fleming et al 6 found the median age at death of SIDS to be 13 weeks within an interquartile range of 7 weeks, 6 days, to 21 weeks, 3 days. Aggregated data from several large studies derived the 5th and 95th percentiles as 3 weeks and 9 months, respectively. IS
Unexpected by Whom? An important part of the definition of SIDS is the lack of antecedent symptoms. This absence of symptoms requires critical evaluation and is a forceful argument for the involvement of a multidisciplinary team in the invest iga tion of SUD!. An experienced paediatrician or health visitor may unearth symptoms whose significance was not appre ciated by carers or police. This might be because of lack of experience of the carer or the non-specific or insidious nature of signs and symptoms. The apparent lack of symp toms may, in a few cases, be explain ed by the velY rapid course of the illness, particularly septicaemia or meningitis. The lack of specific signs of illness in the infant may be another contributory factor. In some cases, the history is deliberately misleading, to conceal either real or perceived inappropriate treatment of the infant. Even after thorough, sensitive and timely ques tioning, it is apparent that some infants exhibit no or only trivial symptoms. la A careful and thorough history is particularly important to the pathologist to enable him or her to include additional investigation pertinent to the particular case 4 and in the overall evaluation of the death. It is not infrequent for families to give a history of a recent mild upper respiratory tract infection or of immunization.
Which Historical Events are Significant? The most significant points to elucidate from the history are previous infant deaths (explained or otherwise) in the immediate or extended family, previous admission of the infant to hospital and any documentation of prior injury to the deceased or to siblings. These and other family events are considered in more detail later in the chapter.
What is a Thorough Necropsy Examination? The purpose of necropsy examination in SUDI is to distin guish between natural and non-natural death, to elucidate a cause of death if one is present and to further the under standing of sudden death in the infant period. Published guidelines include those of Wigglesworth et al,J9 the Royal College of Pathologists,20 ValMs-Dapena et al,21 Krous 22 and Howatson.23 It is impOltant that SUDI are investigated to a consistently high standard. This includes the completion of investigations even in those cases where major pathological abnormality is discovered early in the investigation. How else is the pathologist to understand the significance of his findings in other cases? The need for consistent and thor ough investigation of the family background as well as the post-mortem examination itself were emphasized by an intercollegiate working group24 and are supported by Landi et al, 25 who found that more and appropriate investigations were done in cases certified as SIDS than those considered explained. Centres unable to offer full radiological and microbiological support as well as a high technical standard for processing histological samples, including neuropatho logical samples, should not investigate these deaths. Sadler 26 emphasizes the importance of early and thorough microbio logical investigations. The relative contributions of different components of the investigation of SUDr are evaluated by Berry et al 4 and Mitchell et alY Toxicological investigations have to be justified in many jurisdictions because of a low return of positive find ings. Langlois et al 28 detected drugs in 16 per cent of SUD! investigations with drug administration an important, but unsuspected, factor in the death in 2.6 per cent. Arnestad et al 29 found that toxicological investigations made a positive contribution to the evaluation of four per cent of SUD!.
What is an Adequate Cause of Death? The interpretation of the pathological findings in respect of adequacy of a cause of death is frequently difficult in SUD!. The problem does exist in relation to deaths at other ages, but because of the prevalence of degenerative diseases and the relative infrequency of tmly unexplained sudden death in older children and aduIts 30 it is less discussed. Over emphasis of minor pathological findings, particularly in
Epidemiology I
poorly investigated cases, is a particular problem in this age group.4,]I,]2 The significance, or otherwise, of partiClI lar pathological findings is discussed later in the chapter.
Multidisciplinary Case Review A case review following a SUDI is the usual practice in many areas]] and is rapidly becoming the norm in many countries. Review groups involved the family doctor and health visitor (with midwifery input, if appropriate), a researcher, when formal parental interviewing is normal practice, a patholo gist and a paediatrician. Its purpose is to examine both risk factors and the actions of parents or health professionals, which might usefully be modified either in respect of a par ticular family or in professional practice. Review of the pathology findings encourages consistent investigation and goes some way towards introducing uni formity of interpretation of pathological abnormality. It forms the basis of discussion with each family. It is the ideal venue to develop appropriate local strategies to sup port bereaved families. 24
EPIDEMIOLOGY Epidemiology of SUDI and SIDS It is through its epidemiological features that SIOS is best defined, but the interrelationships of these are complex. Investigation of these complexities demands that study methodology is clearly set out and the objective well defined 34 Many factors are as relevant today as they were in the earliest studies, but there have been some changes over the past five decades. One factor that has resulted in changes in the epidemiology of SIOS is better case defin ition, as efforts are made to reach a consensus about what constitutes an adequate explanation for death. However, there is no doubt that changes have occurred; seasonality (i.e. a higher incidence in winter months) is currently much less marked than two decades ago, whilst maternal smok ing exerts a stronger effect. 35 - J7 Most cases occur between 6 and 20 weeks of age with a median at around 12 weeks. There is a male predominance and a continued correlation with higher birth order and low maternal age. The associ ation with low maternal age and birth order has become stronger in some studies. 38 Low birth weight and pre-term delivelY in the index and previous pregnancies 39 and twin pregnancy are recognized risk factors, as is maternal smok ing during pregnancy.37 Blair et al 37 found that pre-term delivery among victims of SIOS increased from 12 per cent (Q 34 per cent over a 20-year period. Malloy40 looked at SIOS in pre-term infants and concluded that the association was the result of socioeconomic rather than medical characteristics. Mothers of SIDS infants have more
201
miscarriages and still bilths 41 and shorter interpregnancy intervals than control subjects. 42 Other pregnancy-related problems such as urinary tract infection and anaemia are weaker associ-ations. An elevated maternal serum alphafe toprotein level in the second trimester of pregnancy is associated with a higher risk of SIOS .43 Maternal recre ational drug use is also a risk. 6,17 In Bristol, UK, the propor tion of SIOS victims from deprived socioeconomic backgrounds rose from 47 per cent to 74 per cent over a 20-year period. 37 Simultaneous SUDI in twins is most unlikely to be SlOS. It is an unusual event, but likely to have an environ mental cause. A detailed death scene investigation is mandatory. 44 Sequential SUD! in twins is likely to result fro m infection. SIOS victims have lower Apgar scores at birth than con trol infants and a re more likely to be admitted to a special care baby unit. They are also more likely to have dysmor phic features, malformations and positional deformities. Low socioeconomic class, lower parental education level and poor housing are consistent findings in the families of SIDS infants. The increase in unemployment and of cohab itation makes it difficult to compare CUlTent and earlier stud ies, but the CESDI SUDr studies clearly demonstrate that, irrespective of the parameters used - occupation, family income, receipt of income support or employment status an excess of SIOS deaths is seen in families with socio economic problems. 17
Predicting SIDS - and What Else? With the recognition of pregnancy and perinatal factors asso ciated with SIDS, scoring systems have been devised with the aim of identifying those infants at high risk of SIDS with a view to intervening, in the forms of family support, education and surveillance aimed at prevention of SIOS.45,46 Much resource/research has been put into a variety of schemes based on sCOling systems or targeted at families where SIDS had occurred, with only modest return. A much more suc cessful intervention, widely applied, has been the 'Back to Sleep' campaign. SIOS scoring systems are not specific but identify babies with other problems such as respiratory infection, the need for hospital admission and non-accidental injury46 and explained infant deathY Some SIOS risk factors are also predictors of infant homicide.48
Recent Medical History in SUDI and SIDS Interrogation of parents following sudden infant death fre quently elicits a history of symptoms of illness in the baby. It is important to remember that these parents are likely to be searching for something to which they can attribute blame for the death, or which might have provoked action
202 I
Sudden unexpected death in infancy
on their part and, perhaps, have averted death. It is also important to remember that young infants often suffer a succession of minor ailments. This background of frequent, minor morbidity must be taken into account when conside ring the relevance of reported symptomatology. Stanton et al 49 found that snuffles and colds were reported frequently by both SUD! and control groups. However, they found that more serious respiratory symp toms, such as cough or wheeze, were more frequent in the SUD! group. Infants who died were also more likely to have exhibited, in the day or so prior to death, unusual drowsi ness, irritability or excessive crying. They were also more likely to be off feeds, pyrexial or sweaty. Almost 60 per cent of SUD! victims had symptoms in the 48 hours before death , half of which were considered major. More babies who died than control babies had seen a doctor both within 24 hours of death and earlier in the course of perceived illness. Half of the infants with system-related symptoms had evidence of significant terminal disease at necropsy whilst only one quarter of those with non-specific symptoms had signifi cant pathology. 16 In that study, about one-t1Fth of victims of unexpected deaths (no recognized symptoms) had evidence of serious acquired disease at necropsy. In another study, babies dying unexpectedly in the infant period were more likely to have had minor signs of illness in the previous seven days than control infants and to have been seen by their general practitioner. 18 This study found no difference between babies who died and control infants with respect to major signs of illness. Stanton and OakJeyS° found that SUD! victims were more likely to have been admitted to hospital previously and required longer admis sion than control (surviving) babies. Although eight infants in their study were admitted to hospital because of suspicion of abuse, no babies had unexplained apnoea, but 9 of 71 suf fered loss of consciousness - repeated in three cases. In New Zealand, SIDS babies were found to have more severe illness resulting in more GP or hospital contact than control infants.51 The clinical signs that prompted parental action were not specified. A more recent study47 found that the clinical characteristics of explained SUD! and SIDS vic tims were similar. Apparent life-threatening events (ALTEs), defined as an episode in which the infant became apparently lifeless, were reported by parents of both groups (9/61 , 14.8 per cent, and 37/317. 11.7 per cent) and were more frequent than among control subjects. However. only two-thirds of these episodes were brought to medical atten tion and less than one-halfwere seen in hospital. Infants in both groups were more likely to have been unwell than control infants in the 24 hours before death. Signs seen more frequently in the SIDS group were drowsiness. wheezing and reduced fluid intake in the 24 hours prior to death. Neonatal problems and hospital admissions were much more common in the SUD! group than among SIDS babies. In the Netherlands, signs of iJJness were more frequent in explained SUDI than SIDS or borderline SIDS. 52
Table 11.1
Factors in the history causing concern in SUDt
Many GP/accident and emergency attendances without clear evidence of illness Apparent life-threatening events (ALTEsl. especially if more than one and starting at less than 4 weeks of age and if unwitnessed or if the same person finds the baby each time Age > 1 year Epistaxis Seizures - especially unexplained Death occurs during the day when the baby was apparently well earlier ALTE in siblings Previous SUDI in sibling Death on anniversary of previous SUDI Parental involvement with the media after child's death
Adverse Historical Factors (Table 11.1) The occurrence of a previous infant death in the family or in the extended family should alert the pathologist to the possibility of an inherited disorder. which might not have been fully investigated in the earlier death, as well as to the possibility of non-natural death 10.53 More than half of Meadow's non-natural deaths were initially certified as SIDS deaths although the thoroughness of investigation is not set out. A detailed background investigation is required, and it is helpful to retrieve the necropsy and police reports on the previous death prior to commencing examination of the second as well as accessing any clinical notes. A history of unexplained collapse with pallor or cyanosis and apnoea is, for me, a cause for concern in SUDI babies. Concern increases when acute AlTEs have occurred on more than one occasion 10,54 and when episodes commence after the neonatal period. 55 A further cause for anxiety is when investigation of these episodes in a hospital setting fails to find an underlying problem, particularly if ALTEs cease in hospital. 54 .56 It is important that the circumstances under which the episodes occurred are fully investigated. Ariagno et al 57 found that AlTEs in pre-term infants occurred earlier and were more likely to be witnessed by professionals (54 per cent) than those in term infants (23 per cent). Similar episodes described in siblings is a further concem. 53- 55 Byard and Krous 58 list potential natural causes of AlTEs and highlight the need to consider imposed suffocation as a basis for such events. In the past, such events have even been called ' near miss cot deaths', but currently many are thought to be the result of imposed upper airways obstruction. 10,55.59,60 Such a history is most unusual in what, both epidemio logically and pathologically, is a typical SIDS death. Having been somewhat sceptical about the frequency of imposed airways obstruction as the cause of ALTEs, per sonal experience of repeated infant death with or without AlTEs has changed my view. I wonder whether the history
~
-_.
Sleeping enviro nment I
of even a single unexplained ALTE should p reclud e use of the SIDS designation. Unless death was clea rly explained, fulfilling a sp ecific diagnosis, then, in these circumsta nces, death sho uld be considered unascertained. Another unusual symptom in the infant perio d is epi staxis, defi ned as bl eeding from both nostrils.55 This should not be confused with blood-streaked mucus from one nos tril, usually the resul t of self-inflicted injury. Epistax is is particularly concerning when accompanied by choking and/or tra nsient change in breathing patte rn as it raises the possibili ty of imposed ainvays obstruction. 58 ,61 The pres ence of blood-tinged mucu s at the nostrils in a su dden baby death is to be treated suspiciously. A hi story of seizures is an indication fo r particularly careful central nervous system (CNS) exa minati on, which can be done only fo llowing fixation of the brain. Detailed information about pregnancy and the neonatal peliod is essentia l when seizures have been reported. Howeve r, the possibili ty of a non-natural process should not be forgo tten, Seizures (due to induced cereb ral hypoxemia) are probably the most common presentation of fabri cated illness 62 Necropsy examination should include careful h arves ting of sampl es for toxi co logical examination ,63 particul a rly urine, blood and vitreous fluid (only if the eyes do not need to be examined histo logica lly). Unusual findin gs amo ng parents of babies who had been smothered were som atizing diso rders, involvement with arson or medical litigation and actively seeking media publicity in relation to their infant's death.1O
SLEEPING ENVIRONMENT Infan t care practices are culture dependent a nd have evolved gra du a ll y over very ma ny years as a result of fam ily experience and loca l contacts. In recent years, opinions from many sources have been wid ely dissemin ated; fash ion, fads and consumerism are pressures on you ng, often inexpe rienced, mothers who compete with advice from health professionals.
Good practice The reco mmend ed sleeping environment for young infants is a cot with firm mattress, with no lise of soft pillows or bumpers, Whilst a firm, permeable pillow may have had merit when used wi th an impermea ble mattress cover when many infants were put down prone,64 it is less useful when most babies sleep supine. Preferred cot bedding is a sheet wit h lightweight blan ket(s), which can be tucked in at the sides, rather than duvets or sleeping pods. It is recommen ded that the enviro nment be neither too hot or cold (at about 16-18°C) and well ven tilated. Pl acing a sm all infant at the end of t he cot (foot to foot) rather than in the middle, to avoid burrow ing under covers and overheating is current advice.65
203
Placing the infant in the 'supine' sleep ing position 66 and avoidance of prone and, now, 'side sleeping'65 is advised, Avoidance of infa nt restrai nts design ed to maintain the initial sleep ing position is current best practice. More details abo ut safe sleeping and appropriate avoidance of co-sleeping have been formulated. 67 Co-sleeping is con si dered further on p. 204. Encouragement of bed sha ring app lies only to non smoking parents who refrain from alcohol and social drugs. Excessive tiredness and use of prescribed sedatives or tranquillizers a re also contraindications to bed sha ring as they may result in overlaying of the baby. Sleeping in a cot in the parental bedroom reduces the risk of SIOS. 68
Suboptimal sleeping environments Stanton 69 drew atte ntion to overheating of sleeping infants as a lisk factor in cot death. He found a number of contribu tory fact ors likely to promote overheating. These includ ed maternal atti tude, especially duri ng co ld weather, use of daytime, outdoor clothing in bed, swa ddling in blankets, use of folded adult bed blankets on cots, use of bonnets or hoods and a hot environment, for example central heati ng on a ll night. In some cases, pyrexial illness in the infant also seemed to compound the problem, Excessive bedding was found to be an indep endent risk factor in south-west England. 70 Wigfield et al 71 fo und that a low outdoor tempera ture sti ll prompted mothers to increase infa nts' bedding when room tempe rature was adequate a decade after Stan ton's observations were made. The associatio n of prone sleeping position and cot death has been reported in the case-control studies of infant deaths since 1970.72 Reports of the association accu mulated with increasing importance of risk from this practice from several centres, as other risk facto rs for SlDS have changed?O,73-76 Further supp ort for the association has been the marked reduction in SlDS followi ng publicity campaigns to increase professional and public awa reness. Mitchell et al 77 consider that the practice of 'Back to Sleep' accou nts for about 50 per cent of the fa ll in SlDS deaths in New Zealand , Interactions between sleep ing position and bedding have been explored. Pron e sleeping and high tog values of bed coverings were independent risk facto rs in a UK Study.70 The use of quilts was a risk factor for SlDS among supine or side sleepers in Tasmania. 78 An additional adve rse factor for prone-sleeping infants may be the adoption of face straight down or face near-straight down positions. These positions have been shown to introduce an element of airways obstruction that makes death more likely in some infants,79 Further examinati on of the infants' sleep ing environ men t demon stra tes that an increased risk is related to side sleepin g pOSition and to head covering as an impediment to maintaining thermoneutrality.76,8o Living in a household where carers smoke is a risk factor for SIOS. This risk is related to the number of cigarettes smoked in the local
204 I
Sudden unexpected death in infancy
environment. 35 ,75,76 Sleeping in a room separate from par ents was a risk factor for SIDS in a New Zealand study.
Inappropriate sleep surfaces for infants The use of unstable sleep surfaces is unsuitable for small infants. The danger of placing infants prone on surfaces such as adult-size water beds was flagged up by Ramanathan et al. 81 They reported that 6 out of 34 SIDS cases investigated over a short period had been found face down on water beds. Bass 82 reported 13 infant deaths on water beds and drew attention to the water bed design (free flow, both original and modified designs). Eleven of these cases had been originaJly certified as SIDS. Deep-pile sheepskin rugs may present a similar, unacceptable hazard to young infants.83 In these situ ations, the infant is not able to exercise the head movement possible when prone on a firm surface. The danger to infants of the use of polystyrene-filled bean-bags as a resting or sleeping place is clearly recorded. 84 ,85 At least 35 deaths have taken place on these cushions, the majority being initially registered as SIDS. Babies were predominantly placed face down on the soft, deformable surface and found dead after an interval of 0.5-10 hours. Thirty-four of the 35 infants were below the median SIDS age of 13 weeks. Thirteen were only 1 month of age. A combination of factors is thought to be respon sible for these deaths, including rebreathing in a confined area, obstruction of nose and mouth, altered permeability of cushion coverings by wetting with secretions, hyper thermia and the instability of the bag. The use of 'V'-shaped pillows for young infants should also be avoided. 86 Movements during sleep can result in wedging in the pillow angle from which the infant cannot retreat and obstruction of the upper a irway can supervene.
Other unsafe sleeping environments Other unsafe sleeping environments are some types of swing-cradles87 and cots with ill-fitting mattresses,88-90 which may permit wedging of the baby between mattress and cot frame with impedance of respiratory movement. The continued use of broken cots also constitutes an unaccept able hazard for infants,91 who can become entrapped through gaps between broken or loose cot components. Parental modification of cots can also be problematic. Byard et a1 92 reported positional asphyxia in two infants whose parents had introduced an additional mattress into mesh sided cots as the original mattresses were thin. This had the effect of raising the sleep surface so that it abutted onto the mesh section of the side-walls. This area is much more read ily distensible by the infant and permitted fixation of the head between cot wall and mattress. Inappropriate cot design is an additional hazard. Bass 88 reported 10 asphyxial deaths seen over an I8-month period where cot design was
inappropriate for young infants, who were able to push their heads between cot slats or rails. Smialek et al 89 and Gilbert Bamess et aJ83 have described similar cases and a correspond ing problem when babies sleep on adult beds. Deaths in similar circumstances were still occurring 12 years later because families still used old or broken cotS.93 The use of bedside cots, with bars along only one long side of the cot, introduces similar potential for the infant to become wedged between the cot and the parental bed. 94 Bed rails, designed to stop young children from falling out of toddler or adult beds, pose a similar hazard for infants less than 1 year of age. 95 There is a danger of strangulation among older infants who can pull themselves up using cot sides when loose clothing or dummy cords catch on cot p rotru bera nces, 83,88,89,96 Byard et al 90 draw attention to the risk of asphyxial death when infants are left sleeping in car seats, baby bouncer or pushchairs (stroller-prams)97 with ill-fitting restraints, resulting in suspension of the infant. There is a danger of asphyxia when impermeable mattress coverings in cots are loose rather than being closely fitted. This applies even when babies are very young and consequently less mobile. Some hazardous sleeping environments are likely to be appreciated only when timely death scene investigations are conducted. 98 Excessive room heating and probable car bon monoxide or carbon dioxide poisoning were problems discovered by Bass et a[3 when thorough investigations were done. Leaving infants sleeping in poorly ventilated vehicles on sunny days poses a similar hazard. The possibility of accidental asphyxia, when mothers, who are very tired or sedated, have been breast-feeding in bed, has been explored by Bass et al 3 and Byard. 99 The for mer considered that 3 of 26 SIDS cases resulted from acci dental asphyxia in this situation. The latter found that 3 of 28 infants who died unexpectedly during a I-year period in south Australia were in bed with their mother, who had fallen asleep during breast-feeding. Although 'overlying' is no longer considered a major cause of SUDI, circumsta nces may dictate its serious consideration in some cases. 83
Co-sleeping Bed sharing with non-smoking parents is enco uraged by many health carers as a means of facilitating breast-feeding, despite a small increased risk of death among co-sleepers. 100 UNICEF 67 stresses the need for a clean, firm mattress, avoid ance of overheating from excessive clothing and hazardo us situations which might lead to wedging or falls, as well as advising against bed sharing during maternal illness. Concern is expressed by some about the increased propor tion of SUDI found after a co-sleeping night. Luke 101 drew attention to this problem when he found that an increasing proportion of SUDI within his jurisdiction occurred following a co-sleeping night. This was four times higher than usual
Pathological findings in SUDI I
practice 76 in a Scottish study. The increase in co-sleeping on the night of death might be a marker of minor illness in the baby, well recognized in SIDS victims, but the spectre of positional asphyxia cannot be dismissed. Unusual bed shar ing was also reported in another (small) study52 lt is most important that the sleeping environment is examined with great care when an infant is found unexpect edly dead in this situation. There are two potential problems, asphyxiation and overheating. As well as the precise pos ition of the infant - prone/supine - the level of the infant in relation to the parents and covers and the llse of pillows and bed coverings, particularly heavy duvets, should be explored. The state of the mattress and its coverings, particularly the presence of blood-stained secretions and evidence of sweat ing, as well as the relative weights of the parents need to be considered. This is in adclition to a clear history about drug/alcohol ingestion and levels of tiredness. Parental alco hol consumption and co-sleeping on the night that the baby died were frequent findings in two Scandinavian stud ies,I02,103 although this was not a factor in one American study.98 I find that lack of detail about many of these factors hampers evaluation of individual cases. Rushton 104 has expressed similar concerns and draws attention to the inevitable lack of important detail related to relative pos itions of parents and baby during sleeping hours. The loca tion of fixed lividity may be informative in some cases. Sleeping on a sofa with an adult is a particularly haz ardous situation. l05 In the CESDI SUDI studies, this factor contributed the greatest risk, with an odds ratio (OR) of 31 :25. It was confirmed in a Scottish study (OR 66.9, 95 per cent confidence interval [CI] 2.8-159.7) 106 and re-emphasized in a more recent study from BristoL 37 In my experience, shared by other pathologists, risk is enhanced when alcohol has been consumed by the carer. 83
PATHOLOGICAL FINDINGS IN SUDI As the number of unexplained, presumed natural deaths in the infant age group has fallen during the last decade, so the proportion of explained deaths in medicolegal autopsies might be expected to rise, Although the increased range and rapidity and improved accessibility of di,agnostic tests may be expected to enable a pre-mortem diagnosis to be reached in more babies. In a UK multicentre study, 1976-79, 18 per cent of victims of unexpected death had evidence of serious acquired disease,16 although this may not have explained death in every case. In a retrospective study by Czegledy Nagy et aI, 107 there was an explanation for SUDI in 20 per cent of cases. Nine per cent were non-natural deaths. Only nine per cent of SUDI in an Adelaide study were thought to be unexplained. 108 Subjective interpretation of pathological findings is, to some extent, inevitable. The CESDI SUDI investigations during 1993-96 emphasize the effect of sub jectivity in the interpretation of findings. In that study, only 17.8 per cent of the deaths were considered explained after
205
review of pathological findings by an expert panel, although death was considered explained by the examining pathologist in 30 per cent of cases. 4 Parham et al 109 found previously unsuspected pathological abnormality in more than half of babies less than 6 months of age in a review of sudden deaths investigated by the Medical Examiner's Office; death was considered to be explained in 35 per cent of cases. Only 39 per cent of SUDI were considered typical SIDS cases in a South Australian study, no although a further 21 per cent were deemed 'undetermined'. Knowledge of the circumstances of death may alter the pathologist's expect ation of positive pathological findings, but facilities for full investigation should be available at the outset. The findings following the sudden death of older infants and children are discussed in Chapter 12.
The Pathology of Explained Natural Death The range of pathological findings encountered in SUDI are shown in Table 11.2111 (see also Byard and Krous 58 ). Patho logical findings in SUDI are discussed in detail by Byard and Cohle, 112 There are no specific pathological markers for SIDS despite the best efforts of many investigators over the years. INSTANTANEOUS DEATH When infant death is instantaneous, then the type of patho logical findings that are likely come from a more restricted group. Most of the deaths will have a cardiovascular cause. Cardiac malformations, particularly ductus-dependent abnormalities, are the most common group. One-third of babies with cardiac malformations were awake at the time of collapse in one study. 113 These are particularly common in the first month of life. Nevertheless, infants with some malforma tions that are particularly associated with sudden neonatal death, such as hypoplastic left heart syndrome, do survive into the post-neonatal period. A surprising number of infants with isolated ventricular septal defect die suddenly.114 Up to one-third of affected children have arrhythmias. Detailed examination of the cardiac conduction system has shown fibrous interruption of conduction tissueY5 Sudden death in infants who have undergone surgical correction of congeni tal heart disease (CHD) some months previously demands detailed examination of the cardiac conduction system. The distribution of the cardiac conduction tissue is frequently aberrant in CHD, and inflammation around suture material can have fatal consequences. A range of other cardiac pathol ogy has been identified in SUDI llG but, in an individual case, requires careful consideration of its relevance to causation. Coronary arteritis, usually a manifestation of Kawasaki's disease, is encountered in the infant age group, more com monly in infants over 6 months of age. ll7 As well as diffuse thickening of the coronary arteries (Fig. 11.1), aneurysms may be found in the proximal parts of the arteries. Other
206 I
Sudden unexpected death in infancy
Table 11.2
Pathalagical findings in SUDI in different
circumstances af death (after ref 117) Instantaneaus death
Natural causes Cardiac malformation Coronary artery anomalies Coronary arteritis Myocarditis Cardiomyopathy Dysrhythmia Acute laryngitis Laryngeal cysts Laryngeal papilloma Laryngeal/tracheal developmental anomalies Acute intestinal obstruction, e.g. volvulus Non-natural causes Accident Non-accidental injury Foreign body in airway Rapid death, illness recagnized
Respiratory infections, both upper and lower Men ingitis/men ingoencephal itis Septicaemia Gastroenteritis Other infections (including viral) Reye/Reye-like syndrome Intracranial haemorrhage Congenital adrenal hyperplasia
Figure 11.1 Heart of an infant with Kawasaki disease; the coronary arteries are prominent and there is aneurysmal dilatation of the proximal segment of the left coronary artery.
Faund dead in cat
Natural Congenital heart disease Respiratory infection Men ingitis/men ingoencepha litis Septicaemia Seizure disorders Unexplained (up to 80% of deaths in this circumstance) Non-natural Suffocation Other non-accidental injury Intoxication Figure 11.2
coronary artery pathology, such as anomalous OrIgIn and bridging of arteries, is also seen l18 and is described and illustrated in Chapter 12. Cardiomegaly may alert the pathologist to a cardiovascular cause of death. Ischaemic myocardial damage is seen in about 50 per cent of cases. Coronary arterial embolization is rare in early life, but para doxical embolism with myocardial infarction is described in the neonatal period 119,120 and in older infants, when it is most commonly a complication of endocarditis. 121 Several cardiomyopathies can be fatal during infancy. 122 Hypertrophic obstructive cardiomyopathy can present as
Histiocytoid cardiomyopathy.
sudden infant death 123 (see Figs 8.12, p. ]56 and 12.2, p. 228). There is disproportionate hypertrophy of the septum, and in fatal infant cases hypertrophy is severe. 124 The diag nosis is an important one in view of the findings that most cases presenting in early life are familial. 125 Histiocytoid car diomyopathy, which may be visible through the ventricular endocardium as irregular, cream or yellowish patches and as pallor in the underlying myocardium, always requires histo logical confinnation (Fig. 11.2). It has been found in both anatomically normal and malformed hearts. 126,127
Pathological findings in SUDI I
207
Figure 11.3 Subend ocardial fibroelas tosis affecting the left ventricle. Sudden death. No evidence of a more specific disorder.
A number of cardiomyopathies that are fatal in early life exhibit subendocardial Iibroelastosis (sub-EFE), usually in the left ventricle (Fig. 11.3), but sometimes affecting all chambers. The majority of these are plimary, 113 but sub -EFE may be a manifestation of a number of conditions includ in g glycogen storage disease type II (Pompe) and mitochon drial myopathies. It is important to culture fibroblasts and reserve snap-frozen samples of fascia lata or pericardium so that the primary disorder can be elucidated. The dysrhyth mias produced are va rious, but co mplete heart block is com mon. The mothers of some of these infants have connective tissue disorders; investigation of the mother and close sur veillance of subsequent pregnancies may be important. A variety of dysrhythmias are associated with sudden deaths - long QT synd rome is the best recognized. 128 Both dominant and recessively inherited forms have been described and gene abnormalities in at least six different loci have been iden tified. 129 Schwartz et al 130 recorded electrocardiograms (ECG s) in the first week of life on more than 34 000 babies. Twenty-four infants died suddenly in infancy, of whom 12 had prolongation of the QT interva l. No abnormalities were seen in the ECGs of 10 explained deaths. Arnestad et al 131 found molecular abnormalities likely to be associated with sudden death in 19 of 201 (9.5 per cent) cases of sudden infant death in a retrospective study. Whilst mutations can be defined in formalin-fixed, paraffin- embedded t issue, frozen sampl es are much easier to work with.132 Most do not have structu ral abnormalities in the cardiac cond uction system. Sometimes dysrhythmias occur in malformed hearts, particularly those with large ventricular septa l defect l1 4 or conotruncal problems.
Figure 11.4 Larynx opened posteriorly; there is a cyst in the anterior wall at the narrowest po int of the airway.
Mucous cysts or papillomata in the laryn x can achieve sufficient size to obstruct the airway (Fi g. 11.4). A lingual thyroglossal cyst can also result in ainllJay occlusion. 133, 134 Tracheomalacia can produce intermittent (and fatal) ailways obstruction. 135 Posterior lary ngeal cleft (see Fig. 10.15, p, 194) predisposes to massive aspiration. In sum, there may be evidence of prior asp iration in the form of gran ulomata. It is important to remember that foreign bodies may have been removed during attempted resuscitation. Acute intestinal obstruction, such as sma ll intestinal volvuluS136 (Fig. 11.5), is another cause of very rapid demise. Sudden death is rep0l1ed complicating large intes tinal volvulus l3 7 and intussusception in infants.138,139 RAPID DEATH DURING RECOGNIZED ILLNESS
When death occurs rapidly in an infant with a symptomatic illn ess, the most likely pathology is infection. It is important in these circumstances that every effort is made to identify the organism both by culture and by using available molecu lar methods. Few infections have pathological findings suffi ciently specific that a precise diagnosis can be reached on morphological and microscopic findings alone. Histological evidence of infection can be focal but nevertheless signifi cant, for example in the myocardium, where wide sampling
208 I
Sudden unexpected death in infancy
Figure 11.5 Female infant, 3 weeks old: sudden death. There is volvulus of the whole of the small intestine.
is appropriate,140 otherwise the abnormality may be over 100ked. 141 Some infants dying suddenly with myocarditis are asymptomatic 142 or exhibit non-specific symptomatology.14J Evidence of viral aetiology may become more frequent with the availability of molecular techniques. 144 Another important group of diagnoses in this type of death are genetic metabolic disorders (GMDs), particularly the ~-oxidation defects. Profound hypo glycaemia and hyperammonaemia may be triggered by relatively trivial infection. Although classically accompanied by a fatty liver, steatosis may not be apparent to the naked eye. It may also be missed on haematoxylin and eosin (HE)-stained sections as the fat is microvesicular (see Fig. 12.21, p. 247). Frozen sections of liver, kidney and heart are useful for screening. Losty et al 14S dispute this, but the management of their case, not a typical SUDI, was probably responsible for lack of steatosis. The most common of the ~-oxidation defects in Western Europe is medium-chain acyl-CoA dehydrogenase (MCAD) deficiency with a heterozygote frequency of 1:40 in Birmingham 146 but only 1 :74 in the west of Scotland. 147 Other GMDs are individually rare but collectively common; Fitzpatrick l48 describes their presentation in early life. Most centres are unlikely to be able to offer the full range of investigations. 149 Lundemose et aliSO suggest using the [9,1O- JHJ myristic acid oxidation assay in cultured fibro blasts as a method of identifying at least nine disorders of fatty acid oxidation. They found three cases and two car riers of the G985 mutation among 79 sum using this method. More extensive investigation of GMDs should be
Figure 11.6 Total anomalous pulmonary venous drainage, sudden collapse at 4 weeks of age. The aberrant pulmonary vein runs through the diaphragm to join the portal venous system.
contemplated when there is a second sum or parental con sanguinity. ~-Oxidation defects of all types may account for up to five per cent of sum. 151 A test for the most common mutation (G985) is readily available. A toxicological screen is warranted when hepatic steatosis is found. Congenital heart disease is less common in this group but must be carefully sought. Anomalous pulmonary venous drainage may present with non-specific signs of illness. IS2 It is worth remembering that this is the structural cardiovascu lar anomaly most often missed by the pathologist. It is not difficult to demonstrate (Fig. 11.6), but needs to be kept in mind. 1SJ Idiopathic coronary arterial calcification has also been found in symptomatic infants dying suddenly. 154 Sudden death in the infant age group during the course of recognized, non-sudden illness, may, as in the older child, result from pulmonary thromboembolism. 155 It is a diagnosis rarely considered in early life and likely to be missed without sampling of the lungs for histological examination.
FOUND DEAD IN COT
Despite a high probability of death remaining unexplained when an infant's death is discovered on household waken ing, nevertheless a full investigation is warranted. Some deaths will be explained; both congenital heart disease and infections are potential causes of death in this situation. Whilst seizure disorders are another recognized cause of death in this group, diagnostic or even supportive patho logical findings are not common.
Pathological findings in Table 11.3
Causes of sudden neonatal death
Congenital heart di sease - particularly ductus-dependent lesions Infection - both bacterial and viral Birth-related causes - both hypoxic and traumatic Acute intestinal obstruction Genetic metabolic disease
SUDI I
209
Table 11.4 Typical post-mortem findings in SIDS (after ref. 711) External
Well nourished , normally developed baby
Frothy blood-tinged fluid around the nose (500f0)
Hypostatic staining often anterior (indicates face-down position)
Cyanosis of lips and nailbeds
Internal
'Large' th ym us with petechiae (800f0)
Petechiae beneath viscera l pleura
Epi ca rdial petechiae
Full expansion of the lung s
Liquid heart blood (800f0)
Prominent mesenteric lymph nodes and Peyer's patches
Empty bladder (>500/0)
Microscopic
Thickened laryngeal basement membrane Pulmonary congestion and oedema Mild inflammation of the upper respiratory tract Focal fibrinoid necrosis of the voca l folds Persistent haemopoiesis in the liver
Figure 11 .7 Herpes type 1 infection . The mother had oral herpes at delivery. The infant displayed non-specific symptoms followed by sudd en deterioration at 2 weeks of age. Slice of liver shows multiple cream, necrotic foci with a hyperaemic border.
SUDDEN DEATH IN NEONATES
Sudden death in the neonatal period (birth to 28 days of age) is less common than SUDI but is usually explained (Table 11.3). It is particularly important that the full range of inves tigations is completed. Although structural cardiac disorders can be easily demonstrated and there is usually some, albeit rarely specifIc, naked-eye abnormality in cardiomyop athy, the other common causes of sudden neonatal death may show little macroscopic abnormality. Group B streptococcal infection, both septicaemia and meningitis, is an important cause of death in this age gro up. Acute viral infections are important too. Herpesvirus infection has characteristic gross (Fig. 11.7) and microscop ic fIndings even when it has eluded pre-mortem diagnosis. Echoviral infection , particularly serotype 11, results in rapidly progressive and often fa tal ill ness. Pathological changes are less specifIc, but haemorrhagic necrosis of liver and adrenals is a common finding. 156 ,1 57 The infection is more likely to be fatal when acquired from the mother. 158 Other virus serotypes are sometimes involved. Fatal group B Coxsackievirus infection is also usually acquired from the mother. 159 Because the effects of perinatal hypoxia may be appreci ated only at microsco pic level, it is important that a formal neuropathological examin ation is made of the fixed brain.
Failure to do this will result in loss of important info rmation, particularly in respect of timing of ins ult (and may protect, rather than condemn , your obstetric colleagues). Although the majority of in fants with ~ -oxi dation defects will present during late infancy, a few will be ac utely, and perhaps cata strophically, symptomatic in the neonatal period. 16o,161 It is important that a fatty liver is not written off as due to peri mortem hypoxic change. As in later infa ncy, infection may be a predisposing factor - we have encountered such a death complicating sta phylococcal in fection. It is important to consider the possibility of GMDs if only to ensure that appropriate sa mpl es are reserved. 148
Which Post-mortem Findings are Compatible With the Conclusion of SIDS? MACROSCOPIC
Although a conclusion of SlDS is reached after excluding explained causes of SUDl, nevertheless there are some fIndings which, although not universally present nor spe cifIc, might almost be considered reassuring (Table 11.4). Several stud ies have identified an increase in dysmorphic features or minor malformations in SlDS.162-164 Although many SlDS babies appear well nourished at death, poor weight ga in has been identifIed in unexplained infant deaths. 165 A fall in centiles from birth weight to death weight might thus be anticipated, although differences between pre and post-mortem weights should be cautiously interp reted. However, a weight less than the 10th centile at death, particularly if accompanied by signs of dehydration or of
210 I
Sudden unexpected death in infancy
Figure 11.8 Sudden infant death; large numbers of petechial haemorrhages are present in a normal-sized thymus, while the lungs are bulky and overlie the heart.
suboptimal care, for example extensive napkin delmatitis, demands careful consideration of the overall care of the infant. Frothy secretions may commonly be visible at the nares. They are pale in colour but may be blood tinged. Fixed dependent lividity over the face and front of the trunk indicates death in the prone position, but if the body is moved sholily after death then this very useful pOinter may disappear. The appearance of the thymus deserves paliicular atten tion. It is usually of normal size, previously often con sidered large. Reduction in thymic weight correlates with stress, often in the form of infection and indicates that the infant had been unwell for more than 24 hours.166 The cer vical extensions are usually pale in comparison with the thoracic part, which appears relatively congested. Thymic petechiae, particularly when found in large num bers in the thoracic portion (Fig. 11.8), are one of the most consistent findings in SIDS. They are often most numerous on the posterior aspect. Beckwith 167 found them in more than 80 per cent of his cases and considered them the result of respiratOlY obstruction with negative intrathoracic pressure. They were present in large numbers in 49 per cent of SIDS in another study, with a further 28 per cent having a few petechiae. IG8 In SIDS, petechial haemorrhages are frequently promi nent along the course of coronary arteries, seen in 48 per cent by Kleeman et al 1995. 168 Petechiae may be present subpleurally. They are typically small and multiple. They
were present in 61 per cent of Kleeman's SIDS cases. Isaksen and Helweg-Larsen 169 found subpleural pulmonary petechiae more frequently in infants in whom resuscitation had been attempted. In SIDS, when the baby has not been subjected to attempts to resuscitate with oxygen, the lungs are typically fully expanded and fill the chest. They do not collapse on slicing. SUDI is more common among pre-term babies and those who have been admitted to special care baby units. Evidence of residual bronchopulmonalY dysplasia in the form of scar ring or fissUling of the lung is not unusual. These cases have related pulmonalY histological abnormalities. 170 It could be argued that these deaths are, in fact, explained by circula tory decompensation triggered by minor infection. Frequently a small amount of straw-coloured fluid is seen within the pericardial sac. There is usually fluid blood within the heart; it is the result of post-mortem fibrinolysis said to be associated with the suddenness of the death. The foramen ovale is often valvular. The significance of an atrial septal defect has been debated, although Dancea et al ll3 record two cases with signs of right heart overload among their sudden cardiac deaths. Cervical and thoracic lymphadenopathy are unusual findings in SIDS, but mesenteric lymph nodes are frequently enlarged, ref1ecting the level of environmental antigenic stimulation in this age group. Peyers' patches are often prominent and may be hyperaemic. The stomach is often distended, particularly when resus citation has been attempted. Milk curd in the stomach is a common finding and does not necessarily mean that death occurred shortly after feeding. However, an empty stomach may indicate survival for some time after the last feed, but vomiting and gastric emptying during resuscita tion must be ruled out. l7l When the small intestine is also empty, the possibility of gastrointestinal infection and sub optimal care should be considered. Any gastric contents should be retained for analysis. The urinary bladder is frequently empty in SIDS.172 It is such a frequent finding in unexplained deaths that the presence of a full bladder raises my expectation of signifi cant pathology.
HISTOLOGICAL
Thorough histological evaluation is important in SUD!. It was the investigation most likely to yield significant infor mation in a UK multicentre study.4 Pul monary oedema and congestion are frequent findings in SIDS. Minor focal alveolar haemorrhage may accompany the oedema. This is the origin of the frothy fluid seen at the nares. There is often an increase in rOLlnd cell infiltration of alveolar walls. Peribronchial lymphoid aggregates are often present, but these are seen in most cases of infant death and are a marker of prior respiratory infection. 17) The significance of minor inf1ammation is a recurrent problem in SUDI. 32 Focal acute inflammation is frequently
~
Pathological findings in SUDI I
found (60 per cent of cases) in both the upper and lower respiratory tract in SIDS. The former is not likely to be a primary cause of death J74 unless it results in major narrow ing of the airway. The latter provokes the question 'How much pathology is required to equate with a cause of death?'. This is a difficult question and one that has been sidestepped on occasion. 16 Consolidation throughout sec tions from more than two lung lobes is a yardstick I have used. Gregersen et al 175 consider the extent of pulmonary disease a critical factor in deciding whether or not death is explained. Culture of a single bacterial pathogen from the organ in question and blood cultures is likely to influence interpretation of histological findings. Localized inflammation might acquire significance because of its localization, for example in the cardiac con duction system or brainstem. 32 These authors are uncertain of the significance of focal inflammation in both the meninges and myocardium, although an earlier publication appears to accept that significant myocarditis may be focal. 140 Basement membrane thickening in the vocal cord was a frequent finding in SIDS in one study.176 However, in a larger study Krous et al J77 found no difference in laryngeal basement membrane thickening between SIDS cases and a control group. More serious damage in the form of fibri noid necrosis of the vocal fold is seen in some SIDS babies. 17s In the kidney, glomerulosclerosis may be a striking finding. It is not significant and is seen frequently in explained death in this age group. J79 Persistence of haemopoiesis in the liver, which normally disappears soon after birth, is seen in many SIDS cases and has been promoted as evidence of repeated or chronic hypoxia. ISO, lSI Cytomegalovirus (CIvrv) inclusions have been recorded in SIDS babies for many years. Their incidence in salivary gland or mucous glands in the upper respiratory tract is no more frequent in SIDS than in explained SUDI. IS2 However, their association with microglial nodules in the brain stem has provoked the suggestion that, in this circumstance, the effect of crvrv infection is likely to be significant. ls3 Evidence of minor old periventricular leucomalacia is sometimes seen in SIDS babies (Fig. 11.9), present in one fifth of cases in a Canadian study.184 This is perhaps not surprising given the association with SIDS of low birth weight, pre-term delivery and neonatal morbidity. A more frequent finding is disruption of the ependymal lining of the lateral ventricles, which, when recent, results in atten uation of the columnar epithelial lining and, when longer standing, results in ependymal rosettes and attendant glio sis. Brainstem gliosis is a relatively common finding. ISS, 186 A variety of abnormalities of brainstem serotonergic sys tems have been described in SIDS following detailed mor phometric and immunohistochemical studies, including hypoplasia of the arcuate nucleus and subtle gliosis of sev eral brainstem nuclei. These are reviewed by Kinney and Filiano ls7 and Kinney.18s
211
Figure 11.9 SUDI, frozen section of liver. Microvesicular fat is present in zone 1; this amount of lipid is a frequent finding in unexplained infant deaths. Oil Red O.
Figure 11,10 Reperfusion injury in an infant who was on life support for hours after collapse at home. There is a small, old calcified lesion in the periventricular white matter on the right indicative of a pre- or perinatal ischaemic lesion.
A number of changes found at a histological level in SIDS have come to be regarded as markers of 'stress', i.e. non system-specific markers of illness. These changes are seen in many SIDS deaths, in th e thymus, liver, adrenals and the osteochondral interface (most conveniently accessed at the anterior ends of the ribs). A starry-sky appearance in the thymus, perhaps accom panied by minor cortical lymphocyte depletion, is compatible with a SIDS designation, but major cortical lymphocyte depletion or reversal of cellularity betvveen cortex and medulla indicates a longer (and probably more severe) episode of stress, 166 Minor lipid accumulation in the liver (Fig. 11.10) is a marker of minor stress. It is microvesicular and usually more
212 I
Sudden unexpected death in infancy
marked around central veins. Should fatty change be exten sive, the possibility of an inherited metabolic di sease as an explanation for death should be pursued. When there is pan lobular fatty change, particularly when macrovesicular steatosis is also seen (see Fig. 12.21 , p. 247), then the possibil ity of a GMD is high. Concomitant steatosis should be sought in the myocardium and proximal renal tubules and specific investigations for GMDs carried out. It is important to appre ciate the importance of severity when assessing fatty change. r cannot be the only pathologist to have conducted an inves tigation into the death of a second infant from the same fam ily who has been told by the pathologist who investigated the earlier one that he had noted gross hepatic steatosis but thoug ht that a fatty liver was a common finding in SIDS. Lipid depletion of the adrenal cOliex is another non specific marker of stress. 189 It is perhaps worth remembeling that lipid accu mulation in the adrenal cortex is, in fetal life, a marker of chronic anaemia, probably mediated via hypoxia. 190 Loss of regular transition at the osteochondral junction is a fre quent finding in SIDS, but more serious disruption is a marker of chronic growth disturbance. 191
Changes Induced by Attempted Resuscitation Minor excoriation around the mouth and on the tongue and pharynx may occasionally be seen. Resuscitation makes the interpretation of recent bruising aro und the nose, mouth a nd lower jawline problematic. Kaplan and Fossum 192 relate minor injuries in th is a rea to specific resuscitation manoeuvres. Gastric dil atation due to assisted ventilation is a common finding when resuscitation has been attempted. Ventilation with oxygen, which is subsequently resorbed, can result in pulmonary coJi apse even when carried out for a brief period. It is important to be aware of this phenom enon as it obliterates a typ ica l SIDS findin g, one w hich offers potential distinction between SIDS and hypox ia . Prolonged (greater than 12 hours) ventilation, usually instituted when cardiac electrical activity is elicited during resuscitation in hospita l, can produce a number of changes, particularly reperfusion injury. The general level of tissue preservation is very poor, unlike the usual situ ation in infant deaths, when histological examination is usually worthw hile. Not surprisingly, the brain is swollen and extremely soft. Reperfusion cerebral haemorrhagic infarc tion (Fig. 11.9) ca n occur in this situation. It is impOltant not to interpret this as an indicat ion of cerebral hypoxia occurring prior to the term ina l event. Similar changes may be present in the cardiac intraventricular septum (Fig. 1l.11). Pulmonary alveo lar and interstitial haemorrhage is also common. Focal haemorrhagic infarction may be seen throughout the intestine but may be confined to caecum and colon. Unlike necrotizing enterocolitis (NEC), seen particularly in the pre-term infant, all of the lesions are of similar age and have a diffuse margi n. They are often circumferential and
Figure 11.11 Horizonta l slice through the interve ntricul ar septum of an infant; there is extensive haemorrhagic infarction caused by ventilation for about 12 hours post collapse.
not preferentially along the antimese nteric border, as is often seen in the early lesions of NEe. Cardiopulmonary resuscitation (CPR) is an infrequent cause of rib fractures in infants because their ribs are very pliable. Only 1 infant out of 2 11 who unde rwe nt resuscita tion aro und the time of deat h susta ined rib fractures in a North American study.1 9J Bilateral fractures at the sterno chondral junction were found. We have seen similar but unilateral fractures in a single infant - an explained natu ral death following CPR (Fi g. 11.12) - and undisplaced lat eral rib fractures in a further case (Fig. 11.13). Feldman a nd Brewer 194 and Spevak et al 195 found no rib fractures in infants w ho had undergone CPR in their cases. Contrary to popular belief, professionals are more likely to cause injury during resuscitation than are amateurs, irrespective of physical mass or the use of in appropriate resuscitation techniques. On those infrequent occasions when resuscita tion does resul t in rib fractures, bleeding is min imal. Vis ceral injuries are similarly uncommon resuscitation injuries. 193 ,1 96 Focal disrupti o n of cortical bone of the anterio r tib ia during attempted intraosseous transfusion can produce a confusing radiological ap pearan ce and give rise to co ncern about the possibility of fracture.
Pathological Findings which Raise Concern About the Possibility of Non-Accidental Injury (NAI) (Table 11.5) EXTERNAL FINDINGS Petechial haemorrhages in the face or neck are unusual findin gs in SIDS and in healthy babies. In the latter gro up, most occurred below the nipple line. Only 2.5 per cent had
Pathological findings in SUDI I
213
Figure 11.12 (a) A rib cage with recent fractures at the anterior ends of three adjacent ribs; there is little associated haemorrhage. (b) Radiograph of affected ribs. The fractures are accompanied by irregularity of outline at their anterior ends. (c) Internal aspect of the ribs. Fractures are visible close to the osteochondral junction. (d) Histological examination reveals a fracture with haemorrhage but no reactive changes. Sudden collapse, attempted resuscitation, anomalous pulmonary venous drainage found at necropsy.
petechiae in the head and neck, and only 2.6 per cent had three or more petechiae in total. 197 Their presence raises the possibility of deliberate asphyxia (Fig. 11.14). When pre sent over the face, neck and upper chest, petechiae are sug gestive of chest compression.lg8-200 The infant should be carefully examined, searching for bruising around the face and circumoral pallor, tears at the frenulum (recent or healed) and abrasions or bruises on the inside of the mouth and lips corresponding to teeth. Collins 98 found little
evidence of contusions, petechiae or intraoral trauma in overlaying and considered that their presence should incline towards inflicted injurY. Conjunctival petechiae should be very carefully sought. Kleeman et al 168 found small numbers of conjunctival petechiae in 2.4 per cent of their SIDS cases. This has not been confirmed by others. Larger numbers of petechiae were seen in 21.9 per cent of babies dying as a result of trauma, including in five out of six babies dying from
214 I
Sudden unexpected death in infancy
Figure 11.14 Petechial haemorrhages on the neck and front of chest are not a usual finding in unexplained SUDI. Post-mortem rib fractures mid shaft in adjacent ribs. There is only very minor subperiosteal haemorrhage at the fracture lines. Figure 11.13
Table 11.5
Pathological findings causing concern in SUDI
Petechial haemorrhages face/neck Pallor around nose/mouth Torn frenulum Bru ises - even one is serious Heavily blood-stained secretions mouth/nose Blood in pharynx (seen by a doctor) Rib fractures - recent or old Any other injury Blotchy haemorrhages on lung Alveolar haemorrhage> 10% alveoli Siderophages in lung
strangulation. A combination of conjunctival petechiae and acute pulmonary emphysema was found in all of seven babies with asphyxia or strangulation, but not in cases of SIDS, other natural deaths or severe head injury in a study by Betz et al. 201 They considered the combination of two findings useful in distinguishing asphyxial death from SIDS. Development of petechial haemorrhages requires a com bination of hypoxic damage to small blood vessels, a func tioning circulation and an increase in venous pressure, not necessarily in that order. 202 It is unlikely, then, that resus citation will produce petechiae unless circulation and
blood pressure have been successfully restored during the process. 203 Blood or heavily blood-stained fluid issuing from mouth or nose or visualized in the pharynx is also a cause for concern. Nasal haemorrhage was reported in 15 per cent of SIDS cases in a study by Becroft et al,61 based on information extracted from a structured questionnaire. Babies with nasal haemorrhage shared epidemiological characteristics with babies who had pulmonary haemor rhage but not with intrathoracic petechiae. It was found (29 per cent) in babies subjected to imposed airways obstruc tion but not following explained ALTE.55 In a study of 58059 infants under 2 years of age referred to the accident and emergency department from a children's hospital 204 only 16 cases were found, clearly a rare event. Eight of these infants had visible trauma, four had thrombocytope nia and tYro cases were associated with ALTE; retrospective review drew concerns about child protection issues in 44 per cent of cases. Oronasal haemorrhage should be distin guished from the blood-tinged frothy fluid at the nares commonly seen in SIDS. When present, bloody secretions raise the possibility of asphyxia. Krous et al 205 suggest an origin from oronasal mucous membranes. Cutaneous bruises should always cause concern in SUDI. Even a single bruise in a non-mobile infant demands a clear explanation. Such 'minor' injuries must always be taken seriously. 171 Older infants who have some degree of mobility may exhibit an occasional bruise. It is important with any infant injury to consider carefully both the physical abilities of the individual and the explanation proffered by carers.
Pathological findings
Accidental bruises are usually over bony prominences. 206 Bruises of different ages are most concerning.
FRACTURES Fractures are not an expected finding in SIDS and must be clearly explained. Fractures with callus always predate ter minal events, probably by a minimum of 2 weeks, and are incompatible with the use of SIDS on the death certificate. Rib fractures are sometimes seen in extremely pre-term babies. They are accompanied by severe growth distur bance at the osteochondral junction (rickets). In most units, they are encountered less frequently than 20 years ago. 207,2oa However, rickets was described more recently among 39 per cent of very low birth weight babies in one unit, with fractures, most commonly of ribs, occurring in one-quarter of those affected.209 Fractures are unlikely to occur after the baby is weI] enough to leave hospital. Should birth injury be offered as an explanation for a frac ture, the age of the infant and details of mode of delivery are important factors. It is worth remembering that the most common birth-related fracture is clavicular, and a history of difficul ty in delivery of the shoulders should be anticipated. Femoral fractures (usually mid-shaft, sometimes bilat eral) can complicate difficult breech delivery but do not occur during vertex presentation. The most common skull fracture is parietal and linear and runs from the mid-part of sagittal margin of the parietal bone down towards the ear and usually follows instrumental delivery. Rib fractures are exceedingly rare birth injuries and sug gestion of this mode of causation should be viewed most circumspectly. Only nine cases have been reported world wide. Taken together, a pattern emerges of high birth weight, shoulder dystocia, delayed second stage of labour, instru mental delivery and symptoms in the first day or two of life. Symptoms include crepitus of the chest wall, tachypnoea and grunting respiration. That the fractures were unilateral is a useful distinguishing point.210-213
VISCERAL PATHOLOGY Injuries and foreign bodies in the mouth, pharynx or stom ach, particularly in premobile infants, should arouse suspi cion of NAI. 10 Intrathoracic petechiae are seen less frequently in traumatic deaths 168 and among bed-shaling infants 10) than in SIDS cases in all of the three expected sites. In contrast, small numbers of larger haemorrhagic subpleural haemor rhages are seen in some asphyxial deaths. When the asphyx ial insult is very acute, for example a foreign body in the larynx, very few petechiae may be present as death by parasympathetic stimulation may have preceded the onset of mechanical asphyxia. In rabbits, pleural petechiae were read ily produced by three episodes of sublethal airways occlusion but not by a single apnoeic episode.2J3
In
SUDI I
215
Pulmonary congestion and oedema, although seen in histological sections, is not a prominent naked-eye finding in SIDS.
PULMONARY HAEMORRHAGE/SIDEROPHAGES Histological findings causing concern in SUD! are largely confined to the lungs. The abnormalities that have proved most contentious are the presence of widespread alveolar haemorrhage and the presence of haemosiderin within the lung (Fig. 11.15).214 Minor (focal) pulmonary haemorrhage is common in the lungs of deceased infants and may be exacerbated by a long post-mortem interval. 215 Coffin et al 216 found that babies with pulmonary haemorrhage had a high frequency of obstetric and neonatal problems; there was a significant association with hyaline membrane dis ease and haemorrhage elsewhere. There are numerous causes of alveolar haemorrhage in neonates, including hypoxic stress, sepsis and haemostatic problems,216-218 not all of which are easily excluded by necropsy. In older infants, sepsis, heart failure (often sec ondary to CHD) and chronic pulmonary infection, as well as idiopathic pulmonary haemosiderosis (IPH), can give rise to haemorrhage. It is more prominent in babies who have been resuscitated. 2i s Yukawa et al 2J9 looked at the presence and extent of intra-alveolar haemorrhage in a consecutive series of SUD! investigated by a forensic pathology department. They concluded that alveolar haemorrhage involving more than five per cent of alveolar area in histological sections was probably a marker of airways obstruction. It was present in 73 per cent of babies thought to have been subjected to involuntary overlaying (accidental asphyxia) and in 45 per cent of cases in which there were suspicions of deliberate airways obstruction. Every consideration must be given to possible natural causes of pulmonary haemorrhage before asclibing the find ings to imposed airways obstruction. Concerns expressed about the more general application ofYukawa et al's219 find ings are the paucity of typical SIDS cases studied and no sig nificant difference between number of allegedly asphyxiated babies and SIDS cases having greater than five per cent haemorrhage.214 The finding that intra-alveolar haemor rhage was most common in those babies in whom over-laying was a possible factor in their death invokes the possibility that repeated non-fatal hypoxia played an important role in the haemorrhage, drawing a parallel with the development of petechial haemorrhages. The role of chest compression is unclear. 21 9 This group of babies was younger than the rest,214 a factor associated with more frequent, but scanty, pul monary siderophages in another study.220 As an isolated finding, intra-alveolar haemorrhage is not a clear marker for incontrovertible imposed upper airways obstruction but should initiate a critical review of the whole case. 214 Stewart and Fawcett,221 in an uncontrolled study, found pulmonary interstitial haemosiderin in one-half of 24
216 I
Sudden unexpected death in infancy
Figure 11.15 (a) Siderophages are present within the alveoli. (b) Low-power photomicrograph showing the extent of siderophage formation follo wi ng alveolar haemorrage.
consecutive SlDS deaths. Babies with siderophages were usually younger than those without. The autho rs postulated that haemosiderin might be a marker for 'near miss' events (ALTEs). Byard et al222 found more pulmona ry interstitial
haemosiderin in SID S infants with prior ALTEs (33.3 per cent) than in non-traumatized control infants (five per cent). Haemosiderin was found in 18 per cent of SIDS infants without ALTEs, a statistical ly non-significant difference. The authors concluded it was an unreliable marker of a pre vious ALTE. Becroft et al223 consider pulmona ry interstitia l haemosiderin a consequence of normal labour. Becroft and Lockett 224 found large numbers of intra alveolar siderophages in babi es whose deaths were caused by airways obstruction. They proposed that their presence should give rise to suspicion of imposed airways obstruction and that the lungs of all SUD I victims should be stained for iron. Pulmonary siderophages have been demonstrated in repeated airways obstruction. 225 Dorandeu et a1 226 found pul monary siderophages in 11 of 15 infants with non-accidental injury; all had intracrania l haemorrhage and 12 had other injuries as well. Alveolar siderophages are infrequent in SIDS cases. 21B,220.227-229 Krous et al230 found wide variati on in the number of intra-alveolar siderophages in a large group of unexplained infant deaths and were not able to distin guish such infants from babies dying from suffocation. Pulmonary siderophages have been described in leukaem ia.217 They have also been found post-operatively, in serio us sepsis and followin g ventilation, situations which might have caused hypoxic pulmonary damage. 2IB They have been see n in infants from homes with preceding water damage in substandard housing in Cleveland, Ohio,231 sug gesting a role for air-borne contaminants as pulmonary irritants. IPH has been found to occur in infants as well as in older children 232 and appears always to be symptomatic. It has been suggested that IPH can present as SUDI,233 but it is not clea r how confidently imposed airways obstruction was excluded in either study. Haemoglobin breaks down quickly in extravasa ted red cells, which are quickly pha gocytosed. Finely dispersed haemosiderin can be found after 24- 36 hours. Haemosiderin is ab undant at 5 days, and phagocytosed erythrocytes are recognizable. 234 Red cells have disappeared after abo ut 7 days. Inform ation about the persistence of siderophages from the lung is scanty. They are rapidly cleared from large airways (within 2 weeks) and from more peripheral parts of the airway within 4 weeks 235 in infants. Esterley and Oppen heimer 236 found that siderophages had disappeared from the alveoli after massive perinatal pulmonary haemorrhage within 2 weeks. It is reasonable to suppose that, whilst most alveolar siderophages are removed from the lung via the air ways, some macrophages will migrate to septal and pleural connective tissue and might remain there for several weeks after the initiating haemorrhage. ASPIRATION OF FOREIGN MATERIAL It seems likely that many babies respond to an asphyxial
insult by vomiting. Evid ence of major aspiration is a cause for concern. It shou ld be distinguished from the localized
Pathological findings in
SUDI I
21 7
Figure 11.17 Lipid-containing macrophages are present in alveoli at the periphery of a lung lobule.
Figure 11.16 Massive inhalati on in a baby found to have a diffuse neuronal migration disorder. There is brown discoloration of the lungs because of acid lysis of red blood cells.
and usually peripheral foreign body-type granulomata, which are the result of repeated, minor aspiration, often seen in young babies with oesophageal or lary ngea l prob lems. It is most unusual in the absence of a predisposing problem. This may be gastroenteritis, pyloric stenosis, diverse causes of intestinal obstruction, cerebra l palsy or other neurological problem. In a recently investigated case. brown discoloration in lung fissures (Fig. 11.16) due to acid lysis, extensive lipid in airways and terminal air spaces within the lung (Fig. 11.17) and a diffuse neuronal migra tion disorder were observed. The only evidence of cerebral abnormality at necropsy was localized polymicrogyria at one occipital pole.
HEPATIC SIDEROPHAGES
Dorandeu et al 226 found increased haemosiderin in the liver, predominantly in periportal Kupffer cells, in abused infants compa red with control infants who were victims of SIDS or accidental or explained death. They asc ribed this to a chronic increase in red cell breakdown . Increased iron in the liver is not unusual in infants with infection (usually within hepatocytes) and those with inherited erythrocyte a b no rmali ties. 237
Findings of Questionable Significance EXTERNAL
Cutaneous changes are sometimes overinterpreted in infant deaths. Dribbling can result in erythema and abrasions around the mouth or on the upp er chest and, when it leads to excoriation, might be interpreted as an injury. Sometimes excoriation takes place after death when the infant's cheek is bathed in acidic fluid from regurgitated gastric contents; a linear margin points to the artefact. Healing napkin dermatitis (nappy rash) can be pigmented and give rise to suspicion of bruising. A Mongolian blue spot, seen normally, but not always, in infants with pig mented skin, can simulate bruising; its localization around the natal cleft shou ld suggest the true nature of the lesion (see Fig. 8.7, p. 152). Vascular malformations can sometimes mimic injuries. If doubt persists after careful scrutiny, the diagnosis can be resolved by histological examination. Anal dilatation is common post mortem and should not, as an isolated finding, promote suspicion of abuse. McCann et al 238 have made a careful study of the post-mortem appearance of the anus in infants and children. They counsel against overinterpretation of an exposed pectinate line as tears or fissures. They stress the importance of tears and peri anal bruising to a conclusion that sodomy has been commit ted; swabs for DNA studies should be taken nonetheless. INTERNAL
The presence of gastric contents in the oesophagus, phar ynx, mouth and nose is a common necropsy finding in the
218 I
Sudden unexpected death in infancy
, ', •
I'
~,, _.
.
.
'_
-I
.. .. - .
t
Figure 11.19 Histological examination of an area of brown staining of the dura. Abundant iron is demonstrated by Perls' Prussian blue reaction. Iron persists in the dura for many months.
Figure 11.18 Infant aged 4 months. There is brown staining of the dura of tentorium and posterior falx owing to organization of focal subdural haemorrhage sustained at birth.
infant period. Sphincters relax after death, and moving the baby in the course of the removal to the mortuary will read ily disperse stomach con tents into the pharynx and beyond. When resuscitation is attempted, the presence of milk curd in the pharynx is frequently recorded. Attempted resuscita t ion will disperse this further, and milk curd in the upper airways and focally in large intrapulmonalY bronchi should not be interpreted as being a result of inhalation unless it is massive 31 (Fig. 11.16) and there is histological evidence of reaction in the form of extensive (frequently haemorrhagic) oedema and inflammatory cells within the lungs. Signifi cant gastric aspiration in infants will usually be accompan ied by evidence of a predisposing problem. Healing birth injury may give rise to suspicion of NAl. The most commonly encountered injUly is minor subdural haemorrhage, whose appeara nce should be commensurate with the age of the infant, i.e. it wi ll be red-brown and rec ognizable as a haemato ma, prob ab ly 1-2 mm thick, for 2, possibly up to 4, weeks post partum. Case et al23 9 estimate that these small haemorrhages occur in 20-30 per cent of asymptomatic neonates. Later th a n that, brown staining of the dura is apparent (Fig. 11.18). This may persist for sev eral months. Histological examination of the dura will reveal abundant coarsely granular haemosiderin on the surface and within the membrane itself (Fig. ) ).19).
An organizing cephalhaematoma is usually apparent as an irregu lar elevation on the parietal bone, and it may be bilateral. Most cephalhaematoma ta occur during uncom plicated vaginal delivery. A history of a difficult birth is unusual. Cephalhaematomata may be noticed only once bone is laid do wn beneath the elevated periosteum. Its appearance should correlate with post-natal age. Evidence of this new bone formation may persist for many years. The most common fracture susta ined as a birth injury is a mid-shaft fracture of the clavicle. There may be a history of shoulder dystocia, but most will not have this history. A lump over the clavicle may have been noted by the mother or health professionals prior to death. The radiographic an d histological appearance should be compatible with post natal age. Other long-bone fractures are rare birth injuries and occur in well-recognized circumstances. IS3
MICROSCOPIC Basement membrane thickening in the larynx is an incon sequential finding (see p. 211) . Explosive desquamation of bronchial and bronchoalveolar epithelium has been interpreted as evidence of pulmonary infection 240 but is a post-mortem artefact and may be found even when refriger ation of the body has been prompt. Peribronchial lymphoid aggregates are a normal finding in infants. Pancreatic isl ets often appear prominent in SlDS. They are usually sited towards the centre of the lobule and become more generally dispersed with subsequent acinar develop ment. This should not be interpreted as islet hyperplasia. Examination of the pancreas from an expl ained infant death is a useful compalison.
DEATH CERTIFICATION As SIDS is a conclusion reached by exclusion of recognized causes of death and its definition requires that a thorough post-mortem be performed, it is not appropriate to use the term on a dea th certificate unless certification is delayed until investigations are completed. This period is likely to
References I
be longer than the statutory period for death registration (within 8 days in the UK). My practice accords with that of Gilbert-Barness and Barness B in that, when history, circumstances and naked eye findings are typical , I use the terms SUD! or 'sudden unexpected infant death', together with an indication to the registrar that additional information may become available. In recent years, pathologists have become much more likely to use the terms 'unascertained' or 'undetermined' on death certificates, although in the UK practice varies widely.13 This has provoked protest from parents' support groups. However, it is the pa thologist, not the counsellor, whose opinion or competence is called into question when an atypical infant death is reinvestigated after the abuse or death of another family member. 9 Mitchell et al lio have highlighted the swing away from SIDS to 'undetermined'. It is important to consider the potential consequences of the use of SIDS, in effect stating that death is due to natural causes, on subsequent infants in the family.214 The use of 'unascertained' in these circumstances ensures that the case is kept open and might alert medical and support services to the need for vigilance and an open mind when assessing problems in subsequent infants. The importance of multi disciplinary review of unexpected infant deaths has been stressed earlier in this chapter.
REFERENCES
2
3 4
5
6
7
8
9
Carpenter RG, Shad dick Cw. Role of infection, suffocation and bottle feeding in cot death. Br ] Pre/} Med 1965; 19: I. Beckwith JB. Introdu ction: discussion of terminology. In Bergman AB, Beckwith JB, Ray CG (eds) Sudden Infant Death Syndrome. Washington DC: University of Washington Press, 1970, p. 18. Bass M, Kravath RE. Glass L. Death-scene investiga tion in sudden infant death. N Engl] Med 1986 ; 315:100-5. Berry J, Allibone E, McKeever P et al. The parhology study: the contribution of ancillary pathology tests to the investigation of unexpected infant death. In Fleming P, Bacon C, Blair P, Berry PJ (eds) Sudden Une.rpected Deaths in Infancy th e CESDI SUD! Studies 1993-1996. London: The StationeJY Office, 2000, pp. 97-112. Willinger M, James LS, Catz C. Defining the sudden infant death syndrome (SIDS): deliberations of an expert panel convened by the National Institute of Child Health and Human Development. Pediatr Pathol 1991; 11 :677-84. Fleming P, Blair P, Ward-Platt M et al. The case-control study: results and discussion. In Fleming P, Bla ir P, Bacon C, Berry J (eds) Sudden Une.rpecred Deaths in lnfa ncy: The CESD! SUD! Studies. London: The Station ery Office, 2000, pp. 13-96. Bajanowski T, Vennemann M, Rauch E et al. Unnatural causes of sud den unexpected deaths initially thought to be sudden infant death syndrome. Int] Legal Med 2005; 119:21J-16. Gilb ert-B arness EF, Barness LA. Sudden infant death syndrome. Is it a cause of death? Arch Pathol Lab Med 1993; 117: 1246-8. Green MA. Time to put 'cot dearh' to bed 7 BM] 1999 ; 319:697-8.
10 11 12 13
14
15
16 17 18
19
20 21
22
23
24
25
26
27
28
29
30 3I
32
219
Meadow R. UnnalUral sudden infant death. Arch Dis Child
1999; 80:7 -14.
Limerick S. Not time to put cot death to bed. BM] 1999;
319 :698-700.
Bacon CJ. Cot death after CESDI. Arch Dis Child 1997;
76:171-3.
Limerick SR, Bacon CJ. Terminology used by pathologists in reporting on sudden infant deaths. ] Clin Pathol 2004; 57:309-11. Flemming PJ, Blair PS, Sidebottom P, Hayler T. Investigating sudden unexpected deaths in infancy and childhood and caring for bereaved families: an integrated multiagency approach. Br Med] 2004; 328:331-4. Krous HF, Beckwith JB, Byard RW et al. Sudden infant death syndrome and unclassifi ed sudden infant deaths: a definitional an d diagnostic approach. Pediatrics 2004; 114:234-8. Knowelden J, Keeling J, Ni choll JP et a1. A Multicentre Study of Post-neona ta l Mortality. London: HMSO, 1984. Gibson AAM. Current ep id emiology of SIDS. ] Ciin Pathol 1992; 45(Supp1.):7-10. Gilbeli RE, Fleming PJ, Azaz Y, Rudd PI. Signs of illness preceding sudden unexpected death in infants. BM] 1990; 300:1237-9. Wigglesworth JS , Keeling JW, Rushton DI, Berry P1. Technical method. Pathological investigations in causes of sudden infant death.] Clin Patlzol 1987; 40:1481-3. Royal College of Pathologists. Guidelines for Post-mortem Reports. London: Royal Co llege of Pathologists, 1993, pp. 8- 10. Valdes-Dapena M, McFeeley PA, Hoffman HJ et al. Histopathology Atlas fo r the Sudden Infant Death Syndrome. Washington DC: Armed Forces Instit Pathol, 1993. Krous HF. An international standardized autopsy protocol for sudden unexpected infant death. In Rognum TO (ed.) Sudden Infant Death Syndrome. New Trends in the Nineties. Oslo: University of Sca ndinavia Press, 1995, pp. 75-89. Howatson AG. The autopsy for sudden unexpected death in infancy. Current Diagnostic Pathology 2006; 12: 173-83. Report of a Working Group: Sudden Unexpected Death in Infancy. London: Royal College of Pathologists and Royal College of Paediatrics and Child Health, 2004. Landi K, Gutierrez C, Sampson B et al. Investigation of the sudden death of infants: a multicenter anal ys is. Pediatr Del; Patl101 2005; 8:630-8. Sadler DW. The va lue of a thorough protocol in the investigatio n of sudd en infant deaths.] Ciin Pathol 1998 ; 51 :689-94. Mitchell E, Krous HF, Donald T, Byard RW. An analysis of the us efuln ess of specific stages in the pathologic investigation of sudden infant death. Am ] Forensic Med Pathol 2000; 21 :395-400. Langlois NE, Ellis PS, Little D, Hulewicz B. Toxicologic analysis in cases of possible sudden infant death syndrome: a worthwhile exerc ise? Am] Forensic Med Pathol 2002; 23:162-6. Arnestad M, Vege A, Rognum TO. Evaluation of diagnostic tools applied in the examination of sudd en un expected deaths in infancy and early childhood. Forensic Sci Internal' 2002; 125:262-8. Keeling JW, Knowles SAS. Sudden death in childhood and adolescence.] Patho11989; 159:221-4. Arneil GC, Gibson AA, McIntosh H et a1. National post perinatal infant mortality and cot death study, Scotland 1981-82 . Lancet 1985; 1:740-3. Byard RW, Krous HF. Minor inf1ammatory lesio ns and sudden infant death: cause, coincid ence, or ep iphenomena? Pediatr Pathol Lab Med 1995; 15:649-54.
220 I
Sudden unexpected death in infancy
33 Bacon C, Fleming P. Design of the SUDI study. In Fleming P, Blair P, Bacon C, Berry J (eds) Sudden Unexpected Death In Infancy. Th e CESD! SUD! Studies. London: The Stationery Office, 2000, pp. 7-12. 34 Irgens LM. Risk factors for SIDS: Do they exist? In Rognum TO (ed.) Sudden Infant Death Syndrome. Os lo: Scandinavian Uni vers ity Press, 1995, pp. 99-105. 35 Blair PS, Fleming P J, Bensley D et al. Smoking and the sudden infant death syndrome: results of 1993-5 case control study for confidential inquiry into still bil1hs and deaths in inFancy. 8M] 1996; 313:195- 8. 36 Aim B, Milerad J, Wennergren Get al. A case-control study of smoking and sudden inFant death syndrome in the Scandinavian countries, 1992 to 199 5. Arch Dis Child 1998 ; 78 :329- 34. 37 Blair PA, Sidebotham P, Berry P J et a l. Major epidemiolo gica l changes in sudden infant death syndrome: a 20-year population-based study. Lancet 2006; 367: 314-19. 38 Daltveit AK, 0yen N, Skjaerven R, Irgens LM. The epidemic of SIDS in Norway 1967-93: chan ging effects of risk factors. Arch Dis Child 1997; 77:23-7. 39 Smith Gc, Wood AM, Pell JP, Dobbi e R. Sudden infant death syndrome and complications in other pregnancies. Lancet 2005; 366:2107-11. 40 Malloy MH. Sudden infant death syndrome among extremely pre-term infants: United States 1997-1999.
54 55
56 57
58
59
60 61
62 63
] Perinatol2004; 24:101-7. 41
42 43
44
45
46
47
48
49
50 51 52
53
Leach CE, Blair PS, Flemming P J et al. Epidem iology of SIDS and explained sudden infant deaths: CESDI SUDI research group. Ped,:atrics 1999; 104:e43. Golding J, Peters TJ. What else do SIDS risk prediction scores predict? Early Hum Develop 1985; 12:247-60. Smith GC, Wood AM, Pell JP et a!. Second-trimester maternal serum alphaFeto protein levels and the subsequent risk of sudden inFant death syndrome. N Engl] Med 2004; 351 :978-86. Bass M. The Fallacy of the simultaneous sudden infant death syndrome in ('wins. Am] Forensic Med Pat/101 19139 ; 10:200-5. Carpenter RG, Gardner A, McWeeny PM , Emery JL. Multistage scoring system For identifying inFants at risk of unexpected death. ATCh Dis Child 1977; 52:606-12. Golding J, Limerick S, Macfarlane JA. Sudden Infan/ Death: Pattems, Puzzles and Problems. London: Op en Books, 1985, pp. 94-5, 190-5. Ward-Platt M, Blair PS, Fleming PJ et al. A clinical comparison of SJDS and explained sudden inFant deaths: ho w healthy and ho w norm a l? Arch Dis Child 2000 ; 82:98-106. Overpeck MD, Brenner RA, Trumble AC et al. Risk factors for inFant homicide in the United States. N Engl] Med 1998; 339 :1211-16. Stanton AN, Downham MAPS, OakJey JR et al. Terminal symptoms in children dying suddenly and unexpectedly at home. Preliminary report of the DHSS multicentre study of postneonatal mortali ty. 8M) 1978; 2:1249-51. Stanton AN, Oakley JR. Pattern of illnesses beFore cot deaths. Arch Dis Child 1983; 58:878-81. Ford RPK, Mitchel! EA, Stewart AW et al. SIDS, illn ess, and acute medical care. Arch Dis Child 1997; 77:54-5. L'Hoir MP, EngelbeJ1s AC, van Well GThJ et al. Sudden unexpected death in infancy: epidemiologically determined risk factors related to pathological classification. Acta Paediatr 1998; 87:1279-87. British Paediatric Association. Report of a Working Party.
Evaluation of Suspected Imposed Upper Airway ObSfillction.
64
65 66 67 68
69 70
71
72
73 74
75
76
77
British Paediatric Association. London: Roy Soc Med Press, 1994. Rosen CL, Frost JD, Glaze DG. Child abuse and recurrent infant apnea.] Pediatr 1986; 109: 1065-7. Southall DP, Plunkett MCB, Banks MW et al. Covert video recordings of life-threatening child abuse: lessons for child protection. Pediatrics 1997 ; 100:735-6. Berger D. Child abuse simulating 'near- miss' sudden infan t death syndrome.] Pediatr 1979; 95:554-6. Ariagno RL, Guilleminault C, Korobkin R et al. 'Near-miss' for sudden infant death syndrome infants: a clinical probl em. Pediatrics 1983; 71:726-30. Byard RW, Krous HF. Sudden infant death syndrome: overview and update. Pediatr Dellelop Patlwl 2003; 6: 1 12-27. Truman TL. Ayoub CC. Consid ering suffocatory abuse and Munchausen by proxy in the evaluation of children experiencing apparent life-threatening events and sudden infant death syndrome. Child Maltreat 2002: 7:138-48. Galvin HK, Newton AW, Vandevan AM. Update on Munchausen by proxy. Curr Opill Pediatr 2005; 17:252-7. Becroft DMO, Thompson JMD , Mitchell EA. Nasal and intrapulmonary haemorrhage in sudden infant death syndrome. Arch Dis Child 2001; 85:116-20. Barber MA, Davis PM. Fits, faints, or fatal fantasy? Fabricated seizures and child ab use. Arch Dis Child 2002; 86:230-3. Proudfoot A. Is the cause toxicological? Medicine 1999; 27(4):4. Emery JL, Thornton JA. Effects of obstruction to respiration in infants, with particular reference to mattresses, pillows, and their coverings. 8M} 1968; 3:209-13. Foundation For the Study of InFant Death s. Reduce the risks of cot death - an easy guide (Leaflet). FSID, 1996. Foundation For the Study of Infant Deaths. Reduce the ris.~s of cot death (Leafl et) . FSID, 1991. UNICEF UK. Sharing a bed with your baby. London: UNICEF UK,2003. Scragg KR, Mitchell EA, Stewart AW et al. Infant room sharing and prone sleep position in sudden infant dea th syndrome. Lancet 1996 ; 347:7-12. Stanton AN. Sudden inFant death. Overh eating and cot death. Lancet 1984; 2:1199-2 01. Fleming P J, Gilbert R, Azaz Y et a!. Interaction between bedding and sleeping position in the sudden inFant death syndrome: a popul ation based case-control study. 8M} 1990; 30:85-9. Wigfield RE, Fleming PJ, Azaz YEZ et al. How much wrapping do babies need at night? Arch Dis Child 1993; 69:181-6. Beal SM, Finch CF. An overview of retrospective case control studies investigating the relationship between prone sleep ing position and SIDS: } Paediatr Child Health 1991; 27:334-9. Beal SM. Sleeping position and sudden inFant death synd rom e. Med] Aust 1988; 149:562. de Jonge GA, Engelberts AC, Koomen-Liefting AJM, Kostense P J. Cot death and prone sleeping position in The Netherlands. 8M] 1989; 298:722. Mitchell EA, Scragg R, Stewal1 AW et a!. Results From the first year of the New Zealand cot death study. N Z Med] 1991; 104:71-6. Brooke H, Gibson A, Tappin D, Brown H. Case-control study of sudden inFant death syndrome in Scotland, 1992- 5. 8M} 1997 ; 314:1516-20. Mitchell EA , Hutchison L, Stewart AW. The continuing decline in SIDS mortality. Arch Dis Child 2007; 92: 625-6.
Referen ces I
78 Ponsonby A-L, Dwyer T, Couper 0, Cochrane J. Association between use of quilt and sudden infant death syn drome: case-control study. BM] 1998; 316: 195-6. 79 Waters KA, Gonzalez A, Jean C et a!. Face-straig ht-down
and face-near-straight-down posilions in healthy, prone
sleeping infants. ] Pediatr 1996 ; 128:616-25.
80 Fleming PJ, Blair PS, Bacon C et a!. Environment of infants during sleep and risk of the sudden infant death syndrome: results of 1993-5 case-control study for contidential inquiry into still births and deaths in infancy. BM] 1996; 3 13:191 -5. 81 Ramanathan R, Chandra 5, Gilbert-Barness E, Fra nciosi R. Sudden infant death syndrome and water beds (Letter to the editor). N Eng] Med 1988; 3 18 : 1700. 82 Bass M. More on sudden infant death syndrom e and water
beds (Letter). N Engl] Med 1988; 3 19:1415.
83 Gilbert-Barness E, Hegstra nd L, Chandra 5 et a!. Hazards of mattresses, beds and bedding in deaths of infants. Am] Forensic Med Patl101 1991; 12:27-32. 84 Kemp JS, Tha ch BT. Sudden death in infants sleep ing on
polystyrene-tllled cushions. N Engl] Med 1991 ;
324: 1858-64. 85 Gilbert-Barness E, Emery JL. Deaths of infants on
polystyrene-filled beanbags. Am ] Forensic IVied Patllol
1996; 17:202-6.
86 Byard RW, Beal SM. V-shaped pillows and un safe infant
sleep ing.] Paediatr Child Health 1997; 33:171-3.
87 Moo re L, Bourne AJ , BealS e t aJ. Unexpected infant death in associa tion with suspend ed rocking cradles. Am ] Forel1sic
Med PatilO11995; 16:177-80. 88 Bass M. Asphyxial cri b death. N Ellgl] Med 1977; 296:
555-6.
89 Smi a lek JE , Smialek PZ, Spitz WU . Accidental bed deaths in infants due to un safe sleepin g situatio ns. c/in Pediatr 1977; 11:1031-6. 90 Byard RW, Beal S, Bourne AJ. Potentially dangerou s s leeping enviro nments and accide ntal asphyxia in infancy and ear ly childhood. Arch Dis Child 1994 ; 17:497-500. 9) Variend 5, Usher A. Broken cots and infant fatali ty. Med Sci Law 1984; 24:111 - 12. 92 Byard RW, Bourne AJ, Beal SM. Mesh -s ided cots: yet anothe r potentially dangerous infant s leeping environme nt. Forensic Sci Int 1996; 83:105-9. 93 Cassidy SP. Analysis of crib-related deaths for the years
1989 through 1991 (memorandum). Bethesda, MD:
US Consumer Product Safety Commission, 1994.
94 Foundation for the Study of Infant Deaths. Newsletter FSID 1999; 59:12. 95 Nakamura 5, Wind M, Danello MA. Review of hazard s
associated with children placed in adult beds. Arch Pediatr
AdolescMed 1999; 153:1019-23.
96 Moore L, Byard RW. Pathological findings in hanging and
wedging deaths in infants and young chi ldren. Am]
Forensic Med Path ol 1993; 14:296- 302.
97 Bya rd RW, Beal SM, Simpson A et a!. Accidental infant
death and stroller-prams. Med] Aust 1996; 165:140-1.
98 Collin s KA. Death by overlaying and wedging. A IS-year
retrospective study. Am ] Forensic Med Pathol 2001;
22:1 55 - 9.
99 Byard RW. Is breast feeding in bed always a sa fe practice ?
] Paedial,. Child Heal th 1998; 34:413-19.
100 Scragg KR, Mitchell EA. Side sleeping position and bed sharing in the sudden infant death syndrom e. Ann Med 1998; 30:345-9. JOI Luke lL. Sleeping arrangements of sudden infant death syndrome victims in the District of Columbia: a preliminalY report.] Forensic Sci 1978; 23:379-33.
221
102 Rintahaka PJ, Hirvonen J. The epidem iology of sudden infant death syndrome in Finl and in 1969-1980.] Forensic Sci Intern 1986; 30:2 19-33. 103 Norvenius SG. Sudden infant death syndrome in Sweden in 1973- 1977 a nd 1979. Acta Paediatr Sco nd Supp/1987; 333:1-138. 104 Rushton Dl. Co-sleeping and sudden infant death syndro me. Lancet 1997; 349:56-7. 105 Blair PS, Flemin g PJ, Smith IJ et al. Babies sleeping with parents: case-control study of factors influencing the risk of the sud den infant death syndrome. BM] 1999; 319:1457-62. 106 Tappin 0, Ecob R, Brooke H. Bedsharing, roomsharing and sudden infant death syn drome in Scotland: a case control study.] Pediatr 2005; 147:32-7. 107 Czegledy-Nagy E, Cutz E, Becker LE. Sudden death in infants under one year of age. Pedian' Pathol 1993; 13 :671-34. lOB Byard RW, Carmichael E, Beal S. How useful is post-mortem examination in sudd en infant death synd rome? Pediatr Pathol 1994; 14:817-2 2. 109 Parham OM, Savell VH, Kokes CP et a \. Incidence of autopsy findings in unex pected deaths of children and adolescents. Pediatr De)} Pathol 2003; 6: 142-55. 110 Mitch ell E, Krous HF, Donald T, Byard RW. Changing trends in the diagnosis of sudden infant death. Am] Forensic Med PatllOl 2000; 21 :311-14. 111 Berry PJ , Keeling J"I/. The in ves tigation of sudden unex pected death in infancy. In A nth o ny PP, MacSween RNM (eds) Recent Advances in Histopath ology 14. London: Churchill Livingstone, 1989, pp. 151-279. 112 Byard RW, Cohle SD. Sudden Death in Infancy, Childhood and Adolescence. Cambridge: Cambridge Uni versity Press, 1994. IJJ Oancea A. Cote A. Roblicek C et al. Cardiac pathology in sudd en unexpected infant death.] Pediatr 2002; 141 :336-42. 114 Cohl e SD, Balraj E, Bell M. Sudden death due to ven tricular septal defect. Pediatr Develop Pathol 1999; 2:327-32. 115 Smith NM, Ho SY. Heart block and sudden death associated with tibrosis of the conduction system at the margin of a ventricul ar septal defect. Pediatr Cardiol 1994; 15:139-42. 116 Bajanowski T, Ortmann C, Teige K et a\. Pathological changes of the heart in sudden infant death. In t] Legal Med 2003; 11 7: I 93-203. 117 Levi n M, Tizard EJ, Dillon MJ. Kawasaki disease: recent advances. Arch Dis Child 1991; 66:1369-74. 118 Lipsett J, Cohle SD, Berry Pl, Byard RW. Anoma lous cOTOnary arteries: a multicenter pediarric autopsy study. Pediatr Patl101 1994 ; 14: 28 7-300. 119 Berry CL. Myocardial infarcti o n in a neonate. Br Heart ] 1970; 32:41 2- 15. 120 Tillett A. Hartl ey B, Simpson l. Parad oxical embQlism causing fatal myocardial infarcti on in a newborn infant. A rch Dis Child 2001; 35:FI37-B. 12 1 Bya rd RW. Vascular conditions. In Byard RW, Cohle SO (eds) Sudden Death in Infancy, Childhood and Adolescence. Ca_m bridge: Camb ridge Univers ity Press. 1994, pp. 175-261. 122 Gilbert -Barn ess E, Barness LA. Nonmalformative cardiovascular pathology in infants and children. Pediatr Develop Pathol 1999; 2:499-530. 123 Fried K. Beer S, Vure E et al. Autosomal recessive sudden un expected death in children prob ably caused by a cardiomyopathy assoc iated with myopathy. ] Med Genet 1979; 16:341-6. 124 Ed wa rds WO. Cardiomyopathies. HlIm Pathol 1987; 6:625-35.
222 I
Sudden unexpected death in infancy
125 Goodwin Jf. Sudden cardiac death in the young. A family history of sudd en death need s investigation. BM] 199 7; 314:843. 126 Sua rez V, Fuggle WJ , Ca meron AH et al. Foamy myocardial transformation of infancy: an inherited disease. ] Ciin Patl70l 1987; 40:3 29 -34. 127 Shehata BM, Patterson K, Thomas JE et aJ. Histiocytoid ca rdiomyopathy: three new cases and a rev iew of the literature. Pedialr Devel Patl70l 1998; 1 :56-69. 128 Adgey J, Johnston PW, Math ew TP. Sudden cardi ac death - a predictable, avoidable and treatable event? Proc R Coli Physicians Edinb 1998; 28 :350-6. 129 Towbin JA. Molecular genetic basis of sudden cardiac death. Perliatr Ciin N Am 2004; 51: 1229 -55. 130 Schwartz PJ, Stramba-Badiale M, Segantini A et al. Prolongation of the QT interval and the sudden infant deat h syndrome. New Engl] Med 1998; 338:1709-14. 131 Arnestad M, Crolti L, Rognum TO et al. Prevalence of long-QT sy ndrome ge ne vari a nts in sudden infant death syndrome. Circulation 2007; 115:361-7. 132 Ackerman MJ, Tester DJ, Driscoll DJ. Molecular autopsy of sudden unexplained death in the young. Am ] Forensic Med Patlwl 200 I ; 22: 105- 11. 133 Hanzlick RL. Case report: lingual thyroglossal duct cyst causin g death in a four-week-old infant. ] Forellsic Sci 1984; 29 :3 45-8. 134 Kanawaku Y, Funayama M, Nata M, Kanetake 1. Sudden infant death: thyroglossal duct cyst ve rsus environmental factors. Forensic Sci Int 2006; 156: 158-60. 135 McNamarra VM, Crabbe DC. Tracheomalacia. Paed iatr Resp ir Rev 2004 ; 5: 147-54. 136 Wong SW, Gardn er V. Sudd en death in children due to mesenteric defect an d mesenteric cyst. Am ] Fo ren sic Med
Patho11992; 13:2 14-16. 137 Byard RW. Sudden infant death, large intestinal volvu lu s and a duplication cyst of the terminal ileum. Am ] Forens ic Med Pat/wi 2000; 21:62-4. 138 Byard RW, Simpson A. Sudden death a nd intussusception in infancy and childhood-autopsy cons ideration s. Med Sci La w 2001; 41 :41-5. 139 Ng'wa la li PM , Yonemitsu K, Tsuenari S. Fatal intussusception in infancy: an exp erience in forensic autop sy. Leg Med (Tokyo) 2003; 5:181 - 4. 140 Smith NM, Bourne J , Clapton WK, Byard RW The spectrum of presentation at autopsy of myocarditis in infancy and childh oo d. Pathology 1992; 24: 129-3 I. 141 Shatz A, Hiss J , Arensburg B. Myocarditis misd iagnosed as sudden infant death syndrome (SlDS). Med Sci Law 1997; 37: 16-18. 142 de Sa DJ. Isol ated myocarditis in the first year. Arch Dis Child 1985; 60:484-5. 143 Krous HF, Haas E, Chadwi ck AE, Wagner GN. Sudden death in a neonate with idiopa thic eosin ophilic endomyocarditis. PerliMr DelJel Pathol 2005; 8:587-92. 144 Dettm eyer RB, Padosch SA, Madea B. Lethal enterovirus indu ced myo card itis and pancreatitis in a 4-month-old boy. Forensic Sci i M 2006; 156:51-4. 145 Losty HC, Lee P, AJfah am M et al. Fatty infiltrat ion in the liver of medium chain acyl CoA dehydrogenase deficiency. Arch Dis Child 1991; 66:727-8. 146 Matsubaru Y, Narisawa K, Tada K et al. Preva lence of K329E mutatio n in medium-chain acyl- CoA dehydro genas e gen e determ ined from Guthri e ca rds. Lancet 1991; 338: 552-3. 147 Dundar M, Lanyon WG, Con nor JM. Scottish frequency of the common G985 mutation in the medium-chain Acyl-CoA dehydroge nase (MCAD) gene and the role of MCAD
deficiency in su dden infa nt death sy ndrome (SlDS).
] inherit Metab Dis 1993; 16:99 1-3. J 48
Fitzpatrick D. Inborn erro rs of metabolism in the newbo rn: clinical presentation and investigation. ] R Coli PhYSicians Erlillb 2006; 36: 147-51. 149 Emery JL, Variend S, Howat AJ, Vawter GF. Investigation of inborn errors of metabolism in unexpected infant deat h. Lancet 1988; 2:29-3J. 150 Lundemose JB, Kolvraa S, Gregersen N et al. Fatty acid oxidation disorders as primary cause of sudden and unexpected death in infa nts and yo ung children: an investigation performed on cultured fibrobla s ts from 79 children who died aged between 0- 4 years. ] Clin Patl7ol:
Mol Patho11997; 50:212-17. 151 Boles RG, Buck EA, Blitzer MG et aJ. Retrospective biochemical sc reening of fatty ac id oxid ation disord ers in post-mortem livers of 418 cases of sudden death in the first yea r of life.] Pediatr 1998 ; 132:924-3 3. 152 James CL, Keeling JW, Smith JW, Byard RW Total anomalous pulmonary venous drainage associated with fat a l outcome in infancy a nd early ch ildhood - an autopsy study of 52 cases. Pediatr Parhol 1994; 14:6 65 -78. 153 Keeling JW. The perinatal ne cropsy. In Keeling JW, Khong TY (eds) Fetal and Neonatal Pathology, 4th edn. London: Springer-Verlag, 2007, pp. 20-5 3. 154 Byard RW Idiopathic arterial ca lcitlcation and unexpected infant death. Ped Pathol Lab Med 1996 ; 16:985-94. 155 Champ C, Byard RW Pulmonary thromboembolism and unexpected death in infancy. ] Paedialr Child Health 1994 ; 30:5 50-1. 156 Berry PJ , Nagington J. Fata l infec tion w ith echovirus II. Arch Dis Child 1982; 57 :22 - 9. 157 Modlin JF. Perinatal echovirus infection: insights from a litera ture review of 6 1 cases of serious infection and 16 ou tbreaks in nurseries. Re/J Infect Dis 1986; 8:918-26. 158 Wang J, Atchison RW, Walpusk J, Jaffe R. Echovirus hepatic failure in infancy: report of four cases with speculation on the pathogenesis. Pediatr Devel Patl101 2001; 4:4 54 -60. 159 Kaplan MH, Klein SW, McPh ee J, Harp er RG. Group B coxsackievirus infections in infants you nger than three months of age: a serious childhood illness. Rev II/fect Dis 1983; 5:1019-32. 160 Wilcken B, Carpenter KH, Hammond J. Neonatal symptoms in medium chain acyl coenzyme A dehydrogenase deficiency. Arch Dis Child 1993; 69:292-4. 161 Kirk JM , La ing lA, Smith N, Uttley WS. Neonatal presentation of med ium-chain acyl-CoA dehydrogenase deficiency in two families.] Inheri t N/etab Dis 1996; 19:370-1. 162 Vawter GF, Koza kewi ch HPW Aspects of morph o logic va riation among SJDS-v ictims. In Tild o n JT, Roeder LM , Steinschneider A (eds) Sudden Infant Death Syndrome. New York: Acad em ic Press, J 983, pp . 133-44. 163 Molz G, Hal1mann H. Dysmorphism, dysplasia, and anomaly in sudden infant death. N Engl ] /vIed 1984; 311 :259. 164 Molz G, Brodzinowski A, Bar W, Vonlanthen B. Morphologic variations in 180 cases of sudden infant death a nd 180 controls. Am] Forensic Med Patho11992; 13: 186-90. 165 Blair PA, Nadin P, Cole TJ et al. Wei ght gain and sudd en infant death syndrome: cha nges in we ig ht z sco res may identify infants at increased risk. Arch Dis Child 2000; 82:462-9. J 66 va n Baarlen J , Schuurman H-J, Huber J. Acute thymus involution in infancy and childhood: a reliable marker for duration of illness. Hum Pathol 1988; 19: 1155-60.
References I
167 Beckwith JB. Intrathoracic petechial haemorrhages: a clue to the mechanism of death in sudden infant death syndrome? Ann Am Acad Sci 1988; 553:37-47. 168 Kleeman W J, Wiechern V, Schuck M, Trager HD. Intrathoracic and subconjunctival petechiae in sudden infant death syndrome (SIDS). Forensic Sci Int 1995; 72:49-54. 169 lsakasen CV, Helweg-Larsen K. The impact of attempted resuscitation in SIDS: post-mortem findings. In Rognum TO (ed.) Sudden Infant Death Syndrome. Oslo: Scandinavian Unive rsity Press, 1995, pp. 70-80. 170 Werthammer J , Brown ER, Neff RK, Taeusch HW Jr. Sudden infant death syndrome in infants with bronchopulmonary dys plasia. Pediatrics 1982 ; 69:301-3. 171 Berry PJ. Pathological findin gs in SIDS. J Ciin Path 01 1992 ; 45(Suppl.): 11-16. 172 Valdes-Dapena M. The pathologist and the s udde n infant death syndrome. Am J Pathol 1982; 106: 118-3 1. 173 Emery JL. Dinsdale F. Increased incid ence of Jy mphoreticular aggregates in lungs of children found une xpectedl y dead. Arch Dis Child 1974 ; 49:107-11. 174 Krous HF, Nadeau JM, Silva PO, Blackbourne BD. A comparison of respiratolY symptoms and inflammation in sudden infant death syndrome and in acc idental or infli cted infant death. Am J Forensic Med Pathol 2003; 24: 1-8. 175 Gregersen M, Rajs J, Laursen H et al. Pathologic cri teria for the Nordic study of SIDS. In Rognum TO (ed.) Sudden Infant Death Syndrome. Oslo: Scandinavian University Press, 1995, pp. 50-8. 176 Shatz A, Hiss J, Arensburg B. Basement-membrane thickening of the vocal cords in sudden infant death syndrome. Laryngoscope 1991; )01 :484-6. 177 Krous FH. Hauck FR, Herman SM et al. Laryngeal basement membrane thickening is not a reliable post-mortem marker for SIDS. Results from the Chicago Infant Monality study.
Am J Forensic Med Patho11999; 20:221-7. 178 179
180 181
182
183 184
185
186
187
188
Adelson L, Kinney ER. Sudden and unexpected death in infancy and childhood. Pediatrics 1956; 17:663. Valctes-Dapena M, Hoffman HJ , Froelich C, Requeira O. Glomerulosclerosis in the sudden infant death syndrom e. Pediatr Pathol 1990; 10:273-9. Naeye RL. Hypoxemia and the sudden infant death syndrome. Science 1974; 186:837-8. Gilbert-Barness EF, Kenison K, Giulian G, Chandra S. ExtramedullalY hematopoiesis in the liver in sudden infant death syndrome. Arch Pathol Lab Med 1991; 115:226-9. Smith NM, Telfer SM, Byard Rw. A compa rison of the in cidence of cytomegalovirus inclusion bodies in submand ibul ar and tracheobronchial g lands in SIDS and non-SID S autopsies. Pediatr PnthoI1992; 12:185-90. Variend S. Infant morta lity, microglial nodules and parotid CMY-type inclu sio ns. Ea,.ly Hum Deuel 1990; 21 :31- 40 . Takashima S, Armstrong 0 , Becker LE, Huber J. Cerebral white maner lesions in sudde n infant death syndrome. Pediatrics 1978; 62: 155-9. Takas hima S, Armstrong 0 , Becker L, Bryan C. Cerebral hypo perfusion in the sudde n infa nt death syndrome? Brainstem gliosis and vasc ulature. Ann Neuro l 1978; 4: 257-62. Kinney H, Burger P, Harrell F, Hudso n RP. 'Reactive gliosis' in the medulla oblongata of v ictims of the Sudden Infant Deat h Syndrome. Pediatrics 1983; 72: 181-7. Kinn ey HC, Filiano JJ. Brain research in SlDS. [n Byard RW, KrOllS HF (eds) Suddell Infant D~ath Syndrome. Pmblems, Progress and Possibilities. London: Arnold, 2001, pp. 118- 37. Kinney He. Abnormalities of the brainstem serotonergic system in the sudde n infant death syndrome: a review. Pediatr Deuel Pathol 2005; 8:507-24.
223
189 Perez-Platz U, Saeger W, Dhom G, Bajanowski T. The pathology of the adrena l glands in sudden infant death syndrome (SJDS). Inr J Legal Med 1994; 106:244-8. 190 Becker AE, Becker MJ. Fat distribution in the adrenal cortex as an indication of the mode of intra-uterine death. Human Pathol 1976; 7 :495-504. 191 Sinclair-Smith C, Dinsdale F, Emery J. Evidence of duration and type of illness in children found unexpectedly dead. Arch Dis Child 1976; 51 :424-9. 192 Kaplan JA, Fossum RM. Patterns of facial resuscitation injury in infan cy. Am J Foremic Med Pathol 1994; 15:187-91. 193 Bu s h CM, Jon es JS, Co hle SO, Johnson H. Pediatric injuries from ca rdiopulmonary resuscitation . Anll Emerg Med 1996; 28:40-4. 194 Feldman KW, Breweer OK. Child abuse, cardiopulmonary resuscitation, and rib fractures. Pedia trics 1984 ; 73:339-42. 195 Spevak MR, Kleinman PK, Belanger PL et al. Ca rdiopulm o nary resuscitation and rib fractures in infants: a post-mortem radiologic-pathologic study. JAMA 1994; 272:617-18. 196 Price EA , Rush LR, Perper JA, Bell MD. Cardiopulmonary resuscitat ion- related injuries and homicidal blunt abdominal trauma in children. Am J Forensic Med Patho12000; 21 :307- 10. 197 Dow nes AJ, Crossland OS, Mellon AF. Preva lence and distribution of petechiae in well babies. Arch Dis Child 2002; 86:291-2. 198 Perrot U. Masque ecchymotique. Specific or nonspecific indicator for abuse. Am J Forensic Med Patlrol 1989; 10:95-7. 199 Byard RW, Krous HF. Petechial hemorrhages and unexpected infant death. Leg Med (Tokyo) 1999; 1 :193-7. 200 Oehmichen M, Gerling I, Meiilner e. Petechiae of the baby's skin as differentiation symptom of infanticide versus SIDS. J Forensic Sci 2000; 45:602-7. 201 Betz P, Hausmann R, Eisenmenger W. A contribution to a possible differentiation between SIDS and asphyxiation. Forensic Sci Jnt 1998; 91:147-52. 202 Rao V J, Wetli CY. The forensic significance of conjunctival petechiae. Am J Forensic Med Pathol 1988; 9:32-4. 203 Hood I, Ryan D, Spitz WU. Resuscitation and petechiae.
Am J Forensic Med Patho11988; 9:35-7. 204 McIntosh N, Mok JYQ, Margerison A. Epid emio logy of oronasal hemorrhage in the first two years of life: implications for child protection. Pediatrics 2007; 120: 1074-8. 205 Krous HF, Nadeau JM, Byard RW, Blackbolll"ne BD. Oronasal blood in Sudden Infant Death. Am J Forensic Med Parhol 2001 ; 23:346-51. 206 Sugar NF, Taylor JA, Feldman KW, Puget Sound Pediatric Research Network. Bruises in infants and toddlers. Those who don't cruise rarely bruise. Arch Pediatr Adolesc Med 1999; 153 :399-403. 207 Amir J, Katz K, Grunebaum M et al. Fractures in premature infants. J Pediatr Orthop 1998 ; 8:41-4. 208 Koo WW, Sherman R, Succop Pet al. Fracture and rickets in very lo w birth weight infants: co nservative management and outcome. J Pediatr Orthop 1989 ; 9:326-30. 209 Dabezies EJ , Warren PD. Fractures in very low birth weig ht infants w ith rickets. Ciin Orthop Relat Res 1997; 335:233-9. 210 Barry PW, Hocking MD. Infant rib fracture - birth trauma or non-a ccidental injury? Arch Dis Child 1993; 68:250. 211 Bullock B, Schu bert CJ, Brophy PO et al. Cause and clinical characteristics of rib fractures in infants. Pediatrics 2000; 105:e48.
224 I
Sudden unexpected death in infancy
212 Durani Y, Depiero AD, Rib fractu re and birth trau ma, Ann Emerg Med 2006; 47 :2 10,2 15, 213 Campbell CJ, Read DJ(, Lung petechiae - their pathophysiology and diagnostic sig nificance in co t death, Aust Physiol Pharmacol Soc 197 8; 9 : 143, 214 Berry PJ, Intra-al veo lar haemorrh age in sudd e n infant death syndrome: a cause for concern? ) Ciin Pa thol 1999; 52 :553-4, 215 Hanzlick R, Pu lmon ary hemorrhage in deceased infants: ba selin e d ata for fu rther study of infant mo rtali ty, Am) Forensic Med Path 01 2001 ; 22: 188-92, 216 Coffin CM , Schechtman K, Col e FS, Dehner LP, Neonata l and in fantil e pulmo na ry hemorrhage: an autopsy study with clinica l co rrelation , Pediatr Pathol 1993 ; 13:583-9, 21 7 Golde DW, Drew Wl, KJei n HZ et a l. Occult pu lmona ry haemorrh age in leukae mia, BM) 1975; 2: J 66 -8, 218 Keeling JW, Maxwell RS, Busuttil A. Haem osideri n in infant lun gs, ) Pa thol 1998 ; 184:29A, 219 YlIkawa N, Carter N,RlItty G, Green MA. Intra-alveo lar haemorrh age in sudden infant death syndrom e : a cau se fo r concern?) Ciin Pathol 1999 ; 52:581-7, 220 Berry PJ, Lee J, Alveolar s id erop hages a nd the sudd en infant death synd rome,) Pathol 1998 ; 184:27A. 22 1 Stewart S, Fawcett J. Interstitia l haemosid erin in the lungs of sudden infan t death sy ndrome: a histol ogica l hallmark of 'near-m iss' episodes? ) Patlwl 1985 ; 145 :53 -8, 222 Byard RW, Stewart WA, Tel fer S, Beal SM, Assessment of pulmon ary and int ra thymic hemosiderin deposition in sudd en infant death sy ndrom e, Ped iatr Pathol Lab Med 1997 ; 17:275- 82. 223 Becroft DMO, Thompso n JMD, Mi tchell EA. Pulmonary interstiti a l haemosiderin in in fa ncy: a common con seq uence of normal labour. Pedia tr DelJel PatllOl 2005; 8 :448- 52, 224 Becroft DM, Lockett BK, Intra-a lveo lar pulmona ry sid erophages in sudden infant death: a marker for previous impose d suffocation, Pathology 1997; 29 :60-3, 225 Milroy CM. Munchausen syndro me by proxy and in tra al veolar haemosid erin. Int) Legal Med 1999; 1112 : 309-12. 226 Dorandeu A, Peri e G, Jouan H et al. Histolog ical demonstration of haemosiderin deposits in lungs and liver from victims of chronic physica l child ab use. Int) Legal Med 1999; 11 2:280-6 , 227 Fagan DG, Ha emosiderin in pulmonary macrophages from SUD , SIDS and deaths in an fVTVA.) Patho11997; 182:49A.
228 Han zli ck R, Del aney K. Pulm onary haem os iderin in dece ased infants: baselin e data for furth e r study of infant mortality. Am) Forensic lVled Pat/wi 2000; 21:319-22, 229 Schlu ckebier DA. Coo l CD, Henry TE et al. Pulmon ary sid erop hages and unexpected infant death , Am) Forens ic Med Pat/wi 2002; 23:360-3. 230 Krous HF, Wixom C, Chadwick AE et aL Pulmona ry intra alveo lar siderophages in SIDS and suffocation: a San Diego SID S/SU DC project report, Pediatr Deliel Pathol200G; 9:103-J4. 231 Monta na E, Etzel RA, Allan T et aL En vironmenta l risk factors associ ated with pe diatric idiop athic pulmon ary hemorrhage and hemosid eros is in a Cleveland community. Pediatrics 1997 ; 99:3 1-8, 232 Pappas fVlD , Sa rn a ik AP, Meert KL et aL Idiopathic pulmonary hemorrhage in infa ncy. Clinical features and management with high frequency venti lation, Chest 1996; 110 :553 -5. 233 Cutz E, Perrin DG, Vujani c GM, Ackerly C. Idiopathic pulmonary haemos iderosis (IPH) presenting as sudd en unex pected de ath SUD in infancy (Ab stract), Anales Espanoles de Pediatria 1999; 92(Suppl.):52, 234 Mu ir R, Niven J SF. The loca l form ation of bl ood pigments,
) Pa rhol Bacteriol1935; 41:183-97. 235 Sherman JM , Winnie G, Thomassen MJ et al. Time course of hemos iderin production a nd clearance by huma n pulmonary macrophages, Chest 1984 ; 86:409- 1I. 236 Este rley JR, Oppe nheimer EH. Massive pulmonary haemorrhage in the newbo rn, Pathol og ic considerat ions. ) Pediatr 1966 ; 69:3-11. 237 Rushton DI. Liver and gallbl adder. In Kee lin g.JW (ed.) Fetal And Neona tal Pathology, 3 rd edn, Lond on: Springer-Verlag, 2001, pp, 409-39, 238 McCann J , Reay 0 , Siebert J et al. Postmortem perian al findings in children, Am ) Foren sic lVled Path ol 1996; 17:289-98 , 239 Case ME, Graham MA, Handy TC et al. The National Association of Medical Examiners Ad Hoc Committee on Shaken Baby Synd rome, Position paper on fatal abusive head injuries in infan ts and young children, Am ) Foren sic
Med Path o12001; 22:112-22, 240 Bodian M, Hes lop B, Sudde n infan t death syndrome , 1n Siim J-C (ed,) Proceedings of the Eighth 11Iternational Co ngress of Paediatrics, Basel, 1956. Copenhagen: Williams 8: Wilkins, 1960, p. 91.
I
CHAPTER 12
I
SUDDEN NATURAL DEATH IN INFANTS AND CHILDREN Dick Variend
Introduction Cardiovascular causes of sudden death X-linked hypohidrotic (anhidrotic) ectodermal dysplasia Intracranial haemorrhage, neoplasms and malformations Gastrointestinal causes Fatal anaphylaxis Sickle cell disease Haemorrhage as a cause of sudden death Respiratory causes of sudden death Epilepsy and sudden death
225 226 235 236 239 239 240 240 240 242
INTRODUCTION
Sudden death refers to the instantaneous death of an appar ently healthy person, but in practice the definition is extended to include all deaths within 24 hours from the onset of symptoms. The period between the onset of symptoms and the time of death is called the 'terminal event'. The term 'sud den' generally describes the rapidity of the death and the term 'unexpected' describes the surprising nature of the death . There is a wide range in the degree to which death may be expected. Some conditions, most notably cardiovascular in origin, are more prone to cause sudden death, and preced ing knowledge of t heir existence tends to redu ce the level of surprise. By their nature, sudden deaths tend to occur at home, in the ambulance or soon after arrival at hospital. As might be expected, sudden deaths have enormous impact on the family, the community and medical attendants. They are relatively uncommon occurrences but, when they occur, tend to foster suspicion and, in the United Kingdom, are genera lly dealt with by the coroner or procurator fiscal. Many causes of sudden death are age related. Explained sudden deaths in the first year of life are more likely to be due to an infection or congenital abnormality (e.g. congenital
Deaths from acute asthma Diabetes mellitus Genetic metabolic disorders Other bacterial infections Deaths related to obstetric events and premature birth Miscellaneous causes of sudden natural death Sudden unexplained death in older children Sudden natural death in the early neonatal period Sudden death associated with 'intermediate' pathology References
243 243 244 247 248 248 249 249 249 250
heart disease), whereas deaths associated with asthma , epilepsy and diabetes mellitus are more likely to affect the ado lescent years. The terminal event may be non-existent, non-specific or insignificant. Some patients (or their carers) may not fully perceive the severity of the symptoms that precede death, and this may delay summoning medical assistance. A lack of self-perception has particular relevance in conditions such as asthma and hypoglycaemia. Symptoms may be modified by medic atio n. Th e underlying disease mayor may not be symp tom atic. vVhen symptoms are present the terminal event can be variable in duration, depending on the under lying disease. Those dyi ng fro m hypertrophic cardiomyopa thy or congenital aortic stenosis can be expected to have a short terminal event. It might be argued that distinction should be made between sudden death in apparently healthy subjects and in those with recognized preceding illness (e.g. achondroplasia, cardiovascular disease). It is debatable, however, whether the latte r group should be included in the definition of sudden death. The inability of the very young to adequately communi cate symptoms may partly account for their increased vul nerability to sudden death. The discrepancy sometimes
226 I
Sudden natural death in infants and ch ildren
encountered between the account given and the gravity of the post-mortem findings may be explained by failure of the patient (or attend ant) to fully appreciate the seriousness of symptoms . Sudden death is often accompanied by emotional turmoil in the fam ily, and this may affect fam ily members' ability to give a clear account of events leading to death. Relevant information may come to light only after the autopsy has been completed. On the other hand, family members (or med ical attendants) may accurately perceive symptoms but misinterpret their significance. Fo r instance, the parents of a young child may mistake melaena for diar rhoea and not feel the need to ask for urgent medical advice. Severe haemorrhage may thus go unnoticed. A history of chronic disease (e.g. diabetes mellitus, aller gic asthma and epilepsy), w hile clearly very useful, does not always clarify the mechanism of sudden death. The cause of death in such a n individual as determined at post-mortem may be entirely unrelated to the clinically diag nosed entity. Sudden death is well known to be associated with certain syndromic entities (e.g. achondroplasia, La rsen 's syndrome), and knowledge of the relevant pathop hysiology may point the pathologist to the system affected. Deaths occurring during exercise are more likely to be associated with the cardiovascular system and , in such cases, a positive family history or evidence of preceding cardiac dys rhythmia may be helpful. The cause of sudden death at the time of autopsy may be obvious to the naked eye or on microscopic analysis; those without discernible findings require further diligent search or special investigations. After available investigation has been exhausted, those without morphological substrate are suitably referred to as 'physiolo gical deaths' which theo ret ically may cover causes such as cardi ac arrhythmia, con vulsion, laly ngeal spasm or unstable respiratory control. They cover a wide age spa n and may affect the velY yo un g or o ld er child. Medical intervention frequen tly prolongs life following a sudden loss of consciousness or cardiovascular collapse, with death following a period of intensive care. The circu latory collapse and the medical intervention that foUow often produce changes at post-mortem that are likely to mask the picture produced by the initial event. Separation of such ischaemic or reperfusion injuries from changes directly caused by the initial event is important in deter mining the underly ing cause. The results of biochemica l and enzymological studies may be difficult to interpret in such 'lingering' deaths. These cases are generally in cl uded in the definition of sudden death. The distinction between explained or unexplained sudden death often depends on the thoroughness of the autopsy investigation, the experience of the pathologist and avail ab ility of laboratory resources. When facilities are ava il ab le for t he investigation of genetic metabolic disease, the num ber of explained sudden deaths is more likely to increase. In some cases the relevance and significance of certain lesions found at autopsy may be difficult to determine. Pathologists
may, for instance, vary in their interpretation of the severity and/or exten t of certain diseases (e.g. pneumonia or hydro cephalus) and this may have a sig nificant bearing in dete r mining the cause the death in any particular individual. Such t1ndings, on the other hand, may be regarded as coin cidental. Death ma y also be accelerated when infection affects a pa tient with congenital or acqu ired immune deficiency or when the disease is caused by a part icularly vilUlent infec t ive agent.
CARDIOVASCULAR CAUSES OF SUDDEN DEATH Sudden cardiac death in children is far less frequent than in the adult population, and condi t ions that predispose to such deaths in the young are quite different from those preva lent in later years. The preva lence of cardiovascular disease as a cause of sudden death is difficult to determine because of variation in definition of sudde n death and inconsistency in the examination of the conduction system of the heart. In the series of Lambert et al,] the following heart con ditions accounted for 52 per cent of the cases of sudd en death: congenital aortic ste nosis, Eisenmenger's syndrome, cya notic congenital heart disease with pulmonary stenosis and hy pel1rophic obstructive ca rdiomyo pathy. Arrhythmia, hypo xia and coronary insufficiency are recog nized modes of sudden card iac death but, in a substantial proporti o n of the cases, the mechanism is ill-defined. I
Cardiac Malformation Unrecog nized congenital ca rdiac malformation is an impor tant cause of morta li ty in the first year of life. 2 Of 185 cases of infant death from congenital heart disease investigated, 70 per cent had previously diagnosed heart disease; in the remaining 30 per cent, the heart disease was unsuspected or unconfirmed. Of those with unsuspected heart disease about one-half had other severe abnormalities, mainly related to trisomy 13, 18 and 21. Two babies with severe aortic steno sis (both aged 2 months) and interlUption of the aortic arch (aged 6 days) died sud den ly at home. The authors speculated that 200 babi es died each yea r in the UK from unsuspected congenital heart disease. Of these, one-half of the heart defects were unassociated with other malformations. After infancy, death from congenital heart disease is more likely to be due to abnormalities of the coronary arteries, post-operative congenital heart disease, aortic valve stenosis, conduction system abno rmalities and dissection of the aorta complicating Marfan's syndrome, aortic isthmic coarctation or isolated bicuspid aortic valve. 3 The mechanisms of death in this group include co ronary thromboemboli c phenomena,4 fatal cardiac alThyth mias and severe pulmonary vascular dis ease.] Post-operative congenital heart disease is an imp0l1ant cause of cardiac dysrhythmia leading to sudden death;5 the
Cardiovascular causes of sudden death I
defects mostly affected are tetralogy of Fallot, tra nsposition of the great arteries and double-outle t right ventricle. An older age at the time of operation appears to be a significant risk factor for sudden death. Of 20 patients reported by Steinberger et al 6 who died suddenly at the age of less than 12 months, 13 (65 per cent) had a cardiac abnormality, 10 of which were anomalies of the coronary aliery and, of these, 5 were initially consid ered to be sudden infant death syndrome. Congenital aortic stenosis is a rare but well-recognized cause of sudden death,l and preceding symptoms such as syncope, fatigue, shortness of breath a nd dyspnoea on exertion are often present. Dea ths tend to predominate in the second decade of life and are often precipitated by exertional exercise. B Intimal tears and aortic dissection leading to intraperi cardi al rupture and cardiac tamponade is a recognized com plicat ion of undiagnosed aortic coarctation.] Patients with Eisenmenger's syndrome (pulmonary obstructive vascular disease secondary to communication between the systemic and pulmonary circulations) have a moderately high risk of sudden death, with many of the deaths occurring in the sec ond decade. I A ventricular septal defect is a common cardiac anomaly, and the releva nce of its discovery at autopsy in the sudden death of an infant or young child is often difficult to assess. 9.10 Marked cardiac hypertrophy, pleural and intraperi toneal effusions , pulmonary oedema and 'heart failure' cells in the lungs are features of decompensation and implicate the septal defe ct in the cause of death. 9 A multinational study of 186 cases attributed sudden death to ventricular septal defect in two cases (one per cent), suggesting its rarity as a cause. I A ventricular septal defect may also form the basis of Eisenmenger's syndrome, resulting in sudden death much later in childhood or in the second decade of life. I Williams' syndrome is a malformation complex charac terized by typical facies, mental retardation, mild growth deficiency and cardiovascular disease. Patients are at increased risk of sudden death from cardiac deformities that comprise supravalvar aorti c stenosis, bilateral outflow tract obstruction and coronary al1elY stenosis . Myocardial ischaemia, with or without cardiac arrhythmia, from coro nary aliery stenosis is the most likely mechanism of death, II but severe bilateral outflow obstruction may also cause sudden de ath . Abnormal elastin is thought to be responsible for the cardiovascular disease, and the condi tion is linked to the elastin gene on chromosome seven. ll
227
deaths occur in winter months.17 Of 207 cases of sudden death among individuals aged between 1 and 21 years, Neuspiel and Kuller 20 found myocarditiS to be the predomi nant cause of cardiac death. Human myocarditis has been associated with a number of viruses, most commonly of the coxsa ckie group.21 Presenting symptoms are related to the age of the patient, and older subjects may complain of chest pain, but symptoms are more often non-specific, especially in infants.22 Myocarditis may also present with cardiac failure. Pathological findin gs include cardiac dilatation, mottling of the myocard ium and variable opacification of the endo cardium. There is usu ally a diffuse interstitial mononuclear inflammatory infiltrate with a predominance of lymphocytes on microscopic examination (Fig. 12.1).12,16 Interstitial oedema and scattered foci of necrotic muscle fibres are usu ally present. Both ve ntricles and atria may be involved. Viral inclusions are usually absent. The lesion may be foc al or dif fuse and may show a predilection for the endocardium or subepicardial region. Specialized areas such as the conduc tion system may be involved. A subepicardial distribution is often ass ociated with pericarditis. The heart weight is often mildly to moderately increased, suggesting a latent phase of myocarditis despite a short clinical history. 12 However, when involvement is confined to areas such as the conduction sys tem , the heart weight may be normal. 19
Myocarditis Myocarditis is an important cause of mortality in infants l2 and older children,J3-15 and sudden death is a well-recognized presentation. 16,17 This may be related to ventricular asystole, ventricular fibrillation or conduction defects. IB ,19 Infants are more commonly affected than older children, and more of the
Figure 12.1
Microscopic view of the myocardium from a
7-year-old girl who collapsed at home and died soon after arrival in hospital. Extensive mononuclear cellular infiltrat ion of the myocardium is seen.
228 I
Sudden natural death in infants and children
The heart may be specifically affected ('isolated' myocarditis) or cardiac involvement may be part of a gen eralized involvement ('incidental' myocarditis). 12, 17 A prob lem often faced by the pathologist, and for which there is no easy answer, is how many foci of inflammatory cells in the myocardium are sufficient to cause death. 16 Noren et al 17 demonstrated the presence of viral myocarditis as a coincidental finding in two children who suffered violent deaths. This calls for caution in attributing death to viral myocarditis in children who die unexpectedly. It is important to save samples of heart muscle in aJl cases of sudden death for virological investigatio n. Diagnostic methods advocated for clinical investigation are also appli cable at autopsy. These include serology for specific viral antibody, viral culture using tissue or fluid and morphologi cal determination of the vil1Js in infected cells; the last may be au gmented by specific probes for viral antigens or nucleic acid sequences using in situ techniques for probe visualiza tion. 22 Molecular techniques such as polymerase chain reac tion (PCR) have proved valuable in identifying viral genome in formalin-fixed paraffin-embedded tissue.2J - 25 Contraction band necrosis in 'neuro genic cardiomyopa thy' and catecholamine-induced cardiomyopathy may cause minor inflammatory changes that should not be interpreted as myocarditis. 26 ,27 Myocarditis can also occur with any bac terial infection (meningococcus, diphtheria, Staphylococcus aureus, pneumococcus, gonococcus and Haemophilus injlurnzae).28 With the notable exception of diphtheria, car diac involvement in these cases is likely to be incidental. In acute rheumatic fever, pancarditis occurs in 50-75 per cent of children and acute rheumatic carditis may present as sudden death.29 Sudden death in a 6-year-old girl with acute rheu matic carditis complicating by thromboembolic occlu sion of the left anterior descending coronary artery has been described. 4 Acute myocard itis has also been reported in patients with dermatomyositis 30 and Kawasaki's disease.31
Over 50 per cent of cases are inherited as an autosomal dominant trait (familial hypertrophic cardiomyop athy).33,35 The disease may commence in infancy, and death has been recorded in infants as young as 1 year of age. 3 Studies in affected families, however, do not show Signifi cant num bers of infant deaths.32 Genes on five loci on separate chro mosomes are now known to be responsible for the familial disease.32 Of the three genes that have been identified, the best characterized so far is the one that encodes for ~ heavy chain myosin, which is found in abou t 50 per cent of affected families. At least 30 different point mutations are kno wn for this gene. Characteristic of this condition is an increase in ventricu lar muscle mass, with histology marked by myocyte disarray (disorganization), broad and misshap en individual myocytes and circular alignment of myocytes around central foci of connective ti ssue 32 (Fig. 12,2). Interstitial and replacement myocardial fibrosis and acute or subacute myocardial necro sis may be present. 3 The histological changes are maximal in areas of macroscopically thickened wall. The exact mechanism of sudden death in this condition is still unclear, but theories include inappropriate circulatory reflexes leading to haemodynamic collapse,36 arrhythmia an d conduction abnormalities. 37 Cases with the histological char acteristics of HCM may occur in the absence of hypertro phy,16,36 Consequently, extensive histological examination
Cardiomyopathy HYPE RTRO PHIC CARDIOMYOPATHY AND VARIA NTS, Hypertrophic cardiomyopathy (HCM) is a primary myocar dial disease of unknown cause characterized by a hypertro phied, non-dilated left ventricle in the absence of another cardiac or systemic disease,3 2 Symmetrical and asymmetrical forms exist. It is essentially a disease of young subjects, who are often asymptomatic. Clinical findings are ventricular hypertrophy, predominantly involving the ventricular sep tum, with dynamic obstl1Jction to left ventricular outflow, Because of variable phenotypic expression, the incidence of the condition is difficult to assess,33 It is the most common cause of sudden exertional death in yo un g persons, and males are more often affected than females. A combination of young age, a family his tory of sudden death owing to HCM and unsustained ventricu lar tachycardia identifies a subset that is esp ecially prone to sudden dea th. 34
Figure 12.2
Microscopic view of the myocardium showing
myofibre disarray and misshapen myofibres in an 8-year-old girl who died of hypertrophic cardiomyopathy.
Cardiovascular causes of sudden death I
may be required in cases of sudden death to confinn or exclude the di agnosis. A spectrum of disorders is responsible for other cases of HCM that are increasingly attributed to metabolic, genetic and molecular disease (e.g. fatty acid oxidation defects, mitochondrial disease, glycogen storage disease types 2 and 3, Gaucher's disease, GM 1 and GM2 ga ngliosidosis, sialidosis and mannosidosis). Cardiomyopathy occurs in about 25 per cent of patients with Noonan's syndrome J5 and the myocardium in affected patients is histologically indistin guishable from familial HCM.J8 Other associations include Friedreich's ataxia, Turner's syndrome and some disord ers of neural crest tissue. 16 Catecholamine-induced cardiomyop a thy has been reported in patients with phaeochromocytoma. Neonatal cardiomyopathy may follow poorly controlled maternal diabetes. J9 Affected infants are usu ally macrosomic reflecting the growth hormone effect of insulin. An underly ing anatomica l malformation of the heart must be excluded. Ventricular hypertrophy may persist for up to 2 years when such patients are followed by echocardiography.J2 Some vacuolar and hydropic changes of the myocardium are seen microscopically, but the myocardial fibre disarray and inflammatory infiltration that characterize classic HCM do not occur. Cardiac dilatation, which is also sometimes seen, may be secondary to hypo glycaemia and acidosis?9 Sudden death can occur from severe left ventricular hypertrop hy from any cause.J2
229
of the endocardium is seen macroscopically. Twelve cases (seven per cent) of endocardial fibroelastosi s were included in the 186 cases of sudden cardiac death in children sur veyed by Lambert et al. 1 Secondary endocardial fibroelasto sis may be superimposed on cardiac malformation or myocardial disease, and sudden death may be related to the underlying disease. 4J
ARRHYTHMOGENIC RIGHT VENTRICULAR DYSPLASIA (RIGHT VENTRICULAR CARD IOMYOPATHY) Arrthythmogenic right ventricular dysplasia (ARVD) or car diomyopathy is a diso rd er of heart muscle 44 ,45 of unknown prevalence. 46 ,47 The disease is often familial (about 30 per cent of cases) with an autosomal dominant inheritance 46 and appears to be related to abnormal developmen t of the myocardium. 47 A gene for ARVD has been mapped to chro mosome 3p23. 48 Sudden death or a rrhythmia of ri ght ven tricular origin is the usua l mode of presentation. Symptoms ra rely appear before 20 years of age and death often occurs during exertion. 44 ,45 The rate of sudden death is about one per cent of cases per year. 47 The usual macroscopic appearance at autopsy is marked dilatation of the right ventricle with replacement of its free waJl by adipose tissue;49 microscopy confinns the presence of adipose tissue or fibrofatty tissue (Fig. 12.3). Inflammatory
DI LATED CARDIOMYOPATHY Dilated cardiomyopathy, an uncommon disorder in child ren, is characterized clinically by biventricular dilatation and reduced myocardial contractility.4o It acco unts for a dispro portionate number of deaths in paediatric practice, and the condition has been associated with sudden death,5,41 altho ugh the causative mechanism is poorly understood. Macroscopic pathological features are an increased left ventricular mass and decreased left ventricular wall thick ness.J2 Myocyte hypertrophy, myocyte myofibrillaJY loss, interstitial fibrosis a nd increased numbers of interstitial lym phocytes co nstitute the histological picture. Interstitial fibrosis, myocyte hypertrophy and absence of myocyte damage per mits the condition to be distinguished from acute myocarditis. As in hypertrophic cardiomyop athy investigation should be directed at an underlyin g cause,40 but in the majority of children dilated cardiomyopathy is idiopathic. A favoured view is that some cases are a sequel to viral myocarditis. 16,22,42
"\ -'4
\.. . ,
""\
. ...
'
"
1
'
,~
i('
~
'~
ENDOCARDIAL FIBROELASTOSIS Endocardial fibroelastosis is customarily divided into pri mary and secondary types. 29 Most infants with the primary form of the disease die in the first year of li fe and sudden infant death is a common presentation. The affected ventri cle may be constricted or dilated and fibro elas tic thickening
Figure 12.3
The fre e wall of the right ventricular myocard ium is
extensively replaced by adipose tissue consistent with a diagnosis of arrhythm ogenic right ventricular dysplasia.
230 I
Sudden natural death in infants and chi ldren
cells are seen in about 25 per cent of cases, raising the possi bility of an infectious or a genetical ly determined immune process in some of the cases. Heart weight is normal or only moderately increased.45 Similar histological features are rarely seen to involve the left ventricle.47 As some degree of adipose tissue is often seen in the free wall of normal right ventricle, the amount of adipose tissue required for the diag nosis of ARVD is sometimes unclear. 49 ,50 ARVD and Uhl 's anoma ly are now considered to be two distinct morphological entities. 49 Uhl 's anomaly refers to absence of the right ventricu lar parietal myocardium, the endocardium and epicardium being directly apposed; the condition usually presents in neonates and infants with congestive cardiac failure . HISTIOCYTOID CARDIOMYOPATHY
Histiocytoid card iomyopathy (or oncocytic cardiomyopa thy), is a rare condition beset with a multitude of syn onyms, The entity is sometimes regarded as a tumour of the myocardium (see below), 51 Females less than 24 months of age are chiefly affected. 52 The clinical presentation is vari ab le, with congestive cardiac failure, arrhythmias or sud den death all being encountered. In a large revlew,52 68 per cent of the patients were reported to experience cardiac arrhythm ia before their death, Sudden death occurred in 22 per cent of the cases. The aetiology is unknown, but the characteristic myocardia l changes may represent the final common path way of a number of different aetiologies,53 Familial occur rence has been reported S3 At auto psy the heart is hypertrophied with left ventric ular prominence. The myocardium is pale and may show tan -yellow nodules. The endocardium may be thickened. Light microscopy shows groups of large polygonal cells with foamy to granular cytoplasm scattered within the myocardium (see Fig, 11.2, p. 206) . Their distribution may be diffuse, focal or multifo cal. Any part of the heart can be involved and the conduc tion system may be affected predominantly.54 Ultrastnrctura ll y, the swollen cells contain numerous enlarged mitoc hondria and myofibrils and occasional lipid vacuoles that are displaced towards the cell periphery. Sma ll dense bodies may be found within the mitochondria. Accompanying cardiac malformation has been described, and the patient reported by Koponen and Siegel (1996)54 had Peter's anomaly and congenital glaucoma. A defect in complex III of the respiratory chain has been demonstrated in one case, 52 However, the condition differs from mito- .. chondrial cardiomyopathy in several respects: • In mitochondrial cardiomyopathy, all myocytes are affected; whereas in histiocytoid cardiomyopathy, involvement is focal. • In mitochondrial cardiomyopathy, the mitochondria are consistently abnormal in shape. In histiocytoid cardiomyopathy, they are stnrctura lly normal or abnormal but are numerically increased. 52
(a)
(b)
(c)
Figure 12.4 Different abnorma l ang les of origin of the coronary artery wh ich may be associated with sudden death. (al. normal ang le; (b), angle of origin less than 30°; (c), coronary artery is invested in the aortic adventitia as it departs at an acute angle.
Anomalous Coronary Arteries Sudden death associated with anomalous coronary arteries covers a wide age range (neonates, infants, children and young adu lts). Death is commonly associated with exertion and the anomaly is an important cause of sudden death among athletes. 41 Sudden infant death syndrome may be mimicked in the very young,55.56 indicating that death is not always related to exercise. In one autopsy series the inci dence of anomalous coronalY arteries as a cause of sudden infant death was estimated to be 0.4 per cent. 56 One patient reported by these authors had coexisting vascular abnorma l ities. A wide range of anomalies were described: 3,6,41.57,58 (I) an ectopic origin from the wrong coronary sinus, eccentric origin from the correct sinus or a high aortic origin; (2) ostial stenosis, i.e. an internal diameter that is smaller than the artery 1-2 mm from the aorta (this includes a slit-like ostium); (3) an abnomral angle of origin (Fig. 12.4), i.e. the ang le between the aortic lumen and coronary ostial lumen (normally 90 degrees) is between 30 and 45 degrees (proba bly abnormal) or less than 30 degrees (definitely abnormal); or (4) an ostial 'flap' or 'ridge' - a type of ostial stenosis usu ally associated with an acute angle of aortic origi n or origin from the wrong aortic sinus. These various lesions may occur alone or in combina tion, as exemp lifi ed in a 12-year-old child who died sud denly and was shown to have several different anomalies of the left coronary artely. These included a high aortic ori gin, an aortic intramural segment ('intussusception'), inter positio n between the pulmonary artery and ao rta, as well as an intramural course. 58 Lesser degrees of complexity are also descri bed. 55,57 An ectopic origin (anomalous origin from the wrong aortic sinus) is the most common pattern, and origin of the left coronary artery from the right coronary sinus is more
Cardiovascular causes of sudden death I
RCA (a)
8r
LCA
LAD
RCA
(bl Figure 12.5
LAD
Anomalous origin of the left (a) and right (b)
coronary artery from the wrong coronary sinus and their relation ship to the aortic and pulmonary trunks.
common than origin of the right coronary artery from the left coronary sin us. 59 The following, singly or in combina tion, may contribute to ischaemia of the myocardium: a stenosed ostium or one with an acute angle of take-off; a course interposed between the pulmonary and aortic trunk (Fig. 12.5); and investment of the coronary artery in the aortic adventitia as it departs at an acute angle. An origin of one coronary artery from the other close to its com mencement is also reported. 56 Coronary arteries arising from the wrong coronary sinus causes sudden death from ventricular fibrillation owing to poor perfusion of the mus cle mass. Over two-thirds of patients do not have a history of syncope or chest pain ; a history of palpitations may, however, be apparent in old er subjects. Because of the sma ll size of vessels and difficulty with dissection in sma ll babies, the anomal ies are easily over looked, especially when the coronary ostia are normally located. Infants do not usually show myocardial ischaemia on microscopic examination. 56 By contrast, foci of myocar dial necrosis, fibrosis or contraction bands are often found in the ventricular myocardium of older individuals. 3.5.41 . Other anoma lies of the coronary arteries associated with sudden death include aplasia/hypop lasia, aneurysm and fis tula. The significance of intramyocardial tunnelling (intra mural coronary artery) as a cause of sudden death is controversial. 5o Some consider an intramyocardial mid segment of the left anterior descending artelY as a normal variant,6l while others view as significant a deep and long intramyocardial course.57 The latter authors recorded an
231
intramyocardial course from 10 to 25 mm and a depth from 4 to 8 mm in cases of sudden death. While these deaths were mainly of older individuals, one was a 14-year-old boy. Origin of the left coronary artery from the pulmonary trunk (Bland-White-Garland syndrome) is a rare malforma tion. The relative low pressure of the pulmonary artery is inadequate for perfusing the left ventricular myocardium and most affected subjects die in infancy from cardiac failure. 29
Kawasaki's Disease This is a systemic febrile vasculitis with multisystem involvement, whose cause is unknown. Children under the age of 5 years are predominantly affected. 3l Eighty per cent of patients are diagnosed before the age of 5 years. 52 The condition is reported world-wide and, in developed coun tries, males are affected more often than females. It is the most common cause of acqu ired heart disease in children. 53 Kawasaki's disease and infantile polyarteritis nodosum are now considered to be the same entity. An infective aetiol ogy is suggested by the observed clustering of cases. The incidence in Blitain is estimated to be 3.4 per 100000 chi ld ren aged less than 5 years.63 The case fatality in Britai n in 1990 was 3.7 per cent. Six deaths were reported in Britain in 1990; in only one case was the diagnosis made in life. Clinical features are the presence of fever for about 5 days, lymphadenopathy, characteristic changes in the peripheries (erythema of the palms and soles, desquamation of the fin gers and toes and pelipheral non-pitting oedema), bilateral conjunctivitis, buccal inflammation (injected pharynx, dry cracked lips, strawberry tongue, mouth ulcers) and a poly morphous exanthem. 64 Hydrops of the gall-bladder is increasingly recognized. Three-quarters of affected patients show a transient rise in the serum concentration of IgE. 62 Sudden clinical onset is usual. There are four major stages in the development of the disease: 10 • an 'acute febri le' stage which lasts 1-11 days; • a 'subacute' stage which lasts 11 to 21 days and is associated with thrombocytosis; • a 'convalescent' stage from 21 to 60 days; and • a 'healed ' stage. Vasculitis is often widespread but the coronary arteries are always affected 52 (see Fig. ILl, p. 206). The heart is the most severely affected organ, and coronary arteritis with aneurysmaJ formation occurs in 20-30 per cent of cases. 53 Aneurysms more commonly affect the left coronary artery. The prognosis for resolution is favourable unless the aneurysms are 'giant', as these are more often associated with myocardial infarction. 3! Recurrences may occur,65 and coronary aneLllysms/dilata tion and giant aneurysms are more common in this subset. Other organs often affected are the kidneys, spleen, testis, pancreas, adrenal glands and liver. There is correlation between the severity of the autopsy findings and the duration of the illness. Most deaths occur in
232 I
Sudden natural death in infants and children
the subacute and heali n g stages. Sudden unexpected death may occur many years later, following recanalization of thrombosed coronary arteries,lo and sudden death has been reported in a fit athlete during exercise with evidence of pre ceding Kawasaki disease. JI Overlapping features, including the clinical demonstra tion of coronary aneurysms, have been reported bet"feen group A streptococcal infection and Kawasaki's disease. 66
Abnormalities of Cardiac Conduction While myocardial disease, such as hypertrophic cardio myopathy, myocarditis, histiocytoid cardiomyopathy and tumours, may cause sudden death consequent upon involvement of the conduction pathway, abnormalities of the cond uction tract may coexist with certain types of con genital cardiac malformations or the conduction tissue may be damaged during their operative repair. 29 Fibromuscular hyperplasia of the sinoatrial nodal an d atriovent ricul a r nodal arte ri es may cause paediatric sudden death from car diac arrhythmia. 67 On the other hand, fatal arrhythmias may result from intrinsic defects of the conduction tissue that are not obvi ous to the naked eye. 30 .6B For precise diagnostic categoriza tion, such cases require formal examination involving extensive serial sectioning. In practice, this is time-consuming and associated with such a low positive yield that it does not gene rally form part of the routine examinatio n of the heart in cases of sudden death. Conduction defects may thus be easily overlooked and many such deaths are simply referred to as showing no structural cardiac disease. 41 The long QT syndrome (LQTS) is significantly associated with ventricular fibrillation and the risk of sudden death. The condition can be inherited as an autosomal dominant disorder (Romano-Ward syndrome),69 with ab norm ali ties at six different loci identified.1° A recessively inherited syndrome w ith deafness is also recognized .7o Vigorous physical exertion often precedes vent ricula r fibrillation. In these cases there is no discernible abno rmality on histolog ica l examination of the conduction pathway. IS Risk factors for sudd en death in such indi vidua ls include a history of syncope, congenital deafness and a fami ly history of sud den cardiac deathS Preceding symptoms th at suggest a conduction defect include syncope and dizziness. A preceding abnormal elec trocardiogram (ECG) recordin g or a family history of con duction defect is valuable in these cases, and likely to ex plain an arrhythmic cardiac arrest. In their absence, the significance of any histological cha nges of the conduction tract is merely speCUlative. Opportunity for clinical diagno sis in these cases is precluded by the intervention of sudden death of the patient. Accessory pathways of th e conduction system may cause re-excitation and re-entry lea ding to fatal arrhyth mias. The Wolff-Parkinson-White pre-excitation syndrome
may occur in isolation or be associated with other abnor malities, for example Epstein 's anomaly, rhabdomyoma of the heart. These abnormal pathways produce early stimula tion of the ventricle and are associated with a variety of electrocardiographic patterns. Many patients remain asymp tomatic however, but sudden death in this group is well recognized. Heart block may be congenital or result from acquired causes. Marked degrees of atrioventricular (AV) block may produce Stokes-Adams attacks and are rarely caused by endodermal heterotopia of the AV node (mesothelioma of the AV node). Some congenital abnormalities of the conduction system (e.g. simple absence of the short segment of the right bundle branch or the penetrating portion of the main bundle of His) are stable and do not seem to progress. IS Babies born to mothers with systemic lupus erythemato sus are at considerable risk of congenital AV block but the prognosis is rem arkab ly good in such cases, and those affected may not develop syncopal attacks until much later in life. Death among infan ts with preceding congenital heart block or those w h o die in association with maternal connective tissue disease should be tested for anti-La (SS-B), anti-Ro (SS-A) and even anti - UIRNP antibodies. 29 Heart block has been described w ith Kearns-Sayre syndrome, Kartagener's syndrome and certain X-linked myopathies. 10
Occlusive Disease of the Coronary Arteries Infantile arterial calcification is a rare disease mainly affect ing infants,71 with reports of siblings affected. Low levels of plasma cell membrane glycoprotein-l nucleoside triphos phate pyrophosphohydrolase have been demonstrated in an affected infant. 72 The same group subsequently demon strated mutations in the EN??] gene 7J Alieries throughout the body are involved, with the exception of the brain and spinal cord. Of the 62 cases reviewed by Moran (1975),71 10 presented as sudden death, preceded by respiratory distress in 5. The age of death ranged from 2 days to 28 months, with 85 per cent of affected infants dying within 6 months. Clinical diagnosis is possible with rad iological study. The ECG changes, when available, are mostly compatible with myocardial ischaemia. Cardiac enlargement is common and the myocardium is often infarcted. Intra-uterine ultrasono graphic diagnosis has been made in familial cases. 74 Microscopically, there are calc ific deposits with fragmen tation of the internal elastic membrane of arteries along wit h variable degrees of fibrointimal proliferation and luminal narrowing. 71 ·75 An inflammatolY reaction is typically absent. An alieriopathy with features similar to those found in idio pathic infantile arterial calcification has been reported in children with acquired immun e deficiency syndrome (AlDS);76 one of the cases described by these authors showed myocardial infarction associated with thrombus formation in a coronary aneurysm.
Cardiovascular causes of sudden death I
233
Table 12.1 Causes of coronary artery dysplasia in infants and children Fibrous muscular dysplasia Idiopathic arterial calcification Chronic arsenic poisoning Congenital rubella Tuberous sclerosis Neurofibromatosis Acquired immunodeficiency syndrome Homocystinuria Down's syndrome Menkes'syndrome
Fibromuscular dysplasia is a segmental, non-atheros clerotic vascular disease of unknown aetiology mainly causing renal vascul ar disease, but other arteries may be involved. Young adults are usually affected and its occur rence is rare in infancy a nd childhood. Myocardial infarc tion due to fibromuscul ar dysplasia affecting the coronalY artery has been reported as a cause of sudden death in infants and children.77,78 Myocardial infarction with cardiomega ly has been reported in children with chronic arsenic exposure from drinking water. 79 This occurred as part of a ge neralized arterial disease of irregu lar distribution and progressive course, but apparently sparing the arteri es of the lung and brain. Microscopically, there was intimal fibroblastic thick ening of the media of small-sized arteries. The endothe lium, internal elastic lining, media and adve ntitia were uninvolved. Widespread vascular dysplasia in a child with tuberous sclerosis leading to variable luminal nan-ow ing of the coro nary, superior mesenteric, renal and common iliac arteries has been reported;80 these a uthors described a 9-month-old infant who developed an aOliic aneurysm. Fatal haemor rhage followed dehiscence of the aortic graft. Occlusive artelial disease has also been reported in patents with homocysteinuri a,8 J and followin g congenital rubell a syndrome, S2 neurofibromatosis 1,83 Menkes' syn drome 84 and, rarely, Down's syndrome. 85 A list of conditions that may cause abnorm alities of the corona ry arte ries is presented in Table 12.1. Corrado et al 57 referred to the sudden death of a 14 year-old g irl whose coronary ostia were nan-owed by Ta kayasu arteritis.
Tumours of Cardiac Muscle ------Primary cardiac tumours are uncommon but are associated with potentially lethal conditions producing conductive and haemodynamic abnormaliti es that often result in sud den death.
Figure 12.6 The cut surface of the heart sho ws multiple tumours of the myocardium alternating with areas of congestio n from an infa nt who died from cardiac rhabdomyomas.
ENDODERMAL HETEROTO PIA OF THE ATRIOVENTRICULAR NOD E Previously referred to as mesothelioma of the AV node, the lesions are frequently associated with heart block and sud den death. Gross examination of the heart may show a small raised nodule immediately above the septal leaflet of the tri cuspid valve, just antelior to the coronary sinus in the right atrium. Microscopically, the lesion comprises multiple cysts, gland-like structures or nests of epithelioid cells within a fibrous stroma. 86 The tumours may be missed if sections of the AV node are not exa mined in cases of sudden death. These lesions have now been convincingly shown to be due to endodermal inclusions. Associated cardiovascular malfor mation may be present. 87 The entity is sometimes refen-ed to as the 'smallest tumour causing sudden death'. RHAB DOMYOMA Rhabdomyoma is the most common tumour of the cardiac muscle. In 30 per cent of cases, death occurs at birth or in the neonatal period and one-third of cases are associated with tuberous sclerosis.s7 The tumours may be isolated or multi ple (Fig. 12.6), but diffuse infiltrative patterns have been
234 I
Sudden natural death in infants and children
Figure 12.8 The left ventricular myocardium from an 8-month old boy shows a large well-circumscribed tumour with a solid trabeculated cut surface typical of cardiac fibroma.
Figure 12.7 The microscopy of cardiac rhabdomyoma shows 'spider' cel ls chara cte ri zed by ce lls with clear to vacuolated cytoplasm wi th myofibrillar strands extendi ng between nuclear and cytop lasmic membrane and a central or eccentric nucleus. described. The cause of death is outflow tract obstmction or cardiac arrhythmia. Microscopically, large cells with clear cytoplasm and small nuclei (so-called 'sp ider cells') consti tute the microscopic picture (Fi g. 12.7). Most cases of sudden death due to the tumour are sporadic. B6
FIBROMA The seco nd most common cardiac tumo ur, fibroma of the heal1, also ca uses death secondary to outflow tract obstruction or arrhythmia. A well -circumscribed grey-white mass betvveen 3 and 10 cm in diameter is_ the usual gross appearance (Fig. 12.8), cysts may occur. The tumour is most often located in the interventricular septum. Microscopy shows fibroblast-like cells in a colla genous matrix, reminis cent of fibrom atoses of the soft tissues (Fig. 12.9).B6,87 The tumour may occur as a component of Gorlin's syndrome.
MYXOMA Myxomas are rare in infants and children. s7 Most occur as sporadic lesions, arise withi n the left atrium and may reach up to 10 cm in diameter. Su dden death is related to dis turbed cardi ac haemodynamics or systemic emb olization of tumour fragments. Myxomas rarely involve the right side of the heart, where they are capable of causing massive pulmonary embo lism. Bs
-
"
- "'. . Figure 12.9 Histology of the cardiac fibrom a shows fibrobl ast like cells in a coll ag enous matrix closely resemb li ng fibrom atosis.
OTHER CARDIAC TUMOURS Cardiac haeman gioma is a rare cause of sudden death in child re n. A previously asympto matic 13 -year-old girl who was found unresponsive in bed was reported by Krous et a\. B9 Autopsy showed haemangiomatous invo lvement of
X-linked hypohidrotic (anhidrotic) ectodermal dysplasia I
the ventricular conduction system. Death was thought most likely to have been arrhythmic in origin. Swalwe11 90 repolted a benign teratoma of the interven tricular septum causing sudden death in 2.5-year-old girl who had no preceding medical problems. Krous et al 51 analysed 68 cases of tumour-associated sud den deaths in children from the literature. Of those involving the heart, there were 23 cases of oncocytic cardiomyopathy (histiocytoid cardiomyopathy), 12 cases of cardiac fibroma, seven cases of rhabdomyoma, three cases of myxoma and one case each of neuroma, teratoma and mesothelioma. In one case of oncocytic cardiomyopathy, one case of cardiac fibroma and the case of neuroma the lesion was considered incidental and unrelated to the cause of death.
, .
•, I
I '
•,
j
.
(
\.
!
fI
.!.' .I
.."
•
J
f
•
t -
\' ' ,
I -~
.
il Pulmonary thromboembolism is an unusual cause of sudden death in children. Predisposing factors include recent sur gery, congenital heart disease, indwelling venous catheters, sepsis, arteriovenous malformation and occult malignancy.91 Of 17 500 autopsies reviewed by these authors, eight cases (0.05 per cent) showed pulmonary thromboembolism; their ages ranged from 1 month to 13 years. The source of the embolus may not always be demonstrated at autopsy. Unexpected death from massive pulmonary embolism was reported in three infants.92 The underlying conditions were necrotizing enterocolitis , a ventriculoatrial shunt for hydrocephalus and idiopathic arterial calcification. Birth control medication may predispose to pulmonary embolism in teenagers. 59 Sudden death from thromboembolism of the left anterior descending coronary artery complicating acute rheumatic heart disease in a 6-year-old girl and complete atrioventricular canal defect in an 18-year-old girl with Down's syndrome were reported by Stahl et al 4 Thrombosis and pulmonary embolism is a recognized complication of homocysteinuria (cystathionine B synthase deficiency).81 Any blood vessel can be affected, even the intracranial dural sinuses. Microscopically, the arterial lesion comprises marked fibrous intimal thickening, split ting and fraying of the muscle fibres , in the media with increased interstitial collagen. The internal elastic lamina may also be affected. Lipid deposition is not a feature. Potential causes of sudden death in this condition include pulmonary embolism, myocardial infarction and cere brovascular accident.
~,
I!
I
, " I
I
l
,
I
~
,
,.
J
, t ..
£ ~
., I
, G
,,
-/
.1. I
Thromboembolism
~
1
~
\
~
t
•
~ II
[
•
.'
# ~
,
«.
\
\
I , ,.
t
.I
\
235
~
.-
I
"
.
.
("'
\
'0
•
.'..· t
,
,
~
~
I
oJ
Figure 12 .10 Aortic wall with cystic medial necrosis. Death was due to cardiac tamponade secondary to aortic rupture and an undiagnosed aortic coarctation in a 14-year-old boy.
90 per cent of the mortality is related to cardiovascular com plications that include mitral valvular regurgitation or aortic dilatation, dissection and nIptlire. Aortic dissection with nIpture and haemorrhage into the pericardial cavity causing cardiac tamponade is a common mode of death.3.4 1,93 An intimal tear is usually located in the ascending aorta and histological examination of the aortic wall shows cys tic medial necrosis, characterized by fragmentation of the elastomuscular media and accumulation of basophilic material. Spontaneous dissection of the coronary a11ery rarely causes sudden death, 57 but this is not confined to cases of Marfan's syndrome. Cystic medial necrosis may occur with aortic coarctation (Fig. 12.10) and in patients with an isolated bicuspid a011ic valve in the absence of Marfan's syndrome. 3 Patients with Ehler-Danlos type IV syndrome are also at increased risk of vascular nIpture. 8
Marfan's Syndrome
X-LINKED HYPOHIDROTIC (ANHIDROTIC) ECTODERMAL DYSPLASIA
Patients with Marfan's syndrome display a characteristic habitus: tall stature, arachnodactyly, bilateral ectopia lentis, high arch palate, dolichostenomelia and striae distensae. Linkage analysis has shown that the fibrillin gene on chro mosome 15 is associated with Marfan's syndrome. 8 Up to
The condition is characterized by an absence or diminution of eccrine sweat glands, eczematoid rash, thin sparse hair, oligodontia and peg-shaped teeth. The incidence is estimated to be in 1 per 100000 live births.94 The condition carries a sub stantial mortality and morbidity, and death occurs in about
236 I
Sudden natural death in infants and children
Figure 12.11 Mid-sagittal view of brain from a 12-month-old girl with achondropl asi a. There is marked forebrain hydrocephalus and necrosis of the upper cervical cord due to a small foramen magnum.
30 per cent of cases. 95 There is an increased risk of chest infec tion and atopic disease. Because of the deficient number of eccrine sweat glands, patients are unable to perspire and con sequently develop fatal heat intolerance 94 Mucus production is also deficient in the respiratory and gastrointestinal tracts. An inexplicable improvement in the patient's general condi tion occurs with increasing age.
Achondroplasia Achondropl as ia is the most common type of osteochon drodysplasia; it is inherited as an autosomal dominant tra it. Recent studies suggest that sleep apnoea is common among affected individuals,96 involving both obstructive and cen tral mechanisms. Compressive myelopathy and dysfunction at the cervicomedullary junction is caused by a small fora men magnum, and sudden infant and childhood deaths have been reported. 97 Macroscopic or microsco pic necrosis of the spinal tissue at the level of the foramen magnum may be seen (Fi g. 12.11).
Tuberous Sclerosis Complex Tuberous sclerosis is a dominantly inherited condition characterized by genetic heterogeneity. The birth incidence is about 1 in 6000. More than 70 per cent of cases will be new mutations. Early skin manifestations are hypomelanic macules and shagreen patch. Subungual fibromas and facial angiofibromas occur later. Cardi ac rhabdomyomas, giant-cell astrocytomas and renal disease are other patho logical lesions .93 The polycystic renal disease of tuberous sclerosis usually affects infants and young children, while the angiomyolipomas tend to occur at or after puberty. At least 80 per cent of children presenting with cardiac rhab domyomas will have tuberous sclerosis. 93 The incidence of
tumours in patients with tuberous sclerosis approaches 50 per cent,99 but their size and number tend to regress with age. The majority of individuals presenting in childhood h ave epileptic seizures. 98 The preCise mechanism of death is not always clear. Sud den un expected death in infancy has been reported to occur either from cardiac arrhythmia or blood flow obstruction secondary to rhabdomyoma. Patients with tuberous sclerosis may develop Wolff-Parkinson-White syndrome, which may initiate fatal cardiac arrtbythmia and is probably related to the presence of the cardiac tumour. 100 Rarely, a renal angiomyolipoma ruptures with life threatening retroperitoneal haemorrhage. 98 Aortic aneurysm secondary to vascular dysplasia in a child w ith tuberous sclerosis has been reported. so
INTRACRANIAL HAEMORRHAGE, NEOPLASMS AND MALFORMATIOI\IS Sudden unexpected death m ay result fro m intracranial haemorrhage secondary to a generalized bleeding disor der lOI or the haemorrhage may be related to intracranial pathology (Fig. 12. 12). Contrasting with subdural or extradural haemorrhage, sudden unexpected death follow ing parenchymal haemorrhage is more likely to be associ ated with natural disease . Arteriovascular malformations (Fig. 12.13) and aneurysms a re important sources of cata strophic bleeding. 2o ,JOI Intra-crani al a neurysms are found with uncorrected aortic coarctation, familial multiorgan cystic disease or autosomal dominant polycystic kidney disease (ADPKD). While fatal subarachnoid haemorrhage in ADPKD usually presents in adulthood,I02 infants and children are rarely affected. A family history is extremely helpful in making a diagnosis. The cerebral anelllysm may be masked or destroyed by the haemorrhage and, consequently, may be difficult to find at post-mortem examination. Intracranial aneurysms in childhood have been reported to occur as a complication of renal hypertension . 103 The intracranial pathology mayo r may not be clinically apparent prior to sudden death.l04 These authors reported 10 deaths from intracrania l haemorrhage, secondary to tumours (four cases) , berry a neurysm (one case) and vascu lar malformations (five cases); the presence of four vascu lar malformations was assumed at autopsy as their positive identification was difficult owing to the marked tissue destruction. Such haemorrhages may be due to micro angiomas. lOl An additional cause of intracranial bleed in older female patients is eclampsia. Four cases of intracerebral haemorrhage reported by Liv ingston and Brown 101 compbcated a pre-existent coagulopa thy. In two cases this was thrombocytopenia (associated with aplastic anaemia and acute lymphoblastic leukaemia) ; one child had haemophilia and one case was thought to be a late manifestation of haemorrhagic disease of the newborn. Minor trauma was a factor in the case of the haemophilia.
Intracranial haemorrhage, neoplasms and malformations
I
2 37
J8,... Figure 12.12 Cut surface of the cerebellum shows haemorrhage into a tumour which wa s subsequently demonstrated to be low grade astrocytoma. The patient was a 4-year-old girl who had apparently been in good health and had died suddenly.
o
Figure 12.13 Microscopic view of an arteriove nous malformation which caused catastrophic intracerebral haemorrhage in a lO-year old girl. There is vascular prominence with irregular thickening of some of the vessels. Elastic van Gei son stain (EVG). The clinical course of idiopathic thrombocytopenic purpura is rarely complicated by intracranial haemorrhage. The inci dence is estimated to be 1 in 1000 patients. lOS Sudden death from undiagnosed primary intracranial tumours per se rarely occurs. Abu al Ragheb et al (1986)106 found seven cases of sudden death secondary to intracranial
Figure 12.14 The tip of the intraventricular part of a ventriculoperitoneal shunt that had become ensnared in choroid plexus. causing obstruction to cerebrospina l fluid egress in a young child with shunt- treated hydrocephalus. neoplasm in 1055 autopsies. including one in a 6-year-olcl and one in an 18-year-old. Preceding symptoms mayor may not be present and, when present, relate mainly to increased intracranial pressure. epilepsy, focal neurological deficit or psychiauic manifestation. l07 Loss of consciousness may pre cede sudden death. Sudden mass effect, haemorrhage into or adjacent to a tumour l04 obstructive hydrocephalus lO8. 109 and epiIepsyilo are some of the mechanisms responsible for death. An unusual case of spontaneous sub a rachnoid haemor rhage in a 5-month-old girl who died suddenly was reported by Byard et al; III collagen analysis showed virtual absence of type III collagen, diagnostic of type IV Ehlers-Danlos syndrome. The Dandy-Walker malformat ion is a well-recognized but a rare cause of sudden death .112 Autopsy examination may provide the first indication of the presence of the mal formation. The mechanism of death is unclear, but brain stem ischaemia from local pressure abnormalities in the unshunted posterior fossa may be responsible. Sudden death occasionally occurs in children with a colloid cyst of the third ventricle. I08 Owing to its critical location in the anterior part of the third ventricle, the lesion may obstruct cerebrospinal fluid drainage and lead to acute hydrocephalus. Histologically, these are unilocular cysts conta ining eosinophilic debris and lined by cuboidal/columnar epithelium, which may be ciliated. Hydrocephalus may be congenital with and without spina bifida. Acquired hydrocephalus may follow or accompany intraventricular haemorrhage, meningitis or neoplasms. Patients may present with respiratory arrest. 113 Focal and generalized convulsions are common with shunted hydrocephalus and may result in status epilepti 114 Factors precipitating convulsions include infection CUS. and shunt-related complications (Fig. 12.14).
238 I
Sudden natural death in infants and children
Encephalitis and Leucodystrophies Sudden deaths in infancy and childhood are rarely ascribed to encephalitis. The clinical course of herpetic encephalitis in chiJdren or young adults may be rampant, and in the neo natal period severe general ized disease caused by herpes simplex virus infection can also be rapidly fatal. Cerebral involvement as part of a severe disseminated infection in the neonate may a lso occur with coxsackievirus infection, a nd there is predilectio n for involvement of the brainstem and spinal cord. liS Sudden death at home in two siblings associ ated with an encephalitic process has been reported by Howat et al. I1 6 An aetiological agent was not discovered. Shields et a l l1 7 reported sudden death in two female patients secondary to the autosomal recessive form of adrenoleucodystrophy and to Alexander's disease.
Adrenal Hypoplasia/Insufficiency Adrenal hypoplasia in infancy may be secondary to brain malformation or occur in isolation. liS Weakness, weight loss, hyperpigmentation , hypotension, gastrointestinal symptoms and vitiligo are the main symptoms of chronic adrenocortical insufficiency. 119 By contrast, the symptoms and signs of acute adrenal failure are non-specific and are those found in any serious illness. Adrenocortical insuffi ciency presenting later in chi ldhood is more likely to be of autoimmune origin (Addison 's disease) .11 9.120 Favara et al 121 reported severa l deaths in infants, with an age range from a few hours to 24 months, with undiagnosed adrenal hypoplasia without central nervous system malfor mation. There was a strong male predominance. Sudden death was recorded in four of the infants. The morphology of the adrenal glands was variable and all were normally shaped but miniature in size. In a ll cases the combined adre nal weight was less than 2 g. One patient died during induc tio n of anaesthesia. Two infants were born small-for-dates. Pregnancy-induced maternal hypertension was present in just over 50 per cent of mothers whose infants had con genital adrenal hypoplasia. 122 The condition was unsus pected in life, and unexpected clinjca l deterioration occurred in three cases. An 'adrenal crisis' tends to occur in patients with previ ously diagnosed adrenocortical insufficiency and in whom, during intercurrent infection, corticosteroid replacement therapy is not increased. However, Molander lJ referred to a sudden death during exertion from adrenal insufficiency in a previously asymptomatic 19-year-old male. Fibrosis and calcification of the adrenal glands were found at autopsy. Al Sabri et al 120 reported the sudden death of a 12-year-old gi rl with autoimmune Addison's disease. Biochemical abnormalities included hyponatraemia, hyperkalaemia, hypercalcaemia , haemoconcentration, hypoglycaemia and metabolic acidosis. Autopsy revealed depletion and atro phy of the adrenal cortex.
Haemorrhagic Shock Encephalopathy Syndrome Haemorrhagic shock encepha lopathy syndrome is a rare and devastating disorder, with a sudden-onset symptom com plex, usually in previously healthy infants and chil dren. 123 ,124 Symptoms include fever, shock, haemorrhage, diarrhoea and encephalopathy (coma, seizures). Laboratory investigation during hospitalization shows progressive dete rioration of renal function, falling haemoglobin and platelet counts, evidence of disseminated intravascular coagulation, hyper-natraemia, hypoglycaemia, metabolic acidosis, raised serum transaminases and hyperammonaemia. 123 ,124 At autopsy the brain shows oedema, softening and infarc tion.1 24 Hepatic steatosis, hepatic necrosis and small intes t ina l vi llous blunting are other common findings, as are bleeding from the nasotracheal tube and intravenous access sites. Microbiological cultures and toxicological screening are uniformly negative. The aetiology is unknown and the differ ential diagnosis includes heat stroke due to overwrapping, toxic shock syndrome, septic shock, haemolytic uraemic syn drome, Reye's syndrome, metabolic disorders, viral haemor rhagic fevers and poisoning. 124 The age disttibution of the condition is between ) 7 days and) 5 years. The modal age is 3 months, with more than 80 per cent of the patients presenting before 1 year of age. 12S Boys are more commonly affected than girls. There may be a family history of a neurologica l disorder or unexpected death in infancy. The prodromal illness varies from a few hours to several days. Of 33 patients with follow-up information, 22 died, and many of the survivors were neurologically dam aged.125 In my practice the incidence of this condition has declined considerably over the past several years.
Reye's Syndrome This is a rare and serious disorder chiefly affecting chi ldren, characterized by a non-inflammatOlY encephalopathy with hepatic dysfunction in which the diagnosis is made on com bined clinical, biochemical and histological parameters. The internationally accepted case definition is age under 16 years; unexplained non-inflammatory encephalopathy with one or more of t he following: serum hepatic transaminases raised more than three times the upper limit of normal, plasma ammonia level raised to more than three times the upper limit of normal and characteristic fatty infiltration of the liver. 126 In Britain, the epidemiology ofReye's syndrome differs from that in the USA in so far as the mean age is 14 months compared with) 1 years, and there is no clear asso ciation with influenza. 127 The onset of the illness is marked clinically by profuse effortless vomiting, progressing to extreme lethargy and coma. A coagu lopathy and hypog lycaemia are common. Jaundice is rare. Poor outcome is associated with early onset of seizures, profound hypo glycaemia and coma.
Fatal anaphylaxis I
239
At autopsy the brain is oedematous and the liver appears pale due to panlobular fa tty change . These light microscopic findings, however, lack specificity and may be seen in many other conditions. '28 Electron microsco py is widely recommended to confirm the di agnos is, I29 and swo llen and pleomorphic liver cell mitochondria are the ultrastructural hallmarks. 126 The decline in the number of cases of classic Reye's syn drome has been partially attributed to warnings against aspirin exposure durin g viral prodromes in children under 12 years of age and the increased diagnostic awareness of genetic metabolic diseases that mimic the syndrome. ' 27 ,130 Consequently, Reye's syndrome has become an entity comprising a number of 'Reye-like' inherited metabolic disorders,I3 1 includin g p-oxid ation defects, organic acid disorders and urea cycle defec ts. '30 ,'J2 Classical Reye's syn drome has thus become a diagnos is of exclusion, l26 Medium-chain acyl CoA dehydrogenase deficiency is the most common metabolic disorder masquerading as Reye's syndrome. Previous unexplained sibling deaths or similar illness, a previous history of ge netic metabolic disease or unexplained illnesses (e.g. hypoglycaemia, fits) should prompt appropriate investigation. Haemorrhagic shock and encephalopathy syndrome and mitochondrial disease should also be considered in the differential diagnosis,
GASTROINTESTINAL CAUSES Delayed presentation of congenital diaphragmatic hernia has been reported as a cau se of sudden death in two male infants and a 2-year-old girl. 133 The mechanisms of death were mediastinal compression and cardiorespiratory col lapse. One of the infants died in his sleep and was thought to have succumbed from sudden infant death syndrome. The small size of the diaphragmatic defect in these cases probably contributed to th e delayed onset of problems. Small bowel incarceration and infarction as a result of mesodi vertic ular band are reported to have caused sudden dea th in i nfan ts aged 30 and 31 months. lJ4 We have previously encountered a 3-year-old girl whose sudden death was caused by to rs ion of the intestin e in the absence of malrotation (Fig, 12.15).
FATAL ANAPHYLAXIS Acute ana phylaxis is an immediate type IgE antibody mediated hypersensitivity reaction. The clinical spectrum includes hypo tensio n, bronchospasm, angio-oedema and laryngop haryngeal oedem a. IJ5 ,l36 The full-blown clinica l picture may be preceded in some patients by less fulminant symptoms such as itching of the skin, flushin g, generalized warm th or evidence of smooth muscl e contraction, l37 Fatal or near-fatal ana phylactic reactions occur in children and ado lescents,l38 and even infants may be
Figure 12.15 Extensive haemorrhagic infarction of the small bowel secondary to volvulus which caused the sudden collapse and death of a 3-year-old girl.
affected. IJ5 The subjects are often highly atopic, with asthma , allergic rhinitis and atopic dermatitis. Apart from various foods, recognized allergens include drugs, pollens and venom from stinging insects. IJ7 " 39 The offending allergen can be identified in most instances, but occasionally is not found. Mast cell tryptase measurement has been recommended as a useful test for anaphylaxis and has served as a diagnos tic marker at autopsy.lJ5 "'Vbilst tryptase remains stable in samples up to 4 days at room temperature, freezing of sam ples at - 20°C is recommended if the assay cannot be carried out immediately. J40 Elevated levels are not absolutely spe cific for anaphylaxis and may be found in post-mortem serum in other conditions such as sudden infant death syn drome and trauma. 140 Total serum IgE and allergen-specific IgE antib odies may also be usefully measured. l4l The period between onset of ac ute anaphylaxis and death varies considerably, but without medica l interven tion most patients expire within a few hours. 137 An allergic reaction may be conSiderably enhanced in patients receiving p-blockers. l42 Patients with cow's milk protein intolera nce who are challenged with milk following a period of cow's milk protein avoidance are es pecially at risk of developing acute anaphylaxis, Autopsy findings in ac ute anaphylactic deaths are non specific, for example pulmonary congestion and oedema with variable intra-alveolar haemorrhage, Changes may
240 I
Sudden natural death in infants and children
be enhanced by efforts at resuscitation . 1J7 Other findings include increased tracheobronchi al secretions, laryngeal oedema and pulmonary emphysema. An autopsy diagnosis of acute anaphylaxis cannot be made on morphological grounds alone and appropriate clinical information is esse ntial. IJ6 In cases of fatal reaction to insect stings, the site of the sting may be found.
SICKLE CELL DISEASE Sickle cell disease is more common in the black population and the clinical course is generally more severe in homozy gous than in heterozygo us individuals. Most deaths occur in patients with haemoglobin SS, but young subjects with haemoglobin SC are also at risk.143 Polymerization of deoxygenated sickle haemoglobin produces sickling of red blood cells leading to veno-occlusive crises. Predisposing factors include dehydratio n, fever, acidosis and hypox aem ia. Sudden death in infants and older children has been repolted . 144 Generalized convulsions are also recognized. Sudden death in infants with a sickle cell trait may be asso ciated with anaesthesia. Infarction of the small bones of the hands and feet, sickle cell dactylitis, aplastic and haemolytic clises and splenic sequestration are recognized complications. Splenic sequestratio n crisis, the most severe complication, causes death from circulatory collapse due to marked pooling of blood within the spleen. Th is is particularly common in infants under 2 years of age. 144 Acute infection is a com mon precipitating factor. At autopsy the organs are pale except for the spleen, which is enlarged and engorged. Sickled cells in capillaries may be visualized at microscopy, but sickling as such does not necessarily imply that it has occurred ante-mortem. 144 Purulent meningitis, cerebral infarction, cardiomegaly, splenic infarction and bone mar row hyperplasia are other auto psy findings. In the USA the peak incidence of death was found to be between one and three years and the major precipitating event was infection. 145 The causative agent most commonly isolated was Streptococcus pneumoniae. Pneumococcal septi caemia in childhood is observed less frequently because of the widespread use of immunization and antibiotics. 143 The major cause of death in the second decade is a cerebrovascu lar accident. Other causes of death include acute chest syn drome secondary to pUlmonary vasc ular occlusive disease,J46 a condition that tends to affect older subjects and may lead to chronic lung disease and pulmonary hypertension.
HAEMORRHAGE AS A CAUSE OF SUDDEN DEATH Haemorrhage produces death through mass effect and dis tOltion of vital structures (intracra niai), asphyxia (lung) or exsanguination (gastrointestinal or intraperitoneal).
Sources of intracranial ha emorrhage have been dis cussed earlier. One study l5 found 13 out of 169 sudden deaths to be primarily due to haemorrhage; these were mainly intra cra nial haemorrhage and usually secondary to an arteriovenous malformation. One death followed rup ture of a tubal pregnancy. The aetiology of pulmonary haemorrhage has been out lined by Cutz l47 and, broadly, may be separated into pri mary idiopathic pulmonary haemosiderosis and second ary pulmonary ha emorrhage. The latter includes immunolog i cal causes and vascular, infectious and bleeding disorders. Idiopathic pulmon ary haemosiderosis has been recognized as a cause of sudden death.lo Gastrointestinal sources of haemorrhage include oesophageal varices, peptic ulceration and vascular malformations.
RESPIRATORY CAUSES OF SUDDEN DEATH Abnormalities of the Trachea Tracheomalacia may be congenital or acquired. The con genital type is exceedingly rare,148 and cases have been reported in association with Larsen's syndrome, pulmonary vascu lar sling, bronchopulmonary dysplasia and tracheo oesophageal atresia. Tracheal collapse with life-threatening airflow obstruction is a recognized complication following repair of tracheo-oesophageal fistula. 149
Acute Epiglottitis Acute ep iglottitis, an acute life-threatenin g condition, is one of the most serious manifestations of infection with Haemophilus injluenzae type b. The clinical diagnostic cri teria are a red swollen epiglottis (Fig. 12.16), inspiratory stridor or difficulties with swallowing and pyrexia. 150 Chil dren are most commonly affected, but the case fatality is lo w in all age groupS.1 50, 151 Trollfors et ali SO reported six childhood deaths in their series of 485 children. Four died at home or were dead on arrival in hospital. The other two arrived in hospital deeply comatosed, With current immu ni zation schedules, the incidence of acute epiglottis in the Western world may be expected to decrease sharp ly.
Retropharyngeal Abscess Retrophary ngeal abscess is most commonly reported in children less than J years of age. Clinical features include fever, neck swelling, stridor and pharyngeal swelling. In one series about half of the patients were less t han 12 months old, one-third were less than J months of age and three patients presented in the neonatal period. Preceding upper respiratory infection was present in 45 per cent of
Respiratory causes of sudden death I
Figure 12.16 Marked reddening and swelling of the epiglottis in a child whose death was due to acute epiglottitis. patients. 152 Clinical diagnosis can be difficult in infants and young children as the onset of infection may be insid ious with few signs and symptoms. 153 The condition causes acute upper airways obstruction . Stridor is unlikely in children over J years of age. Two . 152 Inone 0 f deaths were reported among t he J 1 patients. these (a J-month-old baby) the diagnosis was made at autopsy. Another mechanism of death in patients with retropharyngeal abscess is rupture of the abscess. leading to pneumonia and haemorrhage. A useful diagnostic method in life and at autopsy is a lateral radiograph of the neck to demonstrate the increased depth of the retropharyngeal space. Other important radio logical features are a visible fluid level in the abscess cav ity or gas in soft tissues. The bacteriology of retropharyngeal abscess has been weLl documented and includes mixed aerobic and anaero bic infections;153.154 others l52 have reported a mixture of Gram-negative bacilli and anaerob es, although Staphylo coccus aureus was the most common single organism iso lated. Three children had pure isolates of Klebsiella.
241
Figure 12.17 A large amount of acute inflammatory exudate and necrotic slough are seen in the lumen of the trachea from a child who died from bacterial tracheitis.
cases. Parainfluenza virus (I, 2 and J) is the most common preceding viral pathogen. 155 Clinical presentation is characterized by fever, barking cough, hoarsness, shortness of breath, cyanosis, stridor and respiratory distress. 156The epiglottis is noted to be normal but the trachea contains copious purulent secretions (Fig. 12.17). Complications included pneumonia, septic shock, adult-type . . 156 respiratory distress syndrome and tOXIC shock syndrome. Rarely a pseudomembrane extends to involve the whole tra , cheobronchial tree, and the oesophagus an d stomac I1. 157 . II y at ns . k . 158 Children with Down's syndrome are especla In a review of 110 patients,155 the mean age was 54 months, the majority being less than J years of age. Males were more often affected than females . More than 80 per cent of patients required endotracheal intubation and res piratory support. Cardiopulmonary arrest occurred in IJ pat ients, four of whom died. Corynebacterium diphtheriae rarely causes tracheitis in the absence of marked supraglottic involvement.
Bacterial Tracheitis
Acute Bronchiolitis
Bacterial infection of the trachea is usually superimposed on preceding viral infection. Staphylococcus aureus and Haemophilus influenzae are responsible for the majority of
Bronchiolitis is an infection of the lower respiratory tract caused by one of a number of different viruses. It is prima rilya clinical diagnosis 159 and the clinical picture is typified
242 I
Sudden natural death in infants and children
and become progressively narrowed and obstructed by sub mucosal oedema and mucus plugs. Air trapping and lobular collapse ensue. 160 The presence of RSV can be determined by viral culture or commercially available methods such as enzyme-linked immunosorbent assay (ELISA), immunofluorescent anti body and enzyme immunoassay tests. The immunofluores cent technique is more sensitive and specific than viral cuI ture. 159
Pulmonary Veno-Occlusive Disease This is a rare and usually fatal condition. 165 The aetiology is unknown. The clinical picture is inconsistent, but patients usually present with progressive or exertional dys pneoa. Sudden infant death has been reported. 166 Histolog ically, the pulmonary veins show total or partial occlusion by intimal fibrosis with or without recanalization. Fresh thrombi may be superimposed. The veins are not uniformly involved and some may be normal. 165 Secondary changes may be observed in pulmonary arteries.
Pulmonary Arterial Hypertension Figure 12.18
Microscopic view showing intraluminal inflammatory
exudate and peribronchiolar inflammation in an infant who died from acute bronchiolitis due to respiratory syncytial virus infection.
by acute wheezing following an upper respiratory illness. Hyperinflation is the most frequently observed radiological abnormality. Respiratory syncytial virus (RSV) is by far the most com monly isolated pathogen. Less common agents include parainfluenza viruses types I and 3, adenovirus, rhinovirus, and Mycoplasma pneumoniae. RSV is a major cause of the disease in infants, whereas the other agents tend to affect older children. '59 The mortality from RSV bronchiolitis in infants who are otherwise healthy is less than one per cent. 160 Some infants are at greater risk of severe or fatal RSV infection, and these include premature infants l61 and those with congenital heart disease, 162 bronchopulmonary dysplasia, 160.1 63 immune defi ciencyl 64 and pulmonary hypertension. 160 Life-threatening complications of RSV bronchiolitis include supraventricular tachycardia and pneumothorax. 159 RSV replicates in epithelial cells. 159 Epithelial necrosis with cilial destruction is the earliest microscopic change and, subsequently, an inflammatory infiltrate composed of lymphocytes, plasma cells and macrophages invades the peribronchial spaces (Fig. 12.18). Airway swelling, slough ing of necrotic debris, loss of cilia and increased mucus pro duction predispose to luminal obstruction. Bronchioles with a diameter ranging from 300/.Lm to 7 5 ~.m are affected, 160
Severe primary or secondary pUlmonary vascular obstruc tive disease is an important cause of sudden death. I ,59 Eisenmenger's pathophysiology is secondary to a large underlying shunt, usually a ventricular septal defect, reversed patent ductus, atrioventricular canal or secundum type atrial septal defect. The age at death spans the first decade of life, but sudden death resulting from Eisen menger's syndrome tends to occur in older children and adolescents. A large number of primary pulmonary dis orders may also lead to pulmonary arterial hypertension .29
EPILEPSY AND SUDDEN DEATH A shorter lifespan among epileptics compared with the gen eral population is well recognized. 167,1 68 Status epilepticus, which is a life-threatening condition, is reported to occur in some 16-24 per cent of children with epilepsy. 169 The risk of death in status epilepticus is directly related to the speed with which the seizure is controlled. 17o Sudden unexpected deaths in epilepsy, in the absence of continuous fitting, take longer to recognize. 170 Such deaths are referred to as sudden unexpected death due to epilepsy (SUDEP) and may be related to the underlying disorder or to the effect of earlier seizures. The majority of these deaths are unwitnessed, 171 and victims are often found in bed. 20,1 67,170 Most deaths are seizure related and no gross or microscopic pathology at autopsy are appar ent, pulmonary oedema and organ congestion usually being the only findings. The underlying pathophysiology of these
Diabetes mellitus I
deaths is uncertain, but proposed mechanisms involve auto nomically related cardiac arrthymias l67 and ictal apnoea .J72 Poor seizure control and poor compliance with an tiepileptic therapy do not ap pear to be significant factors in SUDEP. 173 A toxico logy screen should fonn pa rt of the au topsy proto col and the heart should be examined in detai l. A serum dmg level lower tha n the therapeutic ran ge do es not necessari ly imply fa ilure of compliance, and sh ould not be in terpre ted by the pathol ogist as a factor in the cause of death. 174 Moreover, the relevance of post-mortem blood lev els of anticonvulsants many hours after death is unclear, as so me degrade more rapidly than others.1 6S The possib ility of adverse cardiovascular effects of carbamazipine in SUDEP in children cannot be ignored. 175 When evidence su ggestive of a seizu re is found, for example a bitten tongue, limb injuries or an empty bladder, this sh ould be record ed. However, should this evidence be absent, a seizure canno t be excluded. It is recommended that when children are fou n d wi th thei r head immersed in water, de ath should be recorded as ep ilepsy- rel ated death, rather than SUDEP. Card iac arrhythmias rarely give ris e to 'epilepsy',1 76. 177 and death in such cases may be erron eously attributed . A prolonged QT interval may be responsible and au topsy examination is unlikely to r eveal any abnorm ali ty.
DEATHS FROM ACUTE ASTHMA Asthma is a common disease and, des pite advances in tbe treatment, there is still a hig b ra te of mortality and morbid ity amon g children. Deaths from as th ma are usually as a result of prolonged attacks tha t fail to respond to conve n tional measures, status asthmaticus. Sudden unexp ected death is a lso recognized but occurs less freq uen tly.17B Zach an d Karner l79 reported sudd en deaths in two gi rls, one aged 14 years and one aged 9 years, receiving anti asthmatic medication. In both cases the gi rls ' condition was regarded as moderately severe and well stab ilized. These deaths might be explained by inappropriate percep tion of the severity of the attack by the p atie nt. Of 30 childhood asthma-related deaths reviewed by Car swell,l so the mean age at death was 8.5 years; 19 died at home or in transit to hospital and 11 died in hospita l. One child died within one hour of the sta rt of the attack; three died less than 12 hours after the start and seven died after more than 12 hours. Deaths occurred mainly at night or in th e evening. Only 12 patients in this stud y we re known to be receiving prophylaxis. Unexpected death in asthma patients is especia ll y pro nounced in adolescent an d preadol escent years, and may be related to treatment non-use/abuse, which is ren owned to occur at this age. lSI The increased frequ en cy of deaths from asthma at night or in the early morning has been lin ked to diurnal va ria tion in airflow limitation. ls2 Risk factors for sudden death include the occurrence of previous
243
life- threatenin g asthmatic episodes, hospital admissions for as thm a in the year preceding death, poor access to medical ca re, inadequate medica l ma nagement a nd psy chological and psychosocial problems. IS} Previous stero id treatment, by suppressing the ad ren a l axis, m ay lead to adren al insufficiency and a pred isposition to unexp ected de ath. Recognized mechanisms of death are severe asphyxia du e to airflow restriction IS4 or cardiac arrhythmia from myocardial irri ta bility secondalY to excessive use of ,B2 an tag onists. Hypokalaemia, which may be partly expl ained by ~ 2 antagonists, can cause cardia c a rrhythmia or gene r ali zed muscle weakness, co ntributing to sudden dea th. 135 Life-threa tenin g eve nts are includ ed among the adverse effects of aminophylline. IS) The hi gh mortali ty from asthma in t he 1960s was attrib uted to the excessive use of press urized beta ago nist aerosols, the so-called 'aeroso l hy pothesis'. ls6 Whilst the debate concerning the role of th ese aerosols con tinues, it is difficult to discount fa ilure to promptly and adequatel y treat the asthmatic attack in su ch cases. It has been ques ti oned whether lun g fun ction in such cases is ever normal shortly befor e the fatal attack JB2 Apart from abrupt airway narrowing and medication related deaths, an imp ort ant factor may be u nsuspected pulmonary pathology. lSI Among 13 unexpected asthma deaths in subjects aged between 9 and 19 years, Kravis and Kolski lSI id entified a cause in five cases: pneumothoraces, histiocytic bronchopneumonia, iso lati on of Kl ebsiella pneumoniae, eosinophilic pn eumonia and foc al bronchio litis obliterans. Macroscopic fi n din gs in acute deaths from asthma are bulky, hyperinflated lungs, surgical emph ysema and, rarely, pneumothorax. Microsco py shows intraluminal bronchial and brochiol a r mucinous plu gs, epithelial basement mem brane th ickening, peribronchial smooth muscle hypertro phy, mucus gland hyperpl as ia a nd submucosal cellular infiltration, often with large numbers of eosinophils. 17s Asthmatics are also at increased risk of developing severe or fa tal anap hyl ax is,14 1.142 and this possibility sh ould be considered in the event of sudden death in a patient who suffers from asthma. Thi s is especially relevan t as acu te ana phyla xis may present cl inically as respiratory distress. Mea surement of the semm tryptase level may assist in making the distinction.
DIABETES MELLITUS Despi te modern treatment, in sulin-depe ndent diabetes mel litus (IDDM) in childhood still carries Significant morta l ity.IB7·IBB Most deaths are attributable to metabolic complications, and those related to diabetic ke toa cidosis a re most common, foll owed by hypog lycaemia. IB9 Sudden unexpected death is reported in cases of di abeti c keto ac i dosis,190 and the a dol escent age group is mostly affec ted .
244 I
Sudden natural death in infants and children
Cerebral oedema is an important comp li cation in such cases and carries a poor prognosis. 191 After death , the blood glu cose is raised and the urine contains ketones, but ketones may be absent in the aketotic form of diabetic coma. Post-mortem determination of blood glucose is unreli able. 190 The presence of glucose in the urine may also be misleading as glucose may have been administered intra venously. The site of blood sampling at autopsy has an important influence on the blood glucose level. 192 Thus, samples taken from the right side of the heart often show sp uriously high gl ucose levels as a result of glycogenolysis , espec ially if death is accompanied by an 'alarm' reaction or cardiac massage is performed . Sampling from the periph eral vein, by contrast, often shows a glucose level that is spuriously low due to continuing glycolysis after death. Glucose in the vitreous can be more usefully meas ured, 190.192 but the level decreases rapidly after death. So lon g as the sample has been taken peri mortem into a fluo ride tube, this w ill reflect blood glucose level s at that time and a normal glucose level should exclude hypoglycaemia providing, of course, that the patient has not received glu cose infusion prior to admission. 193 High levels of vitreous or peripheral venous glucose probably do indicate uncontrolled diabetes, especially if concentrations of ketone bodies in the blood or urine are also raised. Measurement of glycated haemoglobin has been suggested as a more reliable indicator. 194 Absence of insulin in ante-mortem blood carries more weight than in a post-mortem sample; by contrast, raised insulin values in post-mortem blood are useful in excluding diabetic ketoac idosis. 190 Histology of the pancreas may provide useful informa tion providing the post-mortem interval is short and a utolysis is minimal. 195 Changes include acinar atrophy, reduction in the number and size of Langerhans islets, insulitis, distortion of islet architecture and histochemical demonstration of a reduced number of insulin-producing B cells. 195 Glycogen may be increased in the renal tubular epithelium (Armani-Epstein nephropathy), but this is not a constant feature. Othe r histo log ical findings are microvesic ular fatty change of the liver and vacuolization of hepatic nuclei; these are, however, non-specific and may be found in other conditions. The 'dead in bed syndrome' constitutes a puzzling group among patients with IDDM. The subjects are without clini cal evidence of late complications. They are usually on insulin and there is often a history of one or more noctur nal hypoglycaemic attacks in the previo us 6 months. 196 They are generally observed to be in good health on the preceding day and are found undisturbed in bed the fol lowing morning. 196 ,197 Children and young adults account for some six per cent of all such deaths under the age of 40 years. The most likely cause is thou ght to be hypogly caemia with associated events such as ca rdi ac a rrhyth mia 198 or respiratory depression. 196 Hypoglycaemia is difficult to confirm and, as the post-mortem often does not
reveal a cause, the diagnosis is usually inferred from t he clinical history.
GENETIC IVIETABOLIC DISORDERS Factors that should prompt consideration of genetiC meta bolic disease in cases of sudden death are parental consan guinity, maternal HELLP (haemolysis, elevated liver enzymes, low platelets) syndrome, previo us Reye-like ill ness or infant deaths in the family (from known genetic metabolic disease or unex plained), a previous acute life threatening even t (near-miss cot death), fasting or recur rent hypo g lycae mia, neonatal hypotonia, dysmorphism, en la rgement of the live r and/or spleen. A pale (fatty) liver or hepatic fibrosis, cardiomegaly and severe brain oedema are important markers of metabolic disease at autopsy. The possible metabolic causes of sudden death in the paediatric group are wide-ranging. 199
Mitochondrial Abnormalities These comprise defects of fatty acid oxidation (FAO) and oxidative phosphorylation (OXPHOS). A Reye-like illness or sudden death is exceptionally reported in association with defects of OXPHOS. In defects of OXPHOS abnormalities are likely to involve a single organ or several systems simulta neously. A diagrammatic scheme of intramitrochondrial ~ -oxidation of fatty acids an d its interrelationship with the respiratory chain complex is shown in Fig. 12.19. 200
FAm ACID OXIDATI ON DEFECTS Enzymes that have been associated with sudden death are shown in Table 12.2. All of these inborn errors are inherited as an autosomal recessive trait. Very long-chain acyl coen zyme A dehydroge nase (VLCAD) deficiency usually presents in early infancy with hypoglycaemia, recurrent vomiting, liver dysfunction, cardiomegaly and a tendency to cardiac arrest.200.201 Presentation as sudden infant death is well rec ognized.202.203 Compared with the other defects, VLCAD is more likely to be associated with cardiomegaly, although sudden death and severe illness in the neonatal period have been reported. 204 Only a few patients with short-chain acyl CoA (SCAD) deficiency are recorded, 202 and its association with sudden death is unclear. 203 Primary carnitine defici ency and carnitine palmitoyl transferase type II (CPT II) deficiency often present with cardiomyopathy, whereas carnitine palmitoyltransferase type I (CPT I) deficiency usually presents with a Reye-like illness, but occasionally cardiac arrhythmia occurs in the neonatal period. 205 Fat oxidation is impaired as the trans fer of long-chain fatty acids across the mitochondrial membrane is dependent on carnitine transporter and carni tine palmitoyltransferase. 202
Genetic metabolic disorders I
245
Long-chain fatty acid
1
Ligase
acyl-CoA
1
Carnitine palmitavl transferases I and /I Carnitine aeVI-earnitine translaease
SCAD, MCAD vLCAD
Enavl-CoA hvdratase -I :lJ
3-hydroxyacyl-CoA : NAD+
~
"
C Z
g 3-Hvdroxvaevl-CoA dehvdrogenose
NADH+H + 1
o z
,:t> m Z
N
-<
3-oxoacyl-CoA
1
':m s Thiolase
acyl(n_2j-CoA
Figure 12.19
Table 12.2
Diagram to show the fatty acid 0-oxidation pathway and the link with the respiratory chain complex. (Modified from ref 200.)
[3-0xidatian of fatty acid defects that have been
associated with sudden death
Very long-chain acyl CoA dehydrogenase deficiency (VLCAD)
Long-chain 3-hydroxyacyl CoA dehydrogenase deficiency (LCHAD)
Mitochondrial trifunctional protein deficiency (TFPD)
Medium chain acyl CoA dehydrogenase deficiency (MCAD)
Carnitine palmitoyltransferase type 1 deficiency (CPTl D)
Carnitine palmitoyltransferase type 2 deficiency (CPT2D)
Plasma membrane carnitine transporter deficiency (PMCTO)
Multiple acyl CoA dehydrogenase deficiency (MADD)
Carnitine acylcarnitine translocase deficiency (CACTO)
Long-chain 3-hydroxyacyl CoA dehydrogenase (LCHAD) deficiency is associated with a high mortality in infancy.200 Cardiomegaly may be severe and hypo ketotic hypogly caemia may be a presenting feature. When the fetus is affected by LCHAD, the later stages of pregnancy may be complicated by the HELLP syndrome in the mother. Disorders of fatty oxidation tend to be clinically silent in the absence of significant physiological stress. 20 ) Meta bolic events such as prolonged fasting and febrile illness
lead to depleted glycogen reserves, initiating ketone body production and muscle fatty acid oxidation. In the UK, medium-chain acyl CoA dehydrogenase (MCAD) deficiency is the most commonly encountered metabolic disorder leading to sudden death in infants and children. 206 It is clinically the least severe of the i3-oxidation defects. 20I,202 According to the review by Touma and Charpentier 207 the average age at presentation is 13 .5 months and the mean age of death is 18.5 months. Patients often present clinically with hypoglycaemia, acute encephalopathy (mimicking Reye's syndrome), sudden death and an acute life-threatening event. 201.202,208,209 One-quarter of patients die with a Reye-like illness or experience sudden infant death. A previous sibling death is often recorded. A preceding viral prodrome is common. The gene for MCAD has been identified and sequenced;209 up to 85 per cent of subjects with MCAD deficiency are homozygous for a single gene mutation of an A-to-G nucleotide at position 985. A recent prospective surveillance of the prevalence of the condition in the UK210 found 1.3 cases per 10 000 births. The precise mechanism of death in MCAD defi ciency is unclear, but hypoglycaernia, cardiac decompensation from fatty infiltration of the myocardium and arrhythmogerric blood levels of acylcamitines have been implicated.
246 I
Sudden natural death in infants and children Table 12.3 Useful tissue samples and their storage recommended for autopsy investigation ofpossible fatty acid i3-oxidation defects
L-..I
1cm
Figure 12.20 Liver from an infant whose sudden death was due to medium chain acyl-CoA dehydrogenase deficiency, showing marked pallor due to fatty in filtration.
At autopsy, pronounced fatty infiltration of the liver (Fig. 12.20), kidney and muscle is usually found and should prompt appropriate investigation. Occasionally, individuals dying with MCAD deficiency do not exhibit significant fatty infiltration of the liver,21I ·212 and its absence should therefore not preclude investigation along these lines. A post-mortem diagnosis can be made by demonstrating: • a characteristic profile of medium-chain dicarboxylicaciduria and hexanoylglycine on urinary organic acid analysis; • a raised plasma cis4 decenoate; • octanoylcarnitine by tandem mass spectrometry (MS/MS) on dried blood spots (or using the original Guthrie screening card); and • ~-oxidation of fatty acids in intact cultured fibroblasts. Urine, blood, vitreous humor, tissues and a skin biopsy are important for establishing a post-mortem diagnosis 20J (Table 12.3). The minimal requirement of urine is 0.1 mL and, even if the bladder seems empty, swabbing the bladder mucosa with a cotton ball may provide a sufficient volume for analysis. Urine (or bladder swab) samples should be stored at -20°e. Organic acid analysis of vitreous humour may be useful in the absence of urine. 193 Liver, skeletal and cardiac muscle should be collected and stored without fixa tive at - BO°C prior to analysis. Blood should be collected into an anticoagulant and centrifuged; blood cells and plasma should be stored separately at - 20°e. With the advent of gas chromatography-mass spectrome try (GC-MS) technology, bile, blood and cerebrospinal fluid can be spotted onto a Guthrie card at post-mortem examina tion for acylcamitine analysis. 206 Vitreous humour has also been used for the diagnosis ofMCAD deficiency. A small skin biopsy sample taken under sterile conditions, as soon as pos sible after death, should be placed in tissue culture medium containing one per cent dimethyl sulfoxide and frozen at -70°C, or sent directly to the culture laboratory if metabolic
Urine (- 20°C) Liver (- BO°C) Skeletal muscle (- BO°C) Cardiac muscle (-BO°C) Blood into anticoagulant (centrifuged) Blood cells - plasma (- 20°C) Small skin sample - tissue culture + 1% dimethyl sulphoxide (Store at -70% or send to culture laboratory straight away if suspicion is high) Whole blood, bile and/or cerebrospinal fluid spotted onto a Guthrie card
disease is strongly suspected. The predominant MCAD defi ciency mutation can be detected by DNA analysis of the stored dried blood spot sample taken for mass neonatal screening of phenylketonuria,2°9 but DNA may also be obtained from frozen tissue samples and cultured fibroblasts. 193 DISORDERS OF OXIDATIVE PHOSPHORYLATION Compared with ~ -oxidation fatty acid defects, deaths due to defects of OXPHOS are more likely to have demonstrable pathology at autopsy (e.g. cardiomyopathy, liver fibrOSis, cinhosis). These disorders are wide-ranging in their clinical manifestation and patients only rarely present as sudden death cases. Isolated myopathy, or a multisystem disorder that includes encephalopathy, liver disease and cardio myopathy are features of the disorder. 2IJ The cardiomyopa thy is usually symmetrical (concentric), without outflow tract obstruction. Death before the age of 1 year from heart failure is the usual outcome.2l3 Histology and electron microscopy of the myocardium reveals swollen myocytes and an increase in the number of mitochondria. The mitochondria may also be morphologically abnormal. The most common defects, according to the review by Guenthard et al,213 were deficien cies of complexes I and lV, in isolation or combination. An Il-month-old previously healthy boy reported by Smeitink et al 214 died following a short Reye-like illness. A defect in the coenzyme Q region of the respiratory chain was found. His 6-month-old female sibling had previously died following a similar illness. An acute apparent life-threatening event (ALTE) was reported as the first sign of respiratory chain complex defi ciency in a 4-month-old girl whose parents were consan guineous. 215 She was found to have markedly reduced complex I activity. By the time of the post-mortem examination, it is usu ally too late to measure the function of the respiratory chain as enzyme activity deteriorates rapidly.19J However,
Other bacterial infections
I
24 7
Table 12.4 Clinical and biochemical indicatians far autapsy investigation of oxidative phosphorylation disorders Recurrent apnoea Reye-like syndrome Aminoaciduria Metabolic acidosis Hypoglycaemia Hyperaminonaemia Dica rboxylicacid uria A metabolic screen with negative resu lts
adenosine triphosphate (ATP) production can be evaluated in cultured fibroblasts and complexes II, III and IV can be measured. It is important to note that secondalY morpho logical abnormalities may be found in the mitochondria of patients with MCAD deficiency and plasma membrane car nitine transporter deficiency.43,2J6 Clinical and biochemical features that should prompt consideration of OXPHOS dis orders as a cause of death are listed in Table 12.4.
OTHER BACTERIAL INFECTIONS Meningococcal infection often has a fulminant onset with a high mortality variously attributed to adrenal haemorrhage, endotoxic shock or a generalized Schwalizman reaction. The disease has two clinical forms: meningitis and meningococ caemia. 217 The main clinical manifestations are fever, rash (Fig. 12.21), vomiting, lethargy, meningeal irritation and cir culatory collapse. The rash may be maculopapular, purpuric or mixed. 2 18 There was a sustained increase in the incidence of meningococcal disease in the 1990s.219 An increased incidence of the condition is observed in late winter and spring. A wide age range is affected, with a peak incidence in the second year. Infection in the first 2 years of life and meningococcaemia with or without menin gitis, compared with meningitis alone, are associated with increased mortality. Other poor prognostic indicators are the onset of petechiae within 12 hours prior to admission, shock, normal or low white cell count, and a normal or low erythro cyte sedimentation rate.217 Autopsy often demonstrates bilateral adrenal haemorrhage (Fig. J 2.22). This is, however, not an invariable or specific finding, as adrenal haemor rhage also may be seen in haemorrhagic shock encephalopa thy syndrome (see above) and other forms of septicaemia. Laboratory findings include isolation of Neisseria meningi tidis in the blood or spinal fluid, or the presence of Gram-neg ative diplococci in the spinal fluid or a petechial scraping in a patient with a typical clinical course. Unsuccessful attempts at isolating the microorganism from the blood and spinal fluid , clinically or at autopsy, are often due to preceding antibiotic treatment. Antigen screening ofbJood, cerebrospinal fluid and urine are other diagnostic measures that may be adopted, as is PCR amplification from blood or cerebrospinal fluid .2J9
Figure 12.21 A widespread maculopapular rash typical of meningococcaemia.
Figure 12.22 Bilateral adrenal haemorrhage secondary to men ingococcaemia.
Sudden unexpected deaths in infants and young children have been ascribed to overwhelming infection as a result of group A 0-haemolytic streptococcus, or Streptococcus pneu111oniae, with the organism being isolated from several
248 I
Sudden natural death in infants and children
sites at autopsy.1J,220 However, such reports generally lack microscopic evidence of sepsis or associated features and a causal association is unclear. Sudden death in infancy due to Haemopi1ilus injluenzae septicaemia has been reported in infants with asplenia, splenic hypoplasia and polysplenia, occurring in isolation or associated with complex heart disease [with or without accompanying visceral abnormalities).221 At autopsy, adre nal haemorrhage was a common finding in these cases. Colonization or infection with Staphylococcus aureus and group A streptococcus is implicated in the toxic shock syndrome. 222 - 224 The condition appears to be toxin medi ated; the exact pathophysiology is unknown and the clini cal course may be rapidly fatal. Identifying the site of infection may be difficult. Both adults and children may be affected. The streptococcal toxic shock syndrome is charac terized by the sudden onset of fever, rash, vomiting and diarrhoea, hypotension, conjunctival injection and straw berry tongue, followed by skin desquamation. Necrotizing fasciitis is a rare, but often fatal, soft-tissue infection with rapidly spreading inflammation and necrosis of the muscle fascia and subcutaneous fat, in some instances involving the epidermis. The condition is life-threatening and carries a high mortality rate. Most cases are a result of polymicrobial or group A streptococcal disease. 225 ,226 Adults and children are affected. Pyrexia, leucocytosis, cellulitis and oedema are the main clinical manifestations. The affected skin becomes progressively warm, erythematous, swollen and painful. Invasive group A streptococcal infection [GAS), with or without 227 necrotizing fasciitis, is the most serious compli cation of varicella infection in children. 22B Progression of GAS may be very rapid, with death at home or shortly after arrival in hospitalYs Some patients with necrotizing fasciitis share many of the features of streptococcal toxic shock syndrome,226 and the onset of streptococcal toxic shock syndrome may be preceded by necrotizing fasciitis.224 Neonatal omphalitis can progress to necrotizing fasciitis, followed by rapid dete rioration and death.229 Mortality is increased by delayed treatment and any underlying debilitating disease.226
DEATHS RELATED TO OBSTETRIC EVENTS AND PREMATURE BIRTH Babies born pre-term are at risk of sudden unexpected deaths at home. The autopsy findings include no anatomi cal changes, coincidental disease and residual pathology from prematurity-related complications. Prolonged tracheal intubation and positive-pressure ven tilation in pre-term babies are associated with a range oftra cheal abnormalities including granulations, pseudopolyps, subglottic stenosis, bronchomalacia and tracheomala cia. 230 - 232 Tracheobronchomalacia may cause life-threaten ing episodes from acute airways collapse in infants with
bronchopulmonary dysplasia.233 The pathophysiology of the condition is not well understood but is thought to be acquired rather than congenital. 231 Significant factors in its production are low gestational age at birth and a high mean airways pressure during ventilation in the first week of life. In our own experience of 40 prematurity-related unex pected post-neonatal deaths, two infants were found to have died from conditions unrelated to the prematurity [RSV-related bronchiolitis and viral myocarditis). Twelve infants displayed abnormalities directly or indirectly as a result of their prematurity; these were post-hypoxic ischaemic encephalopathy, resolving brochopulmonary dysplasia, tracheal stenosis and stricture of the small intestine, alone or in combination. The cerebral lesions included periventricular gliosis, mineralization and haemosiderin deposits, and periventricular cysts. Infants with more extreme degrees of prematurity tended to show more severe pathology, but even in these instances the role of the pathology in the cause of death was often unclear. Pathology was not found in the remaining 26 infant deaths. A limited respiratory reserve in infants with resolving bronchopulmonary dysplasia has been proposed as a mechanism of death in such cases, especially when there was superimposed viral or bacterial infection, producing sudden decompensation of cardiac and/or respiratory func tion. 23o Injury to the pulmonary vascular bed, a feature of bronchopulmonary dysplasia, is commonly associated with pulmonary arterial hypertension and right ventricular hypertrophy.29 From a clinical perspective, however, the picture is quite different, with Tammela and Koivist0 2J4 reporting an absence of fatality in a cohort of 86 low birth-weight babies, 23 with bronchopulmonary dysplasia, after discharge from hospital and followed over a period of 12 months, indicating a low mortality among infants with resolving bronchopulmonary dysplasia.
MISCELLANEOUS CAUSES OF SUDDEN NATURAL DEATH Sudden death has been reported in cases of congenital myotonic dystrophy,235 although the mechanism was unclear. Cardiac dyslythmia is also known to cause sudden death in patients with Emery-Dreifuss muscular dystro phy.2J6 An occult adrenal phaeochromocytoma as a cause of sudden death in a young boy following a kick to the anterior abdomen has been ascribed to a sudden surge of a large amount of noradrenaline into the circulation, produc ing violent myocardial contractility.2J7 The sudden death of an infant who subsequently demonstrated widespread his tological features indistinguishable from Kikuchi-Fujimoto disease has been reported. 238 Encephalopathy in association with acute pancreatitis is welJ recognized in adults and has been rarely reported in chil dren.239 The condition is thought to be a result of damage to
Sudden death associated with 'inte rm ed iate ' pathology I
249
SUDDEN UNEXPLAINED DEATH IN OLDER CHILDREN Many of the reported series of sudden death include a propor tion of older children in whom the cause is unexplained. Thus, 11 of 169 (6.5 per cent) sudden deaths in children and ado lescents aged 2- 20 yea rs reported by Keeling and Knowles l5 were unexplained. However, most of the cases had not been investigated toxicologically or microbiologically, and sampling for microscopy was limited. The study reported by Molander,1 3 from southem Sweden, included four unex plained deaths in 389 sudden natural deaths in subjects between the ages of 1 and 20 years. Similarly, sudden death in 29 of 207 subj ects aged 1-21 yea rs was reported as being of undetermined cause. 20 The subjects were mainly between 1 a nd 4 years of age, but were not discussed in any detail. ' Bangungut' is a con dition that affects yo ung men from t he Philippines, J apan, Vietnam and Thailand. The victims a re excl usively male and usually die mysteriously at night. The mechanis m of death appears to be ventric ular fibrill a tion witho ut underly ing card iovasc ul ar disease. Thiamine deficiency h as been suggested as a contributing factor. 59
SUDDEN NATURAL DEATH IN THE EARLY NEONATAL PERIOD Figure 12.23 The myocardium shows frequent contraction band necrosis. Sudden death followed a kick to the anterior abdomen in a boy with a right adrenal pha eochromocytoma .
brain tissue caused by the circulatin g pancreatic enzymes. The initial clinical picture may be confused with Reye's syn drome. Premature atherosclerotic coronary disease in subj ects aged between 14 and 19 years is rarely described as t he cause of sudden death .41 Precocious coronary atheroscle rotic disease in young patients is alm ost always secondary to type II hypercholesterolaemia, system ic lupus elyth e matosus or juvenil e- onset diabetes. 59 Cerebral disease has long been known to cause electro cardiographic changes and has been linked to overactivity of the sympathetic limb of the autonomic nervo us system. Autopsied cases h ave shown hi stologica l changes of the myocardium ranging from normal muscle to severe necrotic lesions with mononuclear cell 'infiltration;26 con traction band necrosis of t he myoca rdium is a characteris tic finding (Fig. J2.23) . Similar les ions of the myocardium are reported to be ca used by stress, catechol a mine infusion and reperfusion inj u ry. Opening of the calcium channel with infl ux of calcium into the cell and efflu x of potassium causes interaction between the actin and myosin filaments, leading to cell death from hypercontraction, is the likely mechanism. A similar mecha nism may provide an explanation for sudden death occurring with epilepsy, asthma and 'stress' fro m other causes. 26
Hypoxic ischaemic encephalopathy, bacterial infection, undiagnosed cardiac malformation and genetic metabolic disease are important causes of sudden death at this age. Several of the enzyme deficiencies that interfere with ~ -oxi dation of fat may also be responsible for sudden death.240 Disseminated herpes simplex infection has been reported 241 (see Fig. 11.7, p. 209). Maternal evidence of infection may be absent and the disease in the baby may not be clinically manifest. Mercuri et al 242 reported several in fants born with no r mal Apgar scores who subsequently developed seizures after a peliod ranging from 10 hours to 4 days. In all cases, ischaemicjhaemorrhagic les ions, sustained either pe rin a tally or post seizure, were demonstrated on bra in im aging. One of these infants died at 3 days of age.
SUDDEN DEATH ASSOCIATED WITH 'INTERMEDIATE' PATHOLOGY Norman et al 243 have already drawn attention to findings in paediatric a utops ies w hose significance is questionable. Examples are s ma ll ca rdi ac fibrom as, abnorma l cerebra l gyral pattern and atrial septa l defects. The issue has already been raised in relation to premature babies who die unex pectedly at home a nd whose brains reveal changes of a previous hypoxicjischaemic event. Mi ld degrees of hydro cephalus, neuronal heterotopias or a ventricular septal defect are other entities that may be difficult to eva luate in cases in sudden in fa nt deaths.
250 I
Sudden natural death in infants and chi ldren
9 Byard RW, Bourne AJ, Adams PS. Subal1erial ventricular septal defect in an infant with sudden unexpected death: Cause or coincidence? Am J Cardiovasc Pathol 1990; 3:333 -6. 10 Byard RW. Vascular conditions. In Byard W, Cohle SD (ed s) Sudden Death ill Injanty, Childhood and Adolescence. Cambridge: Cambridge University Press, 1994, pp. 175- 261. II Bird LM, Billman GF, Lacro RV et al. Sudden death in
Will iams syndrome: report of ten cases. J Pediatr 1996;
129:926-3 1.
DeSa DJ. Isolated myocarditis in the first year. Arch Dis Child 1985; 60:484-5. 13 Molander N. Sudden natural death in later childhood and adolescence. Arch Dis Child 1982; 57:572-6. 14 Siboni A, Simonsen J. Sudden unexpected natural death in young persons. Forensic Sci lilt 1986; 31 :159-66. 15 Keeling JW, Knowles SAS. Sudden death in childhood and adolescence. J Patliol 1989; 159 :221 -4. 16 Davies MJ. The cardiomyopathies: a review of terminology, pathology and pathogenesis. Histopathology 1984; 8:363-93. 17 Noren GR, Staley NA, Bandt CH, Kaplan EL. Occurrence of myocard itis in sudden death in children. J Forellsic Sci 12
1977; 22: 188-96. 18 Davies MJ, Popple A. Sudden unexpected cardiac death: a
practical approach to the forensic problem. Histopathology 1979; 3:255-77. 19
Figure 12.24
Microscopy of a microglial nodule in the
brainstem of an infant with cytomegaloviral parotitis w ho died suddenly at home.
Cytomegalovirus-related parotitis is found in a small number of infants who die suddenly. These cases may show a few microglial nodules in their brainstem (Fig. 12.24),244.245 but it is unclear whether they have role in the cause of death.
REFERENCES Lambert EC, Menon VA, Wagner HR, Vlad P. Sudden unexpec ted death from cardiovascular disease in children. A cooperative international study. Am J Cardiol 1974; 34:89 - 96.
2 Abu -Harb M, Hey E, Wren C. Death in infancy from unrecognised congenital heart disease. Arch Dis Child 1994; 71:3-7.
3 Basso C, Frescura C, Corrado D et al. Congenital heart disease and sudden death in the young. Hum Pathol 1995; 26(10):1065-72.
4 Stahl J, Santos LD, Byard RW. Coronary artery thromboembolism and unexpected death in childhood and ado lescence. J Forensic Sci 1995: 40:599-601. Klitzner TS. Sudden cardiac death in children. Circulation 1990; 82:629-32.
6
7 8
Steinberger J, Lucas RV, Edwards JE, Titus JL. Causes of sudden unexpected cardiac death in the first two deca des of li fe. Am J Cardiol 1996; 77:992-5. Glew RH, Varghese PJ, Krovetz LJ et al. Sudden death in congenital aortic stenos is. Am Heart J 1969; 78:615-25. McManus BM, Babul S. Pathology of sudden death in th e young. Can J CaTdioil996; 12:1150-6.
Lim C-H, Toh CCS, Chia B-L. Low L-P et al. Stokes-Adams attacks due to acute nonspecific myocarditis. Am Heart J
1975; 90:172-8. 20 Neuspiel DR, Kuller LH. Sudden and unexpected natural death in childhood and adolescence. J Am !VIed Assoc 1985; 254: 1321-5. 21 Woodruff JF. Vira l myocarditis. Am J Patho1 1980; 101 :427-42. 22 Leslie K, Blay R, HaiscJl C et al. Clinical and experimental aspects of vira l myocarditis. Ciin Microbiol ReI! L989; 2:191-203 . 23 Hilton DA, Variend S, Pringle JH. Demonstration of
coxsackie virus RNA in fonnalin-fixed tissue sections from childhood myocarditis cases by ill sirt! hybridization and the polymerase chain reaction. J Pathol 1993; 170:45-51. 24 Lo zinski GM, Davis GG, Krous HF et al. Adenovirus myocarditis : retrospective diagnosis by gene amplification from formalin-fixed, paraffin-embedded tissues. Hum Pathol 1994; 25:831-4. 25 Shimizu C, Rambaud C, Cheron G et al. Molecu lar
indentification of viruses in sudden infant death associated wi th myocarditis and pericarditis. Pediatr lilject Dis J 1995; 14:584-8 . 26 Samuels MA . Neurogenic heart disease: A unifying hypothesis. Am J Cardiol 1987; 60:15-1 9.
27 Frustaci A, Loperfido F, Gentiloni N et al. Catecholamine indu ced cardiomyopathy in multiple endocrine neoplasia: A histolo gic, ultrastructural and biochemical study. Chest 1991; 99 :382-5. 28 Olsen EGJ. Myocarditis. In Olsen EGJ (ed.) The Pathology oj the Heal1. London: The MacMillan Press Ltd, J980; 178-99 . 29 Patterson K, Donnelly WH, Dehner LP et al. The
cardiovascular system. In Stocker JT, Dehner LP (eds) Pediatric Pathology Vol. I. Philadelphia: Lippincott Company, 1992 ; 575 -651. 30 Haupt HM, Hutchins GM. The heart and cardiac conduction system in polymyositis-dermatomyositis: A clinicopathologic study of 16 autopsied patients. Am J Cardiol 1982; 50 :998-1006. 31
Bissenden JG, Hall S. Kawasaki syndrome: lessons for Britain. Br Med J 1990; 300: 1025-6.
Referen ces I
32
33 34
35 36
37 38 39
40 41
42 43
44
45
46
47 48
49
50 51
52
53
54 55
Davies MJ, McKenna WJ. Dil a ted card iomyopathy: an introduct io n to pathology and pathogenes is. Br Ht'Grt J 1994; 72(Supp l.):24. Kelly DP, Stra uss AW. Inherited cardiomyopathies. N Eng l J Med 1994; 330:913-19. Lawson MR. Southwestern Interna l Medicine Conference: hypertrophic cardiomyopathy: current views on etiology, pathophys iology and managem ent. Am J Med Sci 1987 ; 294:191 - 210. Burch M. Hypenrophic ca rdiomyopathy. Arch Dis Child 1994; 71:488-9. McKe nn a WJ, S tewa n JT, Nihoyannopoulos P et al. Hypertrophic cardiomyopathy without hypertrophy : two families w ith myocardial disarray in the abse nce of increased myocardial mass. Br Hea rt J 1990; 63 :287 -90. Gow RM. Sudden cardiac death in the young. Can J Cardiol 1996; 12: 1157-60. Burch M, Mann JM, Sh a rl and M et a l. Myocardial disarray in Noonan sy ndrome. Br Heart J 1992; 68:586- 8. Ha lliday HL. Hypenrophic ca rdiomyopathy in infa nts of poo rl y-co ntroll ed diabetic mothe rs. Arch Dis Child 198 1; 56 :258-63. Burch M, Runcima n M. Dilated ca rd io myopathy. Arch Dis Child 1996; 74:479-81. Maron BJ, Shirani J, PoJiac LC et a l. Sudden death in young competitive athletes: clini cal, demog raphic, and path o logical profiles . J Am !VIed Assoc 1996; 276: 199-204. Huber S. Viral myocarditis - a tale of two diseases. Lab hllJesr 1992; 66:1- 3. Be nnett MJ, Hale DE, Pollitt RJ et al. Endocardial fibroelastosis a nd prima lY ca rnitine deficiency due to a defect in the plasma membran e ca rnitine tran sponer. Ciil1 Cardiol 1996; 19: 243 - 6. Virman i R, Robin owitz M, Clark MA, McAllister HA. Sudden death and partial absen ce of the right ventricular myocardium. Arch Pathol Lab Med 1982; 106:163-7. Thi ene G, Nava A, Corrado D e t a!. Right ventricul ar ca rdiomyopathy and sudd en deat h in young peopl e. N Engl J Med 1988 ; 318:129-33. McKe nna WJ, Thi ene G, Nava A et al. Diagnos is of a rrh y thmoge nic ri g ht ventric ul a r dysplasia/ca rdi o myopathy. Br Heart J 1994; 7 1:2 15-1 8. Fonta ine G. Arrhythmoge ni c rig ht ven tricula r dyspla sia. Curr Opin Cardiol 1995; 10: 16-20. A hmad F, Li D, Karibe A et al. Localisation of a gene responsible for a rrh ythmoge ni c right ventricul ar dyspla sia to chromosome 3p23. Circu lation 1998; 98 :2791 - 5. Gerlis LM, Schmid t-O tt SC, Ho SY, Anderson RH. Dysp lastic cond itions of the right ventricu lar myocardium: Uhl's anoma ly versus arrhythmoge nic right vent ricu lar dysplasia . Br Heart J 1993; 69: 142 - 50. Tansey OK, A ly Z, Sheppa rd MN. Fat in the ri g ht ventricle of the no rmal heart. Histopath ology 200 5 ; 46 :98- 104. Krous HF, Ch adw ick AE, Isaacs H. Tum ors associated w ith sudden infan t and childhood death. Pedia tr Dev Pathal 2005; 8:20-5. Shehata BM, Patterson K, Thomas JE et al. Histiocytoid cardiomyopathy: Three new cases and a review of the literature. Pediatr Dev Parhol 1998; 1:56-69. Suarez V, Fuggle WJ. Cameron AH et al. Foamy myocardia l transformation of infa ncy: a n inherited disease. J Ciill Pathol 1987; 40:3 29 - 34. Koponen MA, Siegel RJ. Hi st iocytoid ca rdi omyopathy a nd sudden death. Hum Pathol 1996; 27:4 20-3 . Ness MJ. McManu s BM. Anomalous right coronary artery origin in otherw ise unexplained infant death. A rch Pat/101 Lab Mcd 1988; 112:626-9.
251
56 Lipsett J. Byard RW, Carpe n ter BF et at. Anomalous co ro nary aneries arising from the ao rta assoc iated with sudden death in infancy and ea rly childh ood. Arch Pa thol Lab Med 1991 ; 115 :770- 3. 57 Corrado D. Thiene G. Cocco P, Frescura C. Non atherosclerotic coronary artery disease and sudd en death in the young. Br Heart J 199 2 ; 68 :60 1-7. 58 Garfia A, Rodrigu ez M, Chava rria H. Garrid o M. Sudden ca rdiac death during exe rcise due to an isolated multiple anomaly of the left co ronary a rtelY in a 12-year- o ld g irl: Clinicopatholog ic fin dings. J Fo rens ic Sci 1997 ; 42:330-4. 59 ScuJly RE, Mark EJ . McNeely W F. McNeely BU. Case reco rds of th e Massachusetts Genera l Hospital . Case 22 -1 989. N Engl J lVled 1989; 320: 1475 - 83 . 60 Thiene G, Pennelli N, Rossi L. Cardiac conduction system abnormaliti es as a possible cause of sudd en de a th in yo ung athletes. Hum Pathol 1983; 14:704-9. 61 Angelini P. Normal and anomalous coronalY arte ries : detinitions and classification . A 111 Heart J 1989; 117 :41 8- 34. 62 Price J . Kawasaki sy ndro me. Br Med J 1984 ; 288:262 - 3. 63 Curti s N. Kawasaki disease. Br Med J 1997 ; 315:322-3 . 64 Ti za rd EJ, Suzuki A. Levin M. Dill on MJ. Clinical as pects of 100 patients wit h Kawasa ki disease . Arch Dis Child 1991 ; 66:18 5- 8. 65 Nakamura Y. Yanagawa H, Ojima T et al. Cardiac sequelae of Kawa saki disease among recurrent cases . Arch Dis Child 1998; 78:1 63 -5. 66 Hoare S. Abinun M. Cant AJ. Ov erl ap between Kawasaki disease and Group A streptococcal infection (Letter). Pediatr /l ljixt Dis J 1997; 16:633 -4. 67 Bya rd RW. Cardiac co nditions. In Byard RW (ed.) Sudden Death in Infancy Childh ood and Adolescellce. 2nd edn. Cambrid ge: Ca mbrid ge University Press. 2004 , pp. 202 - 48. 68 Bharati S. Bauernfiend R, Scheinm an M et al. Congenita l abnormalities of the conducti on system in two patients with tachyarrhythmias. Circulation 1979; 59:593-606. 69 Splawski I, S he n J, Timothy KW et al. Spectrum of mutations in long Q-T syndrome genes KVLQTI. HERG. SCN5A. KCNE I and KNCE2. Circula tiol1 2000 ; 102: 1178-85. 70 Towbin JA. Molecular basis of sudden ca rdiac death. Pediatr Ciin N Am 2004; 51: 1229 - 55. 71 Moran JJ. Idiopathic arterial calcificatio n of infancy: a clin icopatho logic study. Pa t/wi Ann 1975; 10:393-417. 72 Rutsch F. Vainganka r S. Johns on K et al. PC-I nucleos ide triphosphate pyrophosphohydrola se defiCiency in idiopathi c infanti le arte rial calcification. Am J Parha l 2001; 158 :543-54. 73 Rutsch F. Ruf N. Vaingankar S et al. Mutations in ENPP 1 a re associated with 'idiopathi c' infantile arterial calcification. Nat Geller 2003; 34:379-81. 74 Juul S. Ledbetter D. W ight T N, Wood rum D. New in sights into idiopathic infantile arteria l ca lci nos is. Three patient repons. Am J Dis Child 1990; 144 :229-33 . 75 Byard RW. ldiopathic arterial calcification and unexpected infant death . Pediatr Pathol Lab Med 1996; 16:985-94. 76 Joshi VV. Pawel B. Connor E et al. Arteriopathy in children with acquired immune de fi ciency syndrome. Pediatr Pathol 1987; 7:261-75. 77 Siegel RJ. Dunton SF. Systemic occlusive arteriopathy with su dden death in a 10-year- old boy. Hum Parhol 1991; 22:197- 200. 78 Im a mura M. Yokoya ma S. Kikuchi K. Coronary fibromuscular dys plasia presen tin g as sudden infant deat h. Arch Pathol Lab Med 1997; 121 : 159-6 1. 79 Rosenberg HG. Systemic arterial disease and chronic arse ni cism in infants. Arch Pathol 1974; 97:360- 5.
252 I
Sudden natural death in infants and children
80 Rol fes DB, Towbin R, Bove KE. Vascular dysplasia in a chil d with tube rou s sclerosis. Pediatr Pathol 1985 ; 3 :359-73. 81 Mudd SH, Levy HL, Skovby F. Disord ers of transsulfuration. In Sc river CR, Beaudet Al, Sly WS, Va lle D (ed s) The Meta bolic and Molecular Bases of Il1h erited Disease. New Yo rk: McGraw-Hili Inc. , 199 5; 1279- 327. 82 Fo rtu in NJ, Morrow AG, Roberts We. Late vascular
manifesta tions of the Ru be lla synd rom e. Am J Med 1971;
51: 134-40. 83 Hamilton SJ, Allard MF, Friedm a n JM. Cardiac Findin gs in an indiv idual with neurofibromatos is I and sudden dea th . Am J Med Genet 2001; 110:95-9. 84 Uno H, Arya S, Laxova R, Gilb ert EF. Menkes' syndro me
with vascula r and ad ren ergic nerve abn ormalities . Arch
Path ol Lab Med 1983; 107 :286- 9.
85 Fleishe r GR , Buck BE, Co rnfeld D. Prim ary intimal fibroplasia in a child with Down's syndrome. Am J Dis Child 1978 ; 132 :700-3 . 86 Cina SJ, Smialek, JE, Burk e AP et al. Primary ca rdiac tumors causing sudden death: a review of the literature. Am J Forensic Med Pa thol 1996 ; 17:271-81. 87 Siega l GP. Pri mary tumors of mu scl e. In Stocker JT, Askin
FB (eds) Pathology of Solid Tum on il1 Ch ildren. London:
Chapman and Hall Medical, 1998, pp. 161-81.
88 Parker KM , Embry .IH. Sudden dea th due to tri cuspid va lve myx om a with massive pulmonary emb olism in a 15-m onth old male. J Forensic Sci 199 7; 42 :524-6. 89 Krous HF, Chapman AJ, Alts hul er G. Cardiac hema ngioma: a rare (or poss ible) cause of su dden death in child ren. J Foren sic Sci 1978 : 23:375-8. 90 Swalwell CI. Be nign intra cardiac teratoma - a case of
sudd en death. A rch PatilOl Lab Med 1993; 117 :739-42.
91 Byard RW, Cutz E. Sudden and un expec ted dea th in in fa ncy and ch ildhood due to pulm ona lY thro mb oemb olism. Arch Pathol Lab Med 1990 ; 114:14 2-4. 92 Champ C, Byard RW. Pulmonary thromboembolism and
unexpected dea th in infancy. J Pediatr Child Health 199 4;
30:550-1.
93 Pa tton DJ, Ga lliani CA, Johnson WH, Hed lund GL. Sudd en
death in Marfan syndrome. A m Joumal of Roentgenology
1995; 165:1 60.
94 Masse J-F, Perusse R. Ectoderm al dysplasia. Arch Dis Child 1994; 71: 1- 2. 95 Clarke A, Phillips DIM , Brown R, Harpe r PS. Clini cal aspects of X-linked hypohidrotic ectodermal dyspl as ia. Arch Dis Child 198 7: 62:989-96. 96 Waters KA, Evere tt F, Sillen ce D et al. Breathing
abno rm a lities in sleep in achondroplasia. Arch Dis Child
1993; 69:191-6.
97 Bla nd JD, Emery JL. Unexp ected death of ch ildren with
achondroplasia a fter th e perin a tal period . Dev Med Child
Neurol 1982: 24:489-92.
98 Webb DW, Osbo rne JP. Tuberous scl erosis. A rch Dis Child 199 5; 72:471-4. 99 Nir A, Taj ik J , Freeman WK et a l. Tuberous scl erosis and cardia c rh abdomyosarcoma. Am J ·Cardiol 1995; 76:4 19 - 21. 100 O'CalJaghan FJK, Clarke AC, Joffe H et al. Tuberous scle ros is complex a nd Wo lff-ParKinson - Wh ite sy ndrome. Arch Dis Child 1998 ; 78: 159-62. 101 Livingsto n JH , Brown JK. Intrac erebra l haemorrhage after the neon ata l period. Arch Dis Child 1986; 61 :538-44. 102 Gieteling EW, Rinkel GJE. Ch aracteristics of intracranial a neu rysms and subara ch noid haemorrhage in patients with polycysti c kidn ey disease. J Neural 2003; 250:418- 23. 103 Lilova MI , Petkov DL. Intracranial aneurysms in a child with autosom a l. recessive poly cysti c kidney disea se. Pediatr Nephra l 2001 ; 16: 1030- 2.
104 Byard RW, Bo urn e AJ, Hanieh A. Sudden and unexpected death due to hemorrhage from occult central nervo us system lesions. Pedia tr Neurosurg 1991 -92; 17:88-94. 105 Lill ey man JS. Intrac rania l haem orrhage in idiopathic thrombocy topeni c purpura. Arch Dis Child 1994; 71 :251-3. 106 Abu AI Ragheb SY, Koussous KJ, Arm SS. Intracra nial neop lasms associa ted with sudd en death : a rep ort of seve n cases a nd a review of the literature. Med Sci Law 1986 ; 26:270-2. 107 DiMa io SM, DiMaio VJM, Kirkp atrick J B. Sudden , un expec ted death s due to primary intracranial neopl as ms. Am J Foren sic Med Pathol 1980 ; 1:29-45. 108 Bya rd RW, Moo re L. Sudd en and unexpec ted death in childhood du e to a co lloid cyst of the third ven tricle. J Forens ic Sci 1993; 38 :210 -13. 109 Buzzi S, Verdura C, Arlati S, Colecchia M. Sudden deat h in a child due to ra re endo crani al neo fo rm ation . Med Sci Law 1998; 38: 176-8. 110 Leestma JE, Walcz ak T, Hughes JR et al. A prospective study on sudden une xpected death in epilepsy. Ann Neural 1989; 26:195-203. II I Byard RW, Keeley FW, Smith CR. Ty pe IV Ehlers-Danl os sy nd rome presenting as sudd en infant death. Am J Clin Patl101 1990 ; 98:579-82. 11 2 Elterman RD, Bodensteiner JB, Barnard JJ. Sudden unexp ec ted death in patien ts with Dandy-Walker malformation . J Child Neurol 199 5; 10:382-4. 113 Vaishnav A, Ma cKinn on AE. Prog ressive hydrocephalus in teena ge spina bi fid a patients. Z Kindercilinl rg 1986; 41(Suppl. 1):36-7. 114 Stellm a n GR, Banniste r CM, Hillier V. The incidence of seizure disorder in children with acq uired a nd congenital hydrocephalus. Z Kinderchirurg 1986 ; 41( Suppl. 1):38-9. 115 Friede RL. Meningoencephalitic pro cesses in th e perinata l period. Deue!opmental Neuropathology. New York: Sprin ge r Verlag, 1975 ; 178-87. 116 Howat AJ, Smith CML, Variend S. Sudden infant death a nd encephalitis in two sib lings. Dev Med Child Neu ra l 1987; 29:539-40. 117 Shield s LBE, Handy TC, Parker JC, Burns e. Postmoltem diagnosis of leukodystrophi es. J Forensic Sci 1998 ; 4 3:1068 -71. Jl8 Kereny i N. Conge nital ad renal hyp opla sia. A rch Pathol 196 1; 71 :336- 3. 119 Irvine WJ, Toft AD. Diagnosing adrenocortical insufficiency. Practitioner 1977; 218 :539-45. 120 AI Sabri ANI, Sm ith N, Busuttil A. Su dden death due to a uto immun e Addiso n's disease in a 12-yea r-old girl. Int J Legal Med 199 7; 110 :278-8 0. 121 Fava ra BE, Franc iosi RA, Mil es V. Idiopathic adrena l hypop lasia in children. Am J ((in Pa tl10 I 1972 ; 57 :287-96. 122 Brown W, Singer DB. Pregnan cy-indu ced hyp ertension and congenital adrenal hypoplasia. Obstet Gynaecol 1988; 72: 190-4. 123 Bacon CJ , Hall SM. Haemorrh agic s hock encephal opa thy syndrome in the British Isles. A rch Dis Child 199 2; 67:985-9 3. 124 Li ttle D, Wilkin s B. Hemo rrhagic shock and encephalopa thy syndrom e : an unusual cause of sudd en death in children. Am J Forensic Med Pathol 1997; 18 :79-8 3. 125 PHLS Communicab le Disease Surveill ance Centre. Joint British Paediatric Association a nd Communicable Disease Surveillan ce Scheme for haemorrhagic sho ck encephalopathy syndrome surveill a nce report for 198 2-4. Br Med J 1985 ; 290: 1578-9. 12 6 Glasgow JFT, Mo ore R. Reye's sy ndro me 30 years on. Br Med J 1993; 307 :9 50-1.
References I
127 Anon. Rey e's syndrome and aspirin: epidemiolog ical associations and inborn erro rs of metabol ism. Lancet 1987; ii:429-31. 128 Bonnel HJ, Beckwith JB. Fatty liver in sudde n childhood death. Am J Dis Child 1986; 140:30-3. 129 Forsyth BW, Shapiro ED, Horwitz Rl et al. Misdiagnosis of Reye's-like illn ess (Letter). Am J Dis Child 1991 ; 145:964-6. 130 Green A, Hall SM. Investigation of metabolic disorders resemblin g Reye's syndrom e. Arch Dis Child 1992; 67: 13 13-17. 131 Hardie RM, Newton LH, Bruce JC et al.. The chan ging clinical pattern of Reye's sy ndrome 1982- 1990. Arch Dis Child 1996; 74:400-5. 132 Rowe PC, Valle D, Brusilow Sw. Inborn errors of metabolism in children referred with Reye's synd rome . JAm Assoc 1988; 260:3167-70. 133 Byard RW, Bohn DJ, Wilson G et al. Unsuspected diaphragmatic hernia: a potential cause of sudden and unexpected death in in fancy and early childhood. J Pediatr Surg 1990; 25:1166-8. 134 Pfalzgraf RR, Zumwalt RE, Kenny MR. Mesodivelticular band and sudden death in children. Arch Puthol La.b Med 1988; 112:182-4. 135 Pumphrey RSH, Stanw orth SJ. The clinical spectrum of anaphylaxis in north-west England. c/in Exp Allergy 1996; 26: 1364-70. 136 Pumphrey RS, Robert IS. PostmOltem findings after fatal anaphyl actic reactions. J c/in Pa thol 2000; 53 :273-6. 137 Delage C, lrey NS. Anaphylactic deaths: a clinicopathologic study of 43 cases. J Forensic Sci 1972; 17:5 25 - 39. 138 Sampson HA, Mendelson MD, Rosen JP. Fatal and near-fatal anaphyl actic reaction s to food in childre n and adolescents. N Engl J Med 1992 ; 327:3 80-4. 139 Riches KJ, Gillis D, James RA. An autoJlsy approach to bee sting-related death s. Pathology 2002; 34:257-62. 140 Byard RW. Miscellaneous conditions. In Byard RW (ed.) Sudden Death ill Infancy, Childhood and Adolescence, 2nd ed n. Camb ridge: Cambridge University Press, 2004, pp. 469-88. 141 Assem ESK, Gelder CM, Spiro SG et al. Anaphylaxis induced by peanuts. Br i\lIed J 1990; 300: 1377 -8. 142 Patel L, Radivan FS, David T1. Management of anaphylactic reactions to food. Arch Dis Child 1994; 71:370-5. 143 Powars D, Chan LS, Schroeder WA. The variable expression of s ickle cell di sease is genetica lly determined. Semil1
Hematol1990; 27:360-76. 144 Jenkins ME, Scott RB, Bairs RL Stud ies in sickJe cell anaemia. J Pedia tr 1960; 56:30-9. 145 Leikin SL, Gallagher D, Kinney TR et al. Mortality in children and adolescen ts with sickJe cell disease. Pediatrics 1989; 84:500-8. 146 Clini copathologic Conference. Sudden death in a young woman with sickle cell anaemia. Am J Med 1992; 92:556-60. 147 Cutz E. Idiopathic pu lmonary hemosiderosis and related disord ers in infan cy and childhood. In Rosenb erg HS, Bernstein J (eds) Respiratory I1l1d Alimentary Tract Diseases. Basel: Boehm-Hutter Press, 1987, pp. 47-81. 148 Stocker JT. The respiratory tracr. In Stocker JT, Dehner LP (eds) Pediatric Pathology Vol. 1. Philadelphia: Lippincott Company, 1992; 505-73. 149 Schwartz MZ, Fill.er RM. Tracheal compression as a cause of apnea following repair of tracheo-oesophageal fistula: treatment by aortopexy. J Pediatr Surg 1980; 15:842-8. 150 Trollfors B, Nylen 0 , Strangert K. Acute epiglottitis in child ren and adults in Sweden 1981-3. Arch Dis Child 1990; 65:491 -4. 151 Claesson B, Troll fors B, Ekstrom-Jodal B et al. Incidence and prognosis of ac ute ep iglottitis in children in a Swed ish region. Pediatr Infect Dis 1984; 3:534-7.
253
152 Coulthard M, Isaacs D. Retroph ary ngeal abscess. Arch Dis Child 1991; 66: 1227-30. 153 Barrat GE, Koopmann CF Jr, Coulthard Sw. Retropharyngeal abscess - a ten-year experience. Laryngoscope 1984; 94:455-63. 154 Brook J. Microbiology of retropharyngeal abscesses in chi ldren. Am J Dis Child 1987 ; 141:202-4. 155 Donne lly BW, McMillan JA, Wei ner LB. Bacterial tracheitis: Report of eight new cases and review. Rev Infect Dis 1990; 12:729-35. 156 Britto J, Habibi P, Walters Set al. Systemic complications associated with bacterial tracheitis. Arch Dis Child 1996; 74:249-50. 157 McKenzie M, Norman MG, Anderson m, Thiessen PN. Upper respiratory tract infection in a 3 year old girl. J Ped iatr 1984; 105:129-33. 158 Cant AJ, Gibson PJ, West RJ. Bacterial tracheitis in Down's synd rome. Arch Dis Child 1987; 62:962- 3. 159 Pan itch HB, Callahan CW, Schidlow DV. Bro nchiolitis in children. Clin Chest Med 1993 14:715-31. 160 Price JE Acute and long-term effects of vi ra l bronchiolitis in infancy; Lung 1990; 168(Suppl.):414-21. 161 Meert K, Heidemann S, Abella B, Sarn aik A. Does prematurity alter the course of respiratory syncytial virus infection? Crit Care Med 1990; 18 : 1357- 9. 162 MacDonald NE, Breese Hall C, Suffin SC et al. Respiratory syncytial viral infection in infants with congenital heart disease. N Eng! J Med 1982; 307:307-400. 163 Groothuis JR, Gutierrez KM, Lauer BA. Respiratory syncyt ial virus infection in children wit h bronchopulmonary dysplasia. Pediatrics 1988; 82:199 - 203. 164 Hall CB, Powell KR, Ma cDonald NE et al. Respiratory syncytial viral infection in chi ldren with compromised im mun e function. N EI1g1 J Med 1986; 315:77-81. 165 Hastleton PA, Ironside JW, Whittaker JS et al. Pulmonary veno-occlusive disease. A report of four cases. Histopathology 1986; 10:933-44. 166 Cagle P, Langston C. Pulmonary veno -occlus ive disease as a cause of sudden infa nt death. Arch Pat/wi Lab Med 1984; 108:338-40. 167 Jay GW, Leestma JE. Sudden death and epilepsy. Acta Neurologica Scandinavica 1981; 63(Suppl. 82): 11 -16. 168 McGugan EA. Sudden unexpected deaths in epileptics a literature review. Scot iVIed J 1999; 44:137-9. 169 Shields WD. Status epilepticus. Pediatr Clin North Am 1989; 36:383 - 93. 170 Leestma JE. Forensic aspects of complex neural dysfunctions. In Leestma JE (ed.) Forensic Neuropatho logy. New York: Raven Press, 1988, pp. 396-428. 171 Coyle HP, Baker-Brian N, Brown Sw. Coroners' autopsy reporting of sudden unexplained death in epil epsy (SUDEP) in the UK. Seizure 1994; 3:247-54. 172 Nashef L, Walker F, Allen P et al. Apnoea and bradycardia during epilept ic seizures: relation to sudd en death in epil epsy. J Neurol Ne'urosuTg Psychiatry 1996; 60:297-300. 173 Lip GYH, Brodie JVIJ. Sudden death in epilepsy: an avoidab le outcome? J Royal Soc Med 1992; 85:609-11. 174 Brown SW, Mawer GE, Lawler W et al. Sudden death and epilepsy (Letter). Lancer 1990; 335:606-7. 175 Timmings PI.. Sudden unexpected death in epi lepsy: is carbamazipine implicated? Seizure 1998 ; 7:289-91. 176 Gars on A Jr. Medicolegal probl ems in the management of cardiac arrhythmias in children. Pediatrics J 987 ; 79(1):84-8. 177 Brown DC, Godman MJ. Life-threatening 'epi lepsy'. Arch Dis Child 1991; 66:986-7.
254 I
178
Sudden natural death in infants and children
Ch amp C, By ard RW. Sudden death in asthma in chi ldhood .
Forens ic Sci Int 1994; 66:117-27. 179 180 181 182 183 184
185
J 86 187
188
189
190
191
192
193
194
195
196 197
198
199
200 201 202 203
Zach MS , Karner U. Su dden death in asthma. Arch Dis Child 1989; 64: 1446-51. Carswell F. Thirty deat hs from asthma. Arch Dis Ch ild 1985; 60: 25 -8. Kravis LP, Kolski GB. Unexpected death in childhood asthma. Am J Dis Child 1985; 139:558-63. Benatar SR. Fatal asthma. N Engl J IVIed 1986; 314:423-9. Larsen GL. Ast hma in children. N Engl J IVIed 1992; 326:1540- 5. Molfino NA, Nannini LJ. Martelli AN, Sl utsky AS. Respiratory a rrest in near-fatal asthma. N Engl J Med 1991 ; 324:285- 8. Haalboom JRE, Deenstra M. Stmyvenberg A. Hypokalaemia induced by inhalation of fenot erol. Lallcet 18 May 1985 ; 2 :1125-7. Pearce N. Crane J. Burgess C et al. Beta ago nists a nd asthma mOl1a lity: deja vu. Clin Erp Allergy 1991 ; 21 :401 - 10. Warner DP, McKinn ey PA. La w GR, Bod ansky HJ. MOl1ality and diabetes from a population-based register in Yorkshire 1978-93. Arch Dis Child 1998; 78:435-8. Dahlquist G, Kallen B. Mortality in childhood-onset type 1 diabetes: a population-based study. Diabetes Care 2005; 28:2381-7. Edge JA, Ford-Ad ams ME, Dunger DB. Causes of death in children with insulin- depend ent diab etes 1990-96. Arch Dis Child 1999; 81:318-23. DiMaio VJM. Sturn er WQ, Coe JI. Sudden and unexpected deaths after the acute onset o f diabetes mellitus. J Foren sic Sci 1977 ; 22: 147-5 1. Rosenbloom Ai. Kohrman A. Sperling M. Classifica tion a nd diagnosis of diab etes mellitus in children and adolescents. Joul11al of Pediatrics 1981; 99:320-3. Coe JI. Post-mortem bioch em istry of bl ood and vitreous humour in paediatri c practi ce. In Mas on JK (ed.) Paediatric Forensic Medicine and Pa thology. London: Cambridge Universi ty Press, 1989, pp. 191-203. Olpin SE. The metabo lic investigation of sudd en infant death. A/1I1als of Clinical Biochemistry 2004; 41: 282 -93. John WG . Scott KWM, Hawcroft DM. Glycated haemoglobin an d glycated protein and glucose concentrations in necropsy blood samples. J Clin Pa thol 1988 ; 41:415-18. Rozin L, Perper JA. Jaffe R, Drash A. Sudden unexp ected death in childhood due to unsuspected diab etes mellitus. Am J Forensic Med Pa thol 1994; 15:25 1-6. Tattersall RB. Gill Gv. Unexplained deaths of type 1 diabetic patients. Diabetic Medicin e 1991; 8:49-58. Thordarso n H, Sovik O. Dea d in bed syndrome in young diabetic patients in Norway. Diabetic Medicine 1995; 12:782-7. Matyka K, Ford-Adams M. Dunge r DB. Hy poglycaemi a and co unterregulation during childh ood. Horm Res 2002; 57(Suppl. 1):85-90. Emery JL. Howat AJ, Variend S. Vawter GF. Investigat ion of inborn errors of metabolism in un expec ted infant deaths. Lancet 1988; ii: 29 -31. Pollitt RJ. Disord ers of mitochondrial long -cha in fatty acid oxidation. J Inh erited Metob Dis 1995; 18:473-90. Pollitt RJ. Inherited disorders of straight chain fatty ac id oxidation. Arch Dis Child 1987; 62:6-7. Anon. Sudden infant death and inherited disorders of fat oxidation. Lan cet 8 Nov ember 1986; ii: 1073- 5. Benn ett MJ, Hale DE, Coates PM , Sta nley CA. Postmortem recognition of fatty acid oxidation disord ers. Pediatr Pathol 1991; 11:365-70.
204 Largilliere C, Vianey-Saban C, Fontaine M et al. Mitochondrial very long chain acyl-CoA dehydrogenase defiCiency - a new disorder of fatty acid oxidation. Arch Dis Child 1995; 73:FI03-S. 205 Olpin SE. Allen J. Bonham JR et al. Features of Carnitine Palmitoyltransferase Type I Disease. J Inherit lvIetab Dis 2001; 24:35-42. 206 Chace DH. DiPerna JC, Mitch ell BL et al. Electrospray tandem mass spectrom etry for analysis of acylcarnitines in dried postmortem blood specimens collected at autopsy from infa nts with unexp lained cause of death. Clin Chem 2001 ; 47:JJ66-82. 207 Touma EH. Charpentier C. Med ium chain acyl-C oA dehydrogenase deficiency. Arch Dis Ch ild 1992 ; 67:142-5. 208 Roe CR. Millington OS. Maltby DA. Kinnebrew P. Recognition of medium-chain acyl-CoA dehydrogenase deficiency in asymptomatic siblings of children dy ing of sudden infant death or Reye-like syndromes. J Pediatr 1986; 100:13-18. 209 Anon. Med ium chain acyl CoA dehydrogenase deficiency. Lancet J991; 338:544-5. 210 Shortland G. Besley G, Bonham J et al. Newborn screening for medium chain acyl CoA dehydrogen ase deficiency (MCADD): Find ings from a multicentre prospective UK collaborative study. J Inherit Metab Dis 2006; 29(Suppl. 1): 19. 2JJ Losty He. Lee P, AI Faham M et al. Fatty infiltration in the liver in medium chain acyl CoA dehyd rogenase deficiency. Arch Dis Child 199 1; 66:727-8. 212 Bove KE. Letter to the Editor. Pediatr Patlwl 199 2; 12:621-5. 213 Guenthard J. Wyl er F. Fowler B, Baumgartner R. Cardiomyopathy in respiratory chain disorders. Arch Dis Child 1995; 72: 223 -6. 214 Smeitink JAM. Fischer JC. Ruitenbeek W et al. Sudden infant death associated with defect ive oxidative ph osp horylation. Lancet 199 3; 341: 1601. 215 Konstantopouloll V, Sperl W. Wohl gena nnt J et al. Nea r missed sudden unexpected infant dea th (SUID) as the firs t sign of a respiratory chain co mplex I deficiency. J Inh eri t lvfetab Dis 2001; 24(SuppL 1):78. 216 Tripp ME. Katcher ML. Peters HA et al. Systemic carnitine deficiency presenting as familial endocardial fibroelastosis. N El1gl J Med 1981; 305:385-90. 217 Steihm ER. Damrosch DS. Factors in the prog nosis of meningococcal infection. J Pedialr 1966; 68:457-67. 218 Marzou k 0, Thoms on APJ, Sills JA et al. Features and outcome in meningococca l disease presenting with maculopapul ar rash. Arch Dis Child 1991; 66:485-7. 219 Morl ey SL, Lev in M. Bacterial meningitis. Prescr J 1998; 38:129-4l. 220 Sh arief N, Kh an K, Conlan P. Overwhelming sepsis presenting as sudden unexpected death. Arch Dis Child 199 3; 69:381-3. 22 1 Dyke MP, Martin RP, Berry PJ. Septicaemia and adrenal ha emo rrhage in congenital aspl enia. Arch Dis Child 1991; 66: 636 -7. 222 Larkin SM, Williams ON, Osterholm MT et al. Tox ic shock syndrome: clinical. laboratory, an d pathologic findings in nine fatal cases. Annals of internrillvfedicine 1982 ; 96:858-64. 223 Con e LA. Woodard DR. Schlievert PM. Tomory GS. Clinical and bacteriologic observations of a toxi c shock-like syndrome due to Streptococcus pyogenes. N Engl J Mer! 1987; 317:146-9. 224 Steve ns DL. Tanner MH . Winship J et al. Severe g roup A streptococcal infection s associated with a toxic shock-like syndrome and scarlet fever toxin A. N Eng l J Med 1989; 321 :1-7. 225 Stamenkovic I. Lew PD. Earl y recognition of potentially fatal necrotizing fasciitis. N Engl J lvled 1984 ; 310:1689-93.
References I
226
227
228
229
230
231
232
233
234
Chelsom J, Ha lstensen A. Haga T, H0iby EA. Necro tisi ng fasciitis due to g roup A streptococci in western Norway: incidence and cI inical features. Lancet 1994: 344: II J 1-15. Tyrrell GJ, Lovgren M, Kress B, Grimsrud K. Varicella associated invasive group A streptococcal disease in Alberta, Canada - 2000-2002. Ciin Infect Dis 2005; 40: 105 5-7. Vugia DJ, Peterson CL, Meyers HB et aJ. Invasive group A streptococcal infections in children with varicella in Southern California. Pediatr Infect Dis J 1995; 15:146-50. Sawin RS , Schaller RT, Tapper D et al. Early recognition of neonatal abdomina l wall necrotizing fasciitis. Am J Surg 1994; 167:481 -4. Stocker JT. Pathologic features of long-standing 'healed' bron chopulmonaty dysp lasia. Hum Pathol 1986 ; 17:943-61. Miller RW, Woo P, Kellman RK. Tracheobronchial abnormalities in infants with bronchopulmonary dysplasia. J Paediatr 1987; 111 :779-82. Greenholz SK, Hall RJ, Lilly JR, Shikes RH. Surgical implications of bronchopulmonary dysplasia. J Pedia tr Surg 1987; 22:1132-6. Doull lJ, Mok Q, Tasker RC. Tracheobronchomalacia in pre term infants with ch ronic lung disease. Arch Dis Child 1997; 76:F2OJ- 5. Tammela OKT, Koivisto ME. A I-year foHow-up of lo w birth weight infants with and without bronchopulmonalY dysp lasia: hea lth , growth, clinical lung disease, cardiovascular and neurological sequelae. Early Hum Dev 1992 ; 30: 109-20.
255
235 Reardon W, Newcombe R, Fenton I et aJ. The natural history of congenital myotoniC dystrophy: mortality and long term clinical aspects. Arch Dis Child 1993; 68:177-81. 236 Thomas PK, Caine DB, Elliott CF. X-linked scapuloperoneal syndrome. J Neurol Neuroslirg Psychiatry 1972; 35:208-1 5. 237 Primhak RA, Spicer RD, Variend S. Sudden death after minor abdominal trauma: an unusual presentation of phaeochromocytoma. Br Med J 1986; 292:95-6. 238 O'Neill D, O'Grady J, Variend S. Child fatality associated with pathological features of histiocytic necrotizing lymphadenitis (Kikuchi-Fujimoto Disease). Pediatr Pafhol Lab Med 1998; 18 :79-88. 239 Morens OM , Hammar SL, Heicher DA. Idiopathic acute pancreatitis in Children. Am J Dis Child 1974; 128:401-4. 240 Wilcken B, Hammond J, Silink M. Morbidity and mOltality in medium chain acyl coenzyme A dehydrogenase deficiency. Arch Dis Child 1994; 70:410-12. 241 Traisman ES, Young S, Lifschultz BD et al. Sudden death in a neonatal as a result of Herpes simplex infection. J Forensic Sci 1988; 33:267-71. 242 Mercuri E, Cowan F, Rutherford M et aJ. Ischaemic and haemorrhagic brain lesions in newborns with seizures and normal apgar scores. Arch Dis Child 1995; 73 :F67-74. 243 Norman MG, Taylor GP, Clarke LA. Sudden, unexpected, natural death in childh ood. Pediat,. Pafho11990; 10:769-84. 244 Variend S. Infant mortality, microglial nodules and parotid CMY-type inclusions. Ear/y Hum Del) 1990; 21:31-40. 245 Variend S, O'Neill D, Arnold P. The possible significance of cytomegaloviral parotiti s in infant and early childhood deaths. Arch Pat/wi Lab Med 1997; 121 :1272-6.
I
CHAPTER 13
I
RECENT ADVANCES IN PAEDIATRIC TOXICOLOGY Patrice Mangin and Christian Giraud
Scope of the problem Specificity of paediatric toxicology Techniques used in drug testing Special techniques for analysis of volatile substances Alternative specimens for drug testing Pitfalls and limitations of drug screens
256 258 259 262 262 267
During the past decade paediatric toxicology has substan tially advanced in line with all other branches of medical toxicology; such advances should be considered in the context of general forensic toxicology, with particular attention being directed at the various aspects of the sub ject that a re so mewhat unique to the paediatric age group. In this respect, this chapter focuses on several important areas of forensic toxicology that have undergone signifi cant changes in the past decades, but which also are of par ticular interest and specific importance to paediatrici ans and pathologists involved in a medicolegal or forensic practice. After a brief insight into the major issues raised during the last few years in this specialty, this chapter will outline what the paediatrician and/or the forensic pathologist can expect to learn from toxicological investigations in the light of the recent advances in analytical toxicology with reference to drug testing, and the circumstances in which it may be app ropriate to initiate laboratory tests looking for drug use.
SCOPE OF THE PROBLEM The age distribution of children involved in poisonings is bimodaL I Ingestion of, or exposure to, toxic substances by children before the age of 5 is almost invariably accidental, with the peak of instances occurring in the toddler years.2
Specific applications The importance of paediatric toxicology in specific cases Conclusions and future considerations in forensic paediatric toxicology References
271 274 274 275
Poisoning agents are usually those found in the home and may include household chemicals, such as cleansers, hydro carbon fuels, paints and thinners, plants and fungi, and medications that belong to other household members. 3 Poi soning appears to be frequent in this age group: according to the Toxic Exposure Surveillance System compiled by the American Association of Poison Control Centers, 50 per cent of poisoning cases involved children under 6 years of age. In contrast, fatalities in this age group are uncommon, accounting for only about 4 per cent of all poisoning fatal ities. 4 As a result, in any given year, the probability of a forensic pathologist or toxicologist encountering paediatric patients in whom drugs or poisons can be detected is quite lowS Even in these cases, the exact role played by the drug(s) or substance(s) in the death of the child is not always clear; indeed , the presence of the drug or poison may be categorized as either causing or contributing to death by direct toxic effect, or by idiosyncratic or hyper sensitivity reaction, or through drug-induced disease. The second peak of incidence for poisoning occurs in the adolescent years, when drug ingestion is more common. Teenagers are more likely to ingest dangerous substances intentionally and, in this respect, they generally choose pharmaceutica l agents. 6 - 10 Within the teenage age group, girls are more likely than boys to ingest drugs. Agents involved in adolescent poisonings are often those found in the family medicine cabinet, which include aspIrIn, paracetamol (acetaminophen), iron supplements (ferrous
Scope of the problem I
Table 13.1
Substances that are lethal in young children, even in small doses, as reported by Criddle lO
Benzocaine Beta-blockers Calcium channel blockers
Table 13.2
Camphor Clonidine Cocaine
Oi phenoxylate/atropine Lindane Methyl salicylate
Substances most frequently involved in fatal
exposures in adolescents (7 3- 79 years old) as reported by Criddle lO for 2004
1.
2. 3.
4. 5.
6. 7. 8. 9.
10.
257
Analgesics Sedatives, hypnotics, antipsychotics Antidepressants Stimulants and other street drugs Cardiovascular drugs Alcohol and volatile compounds Gases and fumes Anticonvulsants Chemicals Muscle relaxants
sulphate), antidepressant medications, non-steroidal anti inflammatory drugs (NSAlDs), antipsychotic drugs, drugs of abuse and recreational drugs. Inhalants, generally volatile hydrocarbons and carbon monoxide, must also be considered as well as caustic agents. In this age group, fatalities remain rare, with poisoning accounting for less than 10 per cent of the total fatalities. 4 The 2006 annual report of the Swiss Toxicological Information Centre (www.toxi.ch) indicates that the highest number of caJis received for human poisoning involved children under 5 years of age (41.4 per cent). However, the proportion of cases with potential health risk was somewhat higher in adults (35.7 per cent) than in children (21.2 per cent). Boys were more frequently represented among the children, and women were more frequently represented among the ado lescents and adults. Acute accidental intoxications repre sented the largest group, with children ingesting easily accessible household products, pharmaceuticals or plant parts. Acute intentional poisoning was mostly due to attempted suicide and less frequently to drug abuse or criminal behaviour. Small amounts of some drugs can be very toxic to young children; one pill or one mouthful can require aggressive treatment. Table 13.1 lists 12 toxic compounds or classes of substances which are known as 'one pill can kill' substances. 10, II Table 13.2 lists the 10 leading sub stances responsible for teen fatalities. There is a growing concern in the medical community over the maternal use of illicit drugs during pregnancy, since intra-uterine exposure to drugs of abuse and nicotine
Quinidine Su Ifonylu reas Tricyclic antidepressants
is associated with a high incidence of perinatal complica tions.12,13 Approximately 30 per cent of the substance addicted population in the USA is female, and most are women of child-bearing age. Among pregnant women aged 15-44 years, 4.0 per cent had used illicit drugs and 1L8 per cent had used alcohol within 1 month of the 2006 National Survey on Drug Use and Health. The complica tions in pregnant women include an increased incidence of still birth related to abruptio placentae and placenta prae via; a reduction in birth weight, birth length and Apgar scores; intra-uterine growth retardation; and pre-term birth. 14 - 16 After bilih, withdrawal symptoms including hyperactivity, coarse tremor, poor feeding and poor weight gain are present in almost all infants of heroin-, cocaine or amphetamine-addicted mothers. 17 Sudden infant death syndrome (SIDS) is more common among infants of smok ing mothers. IS, 19 In addition, children born to smoking mothers develop more slowly, both physically and men tally, through their teen years.20 Breast-feeding mothers constitute another group in whom there may be a risk of infantile poisoning. Because many drugs, especially lipophilic substances, are excreted into milk, infants are exposed to adverse effects of these drugs if breast fed. During lactation, multiple situations can arise that require maternal pharmacological treatment. Generally, paracetamol (acetaminophen), low-dose aspirin (acetylsalicylic acid) (up to 100 mg/day) and short-term treatment with other NSAlDs, codeine, morphine and propoxyphene are considered compatible with breast feeding and lacking significant danger. 21 However, recently, codeine intake in a breast-feeding mother resulted in a baby's death. Genotype analysis for cytochrome P450 2D6 (CYP2D6) indicated that the mother was classified as a codeine ultrarapid metabolizer. The clinical consequence was a relatively high morphine blood concentration in the baby resulting from metabolism of morphine, leading to central nervous system depressant effect and neonatal death. 22 Diazepam and its metabolite N-desmethyldiazepam enter breast milk. The accumulation of diazepam in breast fed babies whose mothers are taking the drug may result in infantile lethargy and weight 10ss.23 Drugs and substances that require a careful assessment of risk before prescription to breast-feeding women have been published by Ito in 2000. The same author listed the main drugs that are pre ferred for women who are breast-feeding. 24 Attention deficit hyperactivity disorder (ADHD) is the most common neurobehavioural disorder of childhood that
258 I
Recent advances in paediatric toxicology
affects school-aged children, wi t h a prevalence ranging between 4 per cent and 12 per cent of the general popula tion. 25 The traditional therapeutic approach involves the use of psychostimulants such as methylphenidate. As ther apeutic use of methylphenidate has increased, so the risk of accidental intake by to ddlers and of abuse, misuse or use in suicide attempts by pre-teenagers and adolescents has become greater. Many cases of poisoning or abuse have been reported in the scientific 1iterature. 26 ,27 The use of anabolic steroids among teenagers and young adults is becoming increasingly popul ar, particu larly among those involved in sports requiring great strength, such as weight-lifting and body-building. 28,29 All anabolic steroids are derivatives of the male hormone testosterone and all are intended for 'tissue building' and to increase masculini zing effects, Reported psychological side-effects associated with their use include aggression, psychosis and changes in libido, The real incidence of ana bolic steroid consumption is hard to evaluate, but the res ults of the National Household Survey on Drug Abuse indicate that more than 1 million Americans are current or former users.30 In Germany, the estimated number of juve nile users is about 100000.3J In the USA, 2 per cent of college-aged men have been estimated to use anabolic androgenic steroids. In anonymous US high school sur veys, as many as 5-6 per cent of adolescent male athletes and 1-2 per cent of adolescent female athletes have reported experimenting with anabolic steroids in conjunc tion with weight training to improve sports perform ance. 32 ,33 In this context, forensic medicine is involved in several different areas. The forensic pathologist examines doping-associated deaths. The forensic psychiatrist and forensic physician deal with affective and mood disorders including the results of aggressiveness and Criminality, and problems of dependence on anabolic steroids. It is perfectly legitimate to raise concerns about the use fulness of toxicological investigations if poisoning is sus pected in a paediatric case. Indeed, recent literature has confirmed the conclusion that, in most cases, neither com prehensive toxicology nor drugs-of-abuse screening of poisoned patients significantly influences their clinical management. 34 However, these analyses are useful for diagnostic confirmation and are a prerequisite in certain forensic investigations, The utility of focused quantitative serum assays to determine serum levels of particular poisons is more helpful to paediatric clinicians. 35 In a retro spective study of all comprehensive emergency department toxicology screenings performed in paediatric patients, 463 cases were reviewed by Belson et al. 36 In this study, more than 550 toxins were screened, Of 234 positive screening tests, only seven were positive without a documented sus picion of an exposure. In these cases, detection of these specific drugs did not result in a change in medical man agement a nd their presence did not affect the patient's clinical outcome. However, toxicological analyses are jus tified, particularly if the suspected pOison (e.g. carbon
monoxide, CO) can endanger the life of other people. Infants and children are at greater risk of CO poisoning than adults. Fetuses are particula rly vulnerable, because maternal CO crosses the placenta. 37 The rapid identification of the toxic compound may also help in selecting the appropriate antidote, especially if it is potentially toxic. For example, the standard antidote for acetaminophen (para cetamol) toxicity is N-acetylcysteine (NAC) . Intravenous administration of NAC may induce anaphylactoid reac tions. J8 In a few rare cases, administration of the false anti dote may therefore increase the symptoms of toxicity.
SPECIFICITY OF PAEDIATRIC TOXICOLOGY Absorption of a drug from the gastrointestinal tract is slower in infants than in older children an d adults.J9 The skin of infants is more permeable to drug diffusion than in adults, and the blood-brain barrier of infants is also more readily penetrated. 4o Enzyme systems involved in drug metabolism are usually functionally immature at birth, leading to prolonged elimination half-times for drugs. The oxidation P450 and conjugation glucuronidation systems require several weeks after birth to reach maturity. In contrast, the sulphation pathway is more mature and may partially compensate for the deficiencies of the other meta bolic pathways41 For multiple enzyme families, a'develop mental switch' is observed that characterizes the transition between predominant fetal enzyme form to the predomi nant adult enzyme form. For example, CYF3A7 expression dominates in the fetus whereas hepatic CYF3A4 expression dominates in the adult. 42 Differences in pharmacokinetics between newborn children and adults are presented by Soldin and Steele,43 Soldin and Soldin 44 and Alcorn and McNamara. 45 There are also important pharmacokinetic differences between children and adults that influence drug toxicity: (1) the fat compartment of children is generally smaller, providing less storage for lipid-soluble drugs; (2) the propoliion of unbound drugs in the blood is higher in children, indicating that more drugs are available to pro duce a pharmacological effect; and (3) the ratio of liver weight to total body weight in children is 50 per cent higher at 2 years old and 30 per cent higher at 6 years old than for the adult. Drug biotransformation rates are therefore enhanced in children and reduced in neonates compared with adults. In addition to pharmacokinetics differences, pharmacodynamics responses also differ. 46 A final point concerns the marketing of therapeutic sub stances. Only a small minority of approved drugs have received adequate paediatric study in appropriate clinical trialsY Furthermore, a licensed formulation for the admin istration of certain pharmaceutical products to children is often also lacking.48 The situation could be even worse with off-label/unlicensed drugs. Indeed, many cases of adverse events involving off-label/unlicensed prescriptions have been reported. 49 To fill these gaps, new regulations
Techniques used in drug testing I
and patent protections have been issued for new drugs by the US Food and Drug Administration, requiring specific paediatric studies. Another point that deserves consid eration rel ates to the clinical research involving children and therapeutic drugs. Several studies of drugs that bring only very minor therapeutic benefits to neonates and children are ethically questionable. 50 ,51
TECHNIQUES USED 11\1 DRUG TESTING Analysis of biological specimens is the necessary step that determines whether or not an individual has been intoxi cated with illegal drugs, with excessive amounts of prescrip tion drugs or with any other toxic substances. It is important to remember that toxicologists, especially in paediatric rather than adult cases, often do not know what to look for and therefore have to take into account a vast number of toxicologically relevant substances, which makes the task difficult but very challenging. 52.10
Drug Screening The technology of drug screening has greatly improved throughout the last two deca des. Idea lly, a screening test for a drug should be rapid (a short turnaround time). highly specific (without false-positive results), highly sensitive (allowing the detection of a very low concentration of the drug in the specimen), reliable (same result from one lab o ratory to another for a given specimen), easily performed technically and inexpensive. Since none of th e available analytical technology meets all of these requirements per fectly, drug screening is usually achieved by means of the combination of two different techniques; this is done in order either to extend the ambit of the screening test to a larger group of substances not detectable by the other method or to confirm the result from another ana lysis. In practice, screening tests are design ed for max imum sensi tivity at the expense of selectivity. Confirmatory analyses are often just as sensitive (if not more so) but additionally provide the requisit e specificity. It is important, however, to point out that screening procedures must not give false negative results since this inva lidates the testing process, False-positives at this stage are acceptable because all pre sumptive samples wiU subsequently be reanalysed by a confirmatory technique. 52 The methodologies most frequently employed for drug screening are immuno assays and chromatography, and sometimes capillary electrophoresis. Serum (or blood). urin e and oral fluid (OF) (sa liva) can be analysed by these techniques. The pharmacokinetics of distribution and eJ im inati on of dru gs are such that most are detectable in urine for a much longer time tha n in blood or OF so that urine is usually the specimen of choice for screening.
259
IMMUNOASSAY TECHNIQUES Immunoassays are very useful to filter out the negative samples from the positives, a nd thereby reduce the amount of further analytical work required for toxicological inves tigation. 53 Immunoassay techniques are now the most widely used screening pro ce dures in drug testing. Advan tages include high sensitivity, no extraction steps and applicability to a large number of samples. Popular ver sions of these tests are represented by radioimmunoassay (RIA). enzyme-multiplied immunoassay technique (EMIT), enzyme-linked immunosorbent assay (ELISA) and fluores cent polarization immunoassay (FPIA),s4 These techniques use antibodies specific to the drug being assayed and a labelled form of the same drug, The label itself may be a radio active isotope (RIA). an active enzyme (EMIT and ELISA) or a fluorescent label (FPIA), which is incorporated synthetically. A fixed quantity of antibody and labelled drug are added to the test sample. The binding sites on the antibody attract both the labelled drug type and the unla belled drug type in the sample. The amount of labelled drug bound is inversely proportional to the number of unla belled drug molecules present. 53,55 In RIA, a rad ioactively labelled drug competes fo r the same antibody binding site as the unlabelled drug. The ana lytical measurement of radioactivity remaining in the solu tion determines the amount of unlabelled drug in the sample. RIA kits were ava ilable for several classes of drugs of abuse but now have been virtually abandoned, being replaced by non-radioactive immunoass ays. For instance, some larger companies stopped their marketing of most RIA kits during the period 2004-2006. In EMIT, ELISA and FPIA testing, the analytical measure ment is based on an optically detected change, such as ultra violet (UV) absorption, fluorescence or luminescence. These systems avoid the use of radioactive isotopes but also have reduced sensitivity compared with RlAs because the optical signal is measured in the presence of the original biological fluid. An EMIT kit is available for opiates, barbiturates, cocaine, amphetamines, benzodiazepines, methaqualone, methadone, phencyclidine, cannabinoids and lysergic acid diethylamide (LSD) . The advantages of EMIT as a screening technique include its rap idity, its semiquantitation and its ability to detect a number of drugs. Disadvantages include the high cost of the reagents and its susceptibility to both false-positives and false-negatives so that confirmation by an alternative methodology is recommended for positive and even for negative results. 56 Microplate ELISA tests are also competitive immunoassays. They are speCifically optimized for use with a number of different biological matrices. Kits exist for about 20 pharmaceuticals and 'drugs of abuse'.57 An FPIA is both sensitive and specific; however, the equipment is expensive and susceptible to sample alteration, affecting its accuracy. The application of the FPIA technique to foren sic blood work and to alternative specimens such as hair or bile has been described .58 Additional antibody kits are also
260 I
Recent advances in paediatric toxicology
available, expanding the uses of these systems. 55 New immunoassays are available to test the OF from drivers at the roadside and workers at the workplace. For instance, the Cozaft® drug detection system (DDS) test is used to detect impaired drivers suspected of driving under the influence of psychoactive substances. DrugWipe® 5+, developed by Securetec, is another test evaluated for the detection of psy choactive drugs in the oral fluid of suspected impaired driv ers. Improvements are still needed for the similar detection of cannabinoids and benzodiazepines.
SAMPLE PREPARATION Sample preparation is an important step, which still remains the most tedious and time-consuming part of an analytical method. Optimum sample preparation leads to enhanced sensitivity and selectivity while reducing the amounts of interfering matrix components and increasing analyte concentration. Sample preparation represents a major part of analysis that is capable of taking up to 80 per cent of the total time of a complete separation-based ana lytical process, which typically encompasses five steps: sampling, sample preparation, separation, detection and data analysis. Major techniques rely on liquid-liquid, solid-phase and headspace extraction. Recent advances have been made in miniaturization, high throughput, selec tivity, integration and hyphenation and automation of sample preparation. Liquid-liquid extraction (LLE) transfers the target ana Iytes from a liquid matrix into another immiscible liquid phase according to solubility difference. Solid-phase extraction (SPE) is used to extract and concentrate analytes from a liquid matrix by partitioning the compounds between a solid and a liquid phase. An SPE procedure con sists of four consecutive steps: column conditioning, sam ple loading, column washing, and elution of selected analytes. The headspace technique is dedicated to the extraction of volatile compounds. The analyte is parti tioned between the sample and the gas phase in a closed system. Then, the vial is pressurized and the headspace is sampled and injected into the gas chromatograph. In recent years, selective extraction methods with molecularly imprinted polymers (MIPs) and affinity columns with bound antibodies have also been considered. Template molecules are used to produce MIPs by creating cavities in a polymer that will recognize the target molecule. 59 The MIPs somehow mimic natural antibodies. Miniaturiza tion of the SPE methods resulted in the development of solid-phase microextraction (SPME). Other materials cur rently under investigation are restricted-access materials (RAMs).5o The RAMs are used to exclude large molecules, such as proteins, and to extract low-molecular-mass ana Iytes by use of hydrophobic, ionic or affinity interactions, typically from complex blood matrix. They allow the direct injection of plasma or serum samples on to a high performance liquid chromatography (HPLC) column without
rapid clogging of the column, facilitat-ing sample prepara tion and analysis coupling and automation. 51 .52
HIGH-PERFORMANCE LI QUID CHROMATO GRAPHY The separation of non-volatile substances from each other or from other components of an extraction residue can be obtained using HPLC. When a mixture of substances is injected onto the column, each component is partitioned between the stationary phase (column) and the liquid (mobile) phase. Molecules with greater affinity for the col umn spend more time in that phase and, therefore, take longer to reach the detector. The time taken from injection to the peak maximum is known as the retention time. The detector responds in direct proportion to the concentration of material passing through it, hence peak heights and area shown on the chromatogram are directly related to the concentration of each analyte. Good sensitivity and high specificity, depending upon the detection system used, are shown by HPLC. The most common detectors utilize absorption of UV light by the drug. A diode array detector (DAD) will allow a full UV spectrum of the analyte to be obtained. This can be com pared to a standard spectrum of the drug, and identifica tion is then based on both retention time and UV spectrum. The identification of a very broad number of unknown substances in blood specimens can be obtained using HPLC-DAD in combination with a suitable spec trum. 53 .54 High selectivity and sensitivity for molecules with characteristic excitation and emission fluorescence spectra is shown by HPLC coupled with fluorimetric detec tion. In this way, LSD and its metabolite nor-LSD can be quantified in blood by HPLC fluorescence detection down to 20 pg/ml. 55 Pre- and post-column reactions for increased sensitivity have also been described. Owing to its qualities, HPLC can be used to screen a sample for many drugs of abuse. The drawbacks of HPLC include its expense, the high degree of expertise necessary for operation of the equipment, the need for sample extraction and the need for preparation of separate specific columns. 55 New filling column materials have been marketed over the last few years. For example, polar metabolites, such as ecgonine or hydroxyl metabo lites of cocaine, can be analysed without prior derivatiza tion through hydrophilic interaction chromatography (HILlC).57 Similarly, the polar metabolites of morphine (morphine 3- and 6-glucuronide) can be readily analysed on reversed-phase columns, which are compatible with almost 100 per cent water eluent. 58 More recent develop ments in HPLC concern the use of sub-2 p.m particles and mobile phases delivered at high flow rates. The recent com mercialization of porous hybrid organic-inorganic silicon based paliicles with a narrow size distribution in the range of 1.7 pm has enabled a new level of performance, but only through the use of newly developed pumps that permit pumping and injection of liquids at pressures in excess of
Techniqu es used in drug testing I
10 000 pSi. Full implementation of such technology, termed ul tra performance liquid chromatography (UPLC), is further challenged by the requirement for new detectors fast enough to record a minimum number of data-points in a very narrow chromatographic peak (width of a few seconds only).69.70 The use of UPLC columns speeds up analyti cal runs to times of as low as a few minutes for complex mix t ures. Another way to reduce the overall analysis time is to use monolithic supports, which co nsist of a continuous porous silica rod. Separations with monolithic columns can be performed at high flow-rat es, reducing the analysis time to less than 10 min.7!
GAS CHROMATOGRAPHY Gas chromatography (GC) is another method of separating substances of analytical interest. 66 It is one of the most efficient techniques available for separating drugs in body fluids. The separation is carried out on an analytical co l umn containing a station ary phase (liquid or solid, depend ing on temperature), which is maintained at a given temperature inside an oven. The whole GC system com prises six components: gas supply and flow controllers, injector, oven, column, detector and recording device. In drug testing, GC capillary (rather than packed) co lumns are co mmonly used . A com pound is identified by matching its retention time with that of a drug standard under the same cond itions. The use of more sens itive and selective detec tors than flame ionizat ion detectors, such as electron cap ture and nitrogen-phosphorus detectors, has greatly improved the reliabili ty of Gc. The method of choice for volatile compounds a nalysis is GC, which can be easily coupled with mass spectrometlY detectors. Unfortunately, many drugs and poisons conta in polar functional groups and require chemica l derivatization to improve thermal stability, volatility and detection. 72 Other dra wbacks include the time necessary to prepare the specimen (extraction, purification and derivatization), the expense of the equip ment and the requisite expertise of GC technicians. There fore, GC is usu ally reserved for use as a screening and confirmatory technique in the expert laboratory setting. Fast GC is a new development in GC technology. Fast GC allows rapid analyses and reduction of retention time (up to X 10) while main taini ng acceptable analyte reso lution. This is made possible by reduction of column bore size, down to 100 !Lm, increasing the oven temperature ramp up to 120· C/m in, and by using high-pressure ca rrier gas control. Because peaks a re very narrow, fast detectors and data sam pling are also required. Applications include analysis of drugs of abuse in urine, 30 different drugs in oral fluid,73 benzodiazepines in blood 74 and cannabinoids in blood. 75
CAPILLARY ELECTROPHORESIS During the past decade, capillary electrophoresis (CE) emerged as a promising, effective and economic approach
261
for sepa ration of a large variety of substances, including those encountered in forensi c toxicology.76 Capillary elec trophoresis comprises a family of electrokinetic separation tech niques that separate co mpounds based upon differ ences in electrophoretic mobili ty, phase partitioning, iso electric point, molecular size or a combination of one or severa l of these propeliies. Electroki netic capillary methods employed for forensi c toxicology include capillary zone electrophoresis (CZE) and an electrokinetic capillary chro matography technique called micellar electrokinetic capil lary chromatography (MECC). In CZE, a few nanolitres of a sample are applied to the beginning of a fused-silica capil lary filled with buffer. On app lication of a high-voltage direct current field, charged solutes begin to separate and are swept through the capi llary by the combined action of electrophoresis and electro-os motic bulk flow and are on-column detected towards the capillary end . In MECC, the buffer contains charged micelles (dodecy l sulphate micelles), and uncharged and charged solutes separate on the basis of differential partitioning between the micelles and the surrounding buffer and, if charged, also by differ ential charge effects, including electrophoresis. In both techniques, the most commo n detection principles app lied are on-column absorbance and fluorescence. Using on-column multiwave len gth detection, this technology is well suited for toxicological drug scree ning. Compared with HPLC and GC, CE has distinct advantages, including automation , small sample size, minimal sample preparation , use of very small amounts of organic solvents and low cost of capillary columns. 77 ,78 Recent developments concern the co upling of CE with tandem mass spectrometry and the analysis of biofluids focused on less analysed matri ces, such as amni otic fluid, saliva, cerebrospinal fluid, sweat or airway sur face fluid and sputum. 79
Drug Confirmation DlUg confirmatio n is based upon the use of two different techniques, each confirming the other's results. In this resp ect, the combination of two of the previously men tioned techniques adds confidence in the accuracy of the resul t. 55,80 On the other hand, the introduction of the mass spec trometry detection method has revolutionized the analyti ca l tox ico logy in providing the capac ity of a formal identification of unknown compounds. Gas chromatography mass spectrometry (GC-MS), combining GC with mass spectrometry, has become the so - ca ll ed 'gold standard ' of forensic drug testing and doping ana lysis. 8o - 82 In the elec tron impact ionization mode, the sepa rated compounds are bombarded with high-energy electrons, causing them to break apart. The fragments produced are separated on the basis of their mass-charge ratio. Under the same condi tions, a molecule will fragment in exactly the same way evelY time, producing the sa me spectrum of fragm ents.
262 I
Recent advances in paediatric toxico logy
This spectrum, along with the chromatographic retention times, constitutes a virtual 'fingerprint' that is very specific for the particular chemical compound of interest. Detection sensitivity may be as low as to measure the volume of the drug in nanograms per millilitre. The definitive method for drug identification is considered to be GC-MS. Chemical ionization (CI), is a softer ionization method that utilizes a charged reagent gas. Under these conditions of ionization, fewer fragments are formed, increa sing the chance of detecting the pseudomolecular ion. High-tlu'oughput procedures mean that thousands of relevant toxic sub stances can be screened in a single procedure, i.e. the so called systematic toxicological analysis (STA). Urine still remains the standard specimen for compre hensive screening, especially when a general screen for unknown substa nces is required. To detect the largest num ber of xenobiotics , conjugated metabolites are first hydrol ysed and then derivatized prior to GC-MS analysis. Other confirmatory techniques include HPLC coupled with mass spectrometry (LC-MS); this constitutes a straight forward method for the analyses of polar or thermolabile compounds without a derivatization step as required in GC-MS. Gas chromatography or HPLC coupled with tan dem mass spectrometers (GC-MS-MS and LC-MS-MS) enhance the selectivity and the sensitivity by a specific sec ondary fragmentation of selected ions provided by the first fragmentation. All these sophisticated techniques, includ ing GC-MS, can offer valuable information for the identi fication of unknown compounds. Using these hyphenated chromatography and mass spectrometry methods also allows acute or chronic poisonin g to be definitively excluded, which might be sometimes as important as the detection of a toxic substance. Time-of-flight mass spectrometers (qTOF-MS) accu rately determine the mass-charge ratios of ions by measur ing flight time after acceleration in a vacuum tube by a high voltage. qTOF-MS highly improves mass resolution and accuracy allowing exact mass determination. Accurate mass measurement can be used for qualitative identifica tion of unknown molecules. Pavlic et a18 ) used ES1-qTOF MS with mass spectral libraries to identify over 300 different drugs. New and very expensive MS tools which are emerging in toxicology laboratories are Fourier trans form ion cyclotron resonance mass spectrometers (FTMS) and OrbitrapT". These are now considered as the ultimate instruments for high mass resolution a nd accuracy84
determined, namely the calibration model, the accuracy and precision of the method, the low er limit of quantifica tion and the limit of detection. 8s .B6 Participation in external quality control programmes allows laboratOlY performance to be evaluatedY
SPECIAL TECHNIQUES FOR ANALYSIS OF VOLATILE SUBSTANCES Epidemiological studies indicate that a substa ntial number of children and teenagers worldwide experiment with or abuse volatile substances with the intention of experienc ing a euphoric state of consciousness. A large range of vo latile compounds are toxic, especially for young people. They can be classified in three broad classes: the hyd rocar bons (e.g. butane), oxygenated compounds (e.g. butanol, amylnitrite) and halogenated compounds (e.g. halothane, trichloroethylene). In case of suspicion of poisoning with volatile substances, proper sample collection, storage and handling are critical points to guarantee accurate toxico logical results. Samples should be taken rapidly and stored in gas-tight, welI~sea led containers with minimal head space. Storage, transport and handling of t he sample should always occur at low temperature (- 5°C up to 4°C). The headspace technique is the most appropriate and popular method for the extraction of volatile substances in body fluids and tissues. Solid-phase microextraction, cryo genic oven trapping, cryogenic focusing, and purge-and trap extractio n techniques are typical headspace methods. Usually, electron capture detection (ECD), flame ioniza tion detection (FlO) and mass spectrometry (MS) are used for detection and quantification purposes, while MS and Fourier transform infrared (FTIR) spectroscopy are used for identification. 88
ALTERNATIVE SPECIMENS FOR DRUG TESTING The rapid growth and development of drug testing technol ogy has created a number of testing methodologies . that can assess a broad range of biological specimens, including besides urine and blood, hair, oral fluid, sweat, meconium and amniotic fluid. 89
Urine Method Validation Correct interpretation of toxicological findings is only pos sible provided reliable analytical data a re available. There fore, new analytical methods to be used in clinical and forensic toxicology require careful method development and thorough validation. For method validation, a whole bunch of analytical parameters must be considered and
For many reasons, urine is the most common matrix analysed for drug testing. The specimen is easy to collect and is not considered as an invasive sample to request. Large volumes can often be collected , allowing extensive screening and storage for any further analyses by other laboratories if additional expertise is needed. The matrix is one of the simplest that can be used, and it is the easiest to analyse compared with other fluids or tissues. Drugs and
Alternative specimens for drug testing I
metabolites are usually stabl e in frozen urine, allowing long-term storage of positive samples. Drugs are usually metabolized by the liver to form polar metabolites that are conjugated before their elimination from the body via the urine. The metabolism of parent drugs (which are chemically non-polar) is relatively rapid compared with their urinary excretion. Conversely, the uri nary excretion of most metabolites (which usua lly are chemically polar) is more rapid and extensive tha n with the parent drugs. Drug metabolites are therefore frequently found in urine in concentrations mu ch greater than the parent drugs. Therefore, many immuno assay urine screen ing assays for drugs of abuse utilize antibodies to drug metabolites. Urine, however, can be easily adulterated since observed specim en collection is not a common practice. 90 ,91 Addi tionally, most drugs are present in urine for only a few days after consumption so that only recent drug use will be detected. The determination of drug-to-creatinine ratios takes account of the potential dilution of urine. Dilution can result from deliberate adulteration or through fluctuations in fluid intake and elimination. Ulinary excretion profiles are nor malized and smoothed when reported to the creatinine con centration. New intake of drugs can be better predicted by comparing creatine-normalized drug concentrations meas ured in two successive Ulinary voids 92
Blood Blood is co nsidered to be an invasi ve sample. It represents a complex matrix particularly if haemolysed. The detec tion window of the majority of drugs and metabolites to be tested is limited from a few minutes to several hours at most, and much less frequently days. However, the toxic effects or the pharmacological activities of drugs are gen erally correlated with their blood concentrations, which allows the effects of these levels on the living person to be predicted, a nd thus the extent of involvement of the drug in the death to be understood. 9J To assess the significance of drug levels measured in blood, the concentrations are compared with ranges of therapeutic, toxic and fatal levels found in large databases. 94 ,95 These databases compile the results of toxicological investigations of published case reports. They are mainly based on tox icology data obtained from adult patients or fatalities. It is assumed that these compilations can be used to interpret drug levels in blood specimens taken from young adults and teenagers, However, extrapolation to the fetus, neonates and infants is probably more hazardous and prone to error. In the developing chi ld there are rapid changes occurring in terms of organ maturation, changes in body composition and the ontogeny of drug elimination,96 Genotype and the temporal acquisition of drug bio transformation are criti cal determinants of a drug response
263
in infants. For instance, if the O-demethylation of dextro methorphan (reflecting CYF2D6 activity) appears to be well developed by 2 weeks post-natal age, the N-demethylation (reflecting CYFJA4 and possibly also CYP2B6 activities) increases more slowly over the first year of life. 97 All of these changes have a profound effect on the phar macokinetics of drugs across the paedi atric age range. Because of the non-linear nature of these changes, simple allometric scaling methods based on body weight or body surface area often fail in the prediction of drug dosage, especially in neonates and infants.98 Several sca lin g mod els in predicting maintenance doses for children from those used in adults have been suggested. 99 Altered drug metabolism in children may result in adverse drug reactions because of inappropriate high-dose adminis tration. Idiosyncratic adverse drug reactions may also occur. For instance, hepatotoxicity after valproate administration is much more frequent in children. The metabolism of va lproi c acid in children is increased, resulting in the formation of a higher amo unt of a toxic metabolite. In some cases, children are more resistant to drug toxicity than adults. Compared with adults, children appear to be more resistant to the hepa totoxic effects of paracetamol (acetaminophen) overdose. Sul phation, whic h is more active in children, may diminish the formation of toxic metabolites by P450 enzymes. 2J Hence, giving a dose extrapolated from adult dosage may result in unexpected blood levels in paediabic patients. Therapeutic ranges for so me drugs are quite different in adults and chil dren, e.g. therapeutic theophylline concentrations are lower for neonates (5-15 mg/L) than for adults (10- 20 mg/L). 100 Protein binding can be quite different in neonates, producing vely different therapeutic ranges for highly bound drugs such as phenytoin. The blood (pl asma) drug metabolite concentration (or the ratio of parent drug and metabolite) can also be used to dif ferentiate between acute and chronic drug administration. In the case of acute drug administration, the blood (plasma) metabolite concentration in relation to parent drug concen tration can be helpful in determining the approximate length of time between drug administration and death, or specimen collection at autopsy or in a living person. Neonates may also show different ratios of the pa rent drug and metabolite, the hydroxyl metabolite of phenobarbitone and its glucuronide conj ugate being present in much higher concentrations in neonates than in older subjects. 100
Hair Detection of dru gs and their metabolites in hair has gained much attention over the past decade. Drugs incorporated in hair remain in the keratin matrix for a long time, thus open ing a much wider window of detection than there is for drugs in urine. 101-104 Reported drug recovery from the hair of ancient Peruvian mummies suggests that hair analysis may provide inform ation about drug use or exposure
264 I
Recent advances in paediatric toxicology
almost indefinitely under the correct sample storage conditions. l05 However, it has also been reported that a small amount of the entrapped substances may slowly be hydrolysed spontaneously. lOG In addition to the promising diagnostic power of hair analysis, hair sample collection is easy to perform without the embarrassment commonly associated with urine collec tion. Hair samples can be stored or transported without refrigeration, pH control or the preserving agents that are normally needed for other biological samples, such as blood and urine. Hair may also serve as a 'diary' of exposure, producing a drug history. 107 Scalp hair, in the average person, gener ally grows approximately 1 cm per month ; hair that is 3 cm from the scalp would have been formed by the hair-forming cells in the hair follicle approximately 3 months earlier. Therefore, it is theoretically probable that a drug incorpo rated into the hair via the hair-forming cells in the hair follicle would appear 3 cm from the scalp 3 months after consumption. Unfortunately, the interpretation of drug pres ence in hair is not that straightforward. Not all drugs found in the hair may enter through the hair follicle. Drugs may also enter the hair shaft via blood, sweat or sebum, or by diffusion from the skin surrounding the hair. 108 Hair pigmentation 109 and gender differences no also appear to influence the concentration of drugs measured in hair. Several investigations suggest that basic drugs bind to melanin, which explains the higher concentrations gener ally found in dark hair. On the other hand, levels of acidic drugs and metabolites, such as ll-nor-9-carboxy-delta (9)-tetrahydrocannabinol, are unaffected by hair cOIOUr. III ,112 Our understanding about hair as a distribution site for drugs is still limited, and much research is necessary before drug concentrations in hair can be interpreted with accu racy.IJJ Although there is an increasing body of literature being developed about the detection of drugs of abuse in hair, there remain substantial and significant questions about the interpretation of the analytical data. Hair is somewhat unique among tissues in that substances that are incorporated into hair may not have subsequent access to the blood, and are thus retained long after the parent drugs and metabolites have been elimi nated from the body. Although hair (particularly hair bulb) does contain cytochrome P450 enzymes and has the ability to metabolize drugs, it is considerably less able to do so than most other organs. 107 Substances may become permanently bound in hair in the form in which they are distributed into the hair follicle. Cocaine appears to be an exception in that its hydrolysis to benzoylecgonine is thought to occur in the hair shaft. 106 The blood supply to the hair follicle is extremely rich, and the hair-forming cells of the hair follicle are among the most rapidly dividing cells in the body. Drugs in the blood circula tion will be delivered rapidly to the hair follicle and, if they are capable of crossing the cell membrane, will enter the hair forming cells. Parent drug concentrations in hair generally
exceed the concentrations of metabolites even when both the parent drugs and their metabolites are no longer detectable in plasma, blood or urine. 114 ,115 However, it is also possible to detect the metabolites of drugs of abuse in hair, which may provide proof of drug ingestion. II G The finding of parent drugs in the hair may prove to be an advantage in the use of hair testing by forensic toxicol ogists. For example, hair may become the matrix of choice when differentiation between heroin use and prescription use of morphine or codeine is required. IOI , IOJ The finding of greater concentrations of parent drugs than metabolites in hair may result from several mecha nisms. 1l 7 Parent drugs are generally less chemically polar, more lipophilic and better able to cross cell membranes than their metabolites. If substances enter the hair through the hair-forming cells in the hair follicle, then the process would favour the incorporation of parent drugs. Increased lipophilicity may also favour parent drug secretion into sweat and sebum, where it could come into contact with the hair shaft distal to the hair-forming cells. In this respect, Henderson et al 118 administered deuterium-labelled cocaine to known cocaine users and found that in some subjects the drug moved in a distinct band while in others the labelled cocaine appeared first at a site more distal to the scalp. These data suggest that in some individuals cocaine may be secreted rapidly into the sweat or sebum, and then deposited on the hair. It is also possible that the components to which drugs bind in hair favour the binding of parent drugs over metabolites , as has been shown for codeine, which binds to hair in vitro with a greater affinity than morphine. Besides the polarity of drugs, it has been suggested that other chemical characteristics should play an important role for the incorporation of drugs into hair. In this respect, Nakahara et al 11 9 and Nakahara and Kikura 120 have shown that structural changes that increase the lipophilicity of a compound increase its incorporation into hair (for exam ple, an increase in the length of the carbon chain on the nitrogen of amphetamine), whereas structural changes that decrease a drug's basicity will reduce its incorporation. On the other hand, subtle structural changes that influence a drug's lipophilicity and ionization to a cation or anion at physiological pH greatly affect the incorporation of the parent drug and its metabolite into hair. At a given plasma concentration, drugs that form cations at physiological pH appear to be incorporated into hair at greater concentra tions than drugs that form anions. The reason for this observation is not known, but evidence suggests that it may be a result of the binding with melanin (see below). In addition to the impact of changes in the chemical structure of drugs on their incorporation into hair, there is also the possibility that the structure of the hair (fine ver sus coarse) or chemical treatments that affect hair structure are important for the incorporation of drugs. Blank and Kidwell 12 1 demonstrated that the absorption of cocaine into cut hair specimens was greater for thick hair than for fine
Alternative specimens for drug testing I
hair. The hair specimens used in these experim ents were black (thick hair) and brown (fine hair), and it is possible that the difference in pigmentation also influenced the absorption of cocaine. Cirimele et aj122 measured drug con centrations in the hair of a female drug addict who had brown hair with strands of hair bleached with hydrogen peroxide. They found approximately threefold greater con centrations of cocaine and codeine in the brown hair than in the bleached hair from the same person. The role of pigmentation in influencing hair concentra t ions has been clearly demonstrated for several drugs such as cocaine, codeine, methadon e or nicotine. For example, Joseph et al 123 showed that the specific binding of cocaine is 157 times greater in female African black hair than in female Caucasian blonde hair. Furthermore, these authors showed that bleaching black hair decreased the specific binding of cocaine J3-fold. Nakahara et al l1 9 fo und a good correlation between the in vitro melanin affinity and the incorporation ratio for many drugs of abuse. Hold et al 124 showed that stanazolol, an anabolic steroid, is incorporated to a greater extent in pigmented than in non-pigmented hair. If these data clearly show that pigmentation is an important factor in the incorporation of drugs into hair, it must be pointed out that drugs are also incorporated into the hair of albino animals who have non-pigmented hair owing to the absence of ty ros inase. Therefore, one cannot exclude the existence of a racial bias that could contribute to the difference in incorporatin g drugs into hair other than by the degree of pigmen tation of hair. 1l3 Proced ures used in hair testing include a preliminary phase fo r specimen collection and storage, a sample prepa ration phase for decontamination and drug isolation from hair structu re, and a drug identification and quantification phase by instrumental analysis. lOG ,125 Nails can be also used as an alternative to hair analysis for long -term drug detection . It has been established that drugs are incorporated into growing nails at levels similar to those 0 f hair. 126
SPECIMEN COLLECTION Hair samp les are generally collected by cutting, as near as possible to the scalp, in the posterior vertex zone. In some cases, different types of body hair, such as pubic, axillary and beard hai r, can be used. 127 - 129 The amount of hair needed for toxicological analysis is about 30-50 mg. If a segmental hair analysis is required to evaluate drug use history, th e hair sample is generally cut into pieces of }-1.5 cm in length, which represent about 1 month's growth, Hair samples are to be stored at room t emperature in paper, plastic envelopes or in plastic or glass tubes.
SAMPLE PREPARATION Hair specimens are normally cut or pulverized to produce small fragments yielding more surface contact with the
265
extraction medium. However, prior to the extraction step, removing external contamination is mand atory in order to avoid misinterpretin g positive results. The most frequent contaminating factors include passive exposure to dust, aerosol, smoke (when other people smoke drugs), drug powders (by handlin g drug stuff or any contaminated material), and sweat and sebum, which carry drugs actively consumed by the subject. Many decontamination proce dures have been proposed, but all involve washing steps using solvents, detergents and aqueous media, such as buffer solutions, diluted acid solutions or distilled water alone. 125 The other fundam ental preparation phase is the isolation of the drugs from the keratin matri x. In contrast to body fluids, in which drugs and metabolites are dissolved and can be directly extracted and analysed, xenob iotics in hair are, in fact, entrapped into th e solid keratin, Therefore, keratin matrix must be digested, extracted, hydrolysed or dissolved before drug separation and identification, Factors affecting the choice of the most suitable procedure include the chem ical structure of the drug, the system in use for particular dnlg detection, analytical recovery and time of analysis.
DRUG IDENTIFICATI ON Methodologies applied to drug detection and quantifica tio n are simi lar to those usually carried out in laboratories of ana lytical toxicology.130 lmmunochemical and chromatographic methods coupled with mass spectrometry or tandem mass spectrometry to enhance sensitivity and specificity are com monly adopted. 126 ,131 To date, more than 60 pharmaceuticals or drugs of abuse have been reported to be detectable in hair, including, besides the classical drugs of abuse (opiates, cocaine, cannabis, amphetamines, methamphetamine, 3,4 methylenedioxymethamphetamine [MDMA] 3,4-methylene dioxyamphetamine [MDA] and other designer drugs), opioids (i.e. semisynt hetic or synthetic morphine derivatives), hallu cinogens, psychostimulants (including nicotine), barbiturates, benzodiazepines, other sedatives (hypnotics, an tidepressants, neurol eptics), cardiovascular drugs, anti-infectious drugs and other miscellaneo us compounds.67. 131,132
CUT- OFF VALUES Two main threshold values should be mentioned. The first one concerns the analytical thresho ld, which relies on the limits of detection and quantification of the chromato graphic method used for drug analysis. A second cut-off value is used to decide whether an analytical result can be interpret ed with enough reliability with regards to the cir cumstances of use or exposure. This second cut-off value is always higher than the first one. A low cut-off value will be selected to demonstrate a single dru g expos ure while high cut-off values will be chosen in case of repetitive ' abuse to exclude external contamination, Several scientific
266 I
Recent advances in paediatric toxicology
societies have put forward recommendations for cut-off values used for interpretation of hair results {e.g. the SoHT or the Society of Hair Testing) . lOB
Oral Fluid In recent years, OF has attracted much attention as a pos sible alternative to urine for drug testing and, under certain conditions, to plasma in the area of clinical investigation, therapeutic drug monitoring and the assessment of recent drug use. IJJ - 136 Oral fluid is a complex biological matrix consisting of the secretory products of salivary glands (saliva) mixed with other fluids, substances and cellular debris that are present in the oral cavity. In fact, OF has specific advantages over both urine and blood in being readily accessible for sampling. Furthermore , it constitutes a non-invasive sample to collect. Drug concentration-time profiles measured in OF are generally repolied to resemble those determ ined in whole blood, although saliva-plasma ratios may vary considerably depending on the drug prop erties. As such, a positive OF drug test may indicate a recent use, probable intoxication and, in some cases, psy chobehavioural impairment, whereas urine testing cannot. For these reasons, OF drug testing for forensic purposes has been shown to be reliable compared with other means of drug testing. 137 In this respect, it is noticeable that the lit erature on OF testing is expanding at a substantial rate, which is indicative of the increasing interest in this unique biological fluid. 7J ,'38 Testing OF for drugs can provide both qualitative and quan titat ive information about the drug status of an indi vidual. I39 Generally, drug concentrations in saliva are lower than those found ill urine or blood, and the major compound detected is the parent drug, not the metabolites. Initially, oral, intranasal or smoking routes may produce high concentrations of drugs in OF for several hours owing to local contamination of the oral cavity. Thereafter, the concentration of drug in OF is thought to reflect the free fraction of drug in blood. Thus, the m ajor disadvantage of saliva is that, after the contamination phase that follows local administration, many drugs are retained for a shorter period of time than they are in urine. Also, many drugs are weak bases and saliva concentrations may be highly dependent on pH conditions, which itself is dependent on salivary flow. These factors lead to highly variable saliva plasma ratios for many drugs. Although highly sensitive methods of detection a re required, most drugs can be detected in salivary secretions. Therefore, saliva testing is expected to offer many potent ial applications in the general areas of drug screening and forensic investigations. Saliva drug tests can reveal the presence of a pharmacologically active drug in an individ ual at the time of testing. Significant correlations have been found between saliva concentrations of drugs, behavioural
and physiological effects. It can be anticipated that saliva testing for drugs will develop further in the near future, with new domains of application being found as soon as the mecbanisms by which drugs enter the saliva have been c1 ari fied more speci fically. 140 The marketing of new immuno assays (ElAs) specifica lly designed for the detection of drugs in OF has enabled the fast screening and selection of presumably positive samples. Some of these immunoassay kits are commercially available with specific collection devices, e.g. Orasure® microplate, Intercept kits and Cozaft® microplate ElA. Confirmation of presumably positive results is generally performed with LC-MS methods. Atmospheric pressure chemical ionizat ion (APCl) method is the preferred ionization method since it considerably attenuates ion suppression effects due to OF matrix components. 141
Sweat Since 1911, it has been shown that drugs are excreted by the body in sweat, but no one has developed a practical solution to the problem of capturing sweat before testing until recently.142 Occlusive bandages consisting of one to three layers of filter paper or pieces of cotton, gauze or towel were proposed to collect sweat. By using these home made collectors, it was nevertheless possible to identify various drugs including quinine, salicylic acid, antipyrine, ethanol, methadone, phenobarbitone, morphine, cocaine, cannabinoids, methamphetamine and phencyclidine. 142 ,143 More recently, clothes have been shown to retain opiates excreted in sweat, demonstrating that it is possible to obtain evidence about the drug use status of the owners of pieces of c10tbing (e.g. underwear),144 In practice, systematic collection of sweat specimens is difficult bec ause of unequal distribution of sweat glands on different pans of the body.'43 Approximately 50 per cent of the total volume of sweat is produced by the trunk, 25 per cent by the legs and 25 per cent by the head and upper extremities. Sweat is approximately 99 per cent water, tbe most concent rated solute being sodium chloride. Sweat production is irregular and is highly dependent upon an individual's physical activity and emotional state and the ambient temperature and humidity. Furthermore, sweat collection methods used in most studies actually obtain a mixture of sweat and sebum. Drug transport in se bum has not been examined thoroughly.145 Significant advances have been made in recent years to develop a sweat patch technology for the routine collection of sweat samples over an extended period of time. '46 The sweat patch collection device PharmChek T.", marketed by Sudormed, Inc., consists of an adhesive layer on a thin transparent film of surgical dressing to which a rectangular absorbent pad is attached. The sweat patch acts as a speci men container for non-volatile and liquid components of
Pitfalls and limitations of drug screens I
sweat, including drugs of abuse. Non-volatile substances from the environm ent cannot penetrate the transp are nt film, which is a semip ermeable membrane over the pad that allows oxygen, wa ter a nd carbon dio xi de to pass through the patch, leaving the skin underneath healthy. Over a period of several days, sweat saturates the pad and drugs present in sweat are retained. The patch is generally worn over a period of 1-14 days, and drugs and metaboli tes accumul a te over this time period. Patch testing includ es an extraction step followed by a n identification step us ing either immuno assay or GC-MS techniques. 146 The predo minant species fou nd in sweat are parent compounds, not drug metaboli tes. Analyses of duplicate patches in controlled studies have revealed tha t intra-subject variability is low but inter-subject variability is high. 143 Advantages of this type of testing includ e non-invasive sampl e collection (the patch being worn on the back, biceps or chest), ability to carry out normal activities including swimming and showering without removing the patch, and the ability to tell if a patch has been removed and reap plied . 146,147 Thus, w hen applied to drug testing, the method increases the window of drug detection to several days and up to several weeks. Furthermore, the test appears to be very sensitive since the administration of low doses of drugs such as cocaine or opiates produces detectable amounts in sweat after a few days. 147 By offering a cumu lative estimate of drug exposure over a period of several days, this technology is particularly sui table in t he trea t ment and monitoring of substance abusers,148 when it is importan t for medical personnel to obtain information on the behav iour and activities of these patients. Patches can be worn continuously and constitute a record of drug intake during that period .
Meconium and Amniotic Fluid The determination of drugs of abuse in meconium for the purpose of determining maternal drug use during preg nancy is becoming increasing ly popular owing to its ability to provide a larger historical record of drug exposure.149-154 Meconium is a dark green mass of water, cells, mucus, sterol precursors a nd bile pigments which is formed in the fetal gut from swallowed amniotic fluid and sloughed gastroin testinal epithelial cells from 16 weeks to birth; it is dis charged 1, 2 or J days after birth. 155-157 Drugs and their metabolites a re passed into the amniotic fluid from fetal urine, and meconium is though t to reflect the swa llo wed a nd/o r sloughed sta ble metabolites. Analy sis of meconium may therefore reflect in utero exposure to drugs. Therefore, the window of drug exposure that can be detected by meconium testing is about 20 weeks , compared to a 2- to J - day window of de tec tion for a urin e drug abuse screen. 158
267
Since meconium is often available in only sma ll amounts, screening methods require the development of sensitive a na lytical procedures and cl ea n efficient extraction tec hniques for the characterization of drugs in this particular matrix. 152 In this respect, screening and confirmatory procedures by immunoassays and GC-MS respectively have been reported for cocaine and its metabolites, 150, 151,158 cannab is,156 opiates, including methado ne, 159 phencyciidine,157 cotinine, 160 oxy codone l61 a nd benzodiazepines. 162 Generally, the routine anal ysis of meconium fo r drugs of abuse is recommended for cases in which urin e cannot be obtained from the newborn or urinalysis is nega tive for the substances despite a strong suspicion of matern al use of the substances during pregnancy. 149 Liquid chromatography tandem mass spectromehy methods are also ava ilable for the comprehensive analysis of xenobiotics in meconium. 163 Likewise, a similar method has been presen ted for the analysis of 10 amphetamine- , methamphetamine- and MDMA (ecstasy)-related an alytes in human meco nium. 164 Even at low intake, ethanol can cause adverse effects in newborns and later problems in childhood. Fatty acid ethyl esters (FAEEs) are fo rmed in the body by esterificatio n of ethanol with free fatty acids and transesterification of glycerides. They have been detected in human tissues dam aged by ethanol abuse, and also in blood a nd hair. They can be used as ma rkers for ethanol use. In this respect, FAEEs in meconium have bee n reported to be potential bio markers of feta l exposure to alcohol resulting from mater nal alcohol consumption during pregnancy.16 5 According to Moore et al,166 a total FAEE concentration greater than 10 000 ng/g of meconium may indicate that the newborn has been ex posed to significant a mounts of alcohol during pregnancy. Amniotic fl uid, umbilical cord t issue and cord blood have also been suggested as alte rnative samples. 167-169 Amniotic fluid is present throug hou t gestation and is con stantly diluted owing to fetal urin ation . Fetal swa ll owing of a mniotic fluid contributes also to the recircul ation of drug metabolites through the fetu s. Many drugs have been de tected in amniotic fluid. 17o The major disadvan tage of amniotic fluid testing is the difficulty and invasiveness of its collection.
PITFALLS AND LIMITATIONS OF DRUG SCREENS The rapid growth and developmen t of drug testing technol ogy has created a number of testing methodologies that can assess a range of biological s pecimens, not only to pro vide evidence of recent drug use but also to indicate, as far as it may be possibl e, the pattern of drug use (i.e. rou te, fre quency, dose and time of last use), the degree of impa ir ment or the exten t of drug dependence of the individu a l. Many of these specimens, especially in paediatric forensic
26 8 I
Recent advances in paediatric toxicology
medicine a nd pathology, must be used with s pecial atten tion bec ause of the potential for false-positive results owing to the ir contamination during passive env ironmen tal exposure. For the clinician or the pathologist, th e wide variation in the accuracy of drug tests, coupled with t he confusing rubric 'toxic or drug screen', often leads to considerable confusion abo ut what, in fact, the test used is able to detect. All toxic screen methods have in common, ho wever, a basic desig n th at max imizes sensitiv ity while making compromises in specificity. Immuno assays ge nerally are less se nsitive than GC-MS or LC-MS, a lthough all are capable of detecting very low quan tities of drugs. The duratio n of a test positivity depends not only, in part, on the sensitivity of the test, but also on the circumstances of abuse. Multiple doses of a drug taken chronically may redistribute it to deep body compartments w ith slow release back into the blood comp artment, s ig nificantly prolonging the pharmacokinet ics of elimination. This is particularly true for cannabi noids, methaqualone or phencycl idine. Therefore, issues in dru g pharmacokinetics and test ing methodologies can easily lea d to both false-positive and false-neg ative tests, both with significant consequences. Addition all y, forensic pathologists must bear in mind that after death changes may occur due to post-mortem dru g redistribu tion or degradation. The main pitfalls in forensic toxi co logy and interpreta tion of ethanol analysis in post-mortem specimens as well as ana lytical pitfalls in hair have been reviewed by Rich ardso n, 40 Kugelberg and Jones,I71 and Musshoff and Madea.172
Causes of False-positive Drug Screens The two major causes of false - positive drug results concern immunoassay tests and contaminations by envi ronmenta l expos ure to drugs. All immuno assay tests are susceptible to fa lse-pos itive results owing to a ntibody cross-reactivity wi th substances bearing some structural simil arity to the drugs to be detected . 173.174 Included among these are poppy seeds in t he diet (bread), which may conta in opium con ge ners resulting in a drug screen that is pos itive for opiates. Also com mon is the abi li ty of nasal decongestants , such as ephedrine or phenylprop ano lam ine, to produce a uri ne drug screen that is positive for amphetamin es. Newer immunoassays for amph etam ines have begun to use mono clonal antibody assay that theoretically has a high er speci ficity for the amphetamine core. Even with the monoclonal assay, however, cross-reactivity may still occur; phen metrazine and I-ephedrine may produce a posit ive result. Dextromethorphan and diphenhydramine, commonly used over-the-counter agents, also have the potential to p roduce a positive urine test resul t for phencyclidine. Household and commercial adulterants can produce fa lse- negative
results for some drugs a nd false-positive results for others depending on the type of immunoassay that might be used. 173 Finally, cross-reactive su bstances proved to pro du ce false-positive drug screen ing include: 66 • Opiates - chlorprom az in e, codeine, dextrom ethorphan, dihydrocodeine, diph enoxylate, hydromorphone, levorphanol, meperidi ne, oxycodone, d-prop oxyphene. • Ampheta mines - cathinone, diethylpropion, dopamine, ephedrine, p-hydroxya mpheta mine, isoxs upri ne, labata lol , methy lphenid ate, l-meta mpheta min e, ny lidrin, phenmetraz ine, phentermine, pheny lep hrine, phenylpropanolamine, propylhex edrine, pseudoephedrine, ranitidine, sergiline, N-acetylprocain a mide, chloroquine, procainamide. • Barbiturates - gJuthetimide, phenytoin. • Cocaine (benzoylecgonine) - salicylates. • Phencyclidine - am itripty line, chlorp romazine, dextromethorpha n , diphenhydramine, doxylamine, meperidine, thioridazine. • Cannabinoids (THC-COOH) - ibuprofen, promethazin , ribo Ravin. Although it may arguably represent genui ne exposure and therefore not a fa lse-positive result, environmental exposu re to drugs, such as marijuana smoke, may produce a positive urin e specimen. Ho wever, it has been genera lly co ncluded that passive inhalation exposure, in the absence of extreme conditions of ambient exposure, does not result in posi tive urine tests at conve n tiona l cut-off condi tions. 175 The problem of external contamination is more crucial concerni ng hair testing. 172 ,)76 In situ ations in which dru gs are know n to be present in the environment, it is well known that passive exposure can produce posi tive hair analysis resu lts. In this respect, apprecia ble levels of nicotine have been reported in th e unwashed hair of non-smokers which ca me into contact with the hair of smokers. 177 In contrast, cotinine (the ni cot in e metabolite) does appear to be a marker of tobacco use in the smoker population. 17B In a study condu cted by Smith and Kid well 179 on children living in a family in which cocaine (crack) was used, and thus present as smoke in the envi ronment , it has been shown that 85 per cent of the chil dren tested positive for cocaine and benzoyl ecgonine in their hair. In th is study, it was ass umed that children aged 3-10 were unlike ly to be self-admi nisterin g cocaine, so that any cocaine in their hair must have come from pas sive exposure. This was confirmed by the fact that skin wipes obtained from the children by wiping their fore heads with a cotton swab were also positive for cocaine, indicating extensive surface contact, whereas saliva test ing was negative in most of the ch ild ren, showing th at in gestion of coca ine was not likely a so urce of the cocaine in the hair. The problem of hair contam in at io n by environmental exp osure to a drug is more complex; experts disagree about interpretation, with some claiming that external
Pitfalls and limitations of drug screens I
contamination can be removed by washing through several kinds of decontamination procedures, while others contend that washing removes drugs from inside the hair shaft and does not completely remove external contaminants. IBO Caution until better decontamination procedures are established should induce toxicologists to test only short hair lengths from close to the scalp to limit environmental exposure. Of course, such a limitation would reduce the potential advantage of hair analysis to provide long-term drug use history for an individual. Notwithstanding this contro versy, it seems reasonable to point out that not all positive hair analysis results must be interpreted as due to passive exposure. Certainly, most positive results are due to the administration of drugs. Nevertheless, the above example illustrates that interpretation of data must take passive exposure, especially in young children, into consideration to determine the source of drugs in any case of a positive result. 176
Causes of False-negative Drug Screens Equally problematic in drug testing are situations in which a drug test fails to identify the adolescent who has, in fact, been recently abusing psychoactive drugs and in whom a drug test should be positive. The reasons for a false negative test can be divided into three general categories: technological shOitcomings, pharmacokinetic characteris tics and intentional specimen alteration or adulteration.
TECHNOLOGICAL SHORTCOMINGS One of the most common reasons for a negative toxic screen result, despite obvious drug abuse, is that the clinician or the pathologist fails to recognize that the particular screen being used is incapable of detecting the substance in question. 52 For example, some chemicals abused by adolescents, notably sol vents and other inhalants, cannot be detected by routine screening of either blood or urine by commercially available immunoassays. Without prior knowledge of what the test specifically seeks, the diagnosis may be missed. Similarly, although drug screens commonly identify all members of a drug family, structural differences in specific drugs within the family may limit their detectability (e.g. fentanyl and its ana logues in an opiate screen, or clonazepam in a benzodi azepine screen). Finally, drug tests can be falsely negative because of poor quality control of the laboratory. Inaccuracy rates (false positivity rates as high as 1.7 per cent; false neg ativity rates ranging from 17 per cent to 31 per cent) have been reported in some selies.IBI-IB3
PHARMACOKINETIC CHARACTERISTICS The pharmacokinetic characteristics of drugs have an important influence on drug detectability, particularly if
269
the wrong biological specimen is examined or if the wrong technology is used. For example, drugs that have a large volume of distribution have correspondingly low serum or blood concentrations (cannabinoids or LS~). Immunological drug tests designed for urine specimens are therefore gen erally incapable of detecting these drugs in serum or blood. Another pharmacokinetic characteristic, elimination half-life, also has important implications for drug detection capability. With drugs that have a short elimination half life (e.g. cocaine) the parent compound may be undetectable in blood within 8 hours; however, cocaine metabolites may be detected in urine for several days after a significant exposure because of the drug excretory pattern. Another factor to be pointed out is that the pharmacokinetics of cer tain drugs may vary according to age. In younger children the elimination half-lives are shorter and clearances faster. For example, clearance of morphine reaches adult levels by 6 months of age. IB4 Chlorpromazine is metabolized 2-5 times more rapidly by children under 5 years old than by adults. IBs Subsequently, as a general rule, urine remains the best biological specimen for drug testing because it takes advantage of the fact that kidneys are the primary excretory organs for most drugs.
INTENTIONAL SPECIMEN ALTERATION OR ADULTERATION Biological specimens, palticularly urine, can also be inten tionally altered or adulterated to produce a false-negative result. IB6 Methods available to accomplish this are numer ous and varied, ranging from simple dilution of the col lected specimen to substitution of urine that was produced before any drug use occurred or a fluid that resembles urine. 90 ,IB7 Another commonly used method consists of drinking a large amount of fluid, or even using a diuretic, in order to reduce the concentration of drug in urine so that it falls below the detection threshold of the assay and will produce a negative urine. However, adulteration of the collected specimen with chemical agents is the method chosen by many users because it requires little sophistica tion and can be easily accomplished in unobserved collec tion conditions . Such substances include vinegar, lemon juice, bleach, ammonia-based cleaner, crystalline drain cleaner, non-ionic liquid hand soap, methanol, sodium chloride, toilet bowl cleaner, ionic detergents and even whole blood anticoagulated with EOTA IBB Although well designed collection procedures can minimize the OppOItu nity for sample adulteration and laboratory tests exist that may detect certain types of adulterants (temperature, spe cific gravity, pH measurement of the sample and measure ment of urinary creatinine), no system is absolutely fool-proof. IB9 Commercial adulterants are also available: some con tain the fixative glutaraldehyde (UrinAid, ClearChoice). 190 others contain strong inorganic acid (Amber- 13), such as nitrite and chromate (Urine Luck), or they may contain
270 I
Recent advances in paediatric tox 'icoiogy
enzymes, e.g. peroxidase or protease (papain).'91.192 On-site adulterant detection devices that assess the integlity of urine specimens are also commercially available (e.g. Adultacheck 4).193
Post-mortem Changes POST-MORTEM REDISTRIBUTION OF DRUGS The difference in drug concentration in blood samples from different sites represents a major concern for the forensic toxicologist and, in general, blood concentrations after death cannot be interpreted from a knowledge of concen trations in life. This is because, in life, drugs are distributed to different tissues in the body and differences in concen tration between tissues can be built up and maintained. Active processes, which are responsible for maintaining concentration differences, cease after death and concentra tions can change dramatically. The changes in blood are not necessarily uniform. Concentrations in venous blood after death are expected to vary according to the tissues whose blood they drain. For example, tricyclic antidepres sants, narcotic analgesics, local anaesthetics and antihista mines all accumulate in myocardial tissue and may be partially responsible for the observed rise and subsequent high levels found in heart blood post mortem. 194 In this respect, blood in the right ventricle drains from the liver and other body tissues and is likely to differ substantially from the blood in the left ventricle, which drains the lungs. The site and time-dependent variability of post mortem blood and tissues sampling, and the phenomenon of post-mortem redistribution have been documented and reviewed in the recent literaturel95-200 after having been described as 'a toxicological nightmare' by Pounder and Jones in 1990. 201 Several controlled animal studies as well as case reports have been published that address these issues. Studies with acetaminophen (paracetamol) in rabbits demonstrated that post-mortem drug concentrations in blood increased sig nificantly with time for central sampling sites compared with peripheral blood. 202 The post-mortem redistribution of amitriptyline was demonstrated in rats, which suggested the post-mortem drug release from lungs and other drug-rich tissues into the blood. 203 Post-mortem redistribution has also been demonstrated in rats following administration of mor phine, 204 digoxin 205 and secobarbital. 206 Conversely, rats to which cimetidine was given did not show statistically signi ficant differences in post-mortem changes in cimetidine con centrations in tissues compared with blood. 207 Several human case reports have described similar find ings. Site-dependent post-mortem changes in blood and tis sues concentrations of digoxin, 20s cocaine,209 methadone 21O and methamphetamine 2J' have been reported. Post-mortem redistribution of chloroquine,212 dothiepin,213 amitripty line,214 fluoxetine and norfluoxetine,2J5 tranylcypromine 216
and co-proxamol 217 have been reported. In contrast, other case reports have determined that there is little evidence that post-mortem redistribution occurs with tricyclic antidepres sants,2lS trazadone 219 and zopiclone. 220 Similarly, drugs can be expected to diffuse from gastric residue into blood, liver, lung and other nearby organs. Rats killed and then subjected to gastric instillation of amitriptyline showed drug diffusion into the liver from 5 hours post administration at room temperature. 221 Over time, highest concentrations were reached in liver lobes adja cent to the stomach, with significantly lower levels in the right lobe. Similar studies on human cadavers used amitripty line, paracetamol (acetaminophen) , and lithium in quantities representing 10 tablets of each drug. 222 After 48 hours at room temperature, drug diffusion affected the left lobe of the liver and, to a lesser extent, the caudate lobe and variably the light lobe posteriorly (with the cadaver supine). In the lung, the left was more affected than the right and the base more than the apex, although this was not always true. Post-mortem diffusion of ethanol, facilitated by its high hydrosolubility, from the stomach into the blood and the nearby organs has been shown to follow a similar pattern to that observed with drugs 223 It is apparent that drug diffusion (including ethanol) in a cadaver is an important and complex phenomenon that can affect a wide range of organs as well as blood . There fore, knowledge of sampling site and sampling technique is a prerequisite for a valuable interpretation of analytical results. To facilitate this, pathologists should formalize their protocols for toxicological sampling at autopsy. Blood samples should be obtained by needle puncture of the external iliac or femoral vein. This may prove difficult in babies. A liver sample should be obtained from deep within the right lobe and any lung sampling should be taken from the apex rather than the basal lobes.
POST-MORTEM DEGRADATION OF DRUGS Post-mortem degradation of drugs and pOisons is a process that is little understood but may also significantly affect the interpretation of post-mortem toxicological results. First, there is the delay that occurs before the sample has been taken, and then between sampling and initial presumptive drug screening or blood alcohol determination. Second, the subsequent confirmation may not be performed until the case goes to court for trial or may be done many days or weeks after the blood has been taken, especially when toxi cological analyses are carried out by separate institutions or at several laboratories each carrying out isolated tests. Usu ally additives and preservatives have been added but not always, and samples have been stored at 4°C or at -20°C. Degradation before analysis may occur as a result of chem ical or physical decomposition owing to the instability of the drug, which lacks the protective effect of being bound to plasma proteins. 224 For example, diazepam 225 and dilti azem 226 are susceptible to hydrolysis and flunitrazepam 227
Specific applications I
and LSD 228 were reported to significantly degrade within only 24 hours when exposed to sunlight. Chemical instabil ity has been also attributed to the time-dependent loss of drugs, such as chlorpromazine or dothiepin, after death 229 Other drugs, such as paracetamol (acetaminophen), salicy lates, benzoylecgonine, cannabinoids, morphine, buprenor phine and amitriptyline, would be more or less stable. 230-232 Degradation may also derive from enzyme activities linked to bacterial development that takes place after death and may continue in an unpreserved sample after collec tion. 233 Enzyme activities in the blood of healthy living persons are thought to be mainly restricted to esterases, which can affect cocaine, procaine and other esters; this is the reason why an enzymatic inhibitor must be added as a preservative to the blood sample. 224 After death, intestinal bacteria penetrate the intestinal wall, enter the blood and lymph vessels and migrate further throughout the body. At 25°C this transmigration can occur as soon as 5 hours after death. The possible role of bacterial degradation has been assumed to explain, at least partially, the continuous decrease in concentration in blood after death of many drugs such as cocaine or methamphetamine. Bacteria have been shown also to be responsible for the post-mortem metabolic changes of many drugs, including the nitrobenzodiazepines (clon azepam, nitrazepam and flunitrazepam) metabolized to their 7 -ami no meta bolites. 224 ,234, 235
271
even though the predictive value of vitreous humour in estimating blood alcohol concentration appears somewhat questiona ble. 24o ,241
SPECIFIC APPLICATIONS Drug Screening for Doping Agents
In contrast to the degradation of drugs post mortem, cer tain compounds may appear to increase in concentration after death or during storage in unpreserved containers,9] These include cyanide, which is generated by the degrada tion of haemoglobin and ethanol. Ethanol, which tops the list of psychoactive substances encountered in post mortem toxicology, can be produced by post-mortem fer mentation of glucose, though this rarely ex ceeds 0.5 gOfc. Ethanol can, of course, be both formed and destroyed by microbial activity236 A recent comprehensive review of issues relating to the interpretation of ethanol concentra tions in post-mortem specimens was presented by Kugel berg and Jones. llI A chemical preservative, such as sodium fluoride (1-2 per cent w/v), should be added to specimens intended for determination of ethanol, which should pre vent any further production of ethanol. The addition of an insufficient amount of fluoride (such as 2-3 mg/mL in commercially available tubes) may result in very particular cases in the in vitro formation of ethanol, reaching a con centration as high as 3.5 giL. 237 Distinction between ingested ethanol from that formed post mortem can be made by the detection of ethylglucuronide, ethylphosphate and ethylsulphate, three speciflC markers of ethanol intake,2J8,2]9
The misuse of doping agents in sport is concerned with substances registered on an annually updated list, 'The pro hibited list. International sta ndard' (World Anti-doping Code, World Anti-Doping Agency (WADA)) . Depending on the survey methods used, it has been estimated that doping may involve 3-5 per cent of adolescents. For instance, studies carried out in eastern France with 1501 athletes aged 15-19 years found that 4 per cent had used banned substances at least once in their Iife,242 The same study design and setting, i.e. self-questionnaire survey, were used in a 4-year follow-up of doping prevalence among pre adolescent athletes,24J This study disclosed that 1.2 per cent of the respondents entering the cohort and 3.0 per cent 4 years later reported that they had used doping substances at least once during the preceding 6 months. The main sub stances used as doping agents were salbutamol, cortico steroids and cannabis. Cannabis is one of the drugs that is most frequently detected in the context of doping.244 With the ever-increasing number of dietary supple ments used by athletes and the availability of numerous synthetic steroids and recombinant peptide hormones and modulators, detection of performance-enhancing com pounds is increasingly difficult,245,246 and needs more and more sophisticated detection methods such as (fast) chro matography coupled with tandem mass spectrometry247 or carbon isotope ratio determinat ion ,248 The paediatrician might suspect such use when the young athlete presents with recent increase in weight or muscle mass, oedema, gynaecomastia in males, deepening of the voice in females, new acneiform rashes, changes in behav iour (initability, labile temperament, depression), or a new onset of chemical hepatitis. Careful and sensitive question ing, in a non-judgemental but knowledgeable manner, can sometimes induce the adolescent to confess steroid use so that counselling can take place. 52 According to Dawson,249 the patients who should concern the physician the most are not elite athletes but the yo uth who are being increasingly drawn to the use of performance-enhancing drugs. Many health problems may arise because of use and mis use of doping agents, such as intake of counterfeit products, drug-drug interaction, dermatological disorders, and cardio vascular, hepatic, genitourinary, musculoskeletal, neuropsy chiatric, endocrine and haematological adverse side-effects, not to mention infection resulting from the use of contami nated needl es,249,250
If available, vitreous humour, which is more protected from bacterial contamination, can be used instead of blood ,
Over 100 an abolic steroids are widely available on the black market in oral or injectable form. 251 Extensive
POST-MORTEM FORMATION OF DRUGS
272 I
Recent advances in paediatric toxicology
metabolic pathways as well as natural occurrence make the detection of these substances difficult to interpret. Positive tests are usually confirmed by GC-MS. Oral steroids can be detected as long as 4 weeks after last use. Normally, testos terone exists in the body in balance with epitestosterone with ratios of between 1 and 2. Epitestosterone is a natu rally occurring epimer of testosterone, thought to be of extratesticular origin and having no biological activity. A testosterone-epitestosterone (TIE) ratio greater than 6 (the threshold ratio used by the International Olympic Commit tee) is considered as evidence of exogenous androgenic steroid use. To further foil ingenious athletes from taking exogenous epitestosterone to reduce their TIE ratios below 6, a maximum allowable urine concentration of 150 ng/mL epistestosterone has been established. 252 When to use urine testing remains controversial. Although unannounced steroid testing at high-school ath letic events would perhaps diminish such substance use, this testing is prohibitively expensive. Additionally, it would require that a chain of custody be established, and such testing might engender an issue of privacy and consent. Furthermore, besides anabolic agents, peptide hormones, including growth hormone, erythropoietin, adrenocorti cotropin (ACTH), chorionic gonadotrophin and their releas ing factors, are also widely used among athletes. Although detection of these peptides by immunoassay is well estab lished in the clinical laboratory, the acceptability of this method of analysis as definitive proof of administration has yet to be established in forensic toxicologY.246 However, the unequivocal identification of peptide and protein analytes used for doping will very likely rely on complex hyphenated mass spectrometry methods following very specific extrac tion and purification methods. In this respect, immuno affinity purification will precede the analysis with tandem mass spectrometers, such as quadrupole-linear ion trap or linear ion trap-orbitrap analysers. 253
Forensic Applications of Hair Analysis In Children and Teenagers Owing to its ability to provide evidence of drug exposure as well as drug use, hair testing for drugs of abuse has been available for several years in courts and accepted according to the legal conditions of the country or state where trials are pending. Of course, toxicologists are aware that no sample technique or specimen can provide answers to all toxicolog ical questions. More information is obtained with a variety of analytical approaches and different biological specimens. However, hair analysis offers 'a unique perspective on human drug use'254 by providing a wider window of drug detection and may offer advantages over other drug testing methods in terms of accessibility, resistance to post-mortem degradation and stability of the entrapped substances in it. The following examples are given to illustrate the potential value of hair drug test results in forensic practice.
Sachs 103 reported the case of a child who had been brought to a clinic with symptoms of intoxication attributed to clozapine. As clozapine was not freely available, the mother of the child was suspected of having intoxicated her child with the drug prescribed to her depressive sister. One year before, a l-year-old boy belonging to the same family had been found dead. Owing to the absence of any obvious cause of death, the case was ruled as sudden infant death syndrome. After exhumation, the child's hair, the only tis sue available for toxicological analysis, tested positive for clozapine, providing evidence of repeated administration of the drug to the child when he was alive. However, only the qualitative result was of importance to the judge; as the child could not have been under antidepressive treatment, the mother was sentenced to 4 years only because she even tually admitted giving the drug to the child. Lewis et al 255 reported the usefulness of hair testing of children living with suspected drug users (crack) and con sidered to be at risk. Using the hair of the children to deter mine smoke exposure provided 'extra credibility' to the child protective services and allowed them to remove chil dren from dangerous households. Similar hair testing applications have been reported by Smith and Kidwell. 179 Strano-Rossi et al 256 described the application of hair analysis in a judicial case to document coercive heroin administration to a 5-year-old child who was admitted with overdose symptoms to an intensive care unit. The results of toxicological hair analysis were accepted by the court with out criticism; only the possibility of the consumption of antitussive drugs was raised by the defendant's lawyers. However, the presence of monoacetylmorphine (MAMl. the specific metabolite of heroin, allowed the toxicologist to refute this objection. The relative of the child was found guilty and sentenced accordingly. In the same way, Huestis 254 quoted several American court decisions ordering that parents' hair be collected and tested for drugs of abuse (cocaine and marijuana cases) to determine the fitness of the custodial parent in child cus tody disputes or to evaluate their degree of drug addiction in order to determine the best interests of the child in adop tion cases. Ketamine is a rapid-acting dissociative anaesthetic. It has been reported to be abused by an increasing number of young people as a 'club drug', and is often distributed at 'raves' and parties. Teenagers are the major abusers. Keta mine concentrations were found to range between 0.8 and 92.3 nglmg in hair collected in entertainment places from 15 ketamine abusers. 257 Ketamine metabolites were also detected, indicating ingestion of the drug and excluding the hypothesis of external contamination. Methylphenidate (sold as Ritaline®, Medikinet® or Con certa®) is used in the treatment of childhood ADHD. Methylphenidate is hydrolysed in the body into ritalinic acid. Both the parent drug and its metabolites can be detected in blood. However, only the parent compound is present in hair from treated individuals. As hair samples in
Specific applications I
children can be easily collected, in contrast to invasive blood sampling, determination of methylphenidate in hair should be an alternative to check compliance in a wider time window than if using blood. Methylphenidate can be measured in hair by LC-MS. 258 Hair analysis can be very useful also in the case of sus pected drug-facilitated crime or alleged sexual abuse of children and teenagers. Drugs generally involved in such crimes are sedatives, hypnotics and anaesthetics. Alco hol or drugs of abuse, such as cannabis, LSD, gamma hydroxybutyrate (GHB) or ecstasy, are also used. lOB Because the majority of cases are reported to the police after a few days, the hair is very often the only specimen suitable for drug detection. In cases of child abuse, the drug can be administered once or chronically. A repeated administration is easier to identify because drug concentrations are gener ally much higher than after single exposure. The analysis is more challenging with low-dosed drugs, such as buprenor phine. 259 In the case of a single application, very sensitive methods are required for drug detection, in concentrations down to the sub-pg/mg hair range. This was made possible through the introduction of new analytical techniques 2 60 For instance, bromazepam (0.8-28 pg/mg) was measured by tandem mass spectrometry methods in three volunteers after a single ingestion of 6 mg261.262 Recently, a sex offender has been charged with oral genital contact with an 8-year-old girl. The events were reported by the girl, who experienced insomnia and speech disorders. Hairs were sampled 1 month later and bromazepam was detected in the hair segment corresponding to the timing of the reported facts. 263 The usefulness of hair testing was demonstrated in another drug-facilitated sexual assault case of a 9-year-old girl. In that case, diphenhydramine, an antihistamine agent with sedative effects, was discovered in successive hair segments, indicating repeated surreptitious administration. The perpetrator admitted charges at court relating to the abuse of young girls. 264 The same group reported the administration of a phenothiazine derivative with sedative properties (trimeprazine or alimemazine) to two children. The stepmother, who was the perpetrator, did not challenge in court the suggested lise of trimeprazine as a sedative drug. 26 5 Two deaths of babies sedated with methadone by their mothers were reported by Kintz et al 266 Both mothers were former heroin addicts under methadone substitution ther apy. At autopsy, no evidence of violence was noticed. Tox icological investigations by GC-MS in hair and blood demonstrated recent and repeated methadone exposure. Hair (6 cm) from the first 14-month-old baby tested posi (ive at 1.9 and 0.8 ng/mL for methadone and EDDP (methadone metabolite) respectively. In another case, a comatose 5-month-old girl was taken to hospital, where she was declared dead 1I days later. Hair analysis (5 cm) revealed the presence of methadone at a concentration of J.O and 21.3 ng/mL in the root and end hair segments
273
respectively. Many other cases have gone to court having similar circumstances of a baby dying suddenly and unexpectedly. Gamma-hydroxybutyric acid is presumed to be one of the most frequently used substances in drug-facilitated crimes. Because of its very short half-life, GHB is particu larly difficult to detect in blood and urine. Furthermore, GHB is also an endogenous substance found in low con centrations in body fluids. Nevertheless, detection of a single GHB exposure has been reported to be possible. 267
Determination of Fetal Exposure to Drugs of Abuse Maternal self-reported drug history has been shown to be unreliable, as many women who deny use during preg nancy exhibit drug metabolites in their urine. 26B On the other hand, systematic urinalysis during pregnancy is hampered by the sholt elimination half-life of the drugs. A negative result may be a result of deliberate abstinence for several days before biomedical screening. A positive result reflects only exposure during the preceding 1-3 days. A common method of estimating the amount of fetal expo sure to a drug could be provided by measuring the concen tration in the umbilical blood at birth. 169 The problem is tha t the presen ce of drug in umbilical blood reflects only a very recent exposure and lack of drug does not rule out fetal exposure for most of the pregnancy. Measurement of drugs in neonatal urine presents the same disadvantages. Evaluation of drug concentrations in the amniotic fluid measured during pregnancy or at delivery cannot provide information on the duration and degree of fetal expo sure. 167 The same disadvantages are noted with the analy sis of meconium, which is only a qualitative test at the moment of delivery.269 Neonatal hair begins to form at approximately 6 months gestational age; a positive result indicates use during the last trimester. Hair testing can be collected during the first trimester of life, after which time infant hair replaces neonatal hair. Drugs in neonatal hair could originate from deposition from fetal blood or from contamination of hair by amniotic fluid . 170 Hair from the mother can be collected also for drug analysis and results can be compared with those obtained from neonatal hair. Hair analysis may therefore remedy the disadvantages previously mentioned with the other methods, with a wide window of detection ranging from weeks to months, and may provide informa tion concerning the severity and pattern of an individual's drug use when a maternal drug history is not available or is in doubt. Regarding the correlation of drugs in maternal and neonatal hair, concentrations of nicotine and cotinine in paired maternal and neonatal specimens were found to be well correlatedYo Mothers and infants in the smoking groups, both active and passive, had significantly higher
274 I
Recent advances in paediatric toxicology
concentrations of nicotine and cotin in e than did those in the control, non-smoking group. Th is study documents the impoltance of hair ana lysis as a tool for measuring exposure to cigarette smoke. Indeed, passive smoking has been shown to adversely affect the health of infants and children .271 Prenatal cocaine and methamphetamine exposure is now frequently tested because of their relatively high prevalence of use during pregnancy, especially in th e USA. Gestational exposure has been associated with placental abruption and premature labour, as well as with an increase in low birth weight, microcephaly, congenital anomalies and necrotizing enterocolitis. Neonatal hair was found to be a suitable biological marker to demonstrate in utero cocaine and methamphetamine ex posure. 272 ,27J Inter estingly, median cocaine concentratio n was lO-fold higher in the hair of the mothers than in the neonates. 274 In utero alcohol exposure also can be detected by hair an alysis. The retrospective detection of alcohol consu mp tion during pregnancy can be performed also by measuring specific markers of alcohol abuse in the mother's hair. Fatty acid ethyl esters and ethylglucuronide, which are minor metabolites of a lcohol , are suitable markers for the detec tion of heavy alcohol consumption. They may be used in instances of fetal alco hol syndrome or suspicion of regular alcohol intake during pregnancy to objectiv ate alcohol abuse,275 Recently, analysis of vernix caseosa has been proposed as an alternative specimen to hair when hair is not avail able or is too difficult to collect. 276 Vernix caseosa is a thick, white lipid and cell mixture that covers the fetus, starting at about 24 weeks' gestational age. 171 Vernix can eaSily be removed from a newborn's skin with a piece of gauze. Cocaine and/o r three of its metabolites were detected in a 3-cm section of gauze containing verni x obtain ed from three out of five neonates.
THE IMPORTANCE OF PAEDIATRIC TOXICOLOGY IN SPECIFIC CASES Deliberate Poisoning by Health Professionals Beverley Gail Allitt, dubbed 'the Angel of Death', was an English State Enrolled Nurse (SEN) who was convicted of killing four children and injuring five others, in 1991, on the children 's ward of Grantham and Kesteven Hospital, Lincolnshire, w here she worked. 277 She has since become one of Britain's most notorious female serial killers. Her main method of murder was to inject the children under her care, particularly those in intensive care, with insulin or potassium to ca use a cardiac arrest. 278 In an environment such as a children 's wa rd, where staff are dedicated to pro viding the best possible care, the threshold for recogni zing covert acts of excess and inappropriate ad ministration of therapeutic substances is high. It is not, therefore, surprising
that several deaths had occurred before that thresh old was reached in this particular instance. Each individual unex plained death occurring in a hospital environment requires careful investigation, and the careful and thorough exclu sion of poisoning279 (see Chapter 18). Fabricated or induced illness by carers (FlI) was first described in 1977 by Meadows, a paediatrici a n in Leeds . The condition was, at the time, referred to as Munchausen's syndrome by proxy280 Among other manifestations, ill nesses are induced by carers of children (mostly females) by ad ministering substances, smothering, withholding nutri ents and medicines, and by other means, e.g. introducing infectious material into the gut or bloodstream, Commonly used poisons include insulin, salt, antidepressants, anti coagulants and bleach. These chi ldren may present as acute life-threatening events, as unexplained deaths or as baffling clinical problems; the assistance of the tox icologist is of great value in these instances.281 Occasionally, similar prob lems arise when others close to the child but not actual car ers, for a variety of potential reasons, induce illness in a baby by administering therapeutic substances.
Toxicology and Sudden Infant Death Syndrome Among the many hypotheses put forw a rd as explanat ions for sudden infant death syndrome (SIDS), a number involved poisoning, e.g. antimony leaching from cot mat tresses. 282 Another hypothesis that the primalY cause of SIDS is poisoning by toxic gases generated in cot mattresses was first publicized in the media in 1989 by Richardson, an independent consultant on biodeterioration of materials. It was first formally published in the Lancet in March 1990. 283 The theory proposed was that a fungus (Scopulariopsis brevicaulis) sometimes found in the domestic environment could degrade the chemical compounds of phosp horus, arsenic and antimony that may be present in fire ret ardants or plasticizers in PVC cot mattress covers and other cot fur nishings; this wou ld subsequently release the toxic gases phosphine, arsine and stibine 284 The hypothesis was inves tigated by two Independent Expert Groups285,286 that found that Richardson's conclusions could not be substantiated by independent researchers, and thus concluded in 1991 and 1998 that the hypothesis was unfound ed.
CONCLUSIONS AND FUTURE CONSIDERATIONS IN FORENSIC PAEDIATRIC TOXICOLOGY Use of hyphenated mass spectrometlY methods has permit ted the simultaneous measurement of a large spectru m of potentially toxic drugs. Increase in sensiti vity allows a decrease in the size of the sampled specimens. The use of alternative biologica l matrices, such as neonatal hair, meconium, amniotic fluid or oral fluid, extends the field
References I
and time window of investigation. Howeve r, many short co mings ex ist, especially regardi ng the post-mortem redis tribution and stability of drugs in the fetus, neonate and in fant. The neoformation of ethanol in paediatric subjects must be studied and the usefulness of drug exposure mark ers for this particular popul ation should be better charac terized. Basic research is still needed to better characterize the pharmacokinetic and pharmacodynamic aspects of dnlgs in paediatric patients and victims. The know led ge of the developmental patterns of dru g metabolizing activities will facilitate the determin ation of the optimal therapeutic dose ofa drug and its pharm aco logica l consequences in the neon ates. If the mechanisms of drugs deposition in adult hair and other alternative matrices are partially elucida ted, almost eveJY1:hing remains to be done regarding neonates. For instance, the influence of pigmentation or hair type may differ from that known for teenagers and adults. Finally, a la rge database that compiles t herapeutic, to x ic and comatose fatal blood concentrations of drugs from newborns and children should be readily available. The creation of such a database will make easier the interpreta tion of blood concentrations measured in clinical and forensic toxicology.
REFERENCES I 2
3 4
5
6
7
8 9
10 11
12 13
14
Goepp J. Pediatric poisonings. Clill Chem 1996; 42: 1356-60. Ell enhorn MJ, Barceloux DG. Med ica l toxico logy. In Diagllosis alld Treatment oj Human Poisoning. New York: Elsevier, 1988. Fine J, Goldfrank L. Update in medical tox icology. Pediatr
Ciin North Am 1992; 39:103J-51.
Burda A, Burda N. Controversies in tox ico logy. The na ti on's first poison co ntro l center taking a sta nd aga inst acc idental childh ood poisoning in Chicago. Vet HUll! Toxicol 1997; 39:115-19. Hanzlick R. Nationa l association of medical exam iner's
Pediatric Toxicology (PedTox) regis try. Toxicology 1996;
107: 153- 8. Liebelt EL, Shannon MW. Small doses, bi g problems.
A selected review of highly toxic common medications.
Pedia tr Emerg Care 1993; 9:292-7.
Berlin CM. Advances in pediatric pharmacology an d toxicology. Adv Pet/illtr 1993; 40:545-74 . Berlin CM. Advances in ped iatric pharmacology and tox ico logy. Adll Pediatr 1995 ; 4 2 :593-629. Berli n CM. Advances in pediatric pharmacology and toxicology. Adll Pediatr 1997; 44:545-74. Criddle LN!. An ove rvi ew of pediatric poisonin gs. AACN Ad/! Crit Care 2007; 18:109 -18. Eldridge DL, Van Eyk J, Kornegay e. Pedia tri c toxicology. Emerg Med Ciin North Arn 2007; 25:283- 308; abstract vii- viii. Chasn off IJ, Burns WJ, Sc hn o ll SH, Burn s KA. Cocaine use in pregnancy. N Engl J Med 1985 ; 313:666-9. Rubin JD, Ferencz C. Loffredo C. Use of prescription and non-prescription drugs in pregnancy. J Clill Epidemiol 1993; 46:581-9. Chasnoff IJ, Lewis DE, Griffith DR, Willey SD. Cocaine and pregnancy. Clinical and toxicological implicati o ns for the neonate. Ciin Chem 1989; 35: 1276-8.
15
16 17 18 19
20
21
22
23
24 25
26 27
28 29
30
31 32
JJ 34 35
36
37 38 39
40
275
Burkett G, Yasin SY, PaIO\'11 D, et al. Patterns of cocaine
binging. Effect on pregnancy. Am J Obstet Gynecol 1994;
171 :372-9.
Das G. Cocaine a buse and reprodu ction. Int J Ciin Pha/111aCol Ther 1994 ; 32: 7- 11. Sweet AY. Narcotic withdrawal syndrome in the newborn. Pediatr ReI! 1982; 3; 285-91. Bergman A, Weisner L. Relation of passive cigarette smoking to sud den infant dea th syndrome. J Pediatr 1976; 58:665-8. Wa rd SLD, Ba u tista D, Chan L, et al. Sudd en infant death sy nd ro me in infants of substance-abu sin g mothers. J Pediatr 1990 ; 117: 8 76-81. Meyer MB, Jonas BS , Tomascia JA. Perinatal events associa ted with maternal smoking durin g pregnancy. Am J Epidemio l 19 76; 103 :464-9. Bar-Oz B, Bu lk owstein M, Benyamini L, et al. Use of antibio tic and analgesic drugs during lactation. Drug Saj 2003; 26:925-35. Koren G, Cairns J, Chitayat D, et al. Pharmacogenetics of morphin e poisoning in a breastfed neon ate of a codei ne prescribed mother. Lancet 2006; 368:704. Johnson TN. The development of drug metabolizing enzymes and their influence on the susceptibility to adverse drug reactions in children. Tox icology 2003; 192:37-48. Ito S. Drug therapy for breast-feeding women. N Engl J Med 2000; 343:118- 26 . Lopez-Munoz F, Alamo C, Quintero-Gutierrez FJ, Garcia Garcia P. A bibliometric study o f international scientific productivity in attenti on -d eficit hy pera ctivity disorder coverin g the period 1980- 2005. Eur Child Adolesc Psychiatry, 2008 [epub a head of printJ. Klein-Schwartz W. Abuse and toxicity of methylphenidate. CUlT Opin Pediatr 2002; 14: 2 19-23 . Bailey B, Letarte A, Abran Me. Methylphenidate unintentional ingestion in preschool children. Ther Drug Monir 2005; 27:284- 6. Korkia P, Stimson Gv. Anabolic steroid use. Lancet 1993; 341:1407. DuRant RH, Escobedo LG , Heath Gw. Anabolic stero id use, strength training and multiple drug use among adolescents in the United States. Pediatrics 1995; 96:23-8. Yesa lis C, Kennedy N, Kopstein A. Anabolic-androgenic stero id use in the United States. JAtvIA 1993; 2 70: 1217-21. Madea B, Grellner W, Musshoff F, Dettmeyer R. Medico-l ega l aspects of doping. J Ciill Forensic Med 1.998; 5:1-7. Buck ley WE, Yesa lis CE, Friedl KE, et at. Estimated prevalence of anabolic steroid use amo ng male high schoo l seniors. JAMA 1988; 260:344 1-5. Terney R, McLain LG. Th e use of an abo li c steroids in high sc hoo l students. Am J Dis CI1i1d 1990; 144: 99- 103. Hoffman RJ, Nelson L. Rational use of toxico logy testing in children. Curr Opin Pediatr 2001; 13:183 - 8. Madden MA. Pediatric poisonings: recogn ition, assessment, a nd management. Crit Care Nurs Ciin North Am 2005; 17 :3 95 -404, xi. Belson MG, Simon HK, Sullivan K, Geller RJ. The utility of toxicologic a nalysis in children with suspected in ges tions. Pedintr Emerg Care 1999 ; J 5:383-7. Kind T. Carbon monoxide. Pedia tr Rei) 2005; 26:150-1. Whi te ML, Liebelt EL. Update o n a ntid otes for pediatric poisoning. Pedjatr Emerg Care 2006; 22: 740 - 6; quiz 7-9. Heimann G. Enteral absorption and bioavailability in children in relation to age. Eur J Ciin Pharmacol 1980; 18:43- 50. Richardson T. Pitfalls in forensic toxicology. Al1n C/in Biochem 2000; 37 (Pt 1):20-44.
276 I
41 42
43 44 45 46
47
48
49
50
51 52 53
54
55 56
57
58
59 60
61
62 63
Recent advances in paediatric toxicology
Rane A, Tomson G. Prenatal and neonatal drug metabolism in man. Eur J Clin Pharmacol1980; 18:9-15. Hines RN. The ontogeny of drug metabolism enzymes a nd implications for adverse drug events. Phannacol Ther 2008; 118:250-67. So ldin SJ, Steele BW. Mini-rev iew: therapeutic drug monitoring in pediatrics. Clin Biochem 2000; 33 :333 -5. Sold in OP, Soldin SJ. Review : therapeutic drug monitoring in pediatrics. Th er Drug Monit 2002 ; 24: 1-8. Alco rn J, McNamara PJ. Pharmacokinetics in the newborn. Ad/) Dru g Deliv Rev 2003; 55:667-86. Allega ert K, Verbesselt R, Naulaers G, et a J. Developmental pharmacology: neonates are not just small ad ults. Acta Clin Belg 2008; 63:16-24. Ward RM, Benitz WE, Benjamin DK, Jr. , et al. Criteria sup portin g the study of drugs in the newborn . Clin Th e!" 2006; 28: 1385-98. Johnson TN, Rostami-Hodjegan A, Goddard JM, et a J. Co ntribution of midazolam and its I-hydroxy metabolite to preoperative sedation in chi ldren : a pharmacokinetic pharmacodynamic ana lysis . Br J Anoesth 2002; 89:428-37. Cuzzolin L, Atzei A, Fan os V. Off-l abel and unlicensed prescribing for newborns a nd chi ldren in different settings: a review of the literature and a consideration about drug sa fety. Expert Opill Drug Safety 200 6; 5 :703- 18. Shamoo AE, Tauer CA. Ethica lly q ues tion able research with ch ildren: the fenfluramine study. Account Res 2002; 9: 143- 66. Sharav VH. Children in clillical research: a conflict of moral values. A m J Bioeth 2003; 3:W 12-W59. Woolf AD, Shannon MW. Clinical toxicology for the pediatrician. Pediatr Clill North Am 1995 ; 42:317-33. George S. Posi tio n of immunological techn iques in screening in clinical toxicology. Clin Chem Lab Med 2004; 42: J 288- 309. Miyaish i S. lmmunoassays : theolY and in strumenta tion, techniques, applicatio ns. In Townshend A (ed.) Encyclopedia Analytical Sciences. Volume 5. London: Academic Press, 1995, pp. 2049-108. Montagne M, Pugh CB, Fink. JL. Testing for drug use, pan 1: analytical methods. Am J Hosp Pharm 1988; 45:1297-305. Baselt RC. Urin e drug screenin g by immunoass ay : interpretation of results. In Baselt RC (ed.) Adl1allces in Analytical Toxicology. Foster City, CA: Biomedical Publications, 1984, pp. 81-123. Kolbrich EA, Kim I, Barnes AJ, et aJ. Cozart RapiScan Oral Fluid Drug Testing System: an evaluation of sensitivity, specificity, and effic iency for cocaine detection com pared with ELISA and GC-MS following controlled cocai ne administrati on. J Anal ToxicoI2003; 27:407-11. Gjerde H, Christophersen AS, Skuterud B, et al. Screening for drugs in forensic blood samples using EMJT urine assays. Forensic Sci Int 1990; 44: 179-185. Pichon V. Selective sample treatment using molecularly imprinted polymers. J Chromatogr A 2007; 1152:41-53. Souv era in S, Rudaz S, Veuthey JL. Restric ted access ma teria ls and large particle suppo rts for o n-line sa mple preparation: an attractive approach for biological fluids an alys is. J Chro matogr BAnal Technol Biomed Life Sci 2004; 801:141-56. Wille SM, Lamben WE. Recent developments in extraction procedures relevant to analyti cal toxi cology. Ana l Bioana l Chern 2007; 388:JJ8J-9 1. Chen Y, Gu o Z, Wang X, Qiu C. Sample preparation. J Chromatogr A 2008; 1184: 19 1-21 9. Herzler M, Herre S, Pragst F. Selecti vity of substa nce id entifi cati on by HPLC-DAD in toxicological ana lysis using
64
65
66
67
68
69
a UV spectra library of 2682 co mpo und s. J Anal Toxicol 2003 ; 27 :233 -42. Pragst F, Herzler M, Erxleben BT. Systematic toxicological ana lysis by high-performance liquid chrom atograp hy with diode array detectio n (HPLC- DAD). Clin Chem Lab Med 2004; 42:1325-40. Grobosch T, Lemm-Ahlers U. Immun oassay screenin g of lysergic acid die thylamid e (LSD) and its confirmation by HPLC a nd fluorescence detection following LSD ImmunElute extraction. J Anal Toxicol 2002; 26: 181-6. Moore CM. Drug test ing in the 90s. Tn Leikin SB, Paloucek FP (eds) Poisoning alld Toxicology Handbook, 2nd edn. Chicago, IL:Lexi-Comp inc; Hudson (Cleveland) and American Pharmaceutical Associ atio n, 1996, pp. 11 62 -75. Giro ud C, Michaud K, Spo rkert F, et a l. A fata l overdose of cocain e asso ci ated with coin ges tion of marijuana, bup ren orph ine, and fiu oxeti ne. Body flui d an d tissue distribution of coca ine and its metaboli tes determined by hydrophilic interact ion chro matograp hy- mass sp ectrometry (HILlC-MS). J Anal Toxicol2004; 28:464-74. Dienes-Nagy A, Rivier L, Giroud C, et al. Method for quantification of morphine and its 3- and 6- glu curonid es, codeine, codeine glucuronide and 6-m onoacetyJmorphine in human blood by liquid chromatography- electrospray mass spectrome try for routin e analysis in forens ic toxico logy. J Chromatogr A 1999 ; 854:109-18. Swartz ME . UPLC: An introduction and review. Journal of
Liquid Chromatography Et Rela ted Teclmologies 2005; 70
71
72
73
74
75
76
77
78 79
28: 1253-63. Churchwell MI, Twaddle NC, Meeker LR, Doerge DR. Improving LC-MS sensitiv ity thro ugh increases in chromatographic performance: co mparisons of UPLC ES/MS/MS to HPLC-ES/MS/MS. J Chromatogr BAnal Technol Biomed Life Sci 2005; 825 :JJ4-43. Bugey A, Staub C. Appli ca tion of mon o lithic SUppOit S to online extra ctio n and LC-MS analysis of benzodiazepines in whole blood samples. J Sep Sci 2007; 30:2967 -78. Segura J, Ventura R, Jurado C. Derivatization procedures for gas chromatographic-mass sp ectro met ric determination of xenob iot ics in biological samples, with spec ial attentio n to drugs of abuse and dop ing agents. J Chromatogr B BioMed Sci Applications 1998; 713:61- 90. Gunnar T, Ariniemi K, Lillsunde P. Validated toxicological determination of 30 drugs of abuse as optimized derivatives in ora l fluid by long co lumn fast gas chro matogra phy/ electron impact mass spect rometry. J Mass Spectrom 2005; 40: 739-53. Gun nar T, Ariniemi K, Lillsunde P. Fast gas chromatography negative- ion chemica l ionization mass spectrometry with microscale volume sample preparation for the determination of benzodiazepines and alph a-hydroxy metabolites, zaleplon and zopiclone in whole blood. J Mass Spectrom 2006; 41:741-54. Th omas A, Widmer C, Hopfgartner G, Stau b C. Fast gas chromatography and negative-ion chemical ionization tand em mass spectrometry for forensic a nalysis of cannabinoids in whole blood. J Pharl17 Biomed A nal 2007 ; 45:495- 503. Thormann W, Thormann W. Progress of capillary electrophoresis in therapeu tic drug monitoring a nd clinical and forensic toxicology. Th e?' Drug Manit 2002; 24: 222-3 1. Thormann W, Zhang C, Schmutz A. Capillary electrophoresis for drug analysis in body fl uids. Ther Drug Monit 1996; 18:506-520. Vo n Heeren F, Thormann W. Capi llary electrophoresis in clinica l and forensic analysis. Electrophoresis J997; 18 :241 5- 26. Lloyd DK. Cap illary electroph oresis ana lysis of biofluid s with a focus on less commonly ana lyzed matrices. J Chromatogr BAnal Technol Bi0111ed Life Sci 2008; 866:154-66.
References I
80 Smith ML, Vorce SP, Holler JM, et al. Modern instrumental methods in forensic toxicology. J Anal Toxicol 2007; 31 :237-53, 8A-9A. 81 Maurer HH, Peters F1. Toward high-throughput drug screening using mass spectrometry. Ther Drug Monit 2005; 27:686- 8. 82 Maurer HH. Hyphenated mass spectrometric techniques indispen sable tools in clinical and forensic toxicology and
in doping control. J Mass Spectrom 2006; 41: 1399- 4 13.
83 Pa vlic M, Libi se ller K, Oberacher H. Combined use of E51 QqTOF- MS and E51-QqTOF-MS/MS w ith mass-spectral library sea rch for qualitative analysis of drugs. Anal Biaanal C/wn 2006; 386:69-82. 84 Ojanpera S, Pelander A, Pelzing M, et al. Isotopic pattern and accurate mass determination in urine drug screening by liquid chromatography/time-of-flight mass spectrometry.
Rapid Commul1 iVIass Spectrom2006; 20:1161-7. 85 Hartmann C, Smeyers - Verbeke J, Massali DL, McDowall RD. Validation of bioanalytical chromatographic methods. J Pharm Bio111ed Anal 1998; 17:193-218. 86 Peters fl, Drummer OH, Musshoff F. Validation of new
methods. Forensic Sci Int 2007; 165:216-24.
87 Pichini S, Ventura M, Pujadas M, et al. HAIRVEQ: an external quality control scheme for drugs of abuse analysis in hair. Forensic Sci Int 2004; 145:109-15. 88 Wille SM, Lambeli WE. Volatile substance abuse - post
mOliem diagnosis. Forensic Sci Int 2004; 142:135-56.
89 Gallardo E, Queiroz JA. The role of alternative specimens
in toxicological analysis. Biomed Chromatogr 2008; 22:
795-821.
90 Warner A. Interference of common household chemicals in immunoassay methods for drugs of abuse. Ciill Chem 1989; 35:648-51. 91 Jaffee WB, Trucco E, Levy S, Weiss RD. Is this urine really negative? A systematic review of tampering methods in urine drug screening and testing. J Subst Abuse Treat 2007; 33:33-42. 92 Fraser AD, WOlih D. Urinary excretion profiles of 1I-nor-9 carboxy-delta9-tetrahydrocannabinol. Study III. A delta9 THC-COOH to creatinine ratio study. Forensic Sci Int 2003; 137:196-202. 93 Gottshalk LA, Cravey RH. Toxicological and Pathological
Studies on PsychoactilJe Dnlg-inIJolved Deaths. Davis, CA:
Biomedical Publications, 1980.
94 Druid H, Holmgren P. Compilations of therapeutic, toxic, and fatal concentrations of drugs. J Toxicol Clin TO.ricol 1998; 36:133-4; author reply 5-6. 95 Schulz M, Schmoldt A. Therapeutic and toxic blood
concentrations of more than 800 drugs a nd o ther
xenobiotics. Pharmazie 2003; 58:447-74.
96 Blake MJ, Castro L, Leeder J5, Kearns G L Ontogeny of drug metabolizing enzymes in the neo nate. Semill Fetal Neonatal Med 2005; 10:123-38. 97 Blake MJ, Gaedigk A, Pearce RE, et al. Ontogeny of dextromethorphan 0- and N-demethylation in the first year of life. Clin Pharmacol Ther 2007; 81 :510-6. 98 Johnson TN. Modelling approach es to dose estimation in children. Br J C/in Plwrmacol 2005; 59:663 - 9. 99 Johnson TN. The problems in scaling adult drug doses to children. Arch Dis Child 2008; 93:207-11. 100 Walson PD. Therapeutic drug monitoring in special populations. Clin Chem 1998; 44:415-9. 101 Moeller ME, Fey P, Sachs H. Hair anal.ysis as evidence in forensic cases. Forensic Sci In! 1993 ; 63:43-53. 102 Kintz P, Tracqui A, Mangin P. Detection of dnlgs in human hair for clinical and forensic applications. hIt J Legal Med 1992 ; 105:1-4.
103
104
105
106
107
108
109
110
III
112
113
114
115
lJ6
117 118
119
120
121
122 123
277
Sachs H. Forens ic applications of hair analysis. In Kintz P. (ed.) Drug Testing in Hair. Boca Raton, FL: CRC Press. 1996; 211 - 222. Kintz P, Villain M, Cirimele V. Determination of trimeprazine-facilitated sedation in children by hair analysis. J Anal Toxicol 2006; 30:400-2. Springfield AC, Cartmell LW, Aufderheide AC, et a1. Cocaine and metabo lites in the hair of ancient coca leaf chewers. Forensic Sci Int 1993; 63:269-275. Nakahara Y, Kikura R. Hair analysis for drugs of abuse. VII. The incorporation rates of cocaine, benzoylecgonine and ecgonine methyl ester into rat hair and hydrolysis of cocaine in rat hair. ArchilJes Toxicol1994; 68:54-59. Rollins DE, Wilkins DG, Gygi SP, et al. Testing for drugs of abuse in hair. Experimental observations and indications for future research. Forensic Sci Rev 1997; 9:23-36. Pragst F, Balikova MA. State of the art in hair analysis for detection of drug and alcohol abuse. Clin Chim Acta 2006; 370:17-49. Gygi SP, Joseph RE, Cone EJ, et al. Incorporation of codeine and metabolites into hair: role of pigmentation. Drug Metab Dispos 1996 ; 24:495-501. Wilkins DG , Haughey HM , Krueger GG, Rollins DE. Disposition of codeine in female human hair after multiple dose administration. J A/wi Taxieol 1995; 19:492-8. Mieczkowski 1. Assessing the potential of a 'color effect' for hair analysis of 11-nor-9-carboxy-d el ta(9) tetrahydrocannabinol: analysis of a large sample of hair specimens. Life Sci 2003; 74:463-9. Mieczkowski T, Kruger M. Interpret ing the color effect of melanin on cocaine and benzoylecgonine assays for hair analysis: brown and black samp les compared. J Forensic Leg Med 2007; 14:7-15. Cone EJ , Joseph RE. The potential for bias in hair testing for drugs of abuse. In Kintz P (eel.) Drug Testing in Hair. Boca Raton, FL: CRC Press, 1996, pp. 69- 93. Goldberger BA, Caplan YH, Maguire T, Cone EJ. Testing human hair for drugs of abuse. N. Ident ificatio n of heroin and 6-mo noacetylmorphine as indicators of heroine use. J Allal Toxicol 1991; 15:226-31. Harkey MR, Henderson GL, Zhou C. Simultaneous quantitation of cocaine and its major metabolites in human hair by gas chromatography/chemical ionizat ion mass spect rometry. J Anal Toxicol 1991; 15:260-65. Kintz P, Cirimele V, Mangin P. Testing human hair for cannabis. II. Identification ofTHC-COOH by GC-MS-NCI as a unique proof. J Forensic Sci J995; 40:619-22. Henderson GL Mechanisms of drug in corporatio n into hair. Forensic Sci Int 1993; 63:19-29. Henderson GL, Harkey MR, Zhou C, et a l. Incorporation of isotopically-labeled cocaine and metabo lites into human hair. 1. Dose-response relationships. J Analytical Toxicol 199 6; 20:1-12. Nakahara Y, K.i.k.ura R, Takahashi F. Hair analysis for drugs of abuse, X. Effect of physicochemical properties on incorporation rates into hair. Bioi Pharm Bull 1995; 18:1223-7. Naka hara Y, Kikura R. Hair analysis for drugs of abuse. XIII. Effect of struc tural factors on incorporation of drugs into hair: the incorporation rates of amphetamine analogs. Arch Toxicol 1996; 70:841-9. Blank DL, Kidwell DA. External contamination of hair by cocaine: an issue in forensic interpretation. Forensic Sci Int 1993 ; 63: 145-56. Cirimele V, Kintz P, Mangin. P. Drug concentrations in human hair after bleaching. J Anal To.ricol 1995; 19:331-2. Jo seph R, Su TP, Cone EJ. In lIirro binding studies of drugs to hair: influence of melanin and lipids on cocaine binding
278 I
124
125 126
127
128
129
130
131
132 13 3
134
135 136
137
138
139 140 141
142
Recent advances in paediatric toxicology
to caucasoid and africoid hair. J Anal Toxicol 1996; 20:338-44. Hold KJV!, Wilkins DG, Crouch DJ, et al. Derection of stanozolol in hair by negative ion chemical ionization mass spectrometry. J Anal To.ricol 1996; 20:345-49. Chiarotti M, Strano-Rossi S. Preparation of hair samp les for drug ana lysis. Forensi c Sci Rev 1996; 8:111-2 8. Irving RC, Dickson SJ. The detection of sedatives in Ilair and nail sa mpl es using tandem LC-MS-MS. Forensic Sci lilt 2007; 166:58-67. Mangin P, Kintz P. Variability of opiares concentrations in human hair according to their anatomical origin: head, axillary and pubic reg ions. Forensic Sci Jnt 1993; 63:77-83. Offidani C, Strano-Rossi S, Chiarotti M. Dru g distribution in the head, axillary and pubic hair of chronic addicts. Forensic Sci lnt 1993; 63:1105-0. Mangin P. Orug analysis in nonhead hair. In Kintz P (ed.) Drug testing il1 hair. Boca Raton, FL: CRC Press, 1996, pp. 279- 87. Moeller MR, Eser HP. The ana lytical tools for Ilair resting. In Kintz P (ed.) Drug testing ill hair. Boca RalOn, FL: CRC Press, 1996, pp. 95-120. Hegstad S, Khiabani HZ, KJ'istoffersen L, et al. Drug sc ree ning of hair by liquid chromatography-tandem mass spectro merry. J Anal Toxical 2001l; 32:364-72. Tracqui A. Unusu al drugs in hair. In Kintz P (eel.) Drug Testing in Hair. Boca Raton, fL: CRC Press, 1996, pp. 191-210. Idowu OR, Caddy B. A review of the use of sal iva in the forensic detection of drugs and other chemicals. J Forensic Soc 1982; 22:1 23 -235. Caddy B. Saliva as a spec imen for drug analysis. In Baselt RC (ed.) Ad/J in Analytical To.ricology. Vol 1. fo ster City. CA: Biomedical Publi cations, 1984, pp. 198-254. Schramm W, Smith RH, Craig PA, Kidwell DA. Dnlgs of ab use in saliva: a review. J Anal Toxicol 1992; 16:1-9. Dyer KR, Wilkinson C. The detection of illicit drugs in oral fluid: another potential strategy to reduce illicit drug-related harm. Drug Alcohol Re/J 2008; 27:99-107. Dams R, Choo RE, Lambert WE, et al. Oral fluid as an alternative matrix to monilO r opiate and cocaine use in substance-abu se trea tment patients. Drug Alcohol Depend 2007; 87:258-67. Samyn N, Laloup M, De Boeck G. Bioanalytical procedures for determination of drugs of abuse in oral fluid. Anal Bioanal Chem 2007; 388: 1437-53. Cone EJ. Saliva test ing for drugs of abuse. Ann NY ACllri Sci 1993 ; 694:91-127. Kintz P, Cirimele V, Lue\es B. Codeine tesring in sweat and saliva with the drugwipe. lnt J Legal Med 1998; 111 :82- 4. Dams R, Hu estis MA, Lambert WE, et al. Matrix effect in bio-analysis of illicit drugs with LC-MSjMS: influence of ionization type, sample preparati on, and biofluid. J Am Soc Mass Sp ectrom 2003; 14: 1290-4. Kintz P, Tracqui A, Jamey C, Mangin P. Detection of code ine and phenobarbital in swea r collected with a sweat patch. J Anal Toxicol1996; 20:197-201.
143 Cone EJ, Hillsgrove MJ, Jenk.ins AJ, et al. Sweat testing for heroin, cocaine and metabolites. J Alla l Toxicol 1994; 18:298-305. 144 Tracqui A, Kintz P, Ludes B, et al. The detection of opiate drugs in non traditional specimens (clothing): a report of ten cases. J Forensic Sci 1995 ; 40:263-5. 145 Dolan K, Rouen D, Kimber J. An overview of rhe use of urine, hair, sweat and saliva to detect drug use. Drug Alcohol Rev 2004; 23:213-7. 146 Fogerson R, Sc hoendorfer D, Fay J, Spiehler V. Qualitative detection of opiates in sweat by RIA and GC-MS. J Analytical Toxicol 1997; 21 :451-8.
147 Burns M, Basel! RC. Mon itoring drug use with a sweat patch: an experiment with coca ine. J Anal Toxicol 1995; 19:41-8. 148 Kintz P. Brenneisen R, Bundeli P, Mangin P. Swea t testing for heroin and metabolites in a heroin maintenance program. Clin Ch em 1997; 43:736- 9. 149 Moriya F, Chan KJV!, Noguchi n, WU PYK. Testing for drugs of abuse in meconium of newborn infants. J AHal Toxicol 1994; 18:41- 5. 150 Lewis DE, Moore CM, Leikin JB, Koller A. M.econ ium analysis for cocai ne: a va lidarion slUdy a nd comparison with paired urin e ana lysis. J Anal Toxicol 1995; 19:148-50. 151 Oyler J, Darwin WD, Preston KL, et al. Coca ine disposition in meconium from newborns of cocaine-abusing mothers and urine of adult drug users. J Anlll Toxicol 1996; 20:453-62. 152 Moore C. Negrusz A, Lewis D. Determination of drugs of abuse in meco nium. J Chrol11lltogr B Biol11ed Sci Appl 1998; 713: 137-46. 153 Gareri J, Klein J, Koren G. Drugs of abuse testing in meconium. Clill Chim Acta 2006; 366:101 -11. 154 Gareri J, Klein J, Koren G, et al. Drugs of abuse testing in meconium. Cli n Chim Acta 2006; 366:101-11. J 55 Browne S, Moore C. Negrusz A, et al. Detection of cocaine, l1orcocaine and cocaethylene in the meconium of premature neonates. J Forensic Sci 1994; 39:1515-19. 156 Moore C, Lewis D, Becker J, Leikin J. The determinarion of II-nor-69-tetrahydrocannabinol-9-carboxylic acid (THC COOH) in meconium. J Anal Toxicol 1996; 20:50-1. 157 Moore CM, Lewis DE, Leikin JB. The determination of phen cyclidine in meconium using ion trap mass spectromelly. J Forensic Sci 1996; 41: 1057-9. 150 Clark GD , Rosenzweig TB, Ra isys VA, et al. The analysis of cocaine and benzoyJecgonine in meconium. J Anal TO.ricol 1992 ; 16:261-3. 159 Stolk LM, Coenradie SM, Smit BJ. Analysis of methado ne and its primary metabolite in meconium. J Anal Toxicol 1997 ; 21:154- 9. 160 Dempsey D, Moore C, Deitermann D, et al. The detection of cotinin e in hydrolyzed meconium samples. Forensic Sci Jnt 1999 ; 102:167-71. 161 Le NL, Reiter A, Toml illson K, er al. The detection of oxycoclo ne in meconium specimens. J Anlll To.ricoI2005; 29: 54-7. 162 Williamson S, Jackson L, Skeoch C. et al. Determination of the prevalence of drug misuse by meconium ana lysis. Arch Dis Ch ild Fetal Neonatal Ed 2006; 91 :F291-2. 163 Xia Y, Wang P, Banlett MG, er al. An LC-MS - MS method for the co mprehensive analysis of cocaine and cocaine metabolites in meconi um . Anal Chem 2000; 72:764-71. 164 Kelly T, Gray TR, Hu estis MA. Development and va lidation of a liquid chromatography-atmospheric pressure chemical ionization-tandem mass spectrometry merhod for simultaneous ana lysis of 10 amphetamine-, methamphetamine- and 3,4-methylenedioxym ethamp hetamine-related (MDMA) ana lytes in human meconium. J Chrol11a togr BAnal Technol Biomed Life Sci 2001l; 867: 194-204. 165 Moo re CM, Lewis D. fatty acid ethyl esters in meconium: biomarkers for the detect ion of alcohol exposure in neonates. Clin Chim Acta 2001; 312:235-7. 166 Moore C, Jones J, Lewis D, Buchi K. Prevalence of fatty acid ethy l esters in meconium specimens. Clin Che l11 2003; 49: 133-6. 167 Ripple M, Goldberger EA, Caplan YH, et al. Detection of cocaine and its metabolites in human amniotic fluid . J A lla l To.ricoI1992; 16: 328 -31.
168 Moore C, Browne S, Tebbett I, et al. Determination of coca ine and benzoylecgonine in human amn iotic fluid using high flow solid-phase exrraction columns and HPLC. Forensic Sci Jnt 1992; 56:177-81.
References I
169 Winecker RE, Goldberger BA, Tebbett 1, et al. Detection of cocaine and its metabolites in amniotic f1uid and umbilical cord tissue. J Ana! To.rico!1997; 21:97~104. 170 Gray T, Huestis M. Bioanalytical procedures for monitoring in litera drug exposure. Ana! Bioanal Chem 2007; 388:1455~65. 171 Kugelberg FC, Jones AW. Interpreting results of ethanol analysis in postmortem specimens: a review of the literature. Forensic Sci Int 2007; 165:1O~29. 172 Musshoff F, Madea B. Analytical pitfalls in hair testing. Anal Bioana! Chem 2007; 388:1475~94. 173 Liu RH. Comparison of common immunoassay kits for effective application in workplace drug analysis. Forensic Sci Rev 1994; 6:20~57. 174 Green K, lsenschmid DS. Medical review officer interpretation of urine drug test results. Forensic Sci Rev 1995; 7 :42~60. 175 Cone EJ, Johnson RE, Darwin WD, et al. Passive inhalation of marijuana smoke: urinalysis and room air levels of delta9-tetrahydrocannabinol. J Anal To.rico! 1987; 11 :89~96. 176 Kidwell DA, Blank DL. Environmental exposure ~ the stumbling block of hair testing. In Kintz P (ed.) Drug testing in hair. Boca Raton, FL: CRC Press, 1996, pp. 17~68. 177 Haley NJ, Hoffmann D. Analysis for nicotine and cotinine in hair to determine cigarette smoker status. Clin Chem 1985; 31:1598~1600.
178 179
Kintz P, Ludes B, Mangin P. Evaluation of nicotine and cotinine in human hair. J Forensic Sci 1992; 37:72~6. Smith FP, Kidwell DA. Cocaine in hair, saliva, skin swabs and urine of cocaine users' children. Forensic Sci lilt 1996; 83:179~189.
180 Blank DL, Kidwell DA. Decontamination procedures for drugs of abuse in hair: are they sufficient? Forensic Sci Int 1995; 70:13~38. 181 Davis KH, Hawks RL, Blanke RV. Assessment of laboratory quality in urine drug testing ~ a proficiency testing pilot study. JAIVIA 1988; 260: 1749~54. 182 Kelly K. The accuracy and reliability of tests for drugs of abuse in urine samples. Pharmacotherapy 1988; 8:263~75. 183 Badia R, Segura J, Artola A, de la Torre R. Survey on drugs of-abuse testing in the European Union. J Ana! To.ricol 1998; 22: 117~26. 184 McRorie TI, Lynn AM, Nespeca MK, et al. The maturation of morphine clearance and metabolism. Ani J Dis Child 1992; 146:972~6.
185 Furlanut M, Benetello P, Baraldo M, et al. Chlorpromazine disposition in relation to age in children. Clin Pharmacokinetics 1990; 18:329~3J. 186 Mikkelsen SL, Ash KD. Adulterants causing false negatives in illicit drug testing. Clin Chem 1988; 34:2333~6. 187 Cody JI, Schwarzhoff H. Impact of adulterants on RIA analysis of urine for drugs of abuse. J Ana! Toxico! 1989; 13:277~284.
188 Schwarzhoff R, Cody JT. The effects of adulterating agents on FPlA analysis of urine for drugs of abuse. J Anal Toxicol 1993; 17:14~17. 189 Soper JW, Canfield V. Description of a screening assay system to detect adulteration of urine samples. J Anal Toxicol 1996; 20:77. 190 Goldberger BA, Caplan YH. Effect of glutaraldehyde (UrinAid) on detection of abused drugs in urine by immunoassay. Clin Chem 1994; 40: I 605. 191 Cody JT, Valtier S. Effects of Stealth adulterant on immunoassay testing for drugs of abuse. J Anal To.ricol 2001; 25:466~70. 192 Burrows DL, Nicolaides A, Rice P J, et al. Papain: a novel urine adulterant. J Anal To.ricoI2005; 29:275~95. 193 Pearce MR, Tarnai.LD. Performance evaluation of three on-site adulterant detection devices for urine specimens. J Anal Toxico! 2002; 26:464~70.
279
194 Shepherd MF, Lake KD, Kamps MA. Postmortem changes and pharmacokinetics: review of the literature and case report. Ann Pharmacother 1992; 26:51O~14. 195 Hilberg T, Ragde S, Morland J. Postmortem drug redistribution ~ human cases related to results in experimental animals. J Forensic Sci 1999; 44:3~9. 196 Pelissier-Alicot AL, Gaulier JM, Champsaur P, Marquet P. Mechanisms underlying postmortem redistribution of drugs: a review. J Anal To.ricol 2003; 2 7 :533~44. 198 Drummer OH. Postmortem toxicology of drugs of abuse. Forensic Sci Int 2004; 142:!oI~J3. 198 Skopp G. Preanalytic aspects in postmortem toxicology. Forensic Sci Int 2004; 142:75~!OO. 199 Yarema MC, Becker CEo Key concepts in postmortem drug redistribution. C!il1 Toxicol (Phila) 2005; 43:235~41. 200 Drummer OH. Post-mortem toxicology. Forensic Sci Int 2007; 165:199~203. 201 Pounder DJ, Jones GR. Post-mortem drug redistribution ~ a toxicological nightmare. Forensic Sci Int 1990; 45:253~63. 202 Gomez HF, McKinney P, Philipps S, et al. Postmortem acetaminophen pharmacokinetics: an experimental study of site and time-dependent concentration changes. J Forensic Sci 1995; 40:980~2. 203 Hilberg T, Morland J, Bjorneboe A. Postmortem release of amitriptyline from the lungs: a mechanism of postmortem redistribution. Forensic Sci Int 1994; 64:47~55. 204 Koren G, KJein J. Postmortem redistribution of morphine in rats. Ther Drug Manit 1992; 14:461~3. 205 Koren G, MacLeod SM. Postmortem redistribution of digoxin in rats. J Forensic Sci 1985; 30:92~6. 206 Quatrehomme G, Bourre F, Liao Z, Ollier A. An experimental methodology for the study of postmortem changes in toxic concentrations of drugs using secobarbital as an example. J Forensic Sci 1994; 39:1300~4. 207 Imamura T, Nagat T, Kimura K, et al. Pharmacokinetics and postmortem changes of cimetidine in body tissues. Nippon Hoigaku Zasski 1994; 48:75~78. 208 Aderjan R, Mattern R. Validity of digoxin concentrations in blood determined postmortem. ZeitschriJl Jiir Reci1tsmedizin 1980; 86: !3~20. 209 Hearn WL, Keran EE, Wei HA, Hime G. Site-dependent postmortem changes in blood cocaine concentrations. J Forensic Sci 1991; 36:673~84. 210 Levine 8, Wu SC, Dixon A, Smialek JE. Site dependence of postmortem blood methadone concentrations. Am J Forensic
Med Patho11995;
16:97~!oO.
211
Miyazaki T, Kojima J, Yashiki M, et al. Site dependence of methamphetamine concentrations in blood samples collected from cadavers of people who had been methamphetamine users. Am J Forensic Med Patho11993; 14:121~4. 212 Kuhlman JJ, Mayers RW, Levine B, et al. Chloroquine distribution in postmortem cases. J Forensic Sci 1991; 36: 1572~9. 213 Pounder DJ, Hartley AK, Watmough P J. Postmortem redistribution and degradation of dothiepin: human case studies and an animal model. Am J Forensic jl;Ied Pathol 1994; 15:231~5. 214 Pounder DJ, Owen V, Quigley C. Postmortem changes in blood amitriptyline concentration. Am J Patho! 1994; 15:224~30.
215 Pohland RC, Bernhard NR. Postmoliem serum and tissue redistribution of f1uoxetine and norf1uoxetine in dogs following oral administration of f1uoxetine hydrochloride (Prozac"). J Forensic Sci 1997; 42 :812~ 16. 2J 6 Yonemitsu K, Pounder DJ. Postmortem changes in blood tranylcypromine concentration: competing redistribution and degradation effects. Forensic Sci Int 1993; 59: 177~84.
280 I
Recent advances in paediatric toxicology
217 Yonemitsu K, Pounder DJ. Postmortem toxicokinetics of co-proxamol. Jnt J Legal Med 1992; 104:347-53. 218 Hanzlick R. Postmortem tricyclic antid epressa nt concentrations: letha l versus nonlethal levels. Am J Forensic Med Pathol 1989; 10:326-9. 219 Martin A, Pounder DJ. Postmortem toxicokinetics of trazadone. Forensic Sci ll7t 1992; 56:201-7. 220 Pounder DJ, Davies JI. Zopiclone poisoning: tissue distribution and potentia l for postmoltem diffusion. Forensic Sci Jnt 1994; 65:177-83. 22 1 Hilberg T, Bugge A, Beykich KJV!, et at. Diffusion as a mechanism of postmortem drug redistribution: an experimental study in rats. lut J Leg Med 1992; 105:87-91. 222 Pounder DJ, Fuke C, Cox DE, et al. Postmortem diffusion of drugs from gastric residue. Am J Forensic Med Pat/IOI 1996; 17: 1-7. 223 Pounder DJ, Smith RW. Postmortem diffusion of alcohol from the stomach. Am J Forensic Med Pathol 1995; 16:89-96. 224 Skopp G, Potsch L, Kon ig I, Mattern R. A preliminalY study on the stab ili ty of benzod iazepi nes in blood and plasma stored at 4' C.lnr J Leg Med 1998; 111:1-5. 225 Mayer W, Erbe S, WolfG, Voigt R. Beitrage zur Analytik und Stabilitat einiger pharmazeutisch interessanter 1,4 Benzodiazepine. Pharl1lazie 1974; 29: 700-7. 226 Koves EM, Lawrence K, Mayer JM. Stability of diltiazem in w ho le blood: forensic implications. J Forensic Sci 1997; 43:587- 97. 227 Benhamou-Batut F, Demotes-Mainard F, Labat L, Vincon G. Determination of flunitrazepam in plasma by liquid chromatography. J Pharm Biomed Anal 1994 ; 12:93 1- 6. 228 Augsburger M, Mangin P. LSD, Ie phenix des hallucinogenes. Toxicorama 1998; 10:61-6. 229 Ferner RE. Forensic Pharmacology. Oxford, NF: Oxford University Press, 1996. 230 Paterson S. Drugs and decomposition. Med, Sci and Law 1993; 33: 103- 109. 231 Hadidi FA, Ol iver JS. Stability of morphine and buprenorphine in whole blood.lnt J Leg Med 1998; 111:165-7. 232 McCurdy HH, Callahan LS, Williams RD. Studies on the stability and detection of cocaine, benzoylecgonine and I I -nor-69-tetrahydrocannabinol-9-carboxylic acid in who le blood using Abuscreen radioimmunoassay. J Forensic Sci 1989; 34:858-70. 233 Robertson MD , Drummer OH. Postmortem drug metabolism by bacteria. J Forensic Sci 1995 ; 40:382-6. 234 Robertson MD, Drummer OH. Stability of nitrobenzodiazepines in postmortem blood. J Forensic Sci 1998; 43 :5- 8. 235 Robertson MD, Drummer OH. Postmortem distribution and redistribution of nitrobenzodiazepines in man. J Forensic Sci 1990; 43:9-13. 236 Corry JEL. Possible sources of ethanol ante- and postmortem: its relationship to the biochemi stry and microbiology of decomposition. J App! Bacterial 1978; 44: 1-56. 237 Hoiseth G, Kristoffersen L, Larssen B, et al. In uitro formation of ethanol in autopsy samples containing fluoride ions. lnt J Legal Med 2008; 122:63-6. 238 Schm itt G, Adeljan R, Keller T, Wu M. Ethy l g lu curonide: an unusual ethan ol metabolite in hum ans. Synthesis, analytical data, a nd determination in serum and urine. J Anal Toxicol 1995 ; 19:9 /-4. 239 Bicker W, Lammerhofer M, Keller T, et a l. Validated method for the determination of the ethanol consumption markers ethyl glucuronide, ethyl phosphate, and ethyl sulfate in human urine by reversed-phase/weak anion exchange liquid
240
241 242
243
244 245
246 247
248 249 250
251
252
253
254
255
256
257 258
259
260
261
262
chromatography-tandem mass spectrometly. Anal Chem 2006; 78:5 884-92. Cap lan YH, Levine B. Vitreous humor in the eva luation of postmoltem blood ethanol concentrations. J Anal To.ricol 1990; 14:305-7. Pounder DJ, Kuroda N. Vitreous alcohol is of limi ted value in predicting blood alcohol. Forensic Sci lnt 1994; 65:73-80. Laure P, Lecerf T, Friser A, Binsinger C. Drugs, recreational drug use and attitudes towards doping of high school athl etes. Jnt J Sports lVIed 2004; 25:133 - 8. Laure P, Binsinger C. Doping prevalence among preadolescent athletes: a 4-year follow-up. Br J Sports Med 2007 ; 41 :660-3; discussion 3. Sa ugy M, Avois L, Saudan C, et al. Cannabis and sport. Br J Sports Med 2006; 40 (Suppl 1):i13-5. Bowers LD. Analytical advances in detection of performance-enhancing compounds. Clin Chem 1997; 43: 1299-304. Cowan DA, Kickman AT. Doping in sport: misuse, analytical tests, and legal aspects. Clin Chem 1997; 43:1110-13. Thevis M, Schanzer W. Mass spectrometry in sports drug testing. Structure characterization and analytica l assays. Mass Spectrom Rev 2007; 26:79-107. Saudan C, Baume N, Robinson N, et al. Testosterone and doping control. Br J Sports Med 2006; 40 (Suppl l):i21-4. Dawson RT. Drugs in sport - the role of the physician. J Endocl'illo/ 2001; 170:5 5-61. Sjoqvist F, Garle M, Rane A. Use of dop in g agents , palticularly anabolic steroids, in sports and society. Lancet 2008; 37 1:1872-82. Musshoff F, Daldrup T, Ritsch M. Black market in a nabolic steroids - analysis of illegally distributed products. J Forensic Sci 1997; 42:1119 - 25. Catlin DH, Hatton CK, Starcevic SH. Issues in detecting abuse of xenobiotic anabolic steroids and testosterone by analysis of athletes' urine. Clin Chem 1997; 43: 1200-8. Thevis M, Thomas A, Schanzer W. Mass spectrometric determination of insulin s and their degradation products in sports drug testing. Mass Spectrom Rell 2008; 27:35-50. Huestis MA. Technical and legal aspects of drugs of abuse testing in hair. In Nntz P (ed.) Drug testing ill hair. Boca Raton, FL: CRC Press , 1996 , pp. 5-15. Lewis D, Moore C, Morrissey p, Leikin J. Determination of drug exposure using hair: application to child protective cases. Forensic Sci Jnt 1997; 84:123-8. Strano-Rossi S, Offidani C, Chiaroll i M. Application of hair analysis to document coercive heroin administra tion to a child. J Ana/ Toxico/1998; 22: 75-7 . Xiang P, Shen M, ZlllI o X. Hair analysis for ketam ine and its metabolites. Forensic Sci lnt 2006; 162: 131-4. Marchei E, Munoz JA, Garcia-AJgar 0, et al. Development and validation of a liquid chromatography-mass spectrometry assay for hair analySis of methylphenidate. Forensic Sci Tnt 2008; 176:42-6. Kintz P, Villain M, Tracqui A, et al. Buprenorphine in drug facilitated sexual abuse: a fatal case involving a 14-year-old boy. J Anal Toxicol2003; 27:527-9. Negrusz A, Gaensslen RE. Analytica l developments in toxicological investigation of drug-facilitated sexual assault. Anal Bioanal Chem 2003; 376:1192-7. Cheze M, Villain M, Pepin G. Determination of bromazepam, c!onazepam and metabolites after a single intake in urine and ha ir by LC-MS/ MS. Application to forensic cases of drug facilitated crimes. Forensic Sci Jil t 2004; 145:123-30. Deveaux M, Cheze M, Pepin G. The role of liquid chromatography-tandem mass spectro metry (LC-MS/MS) to test blood and urine samp les for the tox icologica l
References I
263
264
26 5 266
267
268
269
270
271
272
273
in ves tigation of drug-facilitated crimes. Ther Drug Manit 2008; 30:225-8. Rey-Sa lmon C, pepin G. (Drug-facilitated crime and sexual abuse: a pediatric observation]. Arch Pediatr 2007; 14: 13J 8-20. Kintz P, Evans J, Vill a in M, et a1. Hair analysis for diphenhyd ramine after surreptitious admin istration to a child. Forensic Sci 1nt 2007; 173: 171-4. Kintz P, Villain M, Cirimele V. Hair ana lysis for drug detection. Thn Drug Manit 2006 ; 28:442-6. Kin tz P, Vi llain M, Dumestre-Toulet V, et a J. Methadone as a chemical weapo n: two fatal cases involv ing babies. TIler Drug Manit 2005; 27:741-3. Kintz p, Cirimele V, Jamey C, Ludes B. Testing for GHB in hair by GCjMSjMS after a single ex posure. Application to document sexual assaul t.] Forensic Sci 2003; 48:19 5-200. Ostrea EM, Chavez CS. Perinatal problems (excluding neonatal withdrawal) in materna l drug addiction : a study of 830 cases.] Pediatr 1979 ; 94:292 - 5. Ostrea EM, Brady M, Gause S, et a J. Drug sc reen in g in newbo rns by meconium ana lysis: a large-scale prospective, epid emiologic study. ] Pediatr 1992; 89: 107 - 13. Klein J, Koren G. Hair analysis - a biological marker for passive smok ing in pregnan cy and childhood. Hum Exp Toxicol 1999; 18 :279-8 2. Kukla L, Hruba D, Tyrlik M. Trends in respiratory morbidi ty of children in relation to their passive smoking exposure. Cent Eur] Public Health 2006; 14:180- 5. Bar-Oz B, Klein J , Karasko v T, Koren G. Comparison of meco nium and neonatal hair ana lysis for detection of gestational exposure to drugs of abuse. Arch Dis Ci1ild Fetal Neonata l Ed 2003; 88:F98-FIOO. Garcia-BoUl'nissen F, Rokach B, Karaskov T, Koren G. Methamphetamine detection in maternal and neonatal hair: impli cations for fetal safety. Arch Dis Child Fetal Neol1a tal Ed 2007; 92.F35 1-5.
281
274 Garcia-Bournissen F, Rokach B, Karaskov T, Koren G. Cocaine detection in maternal and neonatal hair: implic ations to fetal toxicology. Ther Drug Manit 2007; 29:71- 6. 27 5 Pragst F, Yegles M. Determination of fatty ac id ethyl esters (FAEE) and ethyl glucuronide (EtG) in hair: a promising way for retrospective detection of alcohol abuse during pregnan cy? Ther Drug Manit 2008; 30:2 55-63. 276 Moore C, Dempsey D, Deitermann D, et aJ. Feta l cocaine exposure: analysis of vernix caseosa. ] Anal Toxicol 1996; 20:509-11. 277 Marks V, Richmond C. Beverly Allitt: the nurse who killed babies.] R Soc Med 20 08; 101 :110-5. 278 Dyer C. Children's nurse convicted of murd er. EM] 1993; 306:1 43 1-2. 279 Appleyard WJ. Murder in the NHS. EM] 1994; 308:287-8. 280 Repper J. Munchausen syndrome by proxy in health care workers.] Adll Nlirs 1995; 2 1 :299-304. 28 1 Jones DPH, Boots CN. Factitious lJIness by Proxy. In David TJ (ed.) Recel1t AdIJances in Paediatrics J 7. London : Churchi ll Li vingstone, 1999, pp. 57-71 282 Jenkins RO, Craig P J , Goessler W, Irgolic KJ. Antimony leaching from cot mattresses and sudden infant death syndrome (SJDS). Hum E.rp Toxicol 1998; 17 :138- 9. 283 Richardson BA. Cot mattress biodeterioration and SJDS. Lal1cet 19 90; 335:670. 284 Richardson BA. Sudden infant death syndrome: a possible primary cause. ] Forensic Sci Soc 1994; 34: 199- 204. 285 Department of Health. Expert Group to In ves tigate Cot Delllh Theories: To.ric Gas Hypothesis. Chairman, Lady Limerick. Final report. London: Department of Health, 1998. 286 Report of the expert working group enquiring into the hypothesis that toxic gases evolved from matress covers and cot matresses are a ca use of SIDS. Sudden infallt death syndrome (STD S) London: Her Majesty's Sta tionery Office, 1991.
I
CHAPTER 14
I
HEAD AND NECK INJURIES
Robert A Minns and TY Milly Lo
Definition Epidemiology Non-accidental head injury Traumatic birth injury Primary mechanisms of injury to the brain
282 283 294 300 302
DEFINITION
Head injury can be defined as the injury resulting from an external force to the head by penetrating, compression or impact forces, causing damage to the scalp, skull or brain (traumatic brain injury). If brain injury results, there is usu ally an impairment of consciousness and a period of retro grade amnesia. Not all cases of brain damage are associated; however, with a loss of consciousness (e.g. some frontal lobe injuries, extradural haematomas and some penetrating an d compression injuri es) but, conversely, all cases of loss of consciousness indi cate brain injury. Traumatic brain injury in children may result from traumatic birth injury, accidental head injury and non accidental head injury (NAHI). The Gl asg ow Co ma Scale (GCS) developed by Teasdale and Jennett i in 1974 has become a widely used clinical tool for assessment of t he level of co nsciousness in adults. Furthermore, the GCS enab les a basa l initial sco re to be estab lished and thus allows for comparison between suc cessive examinations of the level of consciousness identi fying any clinical deterioration. Many of the responses assessed in the GCS requ ire an adult level of neurodevelop mental function and ca nnot be easily graded in children who are under 10 years of age. 2 Modifications of the GCS or other coma scales that include age- adjusted verbal and motor responses have been described and used to assess the level of consciousness in .chi ldren with neurodevelopm en tal levels who are und er 10 years of age. 2- 8
Secondary mecha nisms of brain injury Injury to the cervical spinal cord Genetic influence on recovery from traumatic brain injury References
307 311 312 313
Yager et al 9 assessed the interobserver variability for six different coma scales used in children an d found tha t the paediatric coma scale had the highest interobserver agreement. In an attempt to standardi ze the way coma is measured in children, the British Paediatric Neurology Association (BPNA) has recommended the Jam es ad apta tion of the GCS as it takes into account the developmental immaturity in small chi ldren, uses the sa me number of points irrespective of the child 's age, and is simple to use. As the co nsciousn ess level decreases, vital mechanisms such as airway protective reflexes (gag, cough and swa llow reflexes) may be compromised, m aki ng the patient vulner able to aspiration 10 and secondary hypoxic ischaem ic injury. The Ad va nced Paediatric Life Suppo rt Guideline suggests that all head-injured chi ldren with a coma score of less than 8 should have endotracheal intubation (to pro tect their airways) and ventilatory suppOl1. Children wi th mild (G CS 13-15) to moderate (GCS 9-12) head injury may have impaired gag reflexes and wou ld also be at risk of developing aspi ration, and may also require intubation and mechanical ventilation. In intubated children, a grimace score has been sug gested to replace the ve rb al component of the GCS assess ment. II This modified paediatric coma sca le for intensive care practice origin ated from the Sharples' adaptation of the James ' adaptation of the GCS. Non-verbal communica tion and verbal language are not totally independent, and an important part of non-verbal commu nication invo lves facial expression and grim ace. Intero bserver reliability was
Epidemiology I
moderate to good, with all components assessed with the grimace score being better than the verbal score in all of the 73 children studied; this suggested that the grimace score may be more useful in intubated patients. II The most common cla ssification for se1!erity of head i njLI1Y in adults is by means of the GCS, I in which a score of 13-15 is designated mild, a score of 9-12 is designa ted moderate and a sco re of 3-8 is designated severe. Modifica tions of the GCS for children are similarly used to assign severity. The duration of unconsciousness or post-traumatic amnesia have also been used to classify severi ty. Post traumatic amnesia may be four times as long as the duration of coma . Severity based on imaging is useful only for severe injury. Therefore, a GCS, which is reall y a mini neurological exa mination , remains the most useful index of severity of brain injury.
EPIDEMIOLOGY Accidental Head Injury Limitations to comparing literature data on head injury include: 1. The interna tional Classification of Diseases (lCD), 9th edition, covers head injury with 10 codes that are not mutually exclusive. 2. The severity of inju ry is difficult to identify from ICD codes with no reference to impairment of consciousness. 3. Application of these codes may be different in different hospitals a nd countries, and coding errors are frequent. 4. Head-injured patients a re frequently transferred to other hospitals, causing difficulties in defining a catchment population. 5. Reports from pathologists or from trauma units do not usually allow denominator identifica tion , thus calcul ation of incidence varies depending on whether the numerator is deaths, admissions or atte ndances. Admission and triage policies vary widely from hospital to hospital. Deaths from head injury at all ages have been falling since 1968 12 in the United Kingdom and in the Uni ted States. 13 In most countries, the peak incid ence is in males in the 15- to 30-year age group. Age-specific admiss ion rates are similar in Britain and in the USA and an overall estimate is 200-300 per 100000 ; 80 per cent are catego rized as mild and 5-10 per cent as severe. The literature perta ining to the incid ence of accid ental head i~UJy in children is seen in Table 14.1, which sum marizes th e nature of the study, the study period, t he age group, inclusion criteria, mea n incidence and mortality rate, along with the main causes for traumatic brain injury (TBI). In Scotland, the incidence of severe TB[ is between 14.5 and 29 per 100000 child ren per year. There are more
283
ARE presentations due to head injury in children less than 10 years than in any other age group. I 8 Head injury is the co mmon est ca use of death in children aged 1-15 years in England and Wales. 41 The mean mortal ity from head injury in children calculated from 18 pub lished studies was 5.1 ± 2.39 per 100000 children per year.9.14-1 6, 18,19,21,24-32,34-39 Although accidental lIlJury mortality rates have fallen in the UK, groups of children who are at greater relative risk remain within the popula tion, such as children residing in less affluent areas. 31,41 [n Scotland, accidents to pedestrians were the leadin g cause of head injury mortality in children. Sample statistics for children with head injury attending ARE at the Royal Hospital for Sick Children in Edinburgh, over two I-year periods, are seen in Fig. 14.1. 42 The peak age group was between I and 2 years of age, with a steady decline in attendance thereafter up to 13 years. Falls accounted for 53 per cent of all attendances, with fails of more than I metre in height accou nting for 12 per cent. Road traffic accidents were responsible for 4.6 per cent of attendances; the remain der (42 per cent) was due to miscellaneous causes.
Non-accidental Head Injury Non-accidental brain injury refers to inflicted TBI (traumatic en cephalopathy), usually by an adult, to infants and young children, sustained as a resul t of deliberate impact (accelera tion or deceleration), head compression, penetrating head injury, repetitive rotational injury (shaken baby syndrome), rotation and impact (shaken impact syndrome) or whiplash (cervicomedullary syndrome). Anyone mechanism or com bination of mechanisms may be present in the individual child. The term 'shaken baby syndrome' has often been used as a generic term for NAHI, and this has caused confusion in lega l proceedings as it implies shaldng as the cause of all NAHls. The term 'non-accidental head injury' is to be pre ferred, as no mechanism of injury is thereby infened. The limitatio ns to obtaining accurate data on NAHI are greater than those for accidental hea d injury data, because of the follo wing factors : diagnostic difficulties, protracted legal processes and outcome, and the mobile nature of this population, making it difficult to asce rtain true, proven, numbers of cases of NAHI. It is essential that information pertaining to suspected cases is not accepted as proven unless its no n -accidental origin has been acknowledged or ha s resulted in a criminal conviction 43 From July 1998 until December 1999, 19 new cases of suspected NAHI were identified in Scotland (12 boys a nd seven girls), giving an annual incidence of 24. 6 per 100000 children under than 1 year of age (95 per cent confidence intervals, 14.9- 38.5). The median age at acute admission was 2.2 months (range 4 weeks to 8.8 months). No child was older than 1 year and 75 per cent of the children were admitted during the autumn and winter months. A higher incidence was found in the urba n areas of greater Glasgow and
Ta bl e 14. 1
Studies of the incidence of accidental traumatic brain injury (TBI) in children
Reference
Country/region
Study
Study period
Age group
Inclusion criteria
Mean incidence
Mortality
Main causes
Main ca use of TBI
for TBI
ofTBI
rate
ofTBI (Ofo)
mortality
----
-
Annegers et al
Olmsted County.
Retrospective
Al l ages,
Evidence of brain
Males 270
1980 14
Overall 35.0
<5 years - falls
Minneso ta, USA
Medical
age-
trauma
Females 116
(age-adjusted)
Gi rls 5-14 yea rs
linkage system
adju sted
Rantaka lJ io,
Finland
von Wend t, 1985 15 Kraus et aI
1965-1974
Prospective
Co ncussion with
From graph: all
From graph:
horse riding,
LaC, PTA,
0-5 years, 180;
0-5 years, 8.0;
cycling
neurologica l signs
boys 5-15 years,
5-15 yea rs, 8.0
All boys 21
Sku II fracture
310;
All girl s 11
girls 5-15 years, 160
Main ly ages 5-14 years
0-14 yea rs
~ 161years
birth cohort California, USA
1986 16
235 cumulative
Consecutive
Children
TBI
0-15 years
Admitted to hospital
13 per cent 'serious
with trauma. GCS, TS,
injuries' with ISS ;;;' 16
180
Review of several cross-sectional studies
Cha n et al
Sydney, Australia
Retrospective
January 198 5
1989 "
(W Metropolitan
Paediatric
to December
Region)
Trauma Study
1986
2
Not given
Pedestrians, MVC occupants, cyclists
ISS noted
90 per cent
Se rious: TS ". 12
involved cranium
Adults,
Evidence of brain
Admission rate 4011
Children constitu te: 40
Children
damage: altered
Admission with brain
50% of ARE attenders,
Serious TB I 17 Brookes et al
Scotland, UK
1990 18
Retrospective
198411985
Case record s for 23 Scottish
< 5 yea rs,
consciousness on
injury 290
31 % of admissions,
ARE
5-11 yea rs,
arriva l or history
Most seve re 5-10
departments
12-14 years
of altered
per ce nt
25% of neurosurgical unit tra nsfers, 200/0 of
consciousness and
14.5-29
MVC 71%
death s on neurosurgical un it, 11 0/0 of operations
am nesia on arrival
for ICH, 750/0 of severe injuri es in children are MVC Yager et al 19909
Virginia, USA
Sha"rpl es et al
N Engl and, UK
1990 19
1978
All ages
Retrospective
1979-1986.
< 16 yea rs
review of HI
in c.
deaths
175 all
5.3 chi ldren
All deaths from head
5.3
HI was major sing le
injury in child re n ICD
Mortality
cause of deaths in
9 codes N800- NB04,
significa ntly related
N850-N854
to depriva tion index:
children (1 5% ) 250/0 deaths 5-1 5 yea rs
HI accidents
53 % of 25 5 fatal
14.0 with greatest
49% di ed before admission
occurred during
deprivation, 0.9 with
Most deaths (76% ) due
play
least deprivation
to MVC: 53% pedestrians, 140/0 cyclists, 100/0
(P < 0.00001)
passengers, 120/135 pedestrian deaths from unsafe child beh avio ur
lepd> cL al
U)A
1990 20
Helrospeclive
Not given
National Pediatri c
44Ofo ha ve HI
Mortality was
Fall s
lower in children
MVC, ped estrian /
Trauma database
Trauma Regi ster,
of children
identified head injury
occupant 3.10/0 of 10098
alone, head plu s
TBI admissions
extracranial inju ry, only and extracranial injury Hansen et al
Denmark,
1991"
Ringkobin g county
0-14 yea rs
Severe TBI
18
Levin et al
USA:
Retrospective,
14 April 1983
,,"15 years
Severe traumati c head
6 months po st
0-4 years: 31 % fall/ju mp;
1992 22
San Diego, affluent
hospital-based
t o 18 April
banded :
inju ry with in 48
injury mortality
26 0/0 pedestrian; 20%
1988
0-4 years
hours of
approximatel y
motor vehicle occupant;
Virginia, inner-city,
urban area;
5-10years
admission,
52 0/0 0-4 years;
17 0/0 assault
suburban , rural ;
11-15
GCS"" 8
180/0 5-10 years;
5-10 years: 49 0/0 motor
Texas, large urban,
years
Gunshot wounds
22Ofo /1-15 years
veh icle occupant; 26%
rural beach resort;
Not give n
pedestrian; 130/0 fall/jump;
excluded
Houston, urba n;
9 0/0 cyclist
New York, inner-city
11 15 years: 55 0/0 MVC occupant; 14% pedestrian ;
12 010 motorcycle; 11% bi cyc le; 4% fall/jump Ald rich et al
USA
Retrospective
1992"
Berney et al
January 1984
""1 6 years
Severe traumati c
to September
HI with diffu se brain
1987
swell ing
Switzerland
;;' 1 hour Lo C
19942 •
53% died
15.2
6.8
Cli nica I/i magi ng evidence of brain contusion Exclud es: concussion, mild HI
Arnarson,
Iceland
1987 -1991
0-14 years
Hal dorsson
Hospita liza ti on
170
fo r HI, all severities
199525 Jennett
Scotlan d, UK
Retrospective
1985
A&E attenders
199626
Attend ers 4011,
5.3
UK age-specific death and adm ission rates
admissions 400, evidence of brain
peak at 15-30 yea rs
damage 290
A&E attenders peak at < 10 years
Gabella et al 1997 27
Colorado, USA
MVC mortality :
TBI defined from ICD
Girls
Age -adjusted
study, residents
9 codes: 800 - 80 1,
0-4 years: 70;
mortality rate s
urban 14.8, rural
only
803-804 (skull
5-14 yea rs:
provided
25.5
fracture), 850-854
60 all areas
Surveillance
1991-1992
All ag es
(intracranial injury)
Boys
Data from hosp ital
0-4 years:
discharge records
20 remote ru ra I,
and death ce rtificates
Causes by age not given
100 all other 5-14 yea rs: 90 metropolitan, 160 rural areas
(Con tinued)
Table 14.1 (Continued) Reference L
Country/region
Study
Eman uelson,
Study period
---
____ . _
SW Sweden
Retrospective
Age group
_._
1987-1991
Inclusion criteria
Mean incidence
Mortality
Main causes
Main cause of TBI
for TBI
ofTBI
rate
ofTBI (%J
mortality
MVC 60<¥0
MVC 75%
-- - -
0-1 7 yea rs
von Wendt
~1
hour of LoC
12
2.6
clinica l,
. 1997 28
Falls 22%
neu rophysiol og ica I,
Sport 7%
imaging evidence of
Missile 3%
brain contusion
Head trauma 20/0 Unknown 60/0
(excludes: con cussion, mild HI) Suom in en
Helsi nki, Finl and
Retrospective
et al 1998 29
1985- 1994
< 16 yea rs
inc.
Durkin et al
New York -
1999 30
N Manhattan
Retro spective
1983-1995
< 17 years
inc.
Severe multiple trauma,
All severe
All severe trauma
MVC (pedestrians
marked impaired
trauma 14-185.6%
4.8
versus vehicle;
MVC 20<¥0
consciousness,
head injury 72%
bicycle versus
Cycl ists 12.7%
admission to ICU,
to neurosurgical
vehicle)
or auto psy
ICU
Major MVC HI = LoC
~
9.3 < 17 years
2.9 < 17 years
1 hour,
Pedestrian s 21%,
Falls 24%
MVC 16%
with lacerati on :!:
(2/3 pedestrians)
haemorrhage Morrison
Scot land, UK
et al 1999 31
Retrospective
1981-1995
0-14 years
Morta lity figures
Regi st rar Genera l,
Injury mortality
MVC mortality 4.8
Injury 140/0 of
11.6
Assau lt 0.7
total child
Falls 0.6
mortality
Not given
Data on Childhood Injury Mortality
o to > 65
Hospitalization rates
1994-1995
In-hospital
In cidence of mild and
Guerrero
years stratified :
for TBI, from National
TBI-related
mortality rate,
moderate TBI decreased
199932
0-4 years,
Hospital Discharge
admissions:
all ages, remained
over study period, but
5-14 years,
Survey, CDC, PN CHS.
0-4 yea rs, 105
constant at 5.3
seve re TBI in creased
15-24 years,
ICD 9 codes
(range 68-142);
25-44 yea rs,
800.0- 801.9,
5-14 years, 75
45-64 years,
803.0-804.9, or
(range 50-100)
;:065 years
850.0-854.1
Thurman,
Benoit R et al 200033
USA
N Virgi nia
Annual surveys
Retrospec ti ve
1980-1 995
1991-1 999
0- 14 years
from 10 to 19
Falls from windows,
3.9<¥0 mortality.
660/0 of 2322 falls resulted
admitted Trauma Centre,
All deaths < 4 years
in head trauma; children
or died
fell onto concrete
";; 4 years more head inju ry but less extremity injury than
10-14 years
rhurman et al
USA, National
ICD 9 codes:
Hospitali zed lBI:
2000 34
Centre Injury
800-801, 803-804,
0-4 ye ars
Prevention,
850-854, and 873
59
Control
for fatalities
5-9 years
Retrospecti ve
1997
0-14 years
= =
50 10-14 years
=
Pre-hospital deaths
Causes different in
2, at all ages
different age groups:
Fatal outcome:
0-4 years: fall s >
7.50/0,0-4 years
transport related>
and 5-9 years;
assault> firearms
9.20f0, 10-14 years
Not given
5-9 and70-years:
transport > falls >
51
assaults/firearm s Reid et al
Minnesota, USA
Retrospective
1993
0-19 years
2001 35
~
Children living in
73.5
19 years
Minnesota with TBI
Incidence peaks were
re sulting in
in groups
< 1 year,
9.3
~
19 years
< 7 year:
fall s,
Metropolitan 6.5,
battering/assau It
non-metropolitan
7-4 and 70- 74 years:
15.4
fall s; MVC
hospitalization or
6-7 yea rs and
death
13-19 years
5-9 years: fall s,
ICD 9 codes:
Incidence in
MVC (pedestrian)
800.00-801.99,
Metropolitan 72.4,
75-79 years: MVC, falls
803.00-804.99,
non-metropol itan
850.00-854.19
areas 76.1
Non-metropolitan MVA
TBI decrea sed as median hou se hold
and cases identified
income and percentage
from death certificates
of high school graduates in census block increa sed Masson et al 2001
Aquitaine, France
Prospective
1996
36
All ages
All admissions via
Overall, 17.3
Overall mortal i ty 5.2
MVC (occupants 53 0/0,
Ove rall highest
(IS-year
emergency service
severe TBI
Mortality:
pedestrians ISO/a, cyclists
mortality: firearm s
bands)
with a diagnosis of
0-14 years, 8.4
0-9 years, 1.3; 10-19
12%, motorcyclists
(but very few
severe head inj ury
severe TBI
years, 2.1
200/0) most common
cases), pedestrians,
Excluded deaths
Low-level falls,
motorcyclists, falls
at scene
high-level fall s
on one level,
At 0-14 years,
cyclists, high - level
mortality from
falls, MVC
MVC-related lBI
occupants
3.9; fall s 3.1 Adekoya et al 2002 37
USA
Surveillance
1989-1998,
o to >75
Mortality from lBI
summary
by year
years (5-year
compiled from death
age bands)
certificates, all States, using ICD 9 codes
Mortality only
Not given
Rates reduced
Racial differences:
0-4 years: Falls
graduall y
At ages 0-19 years,
0.3, MVC 3.6,
1998 mortality:
MVA highest
firearms 0.3, other 2.5
0-4 yea rs, 6.1 ;
among American
5-9 years:
5-9 years, 3.3;
Indian/Alaska Natives
Falls 0.1, MVC 2.8,
10-14years, 5.5;
Firearms lBI deaths
firearms 0.4, other 0.5
15-19 years, 24.6
highest in black
70-74 years:
Americans
Falls 0.1, MVC 3.4,
Fall-related TBI deaths
firearms 2.0, other 0.6
highest in black
75- 79 years:
children at 0- 4 years;
Falls 0.3, MVC
no difference s at 5-14
15.3, firearms
years; highest in white
12.5, other 1.6
chi ldren at 15-19 years old
(Continued)
Table 14.1 (Continued) Reference
Williamso n et al
Country/region
Scotland. UK
2002 38
Study
Study period
HI mortality
1986- 1995
review
inc.
Age group
0-14 years
Inclusion criteria
Mean incidence
Mortal ity
Main causes
Main cause of TBI
for TBI
ofTBI
rate
of TBI (0/0)
mortality
290 child deaths
MVC (ped estrian)
MVC (pedestrian)
data, Registrar
National mortality
from HI over 10
other
bo th sexes 1.2
Gene ral for Scotland
years
M VC (occupants
declined over study
ICD 9 codes for
Mean annu al HI
and cyclists)
external cause of
morta lity 3.0 ; fell
Other MVC
Not give n
inju ry and
from 4.1 in 1986
(occupants,
skull fracture,
to 1.8 in 1995
cyclists) 0.9
intracrania l inju ry
Highest rate 5. 1 in most deprived areas;
Masson et al
Aqui taine,
Prospective
2003 39
SW France
Traum at ic
1996
Falls 0.3 0.3
lowest 2. 1 in most
Other, 0.3 assau lts in infants
All ages,
Hospita I-admitted,
Overa l1 8.5,
All ages, 51.6 0/0 died
All ages MVC 5.2
mas t severe bra i n
0- 14 years 3.2
0-15 yea rs, 42.9<\'0
0- 14 years: MVC
died
76 0Al, gunshot wounds
inj uries (GCS 8 or
Assaults 0.3
affluen t areas
0-14 years
coma
period
le ss), comatose for
10 0Al, falls 40Al,
>
other 10%
24 hours, or died
without waking Andersson et al 2003 40
W Swede n
All ages,
All grades or fatal
Peak incidence
All ages:
Morta lity for seve re
and
lO-year
TBI, ide ntified from
0- 9 years
Fal l at same level
TBI ha s redu ced from
retrospecti ve
age band s
AEtE discharge register
0-9 years 35%
310/0, fall from height
40% in early 1980s
and death registers
10-19 years 20%
27%, MVC 160/0, hi t
to - 10% in 1999
ICD 9 cod es 850-854
Overall, 2.50/0
by object 15<\'0
Prospective
1992- 1993
and 800-804
All ages, 4.0
mode rate or severe TBI, only 1% admitted to hospital
CDCP, NCHS, Centers for Di sease Control and Prevention, National Hospital Discharge Survey; GCS, Glasgow Coma Score; HI, head injury; lCD, International Classification of Diseases; ICU, intensive care unit;
in c., inclusive; ISS, injury severity score ; LaC. loss o f con sc iousness; M VC, motor vehicle collision; PTA, post-traum atic amnesia ; TS, trauma score.
All rate s 100000 chil dren /year.
Epidemiology I 350
289
r
300 rV>
c
.<::> V> V>
E "0
250 200
'"
~
0
~
150
.D
E ::l
100
Z
50 0
I
I
2
I
I
4
3
I
5
I
I
7
6
8
J J J J I~ I 10
9
11
12
13
Age (years)
10 1998-99
•
2002-03
1
Admissions to the Accident and Emergency Department, Royal Hosp ita l for Sick Ch ild ren, Edinburgh, with head injury 1998-1999 and 2002-2003, by age. (From ref. 42, wit h permission.)
Figure 14.1
Incidence of suspected NAHI in Scotland per 100000 chi ldren, aged 0 -12 months 35.0 30.0 0
0
0 0
S2
" "cu u" "g
25.0 20.0
....--
/'
~
~
C>
15.0 10.0 5.0 0.0 1998 95% (119.38.41 ConL int.
Number 57945
1999 (17.5,481)
2000 (16.9,48.1)
2001 (106.37.8)
56629
54028
52027
2002 (121.411) 51 046
Incidence of non-accidental injury in Scotland/ 100000 children aged 0-12 months 43
Figure 14.2
hour. 46 Because birth trauma was often associated with a sec ondary asphyxia, its precise incidence was difficult to obtain . Routine magnetic resonance scanning of newborn infants has demonstrated that clinically silent subdural haemorrhage does occur in infants in 6 per cent of normal vaginal deJiveries. However, these haemorrhages resolve within 4 weeks and remain clinically silent, and the infants develop normally.47 Instrumental delivery was associated with a higher incidence of subdural haemorrhage. 47 It remains to be established whether these haemorrhages may explain minor neurological deficits in children. In a fUlther sma ll study, Holden et al 48 detected small subdural haem orrhages in the falx or tentorium in four out of eight oth erwise normal term infants who were vaginally delivered .
Mild Accidental Head Injury
----------------
Edinburgh. 44 The annual incidence of suspected NAHl in Scot land from 1998 to 2002 has remained constant in a declining population of 57 945 in 1998 to 51046 in 2002, and is shown in the accompanying figure (Fig. 14.2) .43 The incidence of subdural haematoma in infancy in South West England and South Wales was found to be 21 per 100000 children (95 per cent confidence intervals, 7.5-34.4). of which it was estimated that NAHI accounted for 82 per cent. 45
Traumatic Birth Injury The incidence of traumatic birth injury has declined signifi cantly over the last 30 years, from approximately 6 per 1000 live biIths to approximately 0.2 per 1000 live births. This decrease commenced in the 1970s as a result of changes in obstetrics and neonatal care, which adva nced new manage ment principles such as not allowing labour to proceed beyond 24 hours or the second stage to proceed beyond 1
The majority of head injuries in children are mild and these include injuries in children who have had a brief interrup tion of consciousness, with or without vomiting. Clear guidelines have been established to recommend which patients should be admitted, and in all mild head injuries the most important aspect is to recognize those who deteriorate. The most frequent causes are accidents in the home - from falling over, or sholt/high falls. Falls from bunk beds are relatively common, as well as motor vehicle (Fig. 14.3) or pedestrian accidents, non-accidental injury (NA!), SpOlt and playground accidents, and bicycles or equestlian accidents. Local policy reco mmends imagi ng by computerized tomography (CT) scan for patient who have experienced loss of consciousness of 5 minutes or more, together with persist ent vomiting or changed behaviour. The commo nest finding is a linear parietal skull fracture (22 per cent) and small intracranial lesions such as contusions; small extradural or small intraventlicular haemorrhages are seen in fewer than 3 per cent of patients and are usually associated with an
Head and neck injuries
290 I
100
100
90
90
80
80
70
~
c: :::>
60
.2:>
60
50 40
~
50
V\
-R
.s
:~ :::> 0
·c
Q.)
Vl
co
~
0
v co
I::
30
.<:;
u
20 10
3
Figure 14.3
40 30 20
10
0
(a)
70
6
9
20
25
30
35
40
45
50
Speed at impact (mph)
0 20 (b)
30
40
Vehicle speed (mph)
(a and b) Relationship between veh icle speed and li ke lihood of serious injury or survival in Scotland.
overlying fracture and require no surgical action. Plain skull films remain the most common investigative method in most cases of mild head injuries. Hospital policies for admission differ, but most will admit children who have suffered a significant loss of conscious ness (more than 5 minutes) or amnesia, the presence of abnormal neurological symptoms or signs, or skull fractures. Indications Jor CT scanning include loss of consciousness or arrmesia for more than 10 minutes, a deteriorating conscious level, an open skull fracture, seizures and confusion, a signifi cant fall (from higher than 1 metre) or suspected non-accidental aetiology. Continued altered consciousness or deterioration of conscious state, depressed skull fractures, or presence of blood or cerebrospinal fluid (CSF) leaks warrant neurosurgical referral. Mild head injury may cause ischaemic stroke in a sma ll number of children and most will make a good recovery. The vessels involved are the functional end arteries in the striatocapsuJar region (lenticulostriate branches of the mid dle cerebral artery [MCAl).49 The postulated mechanism is a mechanical disruption of flow in the perfo rating branches of the MCA with intimal trauma, and subsequent thrombo sis or arterial spasm. Genetic and other environmental fac tors (chickenpox) may be implicated. 50 Delayed cerebral oedema and fatal coma has been reported after mild or trivial head injUly occurring after a lucid interval. Attacks of familial hemiplegic migraine (FHM) may be triggered by minor head injury and can be accompanied by coma. Mutations in the CACNAIA cal cium channel subunit gene on chromosome 19 are associ ated with episodic symptoms, including co ma. It was, therefore, postulated that the novel S218L mutation in the CACNA lA calcium channel subunit gene is involved in FHM and fatal cerebral oedema after mild head injury.51 Patients with mild head injury and persistent post-con cussive symptoms have been found to have a high inci dence of temporal lobe injury on single photon emission computed tomography (SPECT) and positron emission tomography (PET) scanning, which may explain the fre quent memory disorders. 52
Physiological Monitoring in the Neurointensive Care Unit Modern intensive care management of head injury aims to avoid secondalY physiological insults by optimizing cere bral oxygenation and perfusion while the brain recovers. To ena ble early recognition of these damaging insults and to institute prompt intervention, intensive physiological monitoring is required. Sta ndard physiological monitoring of the critically ill head-injured child includes continuous electrocardiography, oxygen saturation monitoring and measurem ent of arteria l blood pressure via an intra-arterial catheter, as well as the core and peripheral temperatures. Central venous pressure and urine output are monitored continuously to guide ther apy for maintaining euvolaemia in these patients. End-tidal CO 2 levels are measured using a capnometer to provide a trend of the CO 2 removal, but frequent arterial blood gas analyses are also carried out to allow monitoring of arterial oxygenation and maintenance of normocapnia, which is vital for control of intracranial pressure (fCP) . Patients with documented hypoxaemia at any stage of head injury management have a statistica lly poorer out comeY-57 The use of pulse oximetry (Sa0 2) has ena bl ed pre vention and early treatment of hypoxaemia. It has become a widely used non-invasive method of monitoring oxy genation in modern head injUly management. Oxygenation is most accurately determined by arteria l blood gas stud ies, which require arterial puncture or an indwelling alierialline and therefore can be performed only in a hospital setti ng. Over the past few years, the use of a fibreoptic sensor that is capable of measuring the oxygen tension, CO 2 ten sion, pH and temperature in the blood has gained consider able interest and has been adapted for use in the brain tissue. 58 This multipara meter sensor can be inserted into the brain tissue, together with the standard ventricu lostomy catheter and a microdialysis probe, through a triple-lumen intrac ran ial bolt.
Epidemiology I
Many earlier studies of severe head injury have concen trated on the importance of raised ICP, current evidence moves toward emphasising the importance of cerebral perfu sion pressure (CPP), the principal determinant of cerebral blood flo w (CBF). Thus, maintenance of an adequate CPP is essential to maintain cerebral oxygen delivery and prevent cerebral ischaemia. The only way to reliably determine CPP in the brain-injured patient is to continuously measure alierial blood pressure and ICP (cerebral perfusion pressure = mean arterial blood pressure - intracranial pressure). Although ICP monitoring is not universally used , virtu ally all major adult head injury centres in the USA use ICP monitoring in guiding management and it is an integral pali of intensive care in these centres. The practice varies in Britain, with rou t in e ICP monitoring occurring in 48-57 per cent of intensive care units (ICUs) that regularly treat head-injured ad ults. 59 ,GO Similarly, evidence suggests the monitoring and managemen t of children with head injury who are at risk of developing intracranial hype rtension varies considerably in the UK and worldwide. Segal and colleagues G1 conducted a postal survey of the use of ICP monitoring in children with acute neurological illness. All British !CUs that admit five or more children each year with acute neurological illness were contacted. Only one-half of the 70 per cent who responded reported the use of ICP monitoring. In t his survey, children admitted to paediatric intensive care units (PICUs) were more likely to have ICP monitoring performed than those admitted to adult ICUs. 61 There are no published guidelines for the use of ICP monitoring in children, but the major determinan ts in t he decision to initiate lCP monitoring appear to be the admis- . sion GCS score and brain CT scan appearance. GI When ICP monitoring is clinically indicated in a patient, a decision on the type of monitoring device to be used is required. Intracranial pressure monitors that are avaiJable at present consist of a pressure transduction device that can be placed in the following intracranial compartments: ep idural, subdural, subarachnoid, parenchymal or ventricular space. Pressure transduction can be undeliaken t hrough external strain ga uge, catheter tip strain ga uge or catheter tip fibre optic technology. The Camino fibreoptic pressure monitoring device is the most fre quently used system in paediatric neurointensive care. Initial reporting and validation of this type of trans ducing system have indicated its ab ility to measure the brain tissue pressure directly with a rapid response rate to ICP changes. Gambardella's group62 from Italy examined its use in a group of children with severe craniocerebral tra uma, aged 2-16 years, and demonstrated that this method of recording ICP was minimally invasive and that it correlated very closely with the ventriculostomy method, Complications associated with ICP monitOIing include bacterial colonization, infection, haemorrhage, malfunction, obstruction or malposition of the device. Long-telID morbid ity is rarely associated with these complications but they may affect the accuracy of ICP readings and influence the acute
291
management of these critically ill patients. HaemolTh age associated with an ICP device is uncommon, with an overall reported incidence of 1.4 per cent in the adult population 63 7o Fewer patients (0.5 per cent) develop significant haematomas requiring surgical evacuation. 65,67,69 Pople and co-workers 71 examined 303 children who required ICP monitoring and found that only one patient (0.3 per cent) developed intra cranial haemolThage after insertion of the ICP moni tor; this patient was known to have a low platelet count prior to the procedure. Malfunction or obstruction has been repOIied to occur in all types of ICP monitors. The mechanical failure rate of the fibreoptic device used in children with severe brain trauma was reported to be 13 per cent in one series,72 but in another study t he malfunction rate was much lower at 2,6 per cent. 71 Cerebral pe~fusiol1 pressure is the difference between the mean arterial pressure (MAP) and intracra nial pressure (ICP), i.e. CPP = MAP - ICP. The low systolic blood pres sure in infants means that the perfusion pressure is more easily deranged than in adults. The normal cerebral perfu sion pressure is 60-70 mmH g in adults and, when reduced there is a progressive fall in the CBF to the brain. When the CBF falls below 40 mmHg there is absolute reduction in cerebral perfusion and ischaemia a nd infarction will result at flows below 18- 20 mL!100g of brain per minute. Cerebrovascular autoregulatory mechanisms ensure that there is an adequate CBF to meet metabolic demands by changing the cerebrovascular resistance in response to blood pressure changes over the range of 60-160 mmHg. The arterioles constrict when the blood pressure rises and dilate when it falls. 73 Cerebral blood flow also alters in response to the meta bolic demands focally or generally within the brain. Both of these autoregulatory mechanisms are regulated by the small arteries and arterioles. Major blood vessel dimensions are reduced by vasospasm, which can be sufficient to cause infarction. Small vessels have sympathetic innervation, possibly from the locus coeruleus, and respond to meta bolic demands of a paliicular part of the brain (including during sleep). These local flow changes are all influenced by pH , nitric oxide and lactate production. The Cushing phenomenon is a progressive response of systemic arterial pressure in response to an increasing fCP as it increases to levels approaching the diastolic blood pressure. It is mediated via brainstem reflexes, which depend on blood flow to the locus coeruleus. The cardiac output and peripheral resistance are increased, the latter by sympathetically mediated peripheral vasoconstriction and an increase in vasomotor tone. Thus there is a sequence of comp ensatory events to raised ICP from the initial cerebral arteriolar dilatation followed by elevation of the MAP, fol lowing which both MAP and CPP fall in respons e to con tinued ICP elevation. 74 Continuous body temperature monitoring in patients with head injury is a routine practice, as abnormalities in body temperature, paliicufarly pyrex ia, occur frequen t ly after
292 I
Head and neck injuries
brain injury.75,76 Conventional indicators of core temperature include rectal and bladder temperature, but recent advances in intracerebral monitoring have enabled brain temperature to be measured; this is done by using a thermocouple embed ded in the tip of an intraventricular catheter used for rcp monitOling (or a triple bolt inserted into the frontal region encompassing an ICP monitor) , a microdialysis catheter and a Para trend multiparameter (oxygen, carbon dioxide, pH and temperature) sensor. 75 ,77,78 The precise influence of pyrexia on the outcome after brain trauma remains unclear. Pyrexia has been postulated to exacerbate brain injury by increasing cerebral metabolic requirements for oxygen and excitatory neurotransmitter circulation. The clinical use of hypothermia was pioneered in 1938 79,80 and since then several laboratory studies 81 - 84 and small clinical trials 85.86 have suggested its association with an improved recovery. In 1994, a multi centre randomized controlled study was conducted, in the hope of providing definitive evidence of the efficacy of hypothermia in adults with severe brain trauma. The study was halted in 1998 after enrolment of 392 of the planned 500 patients when induced hypothermia, bladder temperature of 33°C, initiated within 6 hours of injury and maintained for 48 hours, failed to improve clinical outcome.87 Furthermore, patients who were in the hypothermic group and more than 45 years old had a higher incidence of poor outcome and had more complica tions than those who were kept nOlIDothermic. 87 Neutropenia and coagulopathy may complicate induced hypothermia. Thus, fuliher studies are required to determine the influence of brain temperature on outcome after head injury. It is not resolved whether hypothermia induced by selective head cooling in newborns merely reduces the extracranial blood flow without alteration of the intracranial circulation.
ELECTROPHYSIOLOGICAL MONITORING Head-injured children requiring intensive care often need to be pharmacologically sedated and paralysed; this masks the motor and other clinical manifestations of seizures and makes the assessment of cerebral function by conventional clinical neurological examination impossible, apart from pupillaly reactions. Furthermore, some children develop abnormal movements without electrical seizure activity after head injUlY and without electrophysiological moni toring they can receive unnecessary treatment. In addition to providing valuable information on the degree of cerebral insult in the critically ill child with an al tered level of consciousness, electrophysiological monitoring of brain activity has been recognized to have prognostic value in children with severe head injury.aa-90 Multichannel electroencephalography (EEG) recording is the gold standard in measuring surface brain activity.91-94 It allows distinction to be made between generalized, local ized, focal or multifocal seizures and to provide an aetio logical and prognostic guide in relation to single, repeated
or multifactorial insults. Tasker and co-workers 95 subse quently used the cerebral function analysing monitor (CFAM) in conjunction with serial multichannel EEG to monitor cerebral function in 54 critically ill comatose chil dren , and demonstrated that CFAM traces could be used to: (1) identify effects of acute or cumulative cerebral insults; (2) recognize unstable patients exquisitely sensitive to standard care procedures; and (3) evaluate seizure control. Although maintaining an adequate CPP is important as it provides a pressure gradient governing the CBF, the ulti mate determinants of the cerebral metabolism are CBF and the oxygen content in the blood. Monitoring of the ICP and CPP alone offers no information on the brain's oxygen delivery and usage. At present, methods to measure cere bral blood flow such as PET, Xenon clearance or SPECT are too cumbersome for use in an ICU, however, jugular venous bulb oximetry and transcranial Doppler ultrasound enable a better understanding of the state of the cerebral circulation and oxygen consumption .
CEREBRAL BLOOD FLOW The CBF depends upon: (I) cerebral perfusion pressure; (2) cerebral vascular resistance (radius of vessel, length of vessel) inversely; (3) viscosity ; (4) venous pressure; and (5) ICP. (CPP = CBF/CVR, CBF = CPP X CVR.) The normal CBF is around 50 mL per 100 g of brain tissue per minute. 96 The CBF is a higher percentage of the cardiac output in small infants than in adults. It is also highest in grey matter (100 mL per 100 g) and decreases, sequentially, through the inferior colliculus, sensory cerebral cortex, motor cerebral cortex, geniculate bodies, superior colliculus, caudate nucleus, thalamus, cerebellum to the cerebral white matter, where the rate of flow is 20 mL per 100 g of brain tis sue per minute. Impairment of CBF due to raised ICP is clin ically seen as distension of scalp and retinal veins, pulsation of the fontanelle and a loud systolic cranial bruit, which appears when the ICP exceeds the diastolic blood pressure and increases in pitch with slowing of the cerebral circula tion, and disappears with a 'carotid stop'. Non-invasive methods of assessing the flow velocities of the intracranial cerebral arteries using transcranial Doppler ultrasound (TCD) were first described in the early 1980s. 97 Flow velocities can be measured in the middle, anterior and posterior cerebral arteries by placing a probe in the temporal area, just above the zygomatic arch. As the cross-sectional area of the arteries cannot be measured directly, flow can not be measured from velocity but the Doppler shift meas ured is inversely proportional to the diameter of the vessel. Thus, providing that all of the other factors remain con stant, the mean flow volume (MFV) will approximate CBF. Vasospasm will result in an increase in flow velocity. The resistive index (RI) or Pourcelot index (S - DIS) is an alternative ratio for assessing flow velocity and is normally of the order of 0.7. It increases with increased ICP and is usually reliable unless there is distortion of the course of the
Epidemiology I MCA. Using transcranial Doppler, MCA spasm was identi fied in 40 per cent of patients after traumatic brain injury and could begin as early as 48 hours post injury.98 Maximal MCA spasm was found between 5 and 7 days after the initial trauma. 98 In another study of head-injured adults, severe MCA spasm identified by transcranial Doppler was confirmed by angiography.99 Daily measurements of the MCA velocity from 121 patients with varying severity of head injury showed an inverse correlation between the severity of head injury and the MCA velocity. 100 In addition, a significantly lower MCA velocity on admission was demonstrated among the head injured patients than among the normal control subjects. 100 High ICP and low CBF were responsible for the low velocities in the intracranial circulation after head injury. FUlihermore, an admission MCA velocity of less than 28 cm per second correctly predicted 80 per cent of the early deaths. 100
293
an increased production of lactate while the production of pyruvate decreases; this causes an increase in the lactate pyruvate ratio. 102 ,103 Glycerol is an integral component of the cell membrane and has been use, therefore, as a marker of cell membrane damage. 104 ,105 Although gluta mate has been proposed as an indirect marker of cell dam age, it is often hard to interpret the dialysate level as the neuronally released glutamate is mixed with the large metabolic pool of glutamate. A reduction in the dialysate glucose is often associated with a decrease in brain oxygen tension suggestive of reduced cerebral perfusion. 106 A marked increase in the adenosine level of the cerebral interstitial fluid has been shown to occur during jugular venous oxygen desaturations suggesting a potential role for adenosine during the periods of secondary insults after brain trauma. 109
IMAGING OF ACCIDENTAL HEAD INJURY MICRODIALYSIS
As the composition of the cerebral interstitial fluid reflects the biochemistry of the neurones and glial cells in the brain, attempts have been made over the past decade to develop the use of cerebral microdialysis to monitor neuro transmitter release and energy metabolism in the brain. By mimicking the function of a blood capilJary, the micro dial ysis catheter enables monitoring of the chemical changes in the interstitial fluid. 101 Because of a concentration gradi ent across the semi-permeable microdialysis membrane, chemical substances within the interstitial fluid diffuse across the dialysis membrane and into the perfusion fluid inside the catheter. 99 Samples are collected in microvials and brought to a bedside analyser as often as necessary. 101 The microdialysis catheter can be implanted during sur gery or when the patient is in the ICU. If implanted during sur gery, the catheter is tunnelated under the scalp and through a small incision through the dura. Ihe membrane of the catheter is then positioned in the penumbra region, usually 1 cm from the border of the evacuated lesion. A second catheter may be placed in the surrounding 'normal' brain tissue through a sep arately drilled hole in front of the intraventricular pressure catheter. Ihis evaluates any further degradation of normal tis sue physiology due to secondary damage. In those patients who do not require surgery the microdialysis catheter is inserted through an intracranial bolt, but this makes it more difficult to reach a position within the brain that is relevant for the interpretation of the brain chemistry. Ihe position of the catheter may be visualized on CT scans. Ihe dialysate lactate, lactate-pyruvate ratio, glycerol, glutamate, glucose and adenosine levels have been investi gated in head-injured patients. 102-108 Ihe lactate-pyruvate ratio has been shown to be a better marker of ischaemia than lactate alone, as an increase in lactate may be the result of hypoxia, ischaemia or hypermetabolism. 102, 103 But, during ischaemia, the neurones become dependent upon the anaerobic production of AlP from glucose, resulting in
A non-contrast CT scan of the head and upper cervical spine and plain cervical spine radiographs should be the first-line imaging in children with accidental head injury, but it is important to recognize that up to 60 per cent of spinal injuries in children may occur without evident bony injury. 110 Computerized tomography scanning of head-injured patients is recommended because it is sensitive to the presence of haemorrhage and best demonstrates bony integrity. It is the usual imaging modality on presentation and also for subse quent scans that are undertaken during the acute admission. The usefulness of magnetic resonance imaging (MRI) for accidental brain injury in children has yet to be defined but, in general, MRI and CI are complementary modalities. Mag netic resonance imaging is better able to depict extra-axial, suboccipital, subtemporal, subfrontal or tentorial haemor rhage and provides a better view of the posterior fossa and brainstem. It is useful in locating shear injury and at follow up, the extent of parenchymal damage. Ihe Marshall CI Score lll is used to classify the CI injuries following head trauma (see Iab.le 14.2). Specific CI scan findings after head injury include dif fuse axonal injury, contusions, focal ischaemia, hydro cephalus, diffuse brain swelling and brain oedema. Diffuse brain swelling is more common after head injury in chil dren than in adults and may be due to hyperaemia, vascu lar congestion, an increased water content or brain (cellular) oedema. It is associated with at least two CI scan appearances: (1) a low-density pattern, suggestive of ischaemia from raised ICP or additional hypoxia in coma tose children and (2) a normal or hyperdense pattern that is more amenable to pressure-reducing measures.
EPIDURAL HAEMATOMAS
Epidural haematomas in adults are usually arterial in origin but may be of venous origin in children. About one third of cases in children are associated with coma, due to
294 I
Head and neck injuries
Table 14.2
The Marshall CTscore
Diffuse injury I (no visible pathology) Diffuse injury II
Diffuse injury III (swelling) Diffuse injury IV (shift) Evacuate mass lesion Non-evacuated mass lesion
No visible pathology seen on CT scan Cisterns are present with midline shift of 0-5 mm and/or: lesi on densities present, no high- or mixed-density lesion > 25 cc, may include bone fragments and foreign bodies Cistern compressed or absent with midline shift of 0-5 mm, no high- or mixed-density lesion >25 cc Midline sh ift >5 mm, no high- or mixed-density lesion >25 cc Any lesion su rgically evacuated High- or mixed-de nsity lesion >25 cc, not surgically evacuated
concomitant parenchymal injury. This improves after surgical remova l of the clot without opening the dura. La rge volumes of blood may be lost, leading to shock and secondary brain injury or death. The risk factors for deterioration include a fracture crossing the middle meningeal artelY, vein or sinus, or an increase in the size of the epid ural haemorrhage.
SUBDURAL HAEMATOMAS Thin subdura l haematomas assoc iated wit h diffuse brain injury or swelling are most common. The indications for decompression include midline shift, persistent raised [CP and no associa ted severe brain swelling. Subdural haemor rhages in the interhemispheric fissure may be tapped in infants via the anterior fontan elle.
CONTUSIONS Focal contusions may not be immediately visible on CT scans but are evident in 50 per cent at follow-up.1I2 They exert a mass effect and contribute to the overa II raised [CP.
in the intensive thera py unit (ITU), paediatric neurology department and by the child protection team. The inappro priateness and inconsistency of these multiple accounts may be an important indicator of an inflicted injury. A traumatic explanation is offered in 10-20 per cent of cases to account for the child's condition. The symptoms at presentation to hospital with thei r rel ative frequencies were: • irritability and crying, 41 per cent; • bruising and su perficia l injury, 41 per cent; • vomiting and a nore x ia , 38 per cent ; • apnoea and respiratolY difficulties, 38 per cent ; • extensor stiffness, due to fit, anoxic rigidity or raised intracranial pressure, 34 per cent; • probable fits, 21 per cent; • pallor or cyanosis, 29 per cent; • history of a potential traumatic event, 26 per cent; • depressed conscious state, 21 per cent.
ON EXAMINATION
SKULL FRACTURES
Scalp Injury
Depressed sku II fractures require repair of the dura to prevent raised pressure, which leads to brain herniation with venous infarction of the cortex. They also predispose to infection. Fractures of the skull base, with leakage of CSF throu gh the nasal cribriform plate or t he ear, result in a compens a tory mechanism that keeps the ICP low until the leakage ceases. The defi nitive rep a ir is undertaken after the acute injUly period. Penetrating injuries (gun shot, knife, darts, needles) to the brain in children are rare in the UK. In those cases when there is a preservation of consciousness, the object is left in position until after sca nning. Debridement is then under taken with pressure monitoring and antibiotic cover.
Bruising and oedema of the sca lp , periorbital bruising or Battle's sig n (bruising over the mastoid) (Fig. 14.4), usually seen in accidenta l head injury, is rarely present, even in the presence of severe intracranial damage.
NON-ACCIDENTAL HEAD INJURY A history of preceding events lea ding to hospital admission is obtained on several occas ions in the AEtE departm ent,
Skeletal Injury Skull fractures are seen in approximately one-quarter of cases and evidence of impact is seen in more than one-half of cases. Signs of impact, particularly of the head, indicate inflicted trauma. As young infants are unable to injure themselves accidentally, they may be an indicator of intent. Clinical evidence of imp act is: • cutaneous or subcutaneous bruising; • subgaleal haemorrhage; • skull fracture; • extradural haemorrh age; • focal subdural haemorrhage; • focal cerebral contusion.
Non-accidental head injury I
295
NEUROLOGICAL PRESENTATIONS OF NON-ACCIDENTAL HEAD INJURY
Figure 14.4 Battle's sign. Prominent bruising behind th e pinna indicates a skull base fracture.
Fifty years ago, the whiplash element of the shake n baby sy ndrome was highlighted by Caffey ll3 and Guthkelch 1l4 but the more obvious supratentorial rotational injuries, i.e. su bdura l haematoma, are easier to see and since then have dominated clinical descripti ons. We propose a classi fica tion based on the clinical pattern of presentation, which should be considered as the tempora l classification of neu rological presentations: I. hyperacute cervicomedullary syndrome (whiplash
shaking injury), 6 per cent;
II. acute encephalopathy of classical 'shaken baby
syndrome' with rotat ion plus or minus impact (fits,
coma, decerebration and central apnoea ventil ation). 53 per cent ;
III. sub acute non - encep halopathic presentation (subdura l haemorrhage, haemorrhagic retinopathy, fractures, bruising). 19 per cent; IV. reCUlTent encepha lopathy - none was recorded in t his series; V. chronic extracerebral presentation (isolated subdural
haemorrhage), 22 per cent. 115
Hyperacute Cervicomedullary Syndrome Overa ll, 53 per cent of cases ha ve no fracture on skeleta l survey. Rib fractures are seen in one-third of cases and fractures of the long bones seen in one-qu arter.
BRUISING AND SKIN TRAUMA Our experience is that some bruising and superficial abra sions occur in more than 40 per cent of cases, often a mixture of new and old bruises. Additionally, abras ions, sc ratches, lacerat ions, hair loss, subconjunctival haemor rhages and bl eedin g from the ears and mouth are present in a small number of cases. A firm adult grasp is necessary in order to shake a child and may result in a pattern of bruises. These must be care fully documented and photographed. Common methods of grasping a chiJd (and subsequent injuries) are: • by the chest - thumb marks at the sides of the nippl es, fractured ribs and retinopathy; • by the arms - spiral fracture of the humerus; • by the legs - bucket handle metaphyseal fracture ; • by the throat - carotid trauma (ipsilateral infarction and loss of ve nous pulsation in t he retina, contralateral hemiplegia, faci al and retina l petech iae); • by the abdomen - finger bruising, retroperitoneal bleeds and bowel bruising; • by the shoulders - finger and th umb bruises; • bruising around t he mouth suggests possible suffocation to prevent the infant screami ng; thumb marks should be carefully sought under the mandib le.
This presentation results from severe shaking forces. Geddes et al" 6 , 11 7 found localized axonal damage at the craniocervi cal junction, in the corticospinal tracts in the low er brain stem and the cervica l cord roots; additionally, t here is the possibility of traum atic thrombosis of the vertebral arteries as they wind through the foram ina of the cervica l vertebrae. Damage to these areas is thought to result from hyperflex ion and hyperextension movements, a 'whiplash' shaking stem injury. These cases are infrequently or only briefl y seen by the clinician (6 per cent) as they a re either dead on a rri val at hospital or die shortly thereafter, but are more frequently seen by pathologists. Th ese young infants, at about 2- 3 months of age, presen t with apnoea as part of this cervi comedullary syndrome, with only 'trivial' subdural bleeding. The apnoea gives rise to severe seconda ry hypoxic brain injury with cerebral oedema but minimal axonal shearing. In this hyperacute presentation, all subdural haemorrhages that are seen on imaging and post-mortem exa mination are sma ll , as a result of ac ute cerebral oedema. Significant sub dural haemorrhages do not become evident until 2- 3 days later. In summary, the presentation of the hyperacute pattern is apnoea, acute respiratOlY fa ilure and death associated wi th severe hypoxic ischaemic dam age found post mortem.
Acute Encephalopathic Presentation This is characterized by seizures, decerebration, homeostatic derangements, bilateral large subdural haematomas and widesp read haemorrhagic retinopathy. Additional findings,
296 I
Head and neck injuries
such as rib fractures, metaphyseal 'corner' or 'chip' fractures, bruising, cuts and cigarette burns, may be present. This is the commonest pattern seen by hospital paediatricians and has been referred to as classical 'shaken baby (or shaking impact) syndrome'. Apnoea and respiratory arrest, grunting respira tions, shallow respirations or choking represent direct medulialY or vertebral artery trauma or the effects of raised ICP. Following admission to hospital, two-thirds of children presenting with 'shaken baby' syndrome develop epileptic seizures, which are often severe and drug resistant, reaching a climax at 24-48 hours post injury, but usually decreasing and ceasing by the fifth day; it is likely that the fits are an epiphenomenon that refiects the extensive brain injury Also, more than two-thirds of these children have docu mented raised ICP. The increased ICP, together with shock and hypotension, further reduce the CPP and increase the risk of secondary ischaemic brain damage. Low cerebral perfusion pressure correlates with long-term handicap. li S
Subdural Haemorrhage Bleeding from torn bridging veins in to the subdural space is the hallmark of non-accidental shaking injUly in the first year of life. Almost 90 per cent of patients suffer subdural haemorrhage, which is frequently bilateral, and in 20 per cent of cases it is associated with subarachnoid haemor rhage. The subdural haemorrhage may be unilateral at presentation. Subdural haemorrhage may be over the con vexity, interhemispheric, subtemporal, suboccipital or in the posterior fossa. The subdural signal intensity on MRI scanning may be different in different intracranial com partments and does not imply repeated bleeding from repeated trauma. By the time of necropsy, only a thin film of subdural haemorrhage may be present over cortical con vexities (Fig. 14.5). Retinal Haemorrhages In shaken baby syndrome, retinal haemorrhages occur in 80 per cent of cases and one-third are unilateral. I1 9 Levin's groupl20 found them in 84 per cent of cases compared to 16 per cent in accidental head injuries - usually high velocity road traffic accidents with side impact. Retinal haemorrhages were extensive and involved all of the reti nal layers. They are thought to be due to vitreous traction. The vitreous humour is attached to the retina at its periph ery and with different inertias this is also the site where shearing forces tear the vitreous attachments from the retina (or ora serrata) and cause disruption of ocular and orbital blood vessel integrity - similar to the shearing forces which cause subdural bleeding and cortical tears. Non-encephalopathic Subacute Presentation With this presentation, there is no acute brain swelling or diffuse cerebral hypodensities or fits, coma or decerebra tion. Subdural haemorrhage, retinal haemorrhage, rib and other fractures and bruising occur in various combinations.
Figure 14.5
Infant, aged 3 weeks, at necropsy. A thin film
of subdural haemorrhage is present over cortical convexities, following inflicted injury. Bilateral rib fractures and bruises were present.
The child may present with a recurrent encephalopathic presentation with 'odd turns' or fits, apnoeic attacks, cyan otic attacks, rigidity or coma.
Non-encephalopathic Isolated Chronic Subdural This is a 'late' presentation in a child with an expanding head circumference, subdural haemorrhage and often little else. There may be vomiting and hypotonicity. Other causes for subdural haemorrhage must be excluded.
DIAGNOSIS OF NON-ACCIDENTAL HEAD INJURY The diagnosis of NARI is made on the history, which is fre quently inconsistent and does not explain the clinical find ings - ophthalmological and radiological features compatible with, and sometimes characteristic of, non-accidental oligin; and the social pathology in a significant number of cases in the form of known risk factors (Table 14.3). Evidence that shaking is responsible for many NARIs emanates from numerous studies and reports of clinical experience, fre quently acknowledgement or confession with deSCription of the mechanism by the perpetrator. Biomechanical and animal studies have confirmed the forces generated by shaking and other mechanisms. Injurious behaviour has been witnessed by closed circuit TV. Consensus statements by scientific bod ies such as the American Academy of Pediatrics state the extensive combined experience of professionals. 121 As the
Non-accidental head injury I Table 14.3
Risk factors for inflicted head injury
Single parent or cohabitating partner Young parental age Past history of abuse to children Drug or alcohol abuse Domestic abuse Parental mental or psychological illness Premature birth Twin s Re cu rrent medical consultations and hospital admissions Past history of social work enquiry or involvement
Table 14.4
Features of the expanded syndrome of non-accidental head injury
Acute encephalopathy Subdural hematoma Age usually less than 1 year Acute cervicomedullary injury Haemorrhagic retinopathy Bruising Sku II fractu re Rib or limb fracture Evidence of malicious injury (bites, cuts, cigarette burns, whip marks) A history that is incompatible or inconsistent with clinical findings Acute cerebral oedema or diffuse cerebral hypodensities Early cerebral atrophy Poor long -term prog nosis
trauma is almost always unwitnessed , it is important that the clinician is careful about attributing a mechanism of injury in a particular case and diagnoses 'inflicted ' or a 'non-accidental' injury (which, however, may be consistent with a shaking mechanism). Although, theoretically, the features in Table 14.4 repre sent the most complete (expanded) syndro me of NAHI, in practice common clinical presentations include a combina tion of several of these components. The component palis do not carry equal weight nor are they independent. The more components present, the more secure the diagnosis. Although a single component, for example subdural haematoma or retinopathy or isolated haemorrhagic retinopathy in a yo ung infant, is still most likely due to NAl, it may be due to other causes. Individual signs and symptoms can have different causes and some features are more characteristic of NAl and so the diagnosis only becomes statistically secure when several components are found in combination. The most secure diagnosis of NAHI is either when there has been an admission from the perpetrator or when the head it\iury is seen in combination
297
with evidence of other malicious injUlY, such as multiple fractures of different ages, skin incision , cigarette burns or repeated beating. Combinations of any two of the following three factors are highly predictive of inflicted head injury (P < 0.001), 122 i.e. head injury plus any two of th e three following : • inconsistent history/physical ex amination; • retinal haemo rrhages; • parental risk factors (alcohol or drug abuse, previous soci al service intervention within the family or a past histolY of child abuse or neglect). Traumatic retinoschisis and retinal folds are thought to be pathognomonic for 'shaken baby sy ndrome', although extensive multilayer retinal haemorrhages are most unlikely to be due to any other cause. It is important to dis tinguish between making a diagnosis of NAHI with raising concern or suspicion of it. Kivlin 123 quotes 'the mere pres ence of any retinal haemorrhages adequate to raise concern of "sh aken baby syndrome", the extent or type of the haemorrhage is less clinically important.' Subdural haemorrhage is more likely to be non-acci dental in origin if it is convex and bilateral, interhemi spheric, and is associated with cerebral tears and ruptured bridging veins that are visible on imaging. Rao and co-workers l24 concluded that hypoxic-ischaemic encephalopathy (HIE) plus interhemispheric subdural haemor rhage was highly specific for the diagnosis of NAI. The inter hemispheric subdural haemolThage and HIE OCCUlTed in 89 per cent of cases. Zimmennan's group 125 suggested that interhemi spheric subdural haemorrhage could be specific for NAI (61 per cent of cases). There is nothing characteristic about the traumatic encephalopathy in NAHl, although post-traumatic seizures occur velY much more frequently in inflicted head injury than in accidental head injury and they abate by the fifth day.1 26 Duhaime et al 127 developed an algo ri thm for determining the probability that an injury was inflicted. It results in a diagnosis of 'presumptive inflicted ' injury or 'suspicious inflicted'injury. The presence of anyone of the features in Table 14.3 in a child with injuries must raise concerns or suspic ion about the possibility of child abuse and should be an indication for investigation. It is imp0!1ant to consider the questio n of intent, which can sometimes be inferred from the clinical findings , for example when the intent was malicious and the perpetrator cannot have been in any doubt about the injury he was causing to the child. The clinical findings supporting this intent include: multiple bruises of different ages and patterns in different sites, repeated admissions with physical injuries, multiple fractures of different ages, and particular types of injuly pat tern, such as bite marks, cuts and cigarette bums. Disciplinary injuries are not associated with the intent to injure and there may be no intention of harming the child at all. The intention to discipline is not premeditated
298 I
Head and neck injuries
or sadistic. AI.though a physician may be sympathetic to the stressful circumstances surrounding such injury, it must be reali zed that without intervention , if the stress is repeated then so may the abuse. Other possible explana tions for the injury include episodic dyscontrol, puerperal psychosis, euthanasia, and Munchausen's syndrome by proxy (fabricated or induced illness).
ACCIDENTAL SHORT FALLS 'Short falls' are sometimes offered as the explanation for a child's injuries and most will report a 'short fall ' «3 feet) from a bed, lounge, changing table or a parent's arm. Most toddlers fall three to five times per week 128 but our own direct observational study of more than 700 child-hours found that 1- to 2-year-olds fell approximately once every 2 hours, hitting their heads in just over 10 per cent of falls but without fracture or serious injury. 129 In infants who are not yet weightbearing, any fall must be initiated by an adult or equipment failure. Fatalities from short falls are extremely rare but do occur: 0.14-0 .22 deaths per 100000 children between aged from 0 to 4 years. IJO Documented fatalit ies and severe brain injuries have been seen as a result of 'infant walker' incidents involving stairs. 131.132 They a lso occur in an older group of children as a result of falls from playground equipement,133 top bunksl J4 a nd 'baby bouncers '. 131 For very short falls, a short impact time is associated with low terminal velocity, which is not able to generate sufficient rotational velocity and thus subdural haematoma. Imp act injuries from linear acceleration a nd deceleration result in skull fracture and extradural ha ematoma without concussion. Shearing, with retinal haemorrhages, subdural haemorrhage, parenchymal inju ry and concussion, results only if the impact is associated with rotational injury. Skull fractures accompany 26 per cent of cases ofNAHI a nd 22 per cent of accidental head injuries. They are clini cally important if depressed, if they extend through the cribriform plate or the petro us temporal bone resulting in CSF leaks and a risk of meningitis, or if they involve the skull base with brainstem injury. They indicate an impact force, which may be accidental or non-accidental in origin. Skull fractures do not heal by callus formation, making dating of an injury difficult; however, if the edges are round and smooth then the injury is more than 2 weeks old. At autopsy, bone edges are heaped , smooth and discoloured by haemosiderin. A skull fracture normally heals in 2-3 months and disappears on radiograph by 6 months.1J5 In small infants, the fracture site may not heal; rather it can form a growing skull fracture as described below. There may be no bruising at all over the skull, even with a severe impact fracture; however, bruising of the aponeurosis may be evident at surgery or autopsy. Equally, fractures may not be seen on radiograph - what appears to be a simple disci plinary shaking injury may be negated at autopsy or sur gery if we ll-defined fractures are seen, suggesting more
severe impact injury. The type of skull fractures reported in a study of 100 consecutive children 127 who were less than 2 years of age revealed that 27 were linear, eight were depressed, three were multiple, one was stellate, three were bilateral and four were basal.
BIOMECHANICS OF SKULL FRACTURE Force is the product of mass multiplied by acceleration . The force of approximately 35 foot-pounds is required to cause an adult skull fracture. With constant mass, force can only be increased by increasing the velocity at impact, i.e. fall from a greater height or harder punch, etc. Falls from a bed result in infrequent skull fractures in young children; those that do occur are unilateral, narrow and do not cause seri ous injury to the child.1J6 A fracture is indicative of force to the head but the force may have been linear or rota tional. Extracranial muscle, hair and skull thickness will diminish the effect of the force on the brain. The membranous skull of the young infant is more elastic than adults and may deform without fr acture. 137 A wider area of impact will dissipate the force and is less injurious or likely to cause fracture. Soft surfaces result in a prolonged impact time and a low resultant fo rce. More elastic infant bones are likely to experience bounce than adult ones and further dissipate force. Skull fractures occur along the lines of leas t cranial rein forcement, i.e. the temporal and parietal regions. Bursting fractures result from compression of the skull. Non-accidental origin of a skull fracture is suggested by a fracture line that: • bra nches; • is stellate; • crosses suture lines; • is bilateral; • is multiple; • is wide at presentation and expands; • is a growing skull fracture; • is a depressed occipital fracture in a child under 3 years of age. Growing skull fractures are seen in both accidental and NAHl, but are thought to be more typical of a non-accidental origin (Fig. 14.6). They occur during the rapid phase of skull growth and most occur in infancy. They may result from trapping of the dura between the fracture margins, meningeal hernia with pulsation of the dura, or as the result of pseudarthrosis. lJs
OTHER SKELETAL FRACTURES The incidence of fractures in NAl varies widely, and is esti mated at between 11 per cen t and 55 per cent. 139 They are commonest in children under 2 years of age. Fractures in babies under 4 months old are nearly all due to abuse. Radiological features suggestive of NAl and their differen tial diagnosis are discussed in Chapter 3.
Non-accidental head injury I
Figure 14.6 Plain lateral radiograph of sku ll with growing fracture caused by the interposition of tissue between bone edges, seen after both accidental and inflicted injury.
Table 14.5
Investigations in suspected non -accidental head
injuries Brain imaging MRI CT USS Dopp ler stud ies Photogra phy Electroen cepha log ra phy Spectrophotometry of cerebrospinal fluid Intracranial pre ssure measurement Coagu lati on screen Inborn errors of metabolism investigations Child Protection Team Paediatrician
Socia l wo rk
Police
Foren sic
INVESTIGATIONS FOR NON-ACCIDENTAL HEAD INJURIES
Table 14.5 lists the impol1ant investigations for all sus pected cases of NAH!. Additional investigations may be necessary to exclude other diagnoses. The Royal College of Radiologists recommends skeletal survey and mandatory CT of the head in infants and children aged 0-2 years l40 and X-ray of the clinically suspic ious area in those aged 3-5 years . Skeletal survey is not generally indicated in ch ildren over 3 years of age. Bone scintigraphy is indicated in children over 2 years if the skeletal survey is equivocal (see Chapte r 3) .
Figure 14.7
299
Ei ght days post admission : T2 transve rse magnetic
resonance imaging sequence showing high intensity white matter bilaterally at the parieto -occi pital and right temporal regions. Other sequen ces confirm posterior and interhemispheric high signal subdural haemorrhages. IMAGING OF NON-ACCIDENTAL HEAD INJURIES (FIGS 14.7-14.10)
Both CT and MR scanning have been used to image the acute non-accidentally head-injured infant for which subdural haematoma is the radiological hallmark. The subdu ral haematomas are most likely to be posterior, i.e. parieto - occipital or interhemispheri c in site. 141 In the early ac ute stage this may simply appear as a thin crescentic density on CT scan. As the haematoma breaks down by fib rinolysis and water is draw into the haematoma to form an effusion, there is marked expansion so that by 7 days it has undergone a significant enlargement and afte r 2 weeks it becomes isodense with brain and may be missed on CT or ultrasound scanning.142 Magnetic resonance imaging findings reflect the pathologica l consequences of a rotational acceleration( deceleration injury. Barl ow et al 143 have identified subdural haemo rrh age in all cases, which may be su btemporal, inter hemispheric, have torn surface veins, differing signal inten sity, compal1mentalized and suboccipital position. Brain contusions, cerebral lacerations, petechial haemorrhage at the grey-w hite matter junction and corpus callosum, focal asphyxial insult, cerebral oedema, intraventricular and sub arach noid haemorrhages were also found . Diffusion-weighted brain MR images were abnormal in all suspected child abuse cases with subdural and retinal haemo rrhages. 142 In CT(MRI comparative studies comparing inflicted with non-i nflicted traumatic brain injuries it has become apparent
300 I
Head and neck injuries
Figure 14.8 (a) Twenty-five days post admission: fluid-attenuated inversion recovery (FlAIR) coronal sequence showing posterior layering within the subdural collection and high signal in the left parieto-occipital cortex. There is widespread encephalomalacia (with large subarachnoid spaces, dilated ventricles and atrophy) and early gliosis. (b) Twenty-five days post admission: T2 transverse magnetic resonance imaging sequence showing left temporal scalp haematoma, widespread high intensity white matter and encephalomalacia more marked on the right. SPECTROPHOTOMETRY Figure 14.11 shows absorbance of oxyhaemoglobin, at 413-415I1m, and bilirubin at 450-460nm, undertaken on the subdural aspirate of a 2-month-old infant who pre sented with retinal haemorrhages, raised lCP and subdural haemorrhage. The methaemoglobin absorbance, when present, is identified at 405 nm. Although timing of the bleed may be more specific with spectrophotometric analy sis of blood-stained CSF, the presence of bilirubin in sub dural aspirates is less precise and indicates that bleeding occurred 24 hours to 3 days earlier.
TRAUMATIC BIRTH INJURY
Figure 14.9 Four months post admission: fluid-attenuated inversion recovery (FlAl R) corona I seq uence showi ng locu lations in secondary subdural collections with rapidly evolving atrophy. that there are more interhemispheric bleeds, more ventricu lomegaly, subdural hygromas and large extra-axial spaces in the inflicted group, whereas axonal shearing injuries and skull fractures are more common in accidental injuries.
Prenatal trauma in pregnancy may result in direct injury to the fetal brain or, more usually, indirect injUly as a result of maternal circulatory or respiratory injury or retroplacental haemorrhage. Direct injuries to the fetal brain are rare because of the protective effects of the amniotic fluid but when they occur they can result in fetal death or later neuromotor disability, such as hemiplegia (see Chapter 9). Birth injuries include: extracranial haematomata (caput succedaneum, subgaleal haemorrhage and cephalo haematoma), skull fractures, intracranial haemorrhage epidural, subdural, subarachnoid (less commonly intra parenchymal or intraventricular), parenchymal contu sions, brainstem and spinal cord injury, and peripheral nerve injury.
Traumatic birth injury
I
301
Figure 14.10
(a) Seven months post adm ission: fluid-attenuated inve rsion recovery (FLAIR) coronal seq uence showing mixed intensity loculations bilaterally, suggestive of low-density clot within the collecti on . Overall brain atrophy. (b and c) Eigh teen months post admission: FLAIR coronal and T2 transverse images demonstrating marked gliosis and residual cystic change particularly in the right hemisphere. Loculations within the subd ural collections are not now visible.
Pericranial and intracranial birth injuri es are illus trated in Chapter 8. Caput succedaneum results from scalp oedema secondary to pressure of the head on cervix. It resolves within a few days. Subgaleal haemorrhage involves bleeding into the sub ga lea l space, often over the whole sca lp . It is an infrequent complication of vacuum extraction or rotational forceps delivery. A large volume of blo od may be lost, thus requir ing resuscitation.
Cephalohaematomata are haemorrhages beneath the periosteum; they limited by its attachment to cran ial sutures and are so metimes associated with skull frac tures (10-2 5 per cent); 145 they are caused by compression to the fetal head and usually resolve without sequ elae. Skull fractures can result from instrumental delivery. They are usu ally simple linear parietal fractures but, if depressed, may impinge on the cerebral cortex, cerebel lum or brainstem.
302 I
Head and neck Injuries
1.8 1.6 1.4 2- 1.2 Q.J u 1 c: co .D 0.8 a 0.6 .D '" <x: 0.4 0.2 0 350
-
a
• • 375
405
415
425
460
500
533
575
Wavelength (nm)
Figure 14.11
direct;
- acceleration;
- deceleration;
indirect;
- linear acceleration/deceleration;
- rotational acceleration/deceleration ;
wh ipl ash/s haking injury ; combi natio ns of the above.
Spectrophotometric graph of a subdural aspirate.
The bilirubin peak indicates that blood has been present for >24 hours.
Epidural haemorrhage is a rare bilih injury. It is usually associated with a skull fracture and is usually extremely small because of the tig ht application of the dura to the skull bones in infancy. Subdural haemorrhages may follow breech delivery, cephalope lvic disproportion, vacuum extraction and pre cipitate delivery, which lacerate bridging veins or dural venous sinuses. They may require surgical intervention or develop as a subdural hygroma. Subarachnoid haemorrhage occurs in full term and, more frequently, in pre-term infants and is usually hypoxic in origin. Choroid plexus haemorrhage or subependymal venous haemorrhage extend into the ventricle system and may subsequently extend into the subarach noid space, around the poste rior fossa structures . There is a risk of sub sequent hydrocephalus about 2 weeks later. Brain contusions resulting from forceps or compressive damage to the head during bilih are rare but are associated with skull fractures. Brainstem and spinal cord injuries result from excessive neck movements, particularly during breech delivelY or other malpresentation. Fractures are not usually evident. InjUlY results from traction or compression of the brain stem and spinal cord, or thrombosis of the vertebral or anterior spinal artery. Lower cord injuries are rare but may result from similar exaggerated spina l movements. Peripheral nerve injuries result either from intrauterine malposition with oligohydramnios, or traction on the neck or upper limbs that produces Erb 's (with or without Horner's syndrome), Klumpke's or other peripheral nerve palsy.
PRIMARY MECHANISMS OF INJURY TO THE BRAIN There are several mechanisms by which head inju ry may be sustained: • penetrating injury; • compression injury; • impact injury;
PENETRATING HEAD INJURY Penetrati ng head injuries damage the brain by laceration, haemorrhage and infection. Children may suffer penetrat ing head injuries as a result of gunshots, scissors, knives, darts and pencils (particularly involving the tonsillar fossa with carotoid artery injury). As they are non-rotational, they a re non- concussive.
COMPRESSION INJURY Compression is rarely a primary mechanism of NAH!. The classical model for compress ion head injury is bilih injury complicating malpresentation or cephalopelvic disproportion. The clinical patterns of traumatic birth injuries have been outlined above . The mechanisms of intracranial dam age in compression injury are: (J) compression (impaired cerebral blood flow, oedema, traumatic asphyxia, coning); (2) moulding (l atera l moulding - sagittal sinus entrapment, tearing of bridging veins, anteroposterior moulding tearing of the vein of Galen, tentorium and the venous sinuses); and (3) traction (cervical spine, brachial plexus, vertebral arteries). Compressional injuries are non-rotational and non shearing; therefore, they are non-concussive until cerebral oedema, raised ICP and surface haemorrhages result (such as subdural and subarachnoid haemorrhages).
IMPACT INJURIES Direct acceleration impact injuries are exemplified by a straight punch and direct deceleration/impact injuries dur ing a fall. Indirect linear acceleration/deceleration injuries are typically from vehicular whiplash injuries an d indirect or direct rotational acceleration/deceleration injuries that commonly result from motor ve hicle collisions. Shaking is a specific mechanism of injury that includes rotational acceleration/deceleration components. In individual cases, there is often more than one of the above mechanisms operating. Most accidental head injuries in children are impact injuries. In acceleration injuries, the force is dissipated in the resultant motion of the head. In deceleration injuries, the force can only be ameliorated by a 'fo rgivin g' surface or bounce. Deceleration injuri es produce the most severe clinical features.
Primary mechanisms of injury to the brain I
SHAKING AS A MECHANISM OF NON-ACCIDENTAL HEAD INJURY IN INFANTS
There are five major components to the syndrome: 1. shaking causing rotation of the brain inside the sk ull ; 2. impact against a soft surface, causing acute deceleration of the brain and vastly increasing the rotational torque forces inside the skull (no external evidence of impact); 3. hyperflexion and hyperextension injury from whiplash, causing repetitive subluxation of the cervical vertebrae; this results in injury to the cervical cord (Fig. 14.12) and/or medulla; 4. shaking with hard impact, causing skull fractures and contusions in addition to the rotational injuries; 5. primary hyp ox ic-ischaemic brain injury resulting from apnoea due to brainstem injury. PREDISPOSING FACTORS
A number of factors predispose the infant to injury from shaking: 1. The young infant has a relatively large and heavy head in relationship to the body size. 2. The neck muscles are weak and there is little head control in early infancy. As the child gets older the neck muscles become stronger so that the head moves with the body; shaking whiplash injuries are rare, therefore, after the second year of life. 3. The extracerebral space is relatively large (physiological craniocerebral disproportion). 4. Th ere is physiological laxity of the meninges in infancy compared with older children, with less anchoring of the brain. 5. In comp lete myelination in the white matter w ith a full complement of neuronal cell bodies in the grey matter means a difference between the specific gravity of the grey and the white matter and a gradient of tangential accelerations that produ ces shearing. BIOMECHANICAL DETERMINANTS OF SHAKII\IG INJURY
An external force has first of all to overcome the neck mus cles and induce anteropos terior movement with rotation. This sets up shearing forces within the skull and within the brain itsel f (rotational shearing forces). As with impact injuries, the amount of damage will depend upon the rate of acceleration/deceleration before the direction is reversed. Soft impact brings the skull to a sudden halt with maximum stretching of the bridging veins, thus the force used is a major determinant of the injury sustained. Sufficient force is necessary to Jift the weight of the infant and accelerate it in space and pOSSibly add additional force by active impaction of the child. The force with which the infant hits the surface will be dependent upon the strength of the individual; the angular forces (i.e. torque), which cause rota tion (see above) do not cause fractures but are of paramount
303
importance in shaking injuries. The forces within the skull resulting from shaking against a soft surface, such as a cot mattress, are signi ficantly less than the forces arising from being shaken against a hard surface, such as a wall. Two biomechanical studies by Duhaime et al 146 and Minns et al 147 found the angular accelerations to be of the order of 1139 rads -2 (mean) and 200-892 rads -2 (range) respectively. The forces exerted by the shaker are dependent on the age, height, gender and strength of the individual shaking, so that young children (3 years) are unable to lift or shake a weight equivalent to a 2-month-oJd child. Acceleration patterns increase gradually to adult values by the age of 10 years. The mean frequency for adult volunteers shaking manikins was 3.5- 2.7 shakes per second; for children was 2.1-1.8 shakes per seco nd. Males shake significantly faster than females (P < 0.0001) but the frequency is independ ent of the load. In order to produce brain damage, a minimum duration of shaking is required. It is likely that relatively fewer shakes would be necessary to induce a whiplash type of stem injulY than those required to produce damage from repetitive rotation. There is also a maximum duration of shaking that is possible before fatigue prevents further shaking. Experimentally determined maximum possible durations for shaking may not be exactly similar to dura tions or rates during situations associated with anger and rage, but for ad ults shaking manikins that a re equivalent in weight to a 2-month-old baby the median duration was 21.5 seconds; for a manikin equivalent to a 5-month-old baby the median duration was 19.75 seconds. Duration sustained was significantly longer in men than in women and was related to the strength of the weakest arm and to the height of the individual. It is, therefore, not possible for adults to shake infants for prolonged uninterrupted periods. The centre of the rotational arc for shaking infants is likely to be in the upper part of the cervical spine, between the craniocervical junction and C2. The posterior excursion of the head during shaking is limited by the spinal joint ang les and ligamentous attach ments, which decelerate the movement more slowly than the sudden deceleration from shaking with impact. It is possible to shake fast with shOlter angles and slower with greater angular excursions. The resultant brain dam age may be a product of the speed and degree of rotation. The pericerebral space in infants can be up to 1 cm in depth to allow for subsequent brain growth. This cranio cerebral disproportion with increased CSF spaces reduces the inertia for any brain movement consequent on cranial movement, and more swirling of the brain is possible with stretching of the bridging veins. Measurements of the sinocortical distance (from the superior sagittal sinus to the cerebral cortex) were made during ultrasound scanning of the brain, showing a change in this space as a result of a change in the child's position, i.e. supine versus sitting.
~===7===A=C=C=ID=EN=T=A=L=H=EA=D=I=NJ=U=R=Y==~~~ll MALPRESENTATION or DISPROPORTION
PEN ETR ATl ON
BRAIN COMPRESSION
PENETRATING INJURY
BRAIN DISTOR TI ON
Tracti on (stretch and tea r) of falx and tentorium
Overriding parietal bones
IMPACT (AcceIera ti onl decel era ti on)
INDIRECT CONTUSION CORTEX
Cervical root pocket hae morrh ages
Tears of vein of Galen and venous Tearing bridg ing veins
Figure 14.12
Mechanisms of brain injury in different situation s. CBF, cerebral blood flow; CPP, cerebral perfusion pressure; EDH, extradural
NON-ACCIDENTAL HEAD INJURY
TYPES OF HEAD INJURY
SHAKING
WHIPLASHING
ROTATION
WHIPLASH
SHAKIN G IN JURY
MECHANISMS OF INJURING
MECHANISMS OF BRAIN INJURY
CERVICOM EDU LLARY PERCU SS ION AND TRACTI ON
ANATOMICAL LEVEL
PRIMA RY INJURY
CHRONIC SDH
ASPHYX IA
INTRADURAL HAEMORRHAGE SPACE
OCCUPATION
RAISED ICP
Art eria l territory
(ACA. PCA. PICA and pericallosa l) infarct
(contd) haemorrhag e; ICP, intracrania l pressure; SAH, subarachnoid haemorrhage; SDH, subdural haemorrhag e.
SECONDARY INJURY
306 I
Head and neck injuries
PATHOPHYSIOLOGY OF SHAKING INJURIES
CHRONIC SUBDURAL HAEMATOMA
There are many intermedialY pathophysiological mecha nisms that will determine the final pattern of brain damage (Fig. 14.13). The esse ntial pathophysiological mechanisms are: • surface shearing - cortical emissary veins - subdural haemato ma; • parenc hymal shearing; - midbrain shearing injury; - shearing of grey-white interface; • cervicomedullary whiplash injury ; - vascular injury: carotid compression and
vertebrals in whiplash;
• ischaemia-hypoxia, secondary shock and cerebra l oedema with raised rcp; • contusion from impact injury.
A chronic subdural haematoma is often clinically silent with only an inappropriately increas ing head circumfer ence. The causes of, and conditions predisposing to, a sub dural haem ato ma, are seen in Table 14.6. A chronic subdural is not a true haematoma but a subdural effusion in that it is mainly water with a haematocrit that is usually less than 10 per cent. It is usually due to failure of the acute haematoma to resolve. An acute haematoma may become cill'onic because of: (1) its large volume; (2) a n increased osmolarity of the subdural haematoma, causing ingress of water; (3) continued fresh bleeding; or (4) a true inflammatory pachymeningitis elicited by red cell break down. It has two components: a liquid low-haematocrit 'haematoma', with evidence of continued fresh bleeding, and a membrane. The membrane, which is vascular and easily bleeds, encapsulates the haematoma and binds it to the dura, where it undergoes degradation and invasion by fibroblasts. Calcification may be detected by 3 months. Incorporation of the haematoma into the dura as a membrane is the basis of the healing process. Brown staining of the dura, resulting from deposition of haemosiderin during haemoglobin break down, persists for many months (see Fig. 11.1, p. 206).
SUBDURAL HAEMATOMA
A haematoma in the subdural space exerts pressure on the underlying brain parenchyma with secondary ischaemia to the immediately underlying white matter and, later, to the grey matter. It also exerts pressure on the overlying calvarium and may cause erosion of the inner table of the skull. An acute subdural haematoma occurs within 3 days of injury a nd is usually associated with severe shock and brain contusion as well as oedema. A subacute haematoma occurs between 3 days and 3 weeks of injury, and a chronic subdural haematoma appears more than 3 weeks from the time of injury.14B
Conditions associated with, or predisposing to, subdural haemorrhage or hygroma
Table 14.6
Figure 14.13 Primary brainstem haemorrhage at pontine level in a 3-year-old girl, who fell 30 feet.
Birth trauma Accidental trauma Non-accidental head injury Aspergillosis Kawasaki's disease Osteogenesis imperfecta Glutaric aciduria type 1 Meningitis (pneumococcal or group B streptococci) Septicaemia Haemophilia/factor 5 deficiency/anticoagulant treatment Idiopathic thrombocytopenic purpura Malignancy Atrioventricular malformation or aneurysm Post-cardiopulmonary bypass Alagille's syndrome Disseminated intravascu lar coagulation Menkes' kinky hair disease Slit ventricle syndrome Hyperosmo lar dehydration from mannitol Wide extra-axial space Prema tu ri ty Physiological Familial 7Autosomal dominant Renal dialysis Severe dehydration Congenital subarachnoid cyst
~
---=-
Chronic subdural haematoma
--
- - --
Secondary mechanisms of brain injury I
INTRADURAL HAEMORRHAGE Intradural haemorrhages are not specific for NAHI ; they are usually microscopic, venous/capillary in origin and equivalent to intradural petechiae, and have been described after fetal, perinatal and later neonatal death and have been put fonrvard as a source of subdural haemorrhage in infants.149 They are likely to be due to agonal anoxia in many cases, although other causes of hypoxia may be involved. The role of intradural haemorrhage as a source of sub dural haemorrhage is an essential component of the 'uni fied theory' of causation of the triad of brain swelling, subdural and retinal haemorrhages, which occurs as a direct result of apnoea in cases when there is no pericranial impact injury or injury elsewhere tha t might be interpreted as inflicted. 149 A recent study of 82 infants with hypoxic-ischaemic encephalopathy from a variety of causes found no concomitant subdural haemorrhage in any case, suggesting that other factors are necessary to produce subdural haemorrhage in infants . 150
307
Strich l5J first described primary shearing of whi te mat ter. Long axons, such as those found in the commissural fibres and long association tracts, can be avulsed by trac tion but axonal damage can also be seen in hypoxic ischaemic damage. Further projection fibres from the cor tex may be sheared at the grey-white interface and be seen as petechiae on MR imaging. 14J It is likely that both white matter shearing and white matter oedema contribute inde pendently to the gross cerebral atrophy of the white matter and ex vacuo ventricular dilatation. The infant's axons may stretch more than fully myeli nated adult fibres l54 but, when sheared by traction forces, they exude axoplasm, which appears as 'retraction balls' on staining, followed by microglial scars after 2 weeks and Wallerian degeneration after 6-8 weeks.1 55, 156 White matter shearing injuries are seen, in particular, in the corpus callosum, the superior cerebellar peduncle and in the midbrain. It has alternatively been suggested that hypoxic ischaemia, resulting in vascular axonal injury, may be the main cause for the axonal damage in shaken baby syndrome. 116,11 7
SUBARACHNOID HAEMORRHAGE Approximately 20 per cent of infants with a NAHI have a significant subarachnoid haemorrhage as well as a subdural haemorrhage. The blood in the CSF may cause ischaemia to the brain itself by inducing arterial vasospasm, and may lead to secondary hydrocephalus requiring CSF diversion.
PARENCHYMAL SHEARING INJURIES Midbrain Shearing Injury The midbrain acts as pivo t upon which the cerebral hemi spheres can rotate, with the immediate effect of concus sion. In impact injuries associa ted with concussion, the sudden coma is due to sudden rotation of the cerebral hemispheres upon the more-anchored midbrain, i.e. a pri mary stem injury (Fig. 14. 14) .1 51 The clinical picture is of dilated pupils, loss of vision, decorticate posture, akinetic mutism and abnormalities of blood pressure, pulse and res piration. The midbrain may be damaged from the primary injury described above or as part of a tentorial herniation from raised ICP with brain stem haemorrhage (Duret haem orrhages). The reticular formation is very oxygen depend ent and consciousness is rapidly lost with lowered oxygen tensions, i.e. hypoxia. White Matter Shearing Injuries The increase in brain weight over the first 4 years of life is mainly due to an increase in white matter myelin. Grey matter is firm and cellular, whereas poorly myelinated white matter is more gelatinous and of slightly different density; t his means that with 'shaking' the white and grey matter swirl at different velocities, giving rise to character istic tears in the parenchyma for each type. 152
SECONDARY MECHANISMS OF BRAIN INJURY In addition to primary brain injuries, patients with head trauma also suffer from secondary brain damage, mostly of ischaemic origin (Fig. 14. 15). These occur in response to the initial trauma, resulting in secondary physiological deran<JeI~' b ments such as hypOXia , low blood pressure, raised ICP, reduced cerebral perfusion, pyrexia and disturbances of global brain oxygen ex traction (Sj02)' These intracranial and systemic secondary physiological insults may arise during the initial resusci tation, or la ter in the reu, and cause inade quate delivery of oxygen and substrates (particularly glu cose), resulting in secondary brain ischaemia and further damaging the already injured brain . I 50,159 Prior to advances being made in acute resuscitation and transport of comatose head-injured patients, up to 30 per cent of these patients were reported to have hypoxaemia and 15 per cent had arterial hypotension on arrival at the ARE department. 159- 164 In 1981, Miller's group 162 reported a high frequency (53 per cent) of raised ICP (levels > 20 mmHg) in head-injured patients during intensive care, and data from the Traumatic Coma Data Bank (TCDB) indicated t hat the incidence of raised ICP among severely brain-injured adults requiring intensive care was even higher. 165 In this study, 654 patients were examined and 72 per cent had ICP insult based upon end-hour ICP recordings. 165 During their intensive care management, head-injured patients may require to have repea t CT scans. Andrews and co-workers l66 examined the number of sec ondary physiological insults occurring in TBI pa tients being transported within the hospital, mos tly between the ICU and CT scanning suite, and found a frequency of
308 I
Head and neck injuries
(b)
Figure 14.14 (a) Haemorrhage in the spinal subdural space. (b) Same case : axonal swellings in cervical nerve roots, beta amyloid precursor protein staining. (Section courtesy of Professor JE Bell, Edinburgh.)
Hae matoma
Cerebral oedema
Impairment of circulation
Rai se d ICP Vasospasm
Infec ti on
Epil epsy
Hydrocephalus
Hypoxia
Hypercarbia
Hypotension
Fever
Ana emi a
Hyponatraemia
Hypoglycaemia
Figure 14.15 Primary and secondary brain injury. DAI, diffu se axonal injury; ICP intracranial pressure; T81, traumatic brain injury.
50 per cent during tra nsit and > 65 per cent in the ICU dur ing th e subsequent 4 hours of continuous monitoring. From the TCDB da ta , a significant increase in mortality and morbidity w as associated with hy poxia or hypotension from injUly throu g h resuscitation in the 717 adults who were st udied after head injury.1 57 The relationship of out come and secondary physiological insults occurrin g during the intensive care management after head injUly was deter mined in a prospective study involving 124 adults by Jones and coll eag ues, 53 who used a computerized data collection system to measure up to 14 clinica lly indicated physiolog ical v ariables (such as rcp, arterial blood pressure, CPP, heart rate, oxygen saturation, and core and periph eral tem perature), minute by minute. 53 Abnormal valu es for each
~-~~--~~~~-~
~
Secondary mechanisms of brain injury I
variable falling outside the pre-set threshold limits as defined by the Edinburgh University Secondary Insult Grading Scheme for ?5 minutes were analysed by insult grade and duration. Ins ults were found in 91 per cent of patients, regardless of the severity of head trauma, their age or the admission injury severity score (ISS).53 The cumulative durations were much greater than previously recorded and the most significant predictors of m0l1ality in this study were durations of hypotensive, pyrexic and hypoxaem ic insults. 53 Hypotensive insults and pupillary response on admiss ion were significant predictors of good versus poor outcome. 53 Only a small amount of research has been done with regard to investigating the clinical aspects and patho physiology in children with head injury apart from a few specialized studies 167, 168 that primarily examined CBF, cerebral hyperaemia, cerebral metabolic rate of oxygen (CMR0 2) and arteriovenous difference of oxygen (AvDOJ, This study found that cerebral hyperaemia was uncommon (7 per cent) a nd that the CMR0 2 and cerebral oxygen extraction was w ithin the normal range in 81 per cent of children with severe head injuries. Cerebral metabolic rate of oxygen and cerebral oxygen extraction fell between the first and third day after the injury, and although the CBF rose throughout the ac u te monitoring time this was not significant. These findings led the investigators to conclude that children were most vulnerable to secondary brain damage short.ly after the injury. In addition, they found that cerebrovascular autoregulation was preserved (signifi cant correlation between CPP and central venous pressure [CVP] , P = 0.0003), the cerebrovascular resistance (CVR) was normal (in 58 per cent), increased (in 32 per cent) and could be impaired in the most severely injured. The ICP was inversely propol1ional to the CBF (P = 0.009) and all parameters (CMR0 2, CPP, 0 2/ partial pressure of arterial car bon dioxide [Paco2], cerebrovenous pH) were independent determinants of CVR. 169 Another complicating factor for neurotrauma research involving children is that the normal ranges for different physiological param eters, such as ICp l70 and blood pres sure,171-174 are age related and change from birth through childhood and adolescence to finally reach adult levels. Pre vious studies by the TCDB 23 grouped all children aged 0-15 years together and used the same end-hour threshold as that used in their adult studies. Jackson's groupl75 attempted to quantify the CPP secondary insults in children using only three different adult CPP thresholds for their patients who were aged 16 years or less. Similarly, Sharples and co-workers used a single threshold value of > 20 mmHg to define raised ICP in their study involving child ren aged 2- 16 years 01d. 166 In a more recent study, the cause and incidence of second ary insults in severely head -injured ad ults and children were examined by Chambers and colleagues. 176 Again, unified threshold values for CPP, ICP and arterial blood pressure were used on their patients, regardless of their age and developmental maturity.176 Using a standard definition
309
of abnormal physiological values, defining raised ICP, hypotension and hypertension at all ages is too simplistic. Jones and co-workers l77 from Edinburgh used age specific physiologica l norms to quantify secondary phys io logical derangements prospectively in 54 children aged < 16 years with TBI. They found that secondary physiolog ical insults occurred frequently in children after head injury, w ith rcp and CPP derangement demonstrated in 95.45 per cent and 77.27 per cent of patients respectively. Furthermore, the duration of CPP derangement was found to predict outcome (dead versus alive, P = 0.003).177 The point at which physiolo gica I derangements become impor tant, i.e. by inducing ischaemi a in terms of magnitude and duration, remains unclear, as does their effect on outcome. A recent two-centre study 178 involving 86 head-injured children, aged from 2 to 15 years, admitted within 24 hours of injury, had prospective time series physiological data downloaded from ICU monitors every minute for 6 physio logical variables (ICP, MAP, CPP, oxygen saturation, tem perature and heart rate) . Critical thresholds of CPP were identified and, ideally, CPP should be kept above 48 mmHg for children aged 2-6 years, above 54mmHg for children aged 7-10 years, and above 58 mmHg for children aged 11-16 years. Duration of CPP was found to be the best pre dictor of survival (P < 0.000l), and a highly significant discriminator between good (Glasgow Outcome Score [GOSs] 4 and 5) and poor outcome (GOSs 1, 2 and 3) (P = 0.005). A new cumulative pressure-time index (CPT) was developed to include severity and duration and the CPT was useful in predicting independent/poor outcome. 179
CEREBRAL OEDEMA Cerebral oedema is defined as an increase in the volume of the w hole (generalized) or part (focal oedema) of the brain due to an increase in the water content. 180 Cerebral oedema is a frequent accompaniment of shaking injuries and has been estimated in as many as 42 per cent of such infants, based on imaging. 181 The frequency is doub led in post mortem studies and Geddes ll6 ,117 confirmed that cerebra l oedema was a major cause of death in 82 per cent with evi dence of hyp ox ic ischaemia in 77 per cent. It takes approx imately 6 hours to appear after the injury.!52 Cerebral oedema has several mecha nisms of production, as see n in Table 14.7.182 The cause of the oedema in the shaking injuries is probably multifactorial, with vascular damage (as in the eye) causing vasogenic oedema, white matter damage. disrupted venous damage from shearing, damage to the blood-brain barrier, high central venous pressure from chest compression, brain necrosis from shock and imp aired perfusion, and secondalY hydro cephalus causing hydrostatic oedema from obstruction of the arachnoid granulations by blood. Cerebral oedema results in a deteriorating level of con sciousness. with signs of brainstem dysfunction in pulse, respiration, blood pressure and pupils. which may progress
310 I
Head and neck injuries
Table 14.7
Mechanisms of production of cerebral oedema
Intracellular (cytotoxic)
Extracellular (vasogenic)
Vasogenic
Hyd rostatic Hydrocephalic Osmotic Iatrogenic Necrotic
Intramyelinitic (myelinac/astic)
Grey matter (especially astrocytes and oligodendroglia) Energy failure of the cell (from hypoxia) causes membrane pumps to fail, i.e. inability to excrete water or exclude sodium, plus intracellular proteolysis causes more idiogenic osmolysi s, finally resulting in hydropic cells White matter due to damage of blood-brain barrier with albumin leakage Damage to the blood-brain barrier (from asphyxia and head inju ry) causes egress of glycine, glutamate, noradrenaline and albumin into the brain; album in degrades to peptides with a strong osmotic effect (cerebrospinal fluid [CSF] and protein similar to blood) At the 'break-point' of cerebrovascular autoregulation an increased perfusion pressure is transmitted to the microcirculation (from anoxic ischaem ic injury/hypercarbia or nitroprusside) Re versa l of the no rma l transepend ymal flow of CSF back into brain parenchymal extracellular fluid Compartmentalized osmotic gradients cause shifts of fluid between: blood/ECF, blood/CSF, CSF/extracellular fluid (ECF), and ECF/in trace llular fluid From inappropriate antidiuretic hormone, hypotonic fluids, disequilibrium syndrome from reducing hypernatraemic states Ischaemia from any cause results in: (1) cellular necrosis and lysosomal rupture with release of vasoactive peptides that are osmotically active; (2) release of thromboplastin, causing thrombosis of the microvenous circulation and increasing vo lume of the infarct; and (3) cell membranes and blood vesse ls be ing necrosed with no possib le osmotic gradient Toxic drug or metabolic effect
to signs of 'coning'. Coma itself maybe due to raised ICP or to primary midbrain injury. Extensor hy pertonus follo wing asphyxia may be due to oedema, but is more often a dys tonia as a result of basal ganglia involvement - i.e. post as phyxial rigidity, which is not influenced by treatment to reduce ICP and which is often normal or secondary to necrotic oedema. I S3 Imaging shows a reduction in lateral ventricular size followed by third ventricle size and, eventually, the ambi ens cistern. The imaging loss of the 'sm iling face' means it is imperative that a lumbar puncture is not performed as there is some existing tentorial herniation. Scans show hydrocephalic oedema as a hypod ense 'bat wing' around the lateral ventricles. It is now possible to measure brain water using MRI technology.
HYPOXIC/ISCHAEMIC INJURY Ninety per cent of children dying from accidental head injuries have evidence of severe hypoxic ischaemic brain damage, 155 whereas 77 per cent of children dying from NAHl are found to have hypoxic ischaemic damage at autopsy. 116, 11 7 The hypox ic/ ischaemic damage may be part of the pri mary injury (apnoea or medullary injUly) or result from secondary insults such as hypotensive shock, raised ICP, decreased perfusion pressure and seizures. The 'big black brain' indicates particularly severe hypoxic ischaemic dam age, which is followed by death or the rapid development of cerebral atrophy. Additionally, suffocation may have occurred to quell the child's crying and hypoxic ischaemia is almost always an agonal event in most infant deaths.
Cerebral Infarction from Impaired Perfusion Brain infarction occurs as a result of hypotension, raised ICP, brain shifts and vascular occlusions, spasm or obstruction. Glucose is the main energy source for the brain: under aerobic conditions 38 molecules of ATP are produced but only two are produced under anaerobic conditions. How ever, under anaerobic conditions the supply of glucose and removal of lactic acid is essential. The ATP produ ced is required for neurotransmitter formation, transpOli mecha nisms, maintenance of membrane pumps and polarization of membran es. The oxygen supply depends on the POz and the haemo globin concentration. Ischaemi a is far more dangerous than hypoxia because there is now no glucose or ketones, or any way of removing lactic acid, so neuronal necrosis rapidly occurs. Ischaemia initially results in swelling of the mitochondria within the neurones and astrocytes as a result of water reten tion. The neurones release potassium, which is collected by the astrocytes; glycogen also accumulates within the astro cytes, ca using further swelling. ls4 Second, there is leakage of protein (and blood - haemorrhagic infarct) through the endothelial cells into the extracellular space. Subsequent breakdown of the red cells causes iron and bilirubin forma tion in the tissues, which is used to date the infarct. Finally, lysosom al rupture, cell death and biochemical breakdown of tissue proteins results in swelling of the infarct (further com pressing capillaries), breakdown of the blood-brain barrier, loss of autoregulation, and sludging (from thromboplastin release) and liquefaction of the tissues.
Injury to the cervical spinal cord I
The necrotic tissue may be absorbed by macrophages leaving a cyst - as in periventricular leucomalacia - or may cause an astrocytic reaction, replacing the dead tissue with the formation of a gJial scar and capillary proliferation. Six to twelve hours after acute hypoxic ischaemia there may be a therapeutic window, possibly due to a biochemical cascade triggering apoptosis or a slowing down of metabo lism caused by adenosine and adrenocorticotrophic hormone.
Patterns of Infarction Hypotension does not result in uniform infarction of the whole brain. 18S The four feeding arteries to the circle of Willis means there are areas where the pressures are equal, resulting in 'no flow', and ischaemia may occur in any major vessel from Jow perfusion pressure. When cerebral blood flow is severely impaired, water shed zones between adjacent vascular territories are suscep tible. In the newborn this is between centrifugal and centripetal arteries, causing periventricular leucomalacia. In the older patient it is between the middle and posterior cere bral arteries (causing infarcts in the pericentral white mat ter of the optic radiations and posterior temporal lobe) or between the anterior and middle cerebral arteries (causing a wedge shaped infarct in the 'leg area'). This susceptibility is because cerebral arterioles are 'end arteries' and do not anastomose. Patterns of infarction also result from brain herniations pericallosal or subfalcine shifts, posterior cerebral (from tentorial herniation) and posterior inferior cerebellar artery (PICA) from foramen magnum coning. There is also selective vulnerability of cell types: neu rones are the most sensitive, whereas microglia and blood vessel cells are the least sensitive. Ischaemia occurs in the brain-neocortex, basal ganglia, hippocampus and cerebel lar Purkinje cells. 186 In head injuries, hypoxic ischaemic damage occurs to these selective areas as well as watershed zone infarction. 12
INJURY TO THE CERVICAL SPINAL CORD Injuries to the spinal cord in children are rare. They occur during birth, and after accidental and non-accidental injuries. In a study by Augutis and Levi 187 the incidence was found to be 4.6 per million children per year (95 per cent, confidence intervals 3.6-5.5); however, when pre-hospital fatalities were excluded the incidence was 2.4 (95 per cent confidence intervals 1.8-3.1). Children's spinal injuries rep resent about 5-10 per cent of all spinal injuries.188.189 In most series, boys predominate. ACCIDENTAL SPINAL INJURY
The aetiology of accidental spinal injury in children is dif ferent from that in adults; falls and pedestrian-related
311
vehicular accidents are most common causes in children, followed by sports-related injuries (e.g. rugby and horse riding). Although the incidence is lower in children than in adults, it carries a worse prognosis and involves a high incidence of severe, permanent neurological disability due to spinal cord involvement. 190 The anatomical and biomechanical characteristics of the child's spine are very different from those of the adult and the cervical spine is particularly at risk because of the high weight and volume head-body ratios, ligamentous laxity and incomplete head control. Younger children have more cervical lesions - the vertebrae are not yet ossified, the unci nate processes are flat and facet joints are horizontal, result ing in a different pattem of cervical spine injury in children. Fractures involve the growth plates. 191 Zabramski et al 192 reported that 71 per cent of children with spinal injuries who were < 10 years old had cervical lesions, and that one-half of these lesions were between the occiput and C2. There are six physical mechanisms that give rise to spinal and spinal cord injUly: I. flexion dislocation; 2. flexion compression; 3. compression burst fractures in which bone fragments are exuded into the spinal canal; 4. spinal extension versus penetrating injuries; 5. distraction-rotation injury. 193 Upper cervical spine injuries take the form of atlanto occipital dislocation, from birth or other trauma, and odon toid fractures that, in small children, are really epiphyseal detachments. Cervical dislocations and compression frac tures are most common in older children. Typical Jefferson fractures, Hangman fractures and cervical strains are exceedingly rare. From these mechanisms of injury, primary spinal cord injlllY results from contusion and compression of the spinal cord, which, within a few hours of injury, results in either cord oedema with ischaemia or occlusion of the intramedullary vessels, resulting in haemorrhagic necrosis. With resolution the necrotic areas leave cavitations or syrinxes in the cord at the level of the injury. 194 Pang and Wilberger llO described spinal cord injury with out radiographic abnormalities (SCrwORA), which is a traumatic myelopathy without evident fractures (due to hyperflexion or hyperextension, or distraction or repetitive flexion-extension injuries), with primary damage to the cord, or cord injury and ischaemia secondary to hypoten sion, or interference with the vertebral artery blood flow (bruising of the adventitia with intramural clots or occlu sion).195 Bollini 196 estimated that almost one-half of such lesions were in the cervical region. Spinal shock with a total flaccid paralysis below the level of the region will last for a variable time after the ini tial spinal injury until cord swelling resolves. Painful stim uli wiU not elicit flexion of the limbs. The neurological findings will point to the classical syndromes of cord
312 I
Head and neck injuries
involvement, i.e. complete, posterior cord or anterior spinal artery territory, central cord, root involvement or Brown Sequard's syndrome cord lesion, whether the spine is stable or unstable. 197 Treatment for the above spinal injuries may take the form of halo distraction or other traction, occipitocervical fixations with bone grafts or progressive traction with transcranial tongs. Spine boards are not advisable because they tend to force the large head into flexion, which can compromise both airway and spine. Spinal cord injuries in children are prone to result in spinal deformity because of imbalance between the paralysed muscles involved in postural control, or assymmetric lesions of growth plates. Syringomyelia is common at the site of the injury but may extend over several segments when the classical symptoms of head and neck pain, made worse on straining, alert one to the di agnos is. Survivors have a high frequency of bulbar and associated brainste m deficits.
column can move in the anteroposterior direction, allow ing a degree of slippage (spondylolisthesis). Whiplash injury can therefore cause cord concussion fro m repe ated partial subluxations, as well as traction with bruising and vertebral artery and anterior sp inal artery lesions. The presentation of infants with cervicomedullary injury is with apno ea and seco ndary hypoxic damage. Death from hypoventilation and apnoea may therefore be from a primary injury to the respiratory centre in the medulla or from associated vaso pa ra lytic shock w ith loss of vasomotor tone, also from medullary injury. The result is severe hypoxic ischaemic damage with oedema. These cases highlight the need for the MRjCT scans to include the cervical spine in all suspected cases and for the cord to be examined at necropsy. Cullen 200 and Swischuk 201 described compression fractures, subluxations, and fracture disloca tions in thoracolumbar levels (and one thoracic level lesion) in 11 in fants suffering NAl.
TRAUMATIC SPINAL INJURIES FROM BIRTH
GENETIC INFLUENCE ON RECOVERY FROM TRAUIVIATIC BRAIN INJURY
Newborns suffer spinal lesions as a result of a traum atic delivery due to breech and other malpresentation. Overall, 70 per cent of neonatal neck injuries result from breech and 30 per cent from cephalic deliveries. 198 Towbin 199 estimated that spinal cord injury was respon sible for 10-33 per cent of neonata l deaths 40 years ago. Additional ligamentous and spi nal cord lesions are often found without fracture at autopsy. These usual spinal injuries in the nevvb orn take the form of atlanto-occipital and atlantoaxial dislocation, frac tures of the odon tOid, and complete transections of the spinal cord. These injuries result in sudden death, tetra pare sis, bilateral Erb 's palsy, diaphragmatic and respiratory paralysis, root pocket haemorrhage, spinal root avulsion, spinal epidural haemorrhage and intraparenchymatous cord haemorrhages. Newborns may be shocked, hypotonic and apnoeic.
NON-ACCIDENTAL SPINAL INJURY Caffey.11J in his origi nal description , labelled a NAHI as a whip lash shaking injury. The pathology of spinal subdural haemorrhages, in farcti ons and root avulsions from non accidental trauma are very similar to the spina l patholo gy of newborns after bilth injury shown by Yates. 19S Both trauma and whiplash are hyperextension injuries. Ged desI14.11S showed epidural spinal bleeding, localized axonal damage to the craniocervical junction and damage to the spinal nerve roots and brainstem in one- third of her fatal cases. The infant head pivots at approx imately the C2 level and lower in the older patient, hence the cervicomedullary damage. The posterior ligam ents and joint capsule are more elastic in the infant and the cervica l spine is exten sile; traction will allow elongation, whereas the spinal cord itself cannot stretch and aV1.llses. The cervical vertebral
Despite vigi lant neurointensive care, outcome remains diverse in head-injured patients. Accumulating evidence over the recent years indicate that genetic factors, particu larly apolipoprotein E (APOE) genotypes, may influence the recovery after brain injury. 202.20J The human APOE gene locus is located on chromosome 19 and has t hree allelic forms - epsion 2 (t:2), t:3 and t:4 with a frequencies of 0.08, 0.77 and 0.15, respective ly, in the general population. 204 lts gene product, apoE, is a 37 kDa glycosylated protein, which forms a major component of the lipid transport system in the brain,20s being synthe sized astrocytes in particular. 206 Teasdale and co-workers 207 examined the relationship of APOE genotypes and functional outcome in 93 head-injured patients and found that patients with APOE e4 allele were more than twice as likely to have an unfavourable outcome (dead, vegetative state or severe disability) 6 months follow ing the initial in su lt as those without it. This association has been confirmed.202.20J The influence of APOE genotypes on the conversion of beta amyloid precursor protein (~APP) to beta amyloid (~ A4) is another postulated mechanism that has gained considerable interest in the last decade. In certain circum stances in predisposed individuals, ~APP may give lise to deposits of :3A4, which is known to be toxic to neurones. 208 Beta amy loid depOSi t ions have been found in up to one third of patients who died from brain trauma. Nicoll's group 209 found a hig her frequency of the APOE e4 allele among thos e who had ~A4 deposition after head injury. The majority of the scientific interest in APOE genotype and brain trauma has focus ed thus far on ad ults and its influence on the recovery of head-injured children remains unknown.
References I
REFERENCES Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale. Lancet 1974; 2 :81-4. 2 Simpson D, Reilly P. Pediatric coma scal e. Lancet 1982; 2:450. 3 Raimondi AJ, Hirschauer J. Head injury in the infant and todd ler. Coma scoring and outcome scale. Childs Brain 1984· 11 :J 2-35. ' 4 Jennett B. Teasdale G, Galbraith S et al. Severe h ead injuries in three countri es.] Neu ral Neurasurg Psychiatry 1977 ; 40 :29 1-8.
5 Hutten loc her PRo Reye's sy ndrome: relati o n of o utco me to
therapy.] Pediafr 1972 ; 30:845-50.
6 Ses hia SS, Seshia MM, Sachdeva RK. Coma in childhood. Dev
Med Ch ild NeuroI 1977; 19:614-28. 7
Morray JP, Tyler DC, J ones TK et al. Coma scale for us e in brain-injured children. Crit Care Med 1984; 12:1018- 20. 8 Gordon NS, Fois A, Ja cob i G et al. The management of the comatose chi ld. Neuropediatrics 1983; 14:3-5. 9 Yager lY, J ohns ton B, Seshia SS. Coma scales in pediatri c pra cti ce. Am] Dis Child 1990; 144:1088-91. 10 Yates DW. RAD Lecture Notes on Emergency Medicin e. Oxford: Blackwell Scientific. II Tatman A, Warren A, Williams A, Powell JE, Whiteh ouse W. Development of a modified paediatric coma scale in intensive care clinical practice. Arch Dis Child, 1997; 77:519-21. 12 J enn ett B, MacMillan R. Epidemiology of head injury. BM] (c/in Res) 1981 ; 232:101 - 4. 13 Kraus JF, Bla ck MA, Hessol N et al. The incidence of acute brain injury a nd se ri ous impairment in a defined population.
25 Arnarson EO, Halld o rsson JG. Head trauma among children in Reykjavik. Acta Paediatr 1995 ; 84:96-9. 26 Jennett B. Epidemiolo gy of head injury. ] Neurol Neurosurg Psychiatry 1996; 60: 362 - 9. 27 Gabella B, Hoffman RE, Marine WW, Stallones L. Urban and rural traumatic brain injuries in Co lorado . Ann Epidemiol 1997; 7:207-12. 28 Emanuelson I, Wendt L. Epidemiology of trauma tic brain injury in children and adolescents in so uth- western Sweden. Acta Paediatr 1997; 86:730-5. 29 Suominen P, Kivioja A, Ohman J e [ al. Severe a nd fatal childhood trauma. Inju/y 1998 ; 29 :425-30. 30 Durkin MS, Laraq ue D, Lubman L, Barlow B. Epidemiology and prevention of traffi c injuries to urb a n ch ildren and adolescents. Pediatrics 1999 ; 10 3 :e7 4. 3 1 Morrison A, Stone DH, Redpath A et al. Trend ana lysis of socioeconomic differentials in deaths from injury in childhood in Scotland, 1981-95. BM] 1999 ; 318:567-8. 32 Thurman D, Guerrero J. Trends in hosp ita lization associated with traumatic brain injury. ]AMA 1999; 282:954-7. 33 Benoit R, Watts DD, Dwyer K et a1. Windows 99: a source of suburban pediatric trauma.] Trauma 2000; 49:477 - 81; discussion 481-2. 34 Thurman D. Traumatic brain inju ry in children and youth as a public health pro b lem: an overv iew. In La ng lois JA (ed.)
Traumatic Brain Inju ry in the United States: Assessing Outcomes in Children. A tlanta, GA: Department of Hea lth and
35
Am] Epidemiol1984 ; 119 :186- 201. 14 Ann ege rs JF, Grabrow JD, Kurland LT, La ws ER Jr. Th e in c idence, causes, an d secular trends of head trauma in Olmsted County, Minnesota, 1935-1974. Neurology 1980; 30:912-19. 15 Rantakallio P, von Wendt L. Trauma to the nervous system and its sequelae in a one -y ea r birth cohort followed up to the age of 14 years.] Epidemiol Community Health 1985; 39:353-6. 16 Kraus JF, Fife D, Cox P et al. Incidence, severity, and external causes of pediatric brain injuly. Am ] Dis Child 1986; 140:687-93. 17 Chan BS, Walker PJ, Cass DT. Urban trauma: an analysis of 1,116 pa ediatric cases.] Trauma 1989; 29:1540-7. 18 Brookes M, MacMillan R, Cully S et a1. Head injuries in accident and emergency departments. How different are children from adults? ] Epidemiol Co mmunity Health 1990 ; 44:147-51. . 19 Sharples PM , Storey A, Ayn sley-Gree n A, Eyre JA. Causes of fatal childhood accid ents invo lving hea d injury in northern region, 1979-36. BM] 1990; 301:1193-7. 20 Tepas JJ JIl, DiScala C, Ram enofsky ML, Barlow B. Morta lity and head injury: the pediatric perspective. ] Pediatr Surg 1990; 25:92-5; discussio n 96. 21 Hansen TB, Pless S, Gravers M. Head injuries among children in the county of Ringkobin g. I. An a nalysis of hospital data. Ugeskr Laeger 1991; 15 3 :2944-6. 22 Levin HS, Aldrich EF, Saydjari C et a1. Severe head injury in children: experience of the Traumatic Coma Data Bank. NeurosurgelJ 1992; 31 :43 5-43; discussion 443-4. 23 AJdrich EF, Eisenb erg HM, Saydjari C et a1. Diffuse brain swelling in severely head-injured child ren. A report from the NIH Traumatic Coma Data Bank. ] Neurosurg 1992; 76:450-4. 24 Berney J , Favier J , Froid eva ux AC. Paediatric hea d trauma: influence of age and sex. 1. Epid emiology. Childs Nerv Syst J994; 10:514-16.
313
36
37
38
39
40
41
42 43
44 45
46
47
Human Services, Centers for Disease Control and Prevention 2000, pp. 31-5. . Reid SR , Roesler JS, Gaichas AM , Tsa i AK. The ep idemiology of pediatric traumatic brain injury in Minneso ta. Arch Pediatr Adolesc Med 2001; 155:784-9. Masson F, Thicoipe M, Aye P et al. Epid emiology of severe brain injuries: a prospective population-based study. ] Trauma 2001; 51 :481-9. Adekoya N, Thurman DJ, White DD, Webb KW. Surveillance for traumatic brain injury deaths: United States, 1989-1998. MMWR Surveill Summ 2002; 51:1-14. Williamson LM, Morrison A, Ston e DH. Trends in head injury mortality among 0-14 year olds in Scotland (1986-95). ] Epidemiol Community Health 2002; 56:285-8. Masson F, Thicoipe M, Mokni T et al. Epidemiology of traumatic comas : a prospective population-based study. Brain Injury 2003; 17 :279-93. Andersson EH, Bjorglund R, Emmanuelson I, Stalhammer D. Epidemiology of traumatic brain injury: a pOjlulation-based study in western Sweden. Acta Neu ral SCOlui 2003; 107:256-9. Roberts I, Power C. Does the decline in child injury mortal ity vary by social class? A co mpariso n of class-specific mortality In 198 1 and 1991. BM] 1996; 313:784-6. Reed lM, Browning JG, Wilkinson AG, Beattie T. Can we aboli sh skull X-rays for head injury? Arch Dis Child 2005; 90:859-64. Minns RA, Millar C, Minns FC et al. Non-accidental head injury in Scotland : Scottish Database. In Minns RA, Brown JK (eds) Shaken Baby Sy ndrome and Other Non-accidental Head Injuri es. London: Mac Keith Press, 2005, pp. 154-84. Barlow KM, Minns RA. Annual incidence of shaken impact syndrom e in young children. Lancet 2000; 356:1571-2. Jayawant S, Rawlinson A, Gibbon F et al. Subdural haemorrh ages in infants: population-based study. BM] 1998; 317:1558-61. Brown JK, Purvis RJ, Forfar JO, Cockburn F. Neurological aspects of perin ata l asphyx ia. Dev Med Child Neural 1974· 16:567-80. ' Whitby EH, Pal ey MN , Smith MF et a1. Low field strength magnetic resonance im ag in g of the neonatal brain. Arch Dis Child Fetal Neo /wta l Ed 2003 ; 88:F203-8.
314 I
Head and neck injuries
48 Hold en KR, Titus MO, Van Tassel P. Cran ial magneti c resonance imagi ng exam ination of normal term neonates : a pilot study.) Child Neurol 1999; 14 :708 - 10. 49 Yonas H, Wo lfson SK Jr, Dujovny Met al. Selective lenti culostriate occlusion in the primate. A hig hly focal cerebral ischemia model. Stroke 1981 ; 12 :567 -72. 50 Shaffer L, Rich PM , Pohl KR, Gan esan V. Can mild hea d injUly cause ischaemic stroke? Arch Dis Child 2003; 88: 26 7- 9. 51 Ko rs EE, Terwindt GM, Vermeu len FL et al. Delayed cereb ral edema a nd fatal coma after minor head trauma: role of the CACNA 1A calcium channe l subunit gene and relation ship with familial hemiplegic migrai ne. Ann Neurol 2001; 49:7 53 - 60. 52 Umile EM, Sand el ME, AJvaria A et al. Dynamic imaging in mild traumatic brain injury: support for the theolY of medial tempo ral v ulne rability. Arch Phys Med Rehavil 2002; 83:1506-13. 53 Jones PA, Andrews P J , Midgley S et al. Measuring the burden of seco ndary insults in head-inju red patients during intensive care.) Neurosurg Anesthesiol 1994; 6:4-14. 54 Jiang JY, Gao GY, Li P, Yu MK, Zhu C. Early indicators of prognosis in 846 cases of severe tra umatic brain injury. ) Nellrotrnuma 2002 ; 19:869-74. 55 Chiaretti A, Pia stra M, Pulitano Set al. Prognosti c factors and outco me of children with severe head injury: an 8-year expe rience. Childs Nerv Syst 2002; 18: 129-36. 56 Chiaretti A, De Bened ictis R, Della Corte F et al. Th e impact of initial management on the outcome of children wit h severe head inju ry. Childs Nero Syst 2002; 18:54 -60. 57 Coats TJ, Ki rk CJ, Dawson M. Outco me after severe head injury treated by an integrated trauma system. ) Acc Emerg Med 1999; 16 :1 82-5. 58 Zauner A, Bullock R, Di X, Young HF. Brain oxygen, CO 2 , pH, and temperature monitoring: eva lu ation in the fe lin e brain. Neurosurgery 1995; 37:1168-76 ; discussion 1176 -7 . 59 Jeevaratnam DR, Menon OK. Survey of intensive ca re of severely head-injured patien ts in the United Kingdom. BM) 1996; 312:944- 7. 60 Matta B, Meno n D. Severe head injury in the United Kingdom a nd Ireland: a survey of practice and impl ica tions for management. Crit Care {"fed 1996 ; 24: 1743-8 . 61 Segal S, Gall ag her AC, Shefl er AG et a\. Survey of the use of intracrani al press ure monitoring in children in the United Kingdom. intensive Care Med 2001; 27:236 - 9. 62 Gambardella G, Zaccone C, Cardia E, Tomasello F. Intracranial pressure monitoring in ch ildren: comparison of ex ternal ventricular device wi th the fiberopt ic system. Childs Nem Syst 1993; 9:470-3 . 63 Piek J, Bock WJ. Continuous monitoring of cereb ral tissue pressure in neurosurgical practice: experienc es with 100 patients. ill tellsil!e Care IVIed 1990; 16: 104-8. 64 Yablon JS, Latner HJ, McCormack TM et a l. Clin ica l experience with a fiberopti c intracrania l pres sure monitor. ) Ciill Moni t 1993; 9:171-5. 65 Shapiro S, Bow man R, Callahan J, Wolfla C. The fiberoptic intra parenc hymal cerebral pressure monitor in 244 patients. Surg Neurol 1996; 45:278-82. 66 Friedman WA, Vries JK. Percutaneous runnel ventri culostomy. Summary of 100 procedures.) Nellrosurg 1980; 53:662-5. 67 Narayan RK, Kishore PR, Becker DP et al. Intracranial pressure: to monitor or not to monitor? A rev iew of our experience with severe head injury. ) Neuro surg 1982; 56:650 - 9. 68 North B, Reilly P. Compariso n amo ng three m ethods of intracranial pressure recording. Neurosurgery 1986; 18:730-2. 69 Param ore CG, Turner DA Re lat ive risks of ventriculostomy infecti on and morbidity. Acta Neurochir (Wi en) 1994; 127 :79-8 4.
70
7J
72
73
74 75
76
77
78
79
80
81
82
83 84
85
86
87
80
89
90
91
Artru F, Terrier A, Gilbert let al. Monitoring of intracranial pressure with intraparenchymal fiberoptic transducer. Technical aspects and clinical reliability. Anll Fr Anesth Real1im 1992; 11 :424 -9. Pople IK, Mahlbauer MS, Sanford RA, Kirk E. Results and complications of intracranial pressure monitoring in 303 children. Pediatr Neuro su rg 1995; 23: 64-7. Jensen RL, Hahn YS, Ciro E. Risk factors for intrac ranial pressure mon itoring in children with fiberoptic devices: a critical rev iew. Surg Neuro/ 1997; 47:16- 22. Meyer JS , Kondo A, Nomura F et al. Cerebral hemodynamics and metabo lism following experimental head injUlY. ) Neuro surg 1970; 32:304-1 9. Kaiser AM , VVhitelaw AG. Hypertensive response to raised intracrani al pressure in infancy. Arch Dis Child 1988; 63:1461 -5. Rossi S, Zanier ER , Mauri I et a l. Brain temperature, body core temperature, a nd intracranial pressure in acute cerebral damage. ) Neurol Neuro surg Psychiatry 2001; 71 :448-54. Albrecht RF II, Wass CT, Lanier WL. Occurrence of potenti ally detrim enta l tem perature alterations in hospita li zed patients at risk for brain injury. Mayo Clin Proc 1998; 73:629-35. Hutchinson PJ , al-Rawli PG , O'Connell MT et al. Head inj ury monitoring using cerebral microdialysis and Paratrend multipara meter sensors. leY/trolvl Neurochir 2000; 61 :88-94. Gupta AK, a l-Rawli PG, Hu tchinson P J, Kirkpatrick P J. Effect of hypothermia on bra in tissue oxygenation in patients with severe head inju ry. Br) Allaesth 2002; 88: 188-92. Smith LW, Fay T. Obse rva tions on human beings with cancer, maintained at reduced temperature of 75- 90 degree Fahrenhe it. Am) Ciin PatllOl 1940; 10:1-11. Fay 1. Ea rly experiences with local and generalised refrigeration of the human bra in.) Neurosurg 1959; 16:239 - 60. Clasen RA, Pandolfi S, Russell J et a!. Hy pothermia and hypotension in experimental cerebral edema. Arch Neurol 1968; 19:472-86. Laskows ki EJ, Klatzo I, Baldwin M. Experimental study of the effects of hy pothermia o n local bra in injury. Neurology 1960; 10:499- 505. Rosom off HL, Gilbert R. Brain vo lum e and cerebrospinal fluid pressure during hypothermia. Am) Physiol 1955; 18 3:19 - 22 . Rosomoff HL, Shulm an K, Raynor R, Grainger W. Experimental brain injury and delayed hypothermia. Surg GYl1ecol Obstet 1960; 110: 27-32. Clifton GL, AJlen S, Barrowdale P et a!. A phase II study of mod era te hypothermi a in severe brain injury. ) Neurotrauma 1993; 10:263-71; discussion 273. Marion DW, Oozeer R, de Vries L et a l. The use of mod erate therapeutic hypoth ermia for patients with seve re head injuries: a preliminary re port.) Neurosurg 1993; 79:354-6 2. Cli fton GL, Miller ER , Choo i SC et al. Lack of effect of induction of hypothermia after acute brain injUly. N Engl ) Med 2001; 344:556- 63. Connell J, Oozeer R, de Vries L et al. Continuous EEG monitoring of neon ata l seizures: diagnostic and prognostic co nsideratio ns. Arch Dis Cilild 1989; 64:452 - 8. Dusse r A, Na valet y, Dev ictor 0, Landrieu P. Short- and long term prognostic value of the electroencephalogram in children with seve re head injury. Electroellcep iIalogr Clin Neurophysiol 1989; 73 :85-93. Altafu ll ah I, Asa ik ar S, Torres F. Status epilepticLls: clinical expe ri ence w ith two spec ial devices for continuous cereb ra l mon itoring. Acta Neurol Scand 199 1; 84:374 - 8 1. Aoki Y, Lombroso C. Progno stic valu e of electroencephalography in Reye' s syndrome. Neurology 1973; 23:333-43.
References I
92 Janati A. Erba G. El ectroencephal ograp hic correlates of
near-drowning encephalopathy in children.
Electroencepi1alogr Clin Nellrophysiol 1982; 53: 182-9l.
93 Pampiglione G. Harden A. Resuscitation after
cardiocirculatory arrest. Prognostic evaluation of ea rly
electroencephal ographic findings. Lancet 1968; I: I 26 1-5.
94 Tasker RC. Boyd S. Harden A. Matthew DJ. Monitoring in
non-traumatic coma. Part II. Electroencephalography.
Arch Dis Ch Ud 1988; 63 :895-9.
95 Tasker RC. Boyd SC. Harden A. Matthew DJ. The cerebral function analysing monitor in paediatric medical intensi ve care: applications and limitations. Intensive Care iVIed 1990; 16:60-8. 96 Lassen NA. Christiensen MS. Physiology of cerebral blood
flow. Br) Anaesth 1976; 40:719-34.
97 Aaslid R. Markwalder TM. Nomes H. Noninva sive transcranial Doppler ultra sound recording of flow velocity in basal cerebral arteries.) Neurosurg 1982; 57:769-74. 98 Weber M. Grolimund p. Seil er RW. Evaluation of
posttraumatic cerebral blood flow velocities by
transcranial Doppler ultrasonography. Neurosurgery 1990;
27:106-12.
99 Martin NA. Doberstein C. Zane C et al. Posttraumatic cerebral arterial spasm: transcranial Doppler ultrasound. cerebral blood flow. and angiographic findings. ) Neurosllrg 1992; 77: 575-83. 100 Chan KH. Miller JD. Dearden NM. Intracranial blood flow velocity after head injury: relationship to seve ri ty of injury. time. neurological status and outcome. ) Neurol Neurosurg Psychiatry 1992; 55:787-9l. 101 Ungerstedt U. Microdialysis - principles and app lications for studies in animals and man.) Intern Med 1991; 230:365-73. 102 Hillered L. Persson L. Neuroc hemical monitoring of the acutely injured human brain . Scand) Ciin Lab /Il1Jest 1999 ; 229(SuppL) :9-18. 103 Hillered L. Persson L. Nilsson P et at Neurom etabolic monitorin g of the ischa emic human brain using microdialysis. Acta Neurochir (Wim) 1990; 102:91-7. 104 Hillered L. Valtysson J. Enblad P. Persson L. Interstitial glycerol as a marker for membrane phospholipid degradation in the acutely injured human brain. ) Neural Neurosllrg Psychiatry 1998; 64:486-9l. 105 Goodman JC. Valadka AB. Gopinath SP et a1. Extracellular lactate a nd glucose alterations in the brain after head injllJY measured by microdialysis. Cnt Care Med 1999 ; 27:1965-73. 106 Hovda. DA, Lee SM. Smith ML et al. The neurochemical and metabolic cascade following brain injury: moving from animal models to man.) Neurofrauma 1995; 12:903-6. 107 Nilsson OG. Brandt L. Ungestedt U. Saveland H. Bedside detection of brain ischemia using intracerebral microdi a lysis: subarachnoid hemorrhage and delayed ischemic deterioration. Neurosurgery 1999; 45: 1176-84; discu ssion 1184-5. 108 Bullock R. Zaune A. Woodward JJ et al. Factors affecting exc itatory amino acid release follo wing severe human head injury.) Neurosurg 1998; 89:507-18. 109 Bell MJ. Robertson CS. Kochanek PM et a1. Interstitial brain adenos ine and xanthine increase during jugular venous oxygen desaturations in humans after traumatic brain injury. Crit Care Med 2001 ; 29:399-404. 110 Pang D. Wilberger JE Jr. Spinal cord injury without radiographic abnormalities in children. ) Neurosurg 1982; 57:114-29. III Marsha ll LF. Marshall SB. Klauber MR et al. A new classification of head injury based on computerized tomog ra phy.) Neurosurg 1991 ; 75(Suppl.):SI4-20.
315
112
Mendelsohn D. Levin HS. Bruce D et al. Late MRI after head injury in children: rel ations hip to clinic a l features and outcome. Childs Nel'/! Syst 1992; 8:445-5 2. 113 Caffey J. On the theo ry and practice of shaki ng infants. Its potentia l residual effects of permanent brain damage and mental retardation. Am) Dis Child 1972 ; 124: 161-9. 114 Guthkelch AN. Infantile subdural haematoma and its relationship to whiplash injuries. BM) 1971; 2 :430-1. 115 Minns RA. Brown JK. Neurological features ofNAHI and whiplash shaken baby syndrome. In Minns RA. Brown JK (eds) Shaken Baby Syndrome alld Other NOll -accidental Head Injuries. London: Mac Keith Press. 2005. pp. 1-105. Jl6 Geddes ]f. Vowles CH. Hackshaw AK et a1. Neuropathology of inflicted head injuIY in children. II . Microscopic brain inju ry in infants. Brain 2001; 124:1299-306. /17 Geddes JF. Vowles CH. Hackshaw AK et al. Neuropathology of inflicted head injury in children. I. Patterns of brain damage. Brain 2001; 124: 1290-8. 118 Barlow KJVI. Minns RA The relation between intracranial pressure and outcome in non-accidental head injulY. De1J
Med Child Neuro11 999; 41:220-5. 119 120
121
122
123 124
125
126
127
128 129
130 13 1 132 133 134
135
Ludwig S. Warman M. Shaken baby syndrome : a review of 20 cases. Ann Emerg Med 1984 ; 13: 104-7. Schloff S. Mulhaney PB. Armstrong DC et al. Retinal findings in children with intracranial hemorrhage. Ophthalmology 2002; 109: 1472-6. Reece RM. Nicholson CE. Inflicted Childhood Neu rotrauma. Bethesda. MD: American Academy of Pediatrics. 2002. Goldstein B. Kelly MM. Bruton D. Cox C. Inflicted versus accidental head injulY in critically injured children. Crit Care Med 1993; 21:1328-3 2. Kivlin JD. Simons KB. Laronitz S. Ruttum MS. Shaken baby syndrome. Ophthalmology 2000; 107:1246-54. Rao p. Carty H. Pierce A The acute reversal sign: comparison of medical and non- accide ntal injulY patients. Ciill Radiol 1999; 54:495-501. Zimmerman RA. Bilaniuk LT. Bruce D et al. Interhemispheric acute subdural hematoma: a comp uted tomographic manifestation of child abuse by shaking. Neuromdiology 1978; 16:39-40. Ewing-Cobbs L. Kramer L. Prasad Met a1. Neuroimaging. physical. and developmental findings after inflicted and noninfJicted traumatic brain injury in young childre n. Pediatrics 1998 ; 102:300-7. Duhaime AC. Alario AJ. Lewander WJ et al. Hea d injulY in very young childre n: mechanisms. injury types. and ophthalmologic findings in 100 hospitalized patients younger than 2 years of age. Pediatrics 1992 ; 90:179-85. Joffe M. Ludwig S. Stairway injuries in children. Pediatrics 1988 ; 82:457 - 6l. Boyles I. Minns RA. How Often Do Toddl ers Fall ancllnjure Themselves? A prospect study. Special Study Module. University of Edinburgh. 2004. Hall JR . Reyes HM. Horvat M et a1. The mOl1ality of childhood falls.) Trauma 1909; 29:1273-5. Claydon SM. Fatal extradural hemorrhage following a fall from a baby bouncer. Pediatr E111erg Care 1996 ; 12:432-4. Stoffman JM. Bass MJ. Fox AM. Head injuries related to the use of baby walkers. Can Med Assoc) 1984; 131:573-5. Plunkett J. Fatal pediatric head injuries caused by short distance fall s. Am ) Forensic Med Patlwl 2001; 22: 1-1 2. Reiber GD. Fatal falls in childhood. How far must children fall to sustain fatal head injury? Report of cases and review of the literature. Am) Forensic Med Patho/1993; 14:201-7. Milhorat TH. Pediatric neurosurgery. Contemp Neurol Ser 1978 ; 16:1-389.
316 I
Head and neck inj uri e s
136 Helfer RE, Siovis TL, Black M. Injuries resulting wh en sma ll children fall out of bed . Pedia trics 1977 ; 50:533-5. 137 Margulies SS, Thibault KL Infant skull and suture properties: measurements and implications for mechanisms of pediatric brain injury. J Biol11ech Eng 2000; 122 :364-71. 138 Scarfo GB, Mariottini A, Tomaccini D, Palma L. Growing sku ll fractures: progressive evolution of brain dam age and effecti ve ness of surgica l treatme nt. Ch ilds Nen) Syst 1989; 5:163-7. 139 Carty H. Training in paedi atri c radiology. Clin Radial 1989; 40:227-8. 140 Royal College of Radiologists. Guidelinesfor Doctors, 5th ed n. London: Royal College of Rad iologists, 200 3. 141 Zi mmerman RA, Bilianiuk LT, Bruce D et al. Computed tomography of craniocerebral injuI)' in the abused chi ld. Radiology 1979; 130:687-9 0. 14 2 Johnson H, Fae rber EN (ed.). Trauma. In CNS Magnetic Resonance Imaging in Infants and Children. Lond on : Mac Keith Press, 1995, pp. 98- I I 5. 143 Barlow KM, Gibson RJ, McPhillips M, Minns RA. Magnetic reso nance imaging in acute no n-accid e ntal head inju ry. Acta Paediatr 1999; 88:734-40. 144 Biousse V, Suh DY, Newman NJ et a l. Diffusion-weighted magnetic reso nance imagin g in shaken baby syndrome.
Am J Oph tha11110 12002; 13 3:249- 55 . 145
146
147
148
149
150
151
152
153 154
155
155 157
Hovi nd KH. Traumatic brain injuries. In Raimondi AJ, Choux M, DoiRocco C (eds) Head Injuries ill the Newborn and Infant. New York: Springer-Verlag, 1986, p. 87. Duhaime AC, Genarelli TG, Thibau lt LE et aJ. The shaken baby synd rome. A clinical, pathological, and biomechanical study. J Neurosurg 1987; 66 :409- I 5. Morison CN, Minns RA. Biomechanics of shaking. In Minns RA, Brown JK (eds) Shaking and Other Non-accidental Head Inju ries ill Children. London: Mac Keith Press, 200 5, pp. 105-46. Choux M, Lena G, Genitori L. Intracranial hematomas. In Raimondi AJ, Choux M, DoiRocco C (eds) Head Iiljuries in the Newborn and Infan t. New York: Springer-Verlag, 1985, p. 203. Gedd es JF, Tasker RC, Hackshaw AK et aJ. Dural haemorrhage in non-traunJatic infant deaths: does it explain the bl eeding in 'shaken baby syndrome'? Neuropathol Appl Neurobiol 20 03; 29: 14-2 2. Byard RW, Blumberg P, Rutty G et al. Lack of evidence for a causal relationship between hypox ic/ischaemic encep halopathy and subdural haemorrhage in fetal life, infancy and ea rly childhood. Pediatr Dev Pathol 2007; 10:348-50. Ommaya AK, Gennarelli TA. Cerebral concussion and traumatic unconsciousness. Correlation of expe rimental and clinica l observ ations of blunt heael injuri es. Brain 1974; 97 :633-54. Jaspa n T, Narbrou gh G, Punt J A, Lo we J. Cerebral contusional tea rs as a marker o f child abuse: detection by crania l sonography. Pediatr Radial 199 2 ; 22:237 -45. Strich SJ. Shearing of nerve fibers as a course of brain damage due to head inju I)'. Lan cet 1961; 2:443-8. Crompton R. Closed head injuI)' in children. In Cromp to n R (ed.) Closed Head InjUly in Children. Lo ndo n : Edward Arno ld , 1986. Adams JH, Mitchell DE, Graha m Dl, Doyl e D. Diffuse brain damage of immediate imp act type. Its relatio nship to 'primaI)' brain -stem damage' in head injuI)'. Brain 1977 ; 100:489-502. Vowl es H, Scholtz CL, Cameron JM. Diffuse axona l injuI)' in ea rly infa ncy. J Clin Pathol 1987 ; 40: 185-9. Ches nut RM, Marshall LF, Klauser MR et al. The role of secondaI)' brain injuI)' in determining outcome from severe head injuI)'. J Trauma 1993; 34:216- 22 .
158 Graham Dl, Ford I, Adams JH. Ischaemic brain damage is still common in fatal non-missile head injuI)'. J Neural Neurosurg Psyclliatry 1989; 52:346-50. 159 Miller JD , Bec ker DP. SecondaI)' insults to the inj ured brain. J R Call Surg Edill 1982; 27:292-8. 160 Gentleman D, Jennett B. Haz ard s of inter-hospital transfer of comato se head -injured pati ents. Lan cet 1981; 2:853-4. 161 Kohi YM, Mendelo w AD, Teasdale GM, Allard ice GM . Extra c ra ni al insults and outcome in patients w ith ac ute head inju I)': relat io nship to th e Gla sgow Coma Sca le. Inju ry 1984 ; 15 :25-9. 162 Miller JD, Bu tte rworth JF, Gud ema SF et al. Furthe r experien ce in the managem ent of severe head injuly. J Ne urosurg 1981; 54:289-99. 163 Miller JD , Sweet RC. Na raya n R, Becke r DP. Early insults to the injured brain. JAMA 1978 ; 240:439-42. 164 Price DJ, MUlTay A. The influence of hypoxia and hypotension Oll recov ery fro m head inju I)'. InjUly 1972; 3:218-24. 16 5 Marmarou A, Anderson RL, Ward JD et al. NIND S Traumatic Coma Data Bank : Intracranial press ure moni toring methodo logy. J Neurosurg 199 1; 75 (Suppl.): 21 -7 . 166 Andrews PJ, Piper IR , Dea rd en NM, Miller JD . Seco nda I)' insults during intrahospital transport o f head -injured patients. Lancet 1990; 335:327-30. 167 Sha'lJles PM, Matthews DS, Ey re JA. Cerebra l bloo d flow and metabolism in children wit h severe head iIJjuI)'. Part I: Relation to age, Glasgow Coma Score, outcome, intracranial pressure, and time after inju I)'. J Neural Nellrosu rg Psychiatry 1995; 58:145-5 2. 168 Muizelaar JP, Ward JD, iVlarmaron A et al. Cerebral blood flow a nd metabolism in severely heael-injured children. Part 1: Relationship with GCS score, outcome, ICP, and PVJ. J Neurosurg 1989 ; 71:63-71. 169 Sharples PM, Matthews DS, Eyre JA. Cerebral blood flo w and metabolism in children w ith severe head injuries. Part 2: Cerebrovascu lar resistance and its determ inants. J Ne urol Neurosurg Psychiatry 1995; 58:153-9. 170 Minns RA. Infectious and para infectious encep halopathies. In Minns RA (ed.) Problems of Intracrallial Pressure in Childhood. London: Mac Keith Press, 1991 , pp. 170- 282 . 171 Report of th e Second Task Fo rce on Blood Pressure Control in Children: 1987. Task Force on Blood Pressu re Control in Children . National Heart, Lung, and Blo od Institute, Bethesda, i\IlaI)'land. Pediatrics 1987; 70:1- 25. 172 Donmall Me. Anthropom et ric and blood press ure studies in children fro m Shetland: reliability of measurement and the use of heig ht- we ig ht indi ces. Acta Paediatr SC(ll1d 198 5; 3 18 (Suppl.) :23-35. 173 Ro cce ll a EJ. Update on the 19~7 Task Force Report on Hi gh Blood Pressure in Children and Adolescents: a working group report from the National High Blood Pressure Educatio n Program. National Hig h Blood Pressure Education Program Wo rkin g Group o n Hypertensio n Control in Children a nd Adolescents. Pediatrics 1995; 98:649 -58. 174 Rosner B, PrineasRJ , Logg JM, Daniels SR. Blood pressure nomog rams for children and adolescents, by height, sex, and age, in the United States. J Pediatr 1993; 12 3 :871-85. 175 J ackso n S, Piper JR , Wagsta ff A, Souter M. Quantification of cerebra l perfu sion pressure (CPP) 'secondaI)' ins ults' in child ren: a study of the effec ts o f using diffe ren t CPP thres holds. British Journal of Neu roSlirgery, 1999; 13: 104. 175 Chambers IR, Treadw ell L, Mendelow AD. The cause and incidence of secon daI)' insults in severely head-injured adu lts and children. Br J Ne urosll rg 2000; 14:424-31. 177 Jo nes PA, A ndrews PJ, Eastman VJ, Minns RA. Traumatic brain injuI)' in chi ldhood: Intensi ve care time-series data and o utcome. Br J Neurosurg 2003 ; 17: 29 -39.
References I
178 Jones PA, Easton VJ, Anderson PJD et al. Traum at ic brain injury in chi ldren: intensive care time series data and outcome. Br J Neurosurg 1982; 17:29-39. 179 Ch ambers IR, Jones PA , Lo TYM et al. Critical thresholds of ICP and CPP related to age in paediatric head inju ry. J Ne urol Nellrosllrg Psych 2007 ; 77:234-40. 180 Pappius HM, Fe ind el W. Dyna mics of Brai l! Edema. Berlin: Spri nge r, 1976. 181 King WJ, MacKay M, Sirnick A. Shaken baby sy ndrome in Canada: clinical cha racteristics and outcomes of hospital cases. Can Med Assoc J 2003 ; 168:155-9. 182 Brown JK, Imam H. Interrelat ions hips of liver and brain with special reference to Reye syn drome. J Inherit Me tab Dis 1991; 14:436- 58 . 183 Seshia SS, PN Chow, Sanka ran K. Coma foJlowin g cardiorespiratory arrest in chi ldhood. Dev Med Child Neurol 1979; 21:143-53. 184 Fujimoto T, Walker JTJ , Spatz M, Klatzo I. Pathophys iol ogic aspects of ischaemic edema. In Pa ppius HM , Feindel W (eds) Dynamics of Brain Edema. Heidelberg: Springer, 1976, pp. 171-80. 185 Bruce DA, Langfil l m , Mi ll er JD et al. Regional cerebral blood flow, intracrani al pressure, and brain metabolism in comatose patients. J Neurosurg 1973; 38:131-44. 186 Dearden NM. ls chaem ic brain. Lancet 1985; 2:255-9. 187 Augutis M, Levi R. Pediatric spinal cord injury in Sweden : incidence, etiolo gy and ou tcome. Spinal Cord 2003; 41 :328-36. 188 Hamilton MG, My les ST. Pediatric sp in al injury: review of 61 deaths. J Neurasurg 1992 ; 77 :705-8. 189 Ham ilton MG, Myles ST. Pediatric spinal injury: review of 174 hosp ital ad miss ions. J Neu rosurg 1992; 77 :700-4. 190 Splendiani A, De Amicis R, Pupill o V et a l. Spi nal trauma in childho od: a four-y ea r experience at a non-specialis t hospital. RiIJista di Neuroradi%gia 2003; 16:455-7. 191 Aufderma ur M. Spinal injuries in juven il es. Necropsy find ings in twelve cases. J Bon e Joint Surg Br 1974; 56B:513-19. 192 Zabramski JM, Hadl ey MN, Browner CM. Pediatric spinal co rd and vertebral column injuries. Ban-ow Neural [nst Q 1986; 2:11 -17. 193 Lena G, Bollini G. Spinal inju ries in children. In Chou x M et al (eds) Pediatric Ne urosurgery. Lond on: ChurchiJi Livings tone, 1999, pp. 313 1-91.
31 7
194 Tator CH, Koyanagi T. Vascular mechani sms in the pathophysiology of human spinal cord injury. J Neurosurg 1997; 86:483 - 92. 195 Yates PO. Birth traum a to the vertebral arteries. Arch Dis Child 1959; 34:436-41. 196 Bollini G. Thoracic and lumbar spine injuries in children. In Fl oman Y, Fa rcy JPC, Arge nson C (eds) TllOra coiuml1ar Spine Fractures. New York: Rave n Press, 1993, pp. 307 - 25. 197 Holdsw oli h F. Fractures, dislocations and fracture-di slocations of the sp ine. J BOl1e Joint Surg (Am) 1971; 52A:1 534-S I. 198 Jones L. Birth trauma and the cervical spine. Arch Dis Child 1970; 45:1 47 . 199 Towbin A. Late nt sp in al co rd and brain stem injury in newborn in fants . Del) }\lIed Child Neurol 1969: 11: 54- 68. 200 Cullen JC. Spina l lesions in bartered babies. J Bone Joint SurgelY (Br) 1975; 57B :364-6. 20 1 Sw ischuk LE. Spine and spin al cord trauma in the battered child synd ro me. Radiology 1969; 92 :733 - 8. 202 Friedman G, Froolll p, Sazbon L et aJ. Apol ipoprotein E-epsilon4 genotype pred icts a poor outcome in surv ivors of traumatic brain injury. Neurology 1999; 52:244- 8. 203 Lichtman SW, Seliger G, Tycke B, Marder K. Apolip oprotein E and functi onal recovery from brain injury followin g posta cute rehabilitation. Neu rology 2000; 55: 1536-9. 204 Utermann G, Langenbeck U, Beisiegel U, Weber W. Genetics of the apolipoprotein E system in man. Am J Hum Genet 1980; 32:339-47. 205 Mah ley RW. ApoJipoprotein E: chol esterol transpOli protein with expandin g role in cell biology. Science 19813; 240:622-30. 206 Boyles JK , Pitas RE , Wil so n E et aJ. Apo lip oprotein E associated with astrocy1ic glia of the central nervous system and with nonmyellnating glia of the peripheral nervous system. J Clin [n/lest 1985; 76:1501-13. 207 Teasdale GM, Ni cho ll JA , Murray G, Fiddes M. Associati on of apolipoprotein E polymorphism with outcome after head injury. Lancet 1997; 350:1069 -7 1. 208 Mattson MP. Cellul ar actions of beta-amyloid precursor protein and its solubJe and fibrillogenic deri va tives. Physiol Rev 1997; 77: 1081-132. 209 Nicoll JA, Robelis GW, Graham DJ. Apo lipoprotein E epsi lon 4 allele is associated with deposition of amyloid beta-protein fo llowi ng head inj ury. Natu re Med 1995; 1:135-7.
I
CHAPTER 15
I
HEAT-INDUCED INJURY OR DEATH
Anthony Busuttil
Introduction House fire deaths
318 318
INTRODUCTION
Deaths in fires are the third leading cause of death in domes tic fatalities among children.! In the USA someone dies in a fire about every 162 minutes and someone is injured in a fire every 32 minutes, and many of these victims are chi ldren ;2 t he USA ranges only sixth in the incidence of deaths from fires among the 25 developed coun tries for which statistics are available.) In retrospect, many of these deaths are found to have been preventable 4 and this raises a substantial pub lic health problem. 5 - 7 Financial and insurance losses are also vely substantial. Thermal injuries are also a common reason for referral of children to the emergency services, with scalding being the commonest reason for such referrals. Heat-induced inju ries also form an important part of the gam ut of the injuries seen in children who have been physically abused by their carers.
HOUSE FIRE DEATHS In the UK, house fires rank second after road traffic fatalities as a cause of accidental deaths in children below the age of 15 years. When a house fire develops, children are more vul nerable, partly because they depend on the ad ults around them to assist them out of the fire as they may be too small and inexperienced to be capable of helping themselves and extricating themselves out of the fire. The installation of smoke alarms in houses has been a major reason for the
The pathologist's role References
319
327
diminution in the incidence of domestic fire deaths since 1987. Poor housing particularly associated with overcrowd ing and poorly built houses present a high lisk of fire deaths. Single-parent families and families from a lower socioeco nomic background, particularly in associatio n with multi deprivation, show a much higher incidence of fire fatalities. The misuse of a lcohol, and contro ll ed substances to a lesser extent, by the adults caring for the child increases the incidence of deaths in domestic fires, given that the adults in such conditions succumb earlier to the fire and, indeed, being a lready disorientated due to the intoxication they are much less able to look after t hemselves and after the child ren whom they are caring for. This asso ciatio n has been shown to o'ccur both in the UK and in the USA. Smokers' materials such as cigarettes, matches and lighters are a very frequent cause of house fires. This has led some countries to ensure that cigarettes lighters are manufactured in a child-proof manner. Deaths in fires usually are the result of smoke inhalation rather than burns. s Although consumption of ambient oxygen by the fire with environmental deoxygenation is important, t he inhalation of carbon monoxide (and carbon dioxide) is more frequently what causes death: thus smoke detectors are valued as preventative measures. Smoke detectors are widely used in several countri es and foun d to be effective. Their use is becoming more widespread in Europe and in the USA,9,lO and they have been sho wn to have an important role in the prevention of conflagrations both in public and private housing. II At present, regul a tions throughout the European Community tend to impose
The pathologist's role I
the use of smoke alarms in public buildings rather than homes where children live but there may be good reason for extending their scope. The banning of upholstery, certain items of furniture and other fire-retardant items whose combustion produces cyanide and other toxic compounds has also gone some way in reducing deaths from such sources. Fires and flames are a significant cause of death in childhood in many countries in Europe. 12 Although there has been a striking fall in the number of children who have died from fire in recent years, the death rate from fires remains unacceptably high. A major component of the reduction has been the fall in the number of deaths from the ignition of clothing following flame-proofing regula tions and the reduction in use of open fires. There have been few specific studies of deaths from fire and flames at the European level. Each year in the USA, an estimated 700 children aged 5 years and under die in house fires. Representing 20 per cent of the fire deaths each year, this age group has a fire risk that is double the national average. Children playing with matches, cigarette I ighters and other fire sources are thought to have started about 91810 fires per year from 1993 through 1997, which resulted in an estimated 338 deaths and 2624 injuries each year. Pre-school children are also the most frequent victims of fires started by children playing with smokers' materials (over 30 per cent of the fires) and they constitute 20 per cent of the total number of fire death vic tims. The United States Fire Administration (USFA) encour ages parents to teach children at an early age about the dangers of playing with fire in an effort to prevent child injuries, fire deaths and future fire setting behaviour. Every year, more than 100000 household fires are reported in the UK and more than 600 people die in these fires. One in ten of these fire victims are children. The num ber of house fires in Scotland is 'markedly different' from England, Wales and Northern Ireland according to the latest government figures. In Scotland 46 per cent of all fires are house fires, compared with 33 per cent in Northern Ireland and 32 per cent in England and Wales. Broken down by fire brigade areas, the figures show that all Scottish brigades recorded that at least 38 per cent of all fires were house fires. Outside Scotland, the largest percentage of house fires was in Greater London. The Home Office research, made available to chief fire officers across the UK, shows that around 75 per cent of all fire casualties are in house fires: 466 people died in fires in the home and another 14600 people were injured. The number of firefighters injured - 655 men and women was the lowest recorded for over 20 years, with over 50 per cent being physical injuries rather than burns. There are more than 100000 fires deliberately started in the UK each year; some of these are targeted at domestic premises for a variety of reasons. In 2003, 111 people died in Scotland because of fires and another 2300 people suffered non-fatal injuries. Misuse of alcohol was a major contributory factor in 41 out of the
319
76 fatal fires; carelessly throwing away cigarettes, matches and lighters caused 39 out of the 76 fatal fires. Smoke alarms were fitted in 44 out of the 76 homes where people died, but only six alarms were known to have operated properly; in the other 38 cases, 15 did not have a battery fitted and three had flat batteries. Fire investigators said that 39 out of the 85 house fire deaths were potentially avoidable if smoke alarms had been fitted and in working order. Of these fatal fires 34 started in the living-room, 21 in the bedroom and 13 in the kitchen. Polyurethane foam filled furniture contributed to 14 fatal fires. Over 42000 children under the age of 15 years were injured in the UK from burns and scalds in 2002, with 28000 of them being under 5 years of age and 95 per cent of the thermal injuries occurring in the home. Over 50 per cent of all severe burns and scalds happened in the kitchen. Hot liquids were the cause of 70 per cent of the heat induced injuries.
THE PATHOLOGIST'S ROLE Key questions to be addressed by the pathologist in fire fatalities are: Can the victim be positively identified? Was the victim actually alive or already dead at the time the fire started? Was the death directly related to the fire and caused by it? What exactly is the cause of death? Why was the victim unable to escape the fire? To respond to these questions the following information must be correlated: (1) the circumstances preceding the death; (2) the anatomy of the fire, i.e. evidence of the origin, development and nature of the fire; and (3) the autopsy findings. When examining a fire it has to be kept in mind that fire artefacts occur to varying extents in any body exposed to a fire, irrespective of whether death was a result of the fire or whether the person was already dead when the fire started. These may be further aggravated at the time of the rescue and recovery of the body from the site of the fire occurs. These 'are: • The 'pugilistic attitude' of the body. Characteristically, the flexor muscles in both the upper and lower limbs, which are on the whole stronger than the extensors, will be damaged by the fire and contract, owing to the heat -induced effects on the component proteins of the muscles. This results in the body assuming a pugilistic (boxer) or 'sit up and beg' posture, with the four limbs solidly flexed. • Loss of fluid from the tissues results in their desiccation and brittleness, not least the skin and of any other exposed underlying tissues that thus tend to crumble on pressure and touch. • Body weight may decrease with a loss of up to 60 per cent, also mainly due to loss of water. • Bone fractures are due to desiccation and direct heat induced effects where they are exposed. The heat of the
320 I
•
•
• •
•
Heat-induced injury or death
fire may result in fractures or render them so brittle that they are fractured during recovery of the body_ Artefact ual fractures are particu lar ly common in the sku ll , w here sometimes they can be difficult to disting uish from ante-mortem injury, Body s hortening because the limbs, particularl y the extremities, may be partly burn ed away an d muscles are fl exed, Distortion of the facial features by bl ackening, skin co ntraction and tightening, heat-induced damage to the facial features and sin gein g of the hair. Accumulation of soot and perhaps th e emergence of some blood-tinged fluid in the facia l openings. Irregular deep splits involving the sk in of the arms, thighs and abdomen du e to ex posure to heat which lead s to heat contraction of the skin. (These may be misinterpreted as tea rs, lacerat ions and cuts inflicted during life.) Accumulation of blood beneath the skull (giving the appearances of an extradural haematoma); this may also be mistaken for an an te-mortem injury.
Distinguishing between bu rns inflicted during life a nd burns inflicted on an already dead body can be difficult, if not impossible, at autopsy, Furthermore, genuine ante morte m nature of burn s may be obscured by the continued ex posure to heat of the body after death. Fire deaths result from: • smoke inhalation; • burns; • heat shock; • any combination of these,
Smoke Inhalation Over 50 per cent of a ll fire fatalities occur as the res ult of the inhalation of soot and gases generated in the course of the fire. The hot air a nd steam generated may burn the fa ce and enter into the upper a ir passages where they cause fur ther damage. This heat-induced effect may cause reflex cardiac death (vagal inhibition). Fires generate a wide variety of noxious gases that, when inhaled, may ca use confusion and disorientation , incapacitation or dea th , often in combination with other factors. With the exception of carbon monox ide, these gases rarely reach lethal levels by themse lves. Inhalation of soot particles dam ages the airways because they are superh ea ted and contain toxic age nts. In building fires, evid ence of soot inhalation at autopsy is very common (90 per cent or more of cases). Autopsy evi dence of soot in the airw ays below the level of the vocal cords is proof that the victim was alive at the time of the fire, Soot in the oesophagus and admi xed with stomach contents impli es the swallo wing of soot and has a similar significance.
Blood taken at autopsy is analysed for carbon monoxi de content and the result expressed as a percentage saturation of the haemoglobin. Levels of up to 10 per cent carboxy haemoglobin saturation may be fo und in the normal popu lation in towns and cities as a res ult of atmospheric pollut ion. A carboxy haemoglobin saturation level of over 50 per cent is generally accepted as sufficient in itself to acco unt for death. Perso ns with pre-existing natural disease of the heart or lungs may succumb to levels as lo w as 30 per cent. In building fires approximately one-half of all victims have a carboxyhaemoglob in level that is sufficiently high to account for death. Carboxyhaemoglobin levels of 20- 30 per cent produce dizziness, headache, nausea and fatigu e; levels of 30-40 per cent produce imp aired judgement, confusion and later unconsciousness, In a fire situation, levels above 30 per cent may explain a failure to escape. The presence of a n elevated carboxyhaemoglobin level and/or soot in the airways beyo nd the vocal cords provid es proof that the victim was a live, but not necessarily conscious, at the time of the fire, Inh aled carbon monoxide and other noxious gases may have an additive effect and accou nt for deaths with carboxyhaemoglobin levels below 50 per cent. With the exception of cyanide estimation, laboratory tests for the presence of these noxious gases is rarely, if ever, per formed on the bodies of fire victims. Cya nide is commonly produced in fires and results from the burning of materials such as wool, silk, horse hair, poly urethane a nd polyaclylo nitrile. Cya nide is a powerful, rapidl y acti ve and generally cytotoxic poison, Normal blood cyanide levels are less than 8 ~mol/L; non-fa tal tox ic effects begin at about 50 i_,mol/L and the fatal threshold is above 100 ~mol/L. Bodies recovered From fires present severa l problems of investigation a nd a close integration of info rmation obtained from the exam inatio n of the scene, the exam ina tion of the body and the histOiY of the deceased is paliicu lady important. Several investigators with different areas of ex pertise a re typically involved . In fire-related deaths a fire investigator (usually an offlcer in the loca l fire brigade) is in attendance and can provide valuable information on the origin, development and nature of t he fire. A full investigation of the circumstances preceding the death requ ires the identification of the victim's past med ical a nd social history. If the identity of a charred body is suspect, then identity wi ll have to be proved scientifically. Wh en specific identi fication is not possible, circumstantia l evidence of identity is usua ll y ava il a ble_ The majority of fire-rel ated deaths are accidental and there is typically abundant coll ate ra l evidence from police and fire brigade investigations to exclude suicide or homi cid e. The yo ung and the elderly are the common victims of accid en ta l fires. The deaths are usua lly the result of care lessness: allowing clothing to bru sh aga inst fires; playing with matches or other ligh ted obj ects su ch as cigarettes; mainta ining faulty electrical and hea tin g appliances, as well as being unable to effecti vely co mba t or escape a fire.
320 I
•
•
• •
•
Heat-Induced injury or death
ftre may result in fractures or render them so brittle that they are fractured during recovely of the body. Artefactual fractures a re palticularly common in the skull, where sometimes they can be difficult to disti nguish from ante-mortem injury. Body shortening because the limbs, palticularly the extremities, may be paltly burned away and muscles are flexed. Distortion of the facial features by blackening, skin contraction and tightening, heat-induced damage to the facial features and singeing of the hair. Accumu lation of soot and perhaps the emergence of some blood-tinged fluid in the facial openings. Irregular deep splits involving the skin of the arms, thighs and abdomen due to exposure to heat which leads to heat contraction of the skin. (These may be misinterpreted as tea rs, lacerations and cuts inflicted during life.) Accumulation of blood beneath the sku ll (giving the appearances of an extradural ha emato ma) ; this may also be mistaken for an ante-mortem injury.
Distinguishing between burn s inflicted during life and burns inflicted on an already dead body can be difficult, if not impossible, at autopsy. Furthermore, genuine ante mortem nature of burns may be obscured by the continued exposure to heat of the body after death . Fire deaths result from: • smoke inhalation; • burns; • heat shock; • any combination of these.
Smoke Inhalation Over 50 per cent of all fire fatalities occur as the result of the inhal ation of soot and gases gene rated in the course of the ftre. The hot air and steam generated may burn the face and enter into the upper air passages where they cause fur ther damage. This heat-induced effect may cause reflex cardiac death (vagal inhibition). Fires generate a wide variety of noxious gases that, when inhaled, may cause confusion and disorientation, incapaci tation or death, often in combination with other factors. With the exception of carbon monoxide, these gases rarely reach lethal leve ls by themselves . Inhal ation of soot particles damages the airways because they are superheated and contain toxic agents. In building fires, evidence of soot inhalation at autopsy is very common (90 per cent or more of cases). Autopsy evi dence of soot in the airways below the leve l of the vocal cords is proof that the vict im was alive at the time of the fire . Soot in the oesophagus an d admixed with stomach contents implies the swallowing of soot and has a similar s ig n ifi cance.
-
---
Blood taken at autopsy is analysed for carbon monoxide content and the result expressed as a percentage saturation of the haemoglobin. Levels of up to 10 per cent carboxy haemoglobin saturation may be found in the normal popu lation in towns and cities as a result of atmospheric pollution. A carboxyhaemoglobin saturation level of over 50 per cent is generally accepted as sufficient in itself to account for death. Persons with pre-existing natural disease of the heart or lungs may succumb to levels as low as 30 per cent. In building fires approximately one-half of all victims have a carboxyhaemoglobin level that is sufficiently high to account for death. Carboxyhaemoglobin levels of 20-30 per cent produce dizziness, headache, nausea and fatigue; levels of 30-40 per cent produce impaired judgement, confusion and later unconsciousness. In a fire situation, levels above 30 per cent may exp lain a failure to escape. The presence of an elevated carboxyhaemoglobin level and/or soot in the airways beyond the vocal cords provides proof that the victim was alive, but not necessalily conscious, at the time of the fire. Inhaled ca rbon monoxide and other noxious gases may have an add itive effect and account for deaths with carboxyhaemoglobin levels below 50 per cent. With the exception of cyanide estimation, laboratory tests for the presence of these noxious gases is rarely, if ever, per formed on the bodies of fire victims. Cyanide is commonly produced in fires and results from the burning of mate rials such as wool, silk, horse hair, polyurethane and polyacryIo nitlile. Cyanide is a powerful, rapidly active and genera lly cytotoxic poison. Normal blood cyan ide levels are less than Bjtmol/L; non-fatal toxic effects begin at about 50f-Lmol/L and the fatal threshold is above 100 f-Lmol/L. Bodies recovered from fires present several problems of invest igation and a close integration of informatio n obtained from the examination of the scene, the examina tion of the body and the history of the deceased is particu larly important. Several investigators with different areas of expertise are typicaJly involved. In fire-related deaths a ftre investigator (usually an officer in the local fire brigade) is in attend ance and can provide valuable information on the origin, development and nature of the fire. A fu ll investigation of the circumstances preceding the death requires the identiftcation of the victim 's past med ical and social history. If the identity of a charred body is suspect, then identity will have to be proved scie ntiftcally. When specific identi fication is not possible, circumstantial evidence of identity is usually available. The majority of fire-related deaths are acc idental and there is typically abundant collateral evidence from police and fire brigade investigations to exclude suicide or homi cide. The young and the elderly are the commo n victims of accid ental fires. The deaths are usually the result of care lessness: allowing clothing to brush against fires; playing with matches or other lighted objects such as cigarettes; maintaining faulty electrical and heating appliances, as we ll as being unable to effectively combat or escape a fire.
~
322 I
Heat-Induced injury or death
Evidence of soot inhalation is very common (approxi mately 90 per cent of cases) in building fires but it is uncommon to find any soot deposition beyond the primary bronchioles. Deposition of soot on the tongue, in the nares, the oropharynx or nasopharynx cannot be taken to imply life during the fire. Deposition of soot below the level of the larynx indicates that the victim was al ive at the start of the fire. This soot that coats the mucosa of the tracheo bronchial tree can be identified through a window cut in the trachea prior to removal of the neck and thoracic struc tures and is most easily seen if the mucus is spread on a white sheet of paper. The presence of soot in the oesopha gus and stomach implies that it has been swallowed and also indicates life at the start of the fire. On microscopy, the soot particles within the tracheobronchial tree lie loosely on the mucosal surface or embedded in the mucus. The par ticles are not incorporated into the tissues and are readily distinguished from anthracitic pigment. Smoke poisoning describes the effects of the various noxious gases other than carbon monoxide that are pro duced by the thermal degradation of both natural and man-made materials (Table 15.1). There are hundreds of such products of combustion. Whereas modern synthetic materials, for example PVC, may have increased this prob lem, natural materials, for example wood, wool, and silk, produce similar gases and the pathological effects of these noxious gases are difficult to separate from direct particu late injury. With the exception of carbon monoxide and hydrogen cyanide, post-mortem analyses for toxic gases are rarely, if ever, performed. Filter masks do not protect against the inhalation of fumes; for this purpose the mask must have a self-contained air supply. These noxious gases rarely reach lethal levels by them selves, but they may cause incapacitation or death in com bination with other factors. Some gases have a synergistic interaction, for exam pie carbon monoxide and hydrogen cyanide, so that non-fatal levels of each in combination may cause death. The concentration of the noxious gas is of importance. Hydrogen cyanide produced from wool, Table 15.1
wood or polyurethane is rapidly fatal at 3000 ppm. Nitro gen dioxide from acrylonitrile is rapidly fatal at 2000 ppm. Hydrogen chloride from PVC is rapidly fatal at 2000 ppm. Hydrogen sulphide from rubber or wool has toxicity that is similar to hydrogen cyanide and is rapidly fatal at 1000 ppm. Acrolein is an aldehyde product of combustion of wood and paper and produces pulmonary oedema after a few seconds exposure at 10 ppm. 14 Carbon monoxide is a colourless, odourless gas that is formed by the incomplete combustion of carbon com pounds and is normally found in the atmosphere in levels well below 1 ppm. Motor vehicle engines are responsible for about 70 per cent of carbon monoxide liberated into the atmosphere. Carbon monoxide poisoning is a distinct cause of inhala tion injury, producing its effects by tissue hypoxia. The affinity of haemoglobin for carbon monoxide is 200-300 times greater than for oxygen so that carboxyhaemoglobin concentration is great even when the carbon monoxide con centration is less than 5 per cent in the inhaled gas. As well as diminishing the oxygen-carrying capacity of the blood, carboxyhaemoglobin (COHb) also alters the dissociation characteristics of the remaining oxyhaemoglobin, making less oxygen available to the tissues. The toxicity of carbon monoxide depends upon: (1) the rate of inhalation of the gas (i.e. concentration of gas in the inspired air and the duration of exposure); (2) physical activity, which influences oxygen requirements; and (3) individual variations in susceptibility. Carboxyhaemoglobin is estimated by spectroscopic methods making use of the fact that oxyhaemoglobin and carboxyhaemoglobin have different absorption spectra. The result is expressed as the percentage saturation and is the ratio of carboxyhaemoglobin to total haemoglobin X 100. The carbon monoxide oximeter is an automated instrument that uses a spectrophotometric method. An alternative gas chromatographic method is both specific and very sensitive. Carbon monoxide liberated from a known volume of blood is estimated on a gas partitioner and the haemoglobin content estimated spectroscopically.
Sources of toxic chemicals produced during conflagrations
Gas
Carbon monoxide, carbon dioxide Nitrogen dioxide Hydrogen chloride Hydrogen cyanide Aldehydes Benzene Ammonia Sulphur dioxide Phenol Acrolein
Source All combustibles containing carbon Cellulose, polyurethanes, acrylonitrile Chlorinated polymers, e.g. polyvinylchloride Wool, silk, nylons, polyurethanes, N-containing plastics Wool, cotton, paper, plasters, phenol-formaldehyde, wood, nylon, polyester resin Petroleum, plastics, polystyrene Melamine, nylon, urea-formaldehyde Rubber, thiokols Phenol-forma Idehyde Wood, paper
The pathologist's role I Table 15.2
The effects on humons of different concentrotions of
carboxyhoemog/obin saturation
Percentage
Effect
0-10
No immediate ill effects; fo und in cigarette smokers or from industrial /motor ve hicle pollution May produce dizziness and shortness of breath on exertion Dizziness, headache, nausea and fatigue Impaired judgement, unconsciousness may occur Unconsciousness and death likely
10-20 20-30 30-40 40-60
The effects of carboxyhaemoglobin at different percent age saturation levels are set out in Table 15.2.
INTERPRETATION OF RESULTS Carbon monoxide at environmental levels of 1000 ppm can kill in about half an hour, and at 5000 ppm is rapidly fatal. About 85 per cent of building fire victims show evidence of carbon monoxide inhalation and approximately 50 per cent of victims have a carboxyhaemoglobin .level suffi ciently high to account for death. The normally accepted fatal carboxyhaemoglobin level is 50 per cent. Ethanol, a central nervous system depressant, might be expected to have an additive or synergistic effect with carbon monoxide but in practice this has not been demonstrated. An elevated carboxyhaemoglobin level is usually but not univer sally (approximately 95 per cent) associated with evidence of soot in the respiratory tract. The carbon monoxide and cyanide produced in fires have an additive effect. A carboxyhaemoglobin level below 10 per cent is not proof that a v ictim was dead prior to the commencement of the fire. Possible explanations may be: (I) little or no carbon monoxide production in the fire due to abundant oxygen allowing complete combustion, for example a forest fire; (2) rapid death in a flash fire; and (3) death following partially successful resuscitation. Carbon monoxide may be produced after death. Blood stained fluid from the tboracic cavity in decomposed bod ies may have COHb levels as high as 80 per cent resulting from the post mortem production of carbon monoxide. 15 Ethylene chloride, found in paint strippers and hairsprays, is metabolized in vivo to carbon monoxide and may pro duce carboxyhaemoglobin levels of up to 40 per cent. Hydrogen cyanide is a product of the thermal degrada ti on of many materials which contain nitrogen, for example wool, si lk, horse hair, polyurethane and polyacrylonitrile. Hydrogen cyanide is a potent toxin with a very rapid action. Ii has a half-life in blood in the order of less than I hour, bei ng metabolized to thiocyanate. Hydrogen cyanide in bl ood is contained principally in the red blood cells and dis appears rapidly from plasma with a half-life of only 15 min :.Ites. Artefactual formation and metabolism of cyanide can
323
occur in post-mortem samples of blood and tissue. Artefac tual formation of cyanide in post-mortem blood is more likely in refrigerated samples than at room temperature. 16
Burns The extent of damage caused by external heat depends on a number of factors. Among these is the applied temperature mammalian tissues only survive within the narrow tempera hire range of 20-44'C; the duration of the interval over which the heat has been applied to the skin or other tissues (e.g. mucosal; and the ability of the sUlface to conduct heat away. Thus a burn may appear after 5 hours of exposure at 44'C but it only take 3 seconds to cause heat-induced dam age at a temperature of 60'C; a child's skin is about 10 times softer than in an adult and it burns 40 times faster. The mor bidity and mortality associated with burns is related to the depth of bum injury and the extent of injury, i.e. the size of the burn relative to that of total body surface (TBS) area. The obvious effect of dry heat is direct physical damage in the shape of a burn and tbe mechanism of injury includes both conduction and radiation. Further prolonged exposure of the area to the heat can result in charring and carbonization of the surface, singeing of the surface hairs and eventually compete destruction of the tissues (cremation). Jackson in 1947 17 described three zones in a burn: the pOint of maximum damage w ith coagulation of proteins locally is the zone of coagulation. This is surrounded by an area of decreased tissue perfusion where the tissue is still potentially sa lvageable, referred to as the zone of stasis. In the outermost pari of t his area, tissue perfusion is increased; there is a n increase in vascularity in this area and it is the zone of hyperaem ia . The older Wilson's tra ditional classification of depth of burn injury is in degrees,18 i.e. first degree, second degree and t hird degree. A first-degree bum involves on ly the epi dermis; it is characterized by erythema and mild pain, the la tter resolving in 48-72 hours. Healing is usually unevent ful and completed in 5-10 days with no residual scarring. Sunburn is the most common first-degree burn. Second degree burns involve the epidermis and a valiable portion of the dermis. Superficial second -degree bums implicate only the upper third of the dermis and are charactelized by blister formation; they are extremely painful but heal in 7-14 days with minimal scaning. A deep second-degree burn extends beyond the upper third of the dermis, but not beyond the dermis itself. These deeper burns are less painful than super ficial second degree burns; heal ing is extremely slow, some times requiring months and usually leading to dense scarring (if the wound is allowed to heal primarily, rather than skin grafted). The fluid loss and metabolic effects of deep second-degree burns are essentially the same as those seen with third degree burns. A/third-degree burn, or full thickness bum, implicates the entire epidermis and dermis. Primary re-epithelialization will not occur and the wound
324 I
Heat-induced injury or death
will require skin grafting. Heat coagu lation of dermal blood vessels leaves the tissue avascular with a characteristic waxy white colour. Prolon ged contact of subcutaneous fat with a flame source produces a leathelY brown, or black, chan-ed appearance. There is charac te ristic lack of pain, due to heat destmction of all nerve endings. Burns are nowadays more frequently classified as either partial thick ness or full thic kness . Partial-thickness wounds contain viable epithelial elements capable of sponta neously re-epithelializing the wo und (first-degree, superfi cia l and deep second-degree burns). Full-thickness burns have no viable epithelial elements and always require cuta neous autografting. Partial thickness burns are usually caused by a brief exposure to heat or contac t with hot liq uids and appear pink to mottled red, wet, covered with ves icl es and bullae, and are painful. Full-thickness burns result from contact with flame, electricity or chemicals. They are often dry and charred, may be translucent, a nd may have thrombosed superficial veins. They are insen sate. Burns wi th chemicals can arise both from direct contact and from fumes (e.g. chlorine), either acting directly on the skin or mucosal surfa ces, or by inhalation. The degree of tissue damage caused by them is dependent on the type of chemical (alkalis are more caustic than acids in general terms), its concentration, its quantity, the duration of contact a nd the extent of penetration; if chemicals are lipid soluble or if proteolysis occurs then this results in a greater depth of penetration and at a quicker rate due to the liquefactive necrosis that they induce, th us exposing deeper tissues to the effects of the chemical. Chemicals that induce burns can be categorized as strong acids, strong alka lis and hydrocarbo ns, wi th hydrocarbons often having penetrating capacity. Heat damage from hot fluids is usually referred to as a scald. Water is the compon ent fluid involved, usually in the kitchen or bathroom, but other fluid s such as oil, molten metal , molten rub ber, other hot liquids and semisolid or liquid foods (e.g. porridge, molten cheese) may be involved; steam as a by-product of the heating process is also very damaging due to its penetrative cap abilities. Scald s do not usually result in charring, carbonization or singeing of body hairs. Their effects are more like that of first-degree burns from dry heat, with superficial redden ing, desquamation and loss of the epidermal layer, swelling and blistering, usually with well-demarcated areas of dam age, with unifo rmi ty all over the site of a rea of contact with the fluid. Thus the severity of skin damage is more uniform in sca lding and from the distribution of the scalds it may be quite possib le to identify th e position of the scalded victim in rel ation to the hot fluid; there may be trickle or splash scalds, and horizontal fluid levels from immersion in hot water. The scald 's severity depends on the duration of exposure of the skin to the hot fluid and its tem perature. Clothing may prevent direct content of the fluid underlying the skin but once it is soaked it may increase the duration of con tact of the hot fluid with the skin's surface. In many
in sta nces co ntact with the hot liquid is therefore momen tary and the large surface area of the skin on to which the hot fluid has made contact a llows for rapid cooling, and thus unless t he temp erature of the water is quite high, there would not be a sufficiently lengthy period of exposure for injury to result from such co ntact. These pheno mena are quite impoltant in attempting to reconstmct how a child has come by his scalds, palticularly in relation to accepting or otherwise the his tory given by the carers when non accidenta l injury (NAI) is suspected. In this respect the temperature of the hot water system in households has to be regulated with care to ensure th at it does not exceed celtain levels l9 and legislation and guide line in terms of plumbing of buildings are in existence.2o Other instances of hea t-induced injury involve fire works, palticularly around the time of bonfire night in November in the UK a nd New Year's Day festivi ties in other countries 2 1,22 Some deaths from overheating have a lso been described in children, particularly infan ts, from the use of electric blankets.
Pathological Changes Beneath the dead burned tiss ue, there is usually a zon e of ischaemia of marginally viable tissue th at is readily con verted to non-viab le tissue (eschar) by any furth er insult, such as hypoxia, decreased blood flow or infection. By this process, a deep seco nd-degree burn frequ ently converts over time to a third-degree burn. Prevention of wound conversion of this type is of major importance in the res us c itation period and thus attention to oxygenation and decreasing the risk of infection are essential elements of the early management of burns. 23 ,24 A determina tio n of the burn surface area in the living is obtained using the 'mle of nines ', each arm constituting 9 per cent, each leg 18 per cent, the anterior trunk 18 per cent, the posterior trun k 18 per cent and the head 9 per cent of TBS area. In assessing irregular areas of burn, it is useful to remember that the surface area of th e person's palm is roughly equivalent to 1 per cent of the TBS. In chil dren under 15 years of age the rel ative body surface area of the head, upper leg and lower leg differs from an adult; this ma y lead to inaccuracy in the estimation of burn injury; use of the Lund Et Browder Burn Diag ram provides age adjusted data accurate for both adults and children. It should be remembered that the extent of burn is often over-estimated, and the depth of burn under-estimated wh en the pati ent is assessed clinically. In addltion to burn depth and extent of injury, other factors determining morbidity a nd morta lity a re the loca tion of the injury, the age of the victim and the presence of injuries or natural disease. Age is a major factor in survival for children under 2 years and adults over 60 years. The dermis acts as a barri er prev,enting loss of body fluid by evaporation an d the loss of excess body heat.
The pathologist's role I
Consequently, loss of water through burn ed skin and loss of heat play a major role in the pathophysiological changes. The skin is also the primary protective barrier against invasive infection so that wound infection is a major cause of mortality and morbidity in the late post burn period. Within the superficial dermis are the nerve endings that mediate pain. Consequently, partial-thickness injuries tha t expose these nerves will be extremely painful, whereas full-thickness burns that destroy the nerves are usually anaesthetic. The loca l and system ic responses to burns fo llow a time course continuum but, for convenience, can be divided in to three periods: (1) up to 48 hours; (2) 2-6 days; and (3) 7 days to wound closure. In the first 48 hours after severe burns, hypovolaemic shock an d shock-induced organ failure (primarily renal failur e) are the major threats to life. Hypovolaemia can also lead to wound convers ion. Al though the exact pathophys iolo gy of the post-burn vasc ular changes and volume shifts remains to be determined, two processes are involved: an increase in microvascular fluid flux into the interstitium both local to the burn and genera li zed (remote from the burn wound) - and a generalized impairment in cell mem brane function, resu lting in cell swelling. With modern therapy, adequate initial volume restora tion is achieved in more than 95 per cent of burns cases. Correction of hypovolaem ia by the intravenous infusion of fluid s is comp licated by ge neralized burn oedema forma tion, wh ich may resu lt in (1) further ischaemic insu lt to already damaged cells; (2) chest wall oedema with resultant increase in the work of breathing; and (3) upper airway oedema. The last two complications may lead to rapidly fatal respiratory failure. Burned skin loses its elasticity. It is less compliant and unable to stretch to accommodate an increase in interstitial oedema. If burns are circumferen tia l, particularly a round the distal extremities, a to urniquet effect is produced by t hem. Initially this impedes venous return , resulti ng in an increase in capillary pressure and further oedema. Impair ment of arterial blood flow may follow, with the potential for distal tissue ischaemic necrosis. Tissue pressure can be decreased by making an incision through the burned tiss ue, i.e. an escharotomy. A chest wall escharotomy may be required to ease respiratory difficulties; incisions are made in the bilateral a nterior axillary lin es from the clavicles to the costal margins; a transverse escharotomy at the level of the costal margins connect ing the two vertical escharo tomies is helpful if the abdomi nal wall is burned to a sig nificant degree. Escharotomy can be a bedside procedure and does not require anaesthesia because full-thickness burns are insensa te. If th e deeper viable tissues are not incised then bleeding should be minim al. Third-degree or full-thickness burns often require exci sion prior to cutaneo us autografting. The two commonly employed techniques of excisi on are tangential excisio n (sequential shaving of no n-viable tissue until a uniformly
325
viable wound bed is obtained) and excision to fascia . The disadvantage of tangential excision is primarily that of mass ive blood loss. Burns also induce a system ic response to them due to the presence of cytokines and other inflammatory media tors if the area of the burns exceeds the 30 per cent surface area. These substances cause systemic effects: 1. increased capillary perm eability that leads to loss of fluid and from proteins within the blood vessels into the interstitial tissues. 2. peripheral and sp lanchni c vasoco nstriction. 3. decreased muscular contractility (probably due to the effect of tumour necrosis factor alpha). 4. systemic h ypotension due to primary and seco ndary fluid losses. 5. fall in systemic blood pressure - hypotension. 6. hypoperfusion of vital organs. 7. bronchoconstriction. 8. adult respiratory distress syndrome. 9. the basal metabolic rate trebl es. 10. No n -specific downregulation of the immun e system affect ing both antibody- and lymphocyte-mediated pathways. NON-ACCIDENTAL BURNS AND SCALDS
Although a significant number of heat-i nduced injuries are truly accidental - although often preventable - they result from temporary lapses in the due care and attention that chil dren require in the home an d elsewhere. Some, however, are due to neglect on the part of the carers, who show inadequate and negligent parenting with failure to protect the child. Others heat-induced injuries are deliberately inflicted. 2s - 27 Sometimes it is eventually suggested by the perpetrato r that these are punitive in nature, allegedly punishing the child for 'bad' behaviour and failure to comply with parental instruc tions, for example in relation to 'potty training' or playing with matches. They may also be sadistic, although less fre quently, in their production, with the perpetrator obtaining pleasure by the infliction of pain and fear on the child. In about 10 per cent of physically abused children bums are a component of the injuries inflicted; about 5 per cent of sexu ally abused chi ldren also show bums. 28 31 Injury induced by liquids is usually referred to as sca lds. These often blister, with peeling of the epidermis when the blisters burst. The damaged area often appears soggy and blanched. Their distribution may follow the contour of the clothes given that the soaking of clothing with hot liquid will increase the duration of the contact between the liquid and the skin and may thus follow the contours of clothing. The liquid will produce splash patterns - pouring and drip pat terns. Domestic hot water, boiling kettles and saucepans, pip in g hot cups of tea and coffee, and baths are frequent causes. Scalds may also be found internally on the lips, gums, cheeks , tongue an d pharynx as a ctlllsequence of swallow ing hot liquids. 32 Large areas of mucosa will peel off, and if
326 I
Heat-induced injury or death
oedema supervenes as a reactive change there may be an obstruction to upper air passages and to air exchange. Contact with metallic, hot, dry surfaces will produce dry or contact burns, such as with contact with central heating radiators, an oven, electric heate r, clothes' irons, curling tongs, fire surrounds, light bulbs, heating grills; serious burns can result even though the contact with such objects is usuall y of a very short duration. These burns may demonstrate a branding type of appearance, with the burn appearing dry and well demarcated, reproducing the pattern of the object that has caused t he burn. Lit cigarettes may leave very characteristic patterned circular or oval marks on the skin and although these may be accidentally induced , particularly in a crowded house hold, they are also often deliberately inflicted on the child. In the latter instances areas of the body that are less easily visible to a medical examiner or observer will be used by the perpetrator: thus such Cigarette burns may be found on the lower back, the nape of t he neck, the buttocks and the legs. They are often multiple, and indeed may also be of a different duration, and if they present in varying healing phases then they are indicative of more than one episode of such trauma. A pattern that is seen occasionally is that the outline of the main bruise is oval in its configuration with a roughly triangular ta il of further, usu ally more superfi ci al, burning radially confluent with the main burn but located at an angle to it. This is due to move ment by the cigarette across the skin or by the child attempting to move away from the point of impact with the lit Cigarette end. This appearance can thus be found in both accidental and delib erately inflicted cigarette burns. Accidental burns are found at sites where the child has handled the lit cigarette or fallen on/brushed against it; thus the back of the hands, the feet and the back are not the usu al sites of an accidental contact. Radiant burns may occur after prolonged contact with the sun or a rather hot tIre. These affect a large anatomical area of the body and are often ch aracterized by marked erythema. All of the surface that is exposed will show heat indu ced injury. Clothing may have a protective effect and thus patterns may be observed in such burns. Chemical agents, not least acids and alkalis, cleaning agents used in the house and also electricity sources will also induce burns. The amount of heat generated and thus the a mount of tissue da mage is equal to 0.24 x (voltage)2 x resistance. 33 Thus voltage is the main determinant of tissue injury, with electric burns thus being divided into low volt age burns (as with t he domestic current) and high-voltage burns. Low-voltage burns are often deep-contact burns at the entry and exit points of the current from the body. When the latter involve voltages of over 1000 volts they are subdivided further into true 'h igh-tension' injuries caused by a high voltage current going through the body and flash injuries caused by tangential exposure to high voltage current with no current actually flowing t hrough the body. These may show extensive tissue losses and often
even involve the loss of limb. There may also be muscle da mage with rhabdomyolysis, which, in turn, may lead to acute renal failure. Contact with voltages of over 70000 volts is invariably fatal. In flash electric burns there is exposure to radiant heat of the surface of the body, often the face or on the arms, but clothing may also be set on fire with further burns ensuing. The depth of a burn is related to: • the temperature of the heat source. • the duration of contact between the skin or mucosa and the warm object. The usual temperature of hot water in the home is about 60°C and contact with water at this temperature results in a full-thickness burn in child at about 10 seconds of con tact. Higher temperatures will cause burns much quicker and at 70°C only 1 second of contact is sufficient for a full thickness burn. About 3-10 per cent of burns in British children are due to NAI.34 In American studies, 9-25 per cent of children who are admitted to burns uni ts have been assaulted. Younger children in the age range 5 months to 6 years are often affected, with a peak noted in older toddlers around the age of] years; about one-third of these children will be repeatedly abused. This is in contrast with other forms of abuse when younger children usually feature more promi nently. In assessing whether a heat-induced injury is accidental or otherwise the presentation of the child and the history given by the carer at medical presentation is impoJta nt, as is the interaction of the carers with medical and nursing staff. If the injury had been inflicted non-accidentally then the appearance of the injury and the history given by the carer may not tally: the carer may even suggest that the injury was not induced by heat or the explanation given for the contact with the warm solid or liquid is incongru ous, for example claiming t hat a two-year-old has opened a tight, large, hot water tap in a bath. Siblings may be blamed and the episode is claimed to be unwitnessed; when the child is asked, there is a different version of events given. Associated concomitant injuries and previous heat induced damage may fe ature in the past medical history in such instances. If t he child has been deliberately immersed into hot water then a clearly delineated tide mark may be seen if the lo we r limbs are involved . Similarly, if the buttocks are immersed then a 'doughnu t' demarcation will be observed on the buttocks. The deliberate infliction of scalds on a child may be the result of chastisement for failure by the child to comply with potty training and in bed wetting. In many instances, there will also be splash marks elsewhere. Children thus non-accidentally assaulted should be exa mined for other signs of abuse and neglect. The differ entiation between accidental and non- acc idental burns is not easy and, unless suspected in all >circumstances, non accidental injUly will be missed?5
References I
BURNS CAUSED BY ELECTRICITY
An electric current is a flow of electrons along the path of least resistance toward a natural ground with any sub stance or object along this path decreasing the rate of flow (resistors) or increasing the rate of flow (conductors). The skin is a natural resistor to electrical flow: dry skin has a resistance of 40 000-100 000 ohms, wet skin approximately 1000 ohms and calloused skin 2000000 ohms. The thin skin and high water content of the skin of children has a resistance much lower than that of an adult. J6 The internal resistance of the body is estimated to be between 500 and 1000 ohms, with bones, tendons, and fat providing most resistance. Nerves, blood vessels, mucous membranes and muscle are the best conductors. With burns resulting from an electric current, the cross sectional area is inversely proportional to tissue damage. The pathway of the current plays an important role in determining injury, with a vertical pathway being more dangerous than a horizontal (hand-to-hand) pathway.J7 Standard household current in the UK is 230-240 volts (alternating current lAC]) at a frequency of 50 hertz; in the USA and Canada it is 110 volts, with a frequency of 60 hertz. Skeletal muscle is stimulated into spasm and tetany by cur rents with frequencies of 40- 110 hertz. Most low- and high tension electrical current is AC. Alternating CUITent produces tetany and the 'locked-on' phenomenon as the fle xo r muscle groups are usually stronger and predominate. As a result, an individual's grasp is uncontrollably locked on to an object in which an electric current is passing, which can increase the length of time the current passes through the body and may result in greater injury. In contrast, direct current (DC) tends to produce a single large muscular contraction t hat often throws the patient away from the source. However, at high voltages, both AC and DC produce similar effects. Contact with high-voltage CUITents results in injuries that are associ ated with arc burns and flash burns. Arc temperahlres may reach up to 5000· C and are usually responsible for the severe thermal injuries. Lightning involves a single massive current impulse that is roughly equivalent to a DC blast of 2000-2 billion volts of extremely short duration (0.1-1 ms).J8 The peak tempera ture in the lightning strike channel is 30000·C. However, the short duration usually precludes serious direct thermal injllly. Four modes of lightning injury have been described: direct strike, side flash (discharge from an object near the victim)' stride potential (enters one foot and exits via the other), and flashover (energy passing outside the body with vaporization of surface water and blast effect to clothing). Lightning may cause full cardiac arrest by inducing either asystole or central apnoea. Massive depolarization of the heart leads to asystole. However, the heart usually restarts spontaneously in normal sinus rhythm. Massive depolariza tion of the brain is believed to stun the respiratory centre, causing a much longer duration of central apnoea. If artifi cial respiration is provided, many patients can survive.
327
Electricity can cause disruption of the body's normal electrical activities, with neurological 39 dysfunction being present in some form, even if only temporalY, in virtually all patients. 4o Transient nerve injuries resulting in tempo rary numbness and tingling are most common. Mass depo larization of the brain may lead to a loss of consciousness amnesia and coma. Spinal cord involvement may result i~ transverse myelitis. Transverse myelitis may have delayed onset and is associated with poor prognosis for recovery. Electrical injuries also may affect the heart 41 and about 25 per cent of patients with electrical injuries have cardiac dysrhythmia. Many of these are benign and transient (e.g. sinus tachycardia , premature atrial ventricular contractions and conduction disorders). Sudden death from an AC elec trical injury often results in ventricular fibrillation, although asystole and other dysrhythmias are common. Ventricular fibrillation is three times more likely to occur if the flow of current is arm-to-arm. Tnle myocardial infarction is rare unless the patient has pre-existing cardiac disease. Another mechanism of injury is related directly to the amount of heat generated by the flow of electrical current through body tissue. At higher voltages, higher tempera hIres are achieved thus resulting in greater direct thermal injury. High-tension voltages cause devastating injuries from huge amounts of internal thermal damage. Vascular injury occurs as a result of vascular spasm, and the heat generated can also cause coagulation and vascular occlusion. Damage to the vascular wall may produce delayed thrombosis and bleeding. Compartment syndrome may develop as a result of acute ischaemic insult to the muscula ture. Renal injuries may occur as a result of rhabdomyolysis. Rhabdomyolysis causes myoglobinuria from massive release of myoglobin, which, on crystallization in the kidney tubules, may cause acute renal failure.
REFERENCES Runya n SA, Casteel C (eds) The State of Home Safety in
Americr: Facts about Unintentional Injuries in the Home, 2nd edn. Washington DC: Home Safety Council, 2004. 2 Centers for Disease Co ntrol and Prevention. Web-based Injury
Statlstlcs Query an d Reporting System (WISQARS) [online). 2005. National Center for Injury Prevention and Control, Centers for Disease Control and Prevention. www.cdc.gov/n cip c/w isqars 3 Internation al Association for the Study of Insurance Economics. World Fire Statistics: Information Bulletin of the World Fire Statistics. Geneva: The Geneva Association, 2003. 4 Ma lon ee S, Istee G, Rosenberg M et al. Surveillance and prevention of reside ntial - fire injuries. N Engl J iVIed 1996; 335:27-32. 5 Sorensen B. Prevention of burns a nd scalds in a developed counBy. J Trauma 1976; 16:249-58. 6 Linares AZ , Linares HA. Burn prevention programmes for children: are they effective? Burns 1979; 6:26-9. 7 MacK ay A, Rothman K. The incidence and severity of burn injuries following Project Burn Prevention. Am J Publ Health 1982; 72:248-5 2. >
328 I
8 9
to
11
12
13
14 15
16 17 18 19 20
21
22
23
Heat-induced injury or death
Istre GR, McCoy MA, Osborn L et al. Deaths and injuri es from hou se fires. N Engl } Nled 2001; 344: 1911-16. McLoughlin E, Marchione M, Han ger L et al. Smoke detecto r legislation: its effects on owner-occupier homes. Am } Publ Health 1985; 75:858-62. Ahrens M. US Experience with Smoke Alarms and Other Fire Alanns. Atlanta, GA: Quincy [MA) National fire Protection Associatio n, 2004. Mill er R, Reisinger K, Blatter M, Wu cher F. Pediatric co unselling and subseq uent use of smoke desecraters. Ani J Publ Health 1982; 72 :392-3. Carlson A, Uden G, Karlsson ED. Burns injuries in sma ll children, a population-based study ill Sweden . } Clin Nursing 2006; 15:129-34. Anderson RA, Watso n AA, Harland WA. fire deaths in the Glasgow a rea: general cons iderat ions and patho logy. Med Sci Law 1981; 21: 175-83. Napier DH. Haza rdou s materials and· the gases they produce. Med Sci Law 1977; 17 :83-90. Kojima T, Nishiyama Y, Yashiki M, Une I. Post mortem formation of carbon monoxide. Forensic Sci Int 198 2; 19: 243-4. Anderson RA, Ha rl a nd WA. Fire deat hs in the Glasgow area: the role of hydrogen cyanide. Med Sci Law 1982 ; 22:35 -40. Hettiaratchy S, Dz iew ulski P. ABC of burns: pathophysiology a nd types of burns. BM} 200 4; 828:1427 -9. Knight B. Burns and sca lds. In Kn ight B [ed.) Forensic Pathology, 2nd ed n. London: Arno ld, 1996, pp. 305- J 7. Weaver AlVI, Himel MHM, Edli ch RF. Immersion scald burns: strate gies for preventio n. J Emerg lVled 1993; 11 :39 7-402. Erdmann T, Felman K, Rivara F et a!. Tap water burn prevention: the effect of legi slati on. Pediatrics 1991: 88:572-7 . Sm ith GA , Knapp Jf, Barnett TM, Shield s BJ. The rockets· red glare, the bomb s bursting in the air : fireworks- related injuries to children. Pediatrics 1996; 98: 1-9. D'Regenio P, Cafaro L, Santon3sras i F, Taggi F et al. Capodanno Senza Danno: the e ffects of an intervention program o n fireworks injuri es in Nap les. Am J Pub' Healtlr 1996; 86:84-6. Walker AR. Emergency department management of house fire burns and carbon monoxid e pOiso ning in chi ldren. Cur Opin Pediatr 1996; 8:239 -42.
24 Wo lf SE, Rose JK, Desai MH et a!. Mortality determinants in massive pediatric burns. An analysis of 103 chil.dren with ;'80% TBSA burns [;'700/0 full thickness). Ann Surg 1997; 225: 554-65, discussion 565-9. 25 Ayoub C, Pfeifer D. Burns as a manifestation of ch ild abuse and neglec t. Alii J Dis Child 1979; 133:910- 14. 26 Sto ne NH , Rinldop L, Humphrey CR et a1. Chi ld abuse by burnin g. Surg Clin North Am 1970; 50:141 9- 24. 27 Showers J, Ga rrison KM. Bu rn abuse; a four -yea r study. J Traumn 1988; 28: 1581 - 3. 28 Hight DW. Bakalar HR , Lloyd J. Inflicted burns in children: recognition and treatment. JAMA 1979; 242:517 - 20. 29 Keen JH, Lend rum J, Wolman B. Inflicted burns and scalds in children. BM} 1975; 4:268-9. 30 Hobbs CJ. When are bum s not accid enta l? Arch Dis Child 1986; 61:357-61. 31 Lenoski IF, Hunter KA. Speciflc patte rn s of inflicted burns injuri es . J Trauma 1977 ; 17:842. 32 Canady JW, Thomps on SA, Bardach J. Oral commissure burns in chi ldren. Plost Recon struct Surg 1996; 9 7:738-44, discu ss ion 745, 74 6-55. 33 Luce E. Electrical burns. Clin Plast Surg 2000; 27:133 - 43. 34 Andronicus M, Oates RK, Peat J et a1. No n-accid enta l burns in children. Bums 1998; 24:552 - 8. 35 Benger JR, McCab e SE. Burns and scalds in pre-school children atte ndin g accident and emergency: accid ent or abuse? Emerg Med J 200 1; 18:172-4. 36 Ga rcia CT, Smit h GA, Cohen DM, Ferna ndez K. Electrica l injuri es in a pediatric emergency depa ltment. A nn Emug IVIed 1995; 26:604-8. 37 Nguyen BH, Mac Kay M, Bailey B, Klassen TP. Epidem iology of electrical- and lightning-related deaths and injuri es amo ng Canad ian ch ildren and you th. Injurv Pre" 2004; 10:122 - 4. 38 Jain S, Bandi V. Electrica l a nd li g htning injuri es. Crit Care Clin 1999; 15:319 - 31. 39 Lee RC. Injury by electrical forces: pathophysiology, manifestations, and thera py. CUlT ?robl Slirg 1997 ; 34:677-764. 40 Rai J, Jeschke MG, Barro w RE , Herndon DN. Electrica l injuries: a 30-year review. J Trauma 1999 ; 46:933 - 6. 41 Bailey B, Gaudreau! t p, Thivierge RL, Turgeon JP. Cardiac mon itoring of children with hou seho ld electrical injuries. Ann Emerg Med 1995; 25:612-17.
I
CHAPTER 16
I
ASPHYXIAL DEATHS IN CHILDREN Anthony Busuttil
Abuse of inhalants (solvent abuse)
Overlaying and wedging
329 330 330 330 330 331 331
Strangulation
331
References
Petechiae Scene of death Traumatic asphyxia in children Entrapment asphyxia Foreign body inhalation Plastic bag asphyxia
The term 'asphyxia' comes from the Greek sphygmos (meaning absence of a pulse) and is applied to pathological unnatural states in which the body is deprived of oxygen while there is a concomitant excess of carbon dioxide, i.e. a hypoxaemia (hypoxia) and hypercarbia (hypercapnoea) together. This may result in loss of consciousness with the coma thus produced leading to death. Children die from mechanical asphyxia in many situations and exhibit simi lar pathological changes that one would expect in an adult dying in a similar situation. In the international classification of diseases, accidents caused by asphyxia are classified under E91O-E913: • E910 = Accidental drowning and submersion. • E911 = Inhalation and ingestion of food causing obstruction of respiratory tract or suffocation. • E912 = Inhalation and ingestion of other object causing obstruction of respiratory tract or suffocation. • E913 = Accidental mechanical suffocation.
PETECHIAE This pathological feature has long been recognized in the forensic context. As a finding of medicolegal import, petechiae were first described by Ambrose Tardieu in 1855. 1 It was, at one stage (and perhaps by many still is),
Hanging by a ligature Drowning and near drowning Imposed airways obstruction Reverse suspension Chemical asphyxia Prevention
332 332 332 333 333 333 333 334
considered to be pathognomonic and an almost sine qua non of asphyxia. These are haemorrhages arsing from tiny blood vessels, probably mostly venules (capillary haemorrhages may be too small to be visible naked eye being 1-2 mm lesions or smaller), which rupture when the pressure within them is increased and the endothelial lining is distended beyond its innate elasticity. Contribution by hypoxaemia is likely in that the vessel wall is also affected by a decrease in oxygen tension, becoming leaky and pennitting the extravasation of red blood cells. These lesions tend to occur at sites where the connective suppot1ing tissue is loose, provided that the pres sure is locally raised in the specific area. 2 Thus an applica tion of a sphygmomanometer cuff at a level sufficient to occlude, selectively, the venous supply for a period will result in petechiae formation distal to the edge of the cuff. Another cause of petechial haemorrhages is a blood clotting problem. A low platelet count from any cause, for example idiopathic thrombocytopenic purpura, aplastic anaemia, or leukaemia, may produce petechiae. Excessive capillary fragility such as Henoch-Schonlein purpura may have a similar effect (see Chapter 4). Petechiae may also be found when there is toxic damage to the endothelium, particularly in association with dissemi nated intravascular coagulopathy, as in coliform and meningococcal septicaemia. In septicaemia, petechiae tend to
330 I
Asphyxial deaths in children
be widespread. They are also found in situations when microembolic phenomena occur, such as fat and air embolism, amniotic fluid embolism and sub-acute bacterial endocarditis. In asphyx ia, these haemorrhages tend to be dis tributed externally above the level of the obstruction and not below the level, and are thus commonly found: on the palpe bral and bulbar conj unctivae ; on the face - mainly around the eyes; behind the ears (pinnae); and on the mucosal sur faces of the lips and nose. In children it is not uncommon to find these haemolThages in a single site only and not neces sarily the eye. 3,4 They may be present over the face a nd ante rior chest wall in still births and in early neonatal deaths of babies who die as a result of acute hypoxic/ischaemic insult, particularly retroplacental haemorrhage (see Fig. 10.10). They also occur naturally in normal children, albeit in sma ll numbers. In a total of 116 children under the age of 1 yea r, who were fully examined in child surveillance clin ics in Newcastle-upon-Tyne, Downes et al fo und that 27.6 per cent of children had one or more petechi ae, 8.6 per cent two or more, and 2.6 per cent had more than two. s Petechiae can also occur if the child has been experienc ing pronounced and recurrent Valsalva manoeuvres, as with coughing (for example pertussis infection , bronchioli tis, vomiting [gastroenteritis], crying, straining, temper tantrums). This may also explain why petechiae may be discovered after active cardiopulmonary resuscitation. 6 They a lso occur interna ll y with any form of hypoxaemia as the final common pathway for death. They are character istic of the sudden infant death syndrome, in which they are found on the thymus, the epicardium and the pleurae. 7 - 9
environmental catastrophes such as earthquakes and gas explosions, will show features above the level of obstruc tion associated with t his condition. Another accidental asphyxial cause of death in yo ung children is being run over at low speed by reversing cars, not infrequently driven by one of the other members of their own family and sometimes in their own driveway. The child's short stature precludes their visibility to the driver with a consequent accidental knocking over of the child who, having a low centre of gravity, will tend to fall below the ve hicle rather than being lifted up and over the car. Many children surv ive this type of insult, provided that the duration of appl ication of the compressive force is not prolonged, the velocity of imp act is low and the weight of the object trapping them is not considerable. 13 - IS Very often they present with features of cerebral anoxia and convulsions. They frequently have visceral injuries and extensive soft tissue crushing injury. Drag marks in the shape of directional scuff abrasions are often found on the body, indicating the direction in which the vehicle has moved over the child's body. Many of these children are yo unger than J years and boys predominate. This matter is a serious public health issue. In Aus tralia, 16 this type of accident acco unts for 8 per ce nt of pae diatric pedestrian fatalities, in New Zealand 10.7 per cent and in the USA 20 per cent. 17 ,18 Ultrasonic transceivers located on the rear bumpers have been introduced as prox imity warning devices.
SCENE OF DEATH
The inquisitive nature of children during play actIVIties may result in their entrapment in objects and spaces from which they are not able to extricate themselves and thus they die as a consequence. Discarded chest freezers, fridges, old safes, large tnmks and suitcases adopted by children as play areas.can cause such problems. Entrapment in the luggage compaliments of cars (car trunks, boots) may have a similar effect. In the latter, there may also be the added effect of heat if the vehicle has been parked in a particularly hot day at sites where the sun is shin ing directly on to it; heat stroke (hypertherm ia) may develop in such cases in addition to asphyxial changes. Such deaths can occur if the outside temperature exceeds 29.5"C. Cars parked in direct sunlight can reach interna l temperatures of 55°C to 78°C when the outside temperature is 27 -38°C. The less well ventilated the space in which the child is trapped, the more likely is it that excessive internal temperatures are reached . This temperature rise occurs within about 15 min utes of co nsta nt exposure to tllis temperature. 19,20
As with aU other deaths it is essential that the death scene is thorou gh ly examined 10 in all cases when asphyxia appears to have been the mode of death. In this respect, if the baby has already been moved it maybe usefu l to seek a reconstruction of the in cident of w here exactly the dead child was found, by using dolls that can be manipulated by the carers in line with their recollection of events. I! These interactions with the family may be very emotionally fraught and have to be carried out with great sensitivity by police officers or others who have been specia lly trained . 12
TRAUMATIC ASPHYXIA IN CHILDREN In this situation the thoracic cavity is transfixed and no respiratolY movements are thus possible. There are classic signs of congestio n above the obstruction, central cyanosis and petechial haem orrha ge formation also above the obstructio n, usually ending at about the level of the clavi cles. Children who are trapped under masonry and other items, for example in explosions, in the course of crowd stampedes, in the course of warfare and as a result of other
ENTRAPMENT ASPHYXIA
FOREIGN BODY INHALATION An aspirated foreign object, which is splid or semisolid, can lodge in the lalynx, trachea or main bronchi of a child. If the
Strangulation I
object is large enough to occlude the airway completely, it will lead to immediate asphyxia by preventing any gaseous exchange in the lungs and death results within minutes. Pas sage of the object beyond the calina may still be quite dan gerous and can cause serious breathing problems; it may also result in death from bronchospasm in those who are susceptible. Common objects include seeds, nuts, bone frag ments, nails and screws, small toys and pins. 2J.22 As the angles of bifurcation of the main stem bronchi are acute, for the first 15 years of life foreign bodies may find their way into either side of the lower airways.23 Once aspi rated, objects may subsequently change position or migrate distally, particularly after spontaneous or external attempts at removal of the object, after thumping on the back of the patient or attempting to make him or her cough or retch. This may cause delayed obstruction. Inhaled vegetable mate rial may swell over subsequent hours or days, and cough, stridor, wheeziness, breathlessness and cyanosis may ensue. Other objects, such as peanuts and other organic foreign bodies, may, in addition, excite an acute inflammatory response with tissue swelling, which makes the level of asphyxia gradually worse; children aged 1-3 years are more at risk and death results at a rate of 0.7 per 100000 per annum in the USA. This is due to the tendency of young children to put everyth ing into their mouths and the way in which they chew food. Because their molars are unerupted they tend to use their incisors; objects a nd fragments of food are then propelled posteriorly, thus exciting a reflex reaction and a tendency to inhalation. 24 ,25
PLASTIC BAG ASPHYXIA In the later 1950s, polythene bags started to be used for packaging and as supermarket carrier bags. It was soon reported that children had died directly as a consequence of these plastic bags being placed over the head, although not necessarily secured around their neck. 26 In most countries, it is now obligatory by law that all polythene bags used for wrapping have holes intrinsically cut into them close to the base to allow air to enter if they are slipped over the head; however, this legislation does not cover such household items such as rubbish bin bags, carrier bags etc., and the role of child carers cannot be overestimated with respect to the prevention of these deaths. The imm ediate cause of death in these deaths was at one time thought to be due to a decreasing oxygen concentra tion within the bag and rebreathing of carbon dioxide, with consequent narcosis and loss of consciousness. Physical obstruction of the nose and mouth may also have aided and abetted this mode of death. In addition, it was pro posed that the bag becam e electrostatically charged and adhered to the face, aided by condensation of water vapour from the expired air. However, as there is a dearth of typi cal asphyxial features and the persons concerned looked pale and placid, it was postulated that other phenomena
331
were at play. Knight 27 suggested an overstimulation of the sympathetic nervous system (resulting in arrhythmias, par ticularly ventricular fibrillation) caused death. In some of these cases in older children, a gas or a vapour, for example solvent vapour, may have been intro duced into the polythene bag and this may also have influ enced directly the mode of death.
OVERLAYING AND WEDGING Overlaying is the accidental death of a child by smothering resulting from a larger individual sleeping on top of the baby in the course of deep sleep or sleep induced by extraneous intoxication . This has been a well documented condition since biblical times 28 and tended to be much commoner in Victorian times 29 when parents in a state of inebriation may have taken to bed a large number of their children. 3o This hazard is greatest in infants below the age of 5 months but can occur in children of up to 2 years of age. 3J ,32 The rolling over by the adult to co me to lie on top of the child exerts pressure on the child 's face against bedding, the mattress or into the body of the sleeping adult or a co-sleeping older child. Owing to the pressure being exerted on the child's chest, he or she is unable to cry out and attract attention. Some of these babies show no pertinent clinical signs, often not even petechiae. 33 Some babies show the presence of contusions and abrasions, as well as an unusual distribution of the lividity demonstrati ng the points of compression. Indeed, it is stated that if a child who appears to have died of overlaying shows other exter nal blunt force injuries, it is likely that the child has been the subject of non-accidental injury.34,35 In wedging, the child may be wedged by the bed's co-occupants against a wall, between the mattress and the wall, the bed frame or an adjacent piece of bedroom furniture, the cot sides, railing in the cot sides, the head- or footboard of the bed, other co-sleepers including other children etc., with similar consequences. 36,37 Another hazard has been associ ated with co-sleeping on waterbeds. Kirchne.-J7 identified 515 such deaths in children below the age of 2 years during a 7 year period. These included 121 deaths due to overlaying by a parent, sibling or by other adult - with 77 per cent of the deaths involving children younger than 3 months; 394 deaths due to wedging in beds - 296 on regular ad ult beds, 79 on adult waterbeds - with onJy two involving previous alcohol or drug abuse, 10 on adult day beds, nine in adult-sized beds fitted with bed rails. 38 The larger the number of persons occu pying the beds and the higher the weight of the occupants, the greater is the risk involved. Unsafe sleeping environments for babies are further discussed in Chapter 11.
STRANGULATION Non-intentional, accidental self-sJrangulation of young children with loose wires, cords and other potential ligatures
332 I
Asphyxial deaths in children
found commonly around the house and often in close proximity to their beds are well documented. 39 - 4 ! Entangle ment in such cords was responsible for 14.3 per cent of Amer ican childhood deaths. Other causes were plastic bags, bedding (non-plastic), cords of blinds or curtains, cords by which dummy teats (pacifiers) were attached to them and other types of cord. 32 An intravenous tubing set was also reported as hav ing caused such a fatality in a I-month-old boy and a near fatality in an 8.S-month-old boy,42 and the wiring of an apnoea monitor, which was non-fatal, in another baby.43
HANGING BY A LIGATURE Deaths by hanging in children below the age of 14 years are uncommon. Much of the published literature relates to individual case reports, with no epidemiological studies being undertaken. 44 - 46 Cases usually occur against a back ground of unhappiness or frank depression. A history of poor achievement at school, parental and peer bullying and similar adverse effects come to light after death when a psychological review is carried out. There may have been previous attempts at self-harm. Autoerotic unintentional deaths from suspension have been reported in children aged 9 years and above. 47 - 49
the low oxygen tension causing endothelial cellular dam age and hyperpermeabili ty with leakage of body fluids into the extracellular space. 51 Hyponatraemia may also develop if large quantities of fresh water are swallowed or if the syndrome of inappropriate antidiuretic hormone secretion occurs. Inhalation of 13 mL/kg of fluid can result in signifi cantly impaired gas exchange. Fresh water, being hypo tonic, damages type 2 pneumocytes and causes a disruption of alveolar surfactant. Sea water draws fluid from the blood into the alveoli and thus dilutes surfactant, resulting in lower residual pulmonary functional capacity and pulmonary oedema, with development of an acute res piratory distress syndrome. There may also be plugging of small airways by debris, which increases airway resistance. Release of inflammatory chemical mediators from the lungs leads to local vasocon striction. Biochemical changes may lead to cardiac dys rhythmia or asystole due to hypoxaemia. There may also be metabolic acidosis and acute pulmonary coextension. In the USA, for each death from drowning (about 1500 children per annum) an additional four hospital admissions per drowning death occur after submersion incidents. 52 ,53 Drowning and near drowning are further discussed in Chapter 18.
DROWNING AND NEAR DROWNING IMPOSED AIRWAYS OBSTRUCTION Drowning is defined as death from asphyxia within 24 hours of submersion in water. Near-drowning refers to sur vival for 24 hours after a submersion episode. Drowning by definition is fatal but near drowning may also be fatal. 50 Drowning in children is often classified according to the temperature of the water into which immersion has taken place: warm-water drowning at temperatures greater than or equal to 20°C, cold-water drowning in temperatures less than 20°C, and very cold-water drowning when the tem perature is less than or equal to 5°C. When a person is submerged in water and attempts to breathe, he or she may inhale or aspirate water into his or her airways, and laryngospasm may develop, thus leading to deceased oxygenation, hypoxaemia and brain death. Vomiting may occur at this time and gastric contents may also be aspirated into the airvvays, further complicating matters. About 10 per cent of persons who drown do not inhale water or gastric contents. In near drowning the cardinal features are those of hypoxaemia of the internal organs. This may affect the brain, but also and perhaps more importantly in the sur vival situation, the lungs. Most individuals will aspirate less than 4 mL/kg of body weight; 11 mL/kg body weight is required for the aspirated fluids to alter the blood volume and about twice this level is needed to alter electrolyte levels in the blood. Thus fresh water has earlier effect than salt water. In most patients who survive, hypovolaemia is a frequent finding owing to
One form of child abuse that has been shown to occur under the age of I year is for the parent, usually the mother or another carer, to obstruct the mouth with a soft object such as a pillow, clothing or by pressure against her breast or in some other way. The child does not seem to struggle too much or too ovelily and develops hypoxia and cyanosis; he or she may also demonstrate convulsions due to cerebral hypoxia and, eventually, a respiratOlY and cardiac arrest. This may occur when a mother with psychosocial problems is attempting to attract attention to herself, formerly referred to as Munchausen's syndrome by proxy. Coveli video surveillance has been used to document such abuse, with a strong debate as whether this is ethical and legal. 54 Whether the end (namely convincing the Crim inal Court of the abuse by the compelling evidence of a video) justifies the means is dubious in the mind of many of the professionals involved in such cases. This form of surveillance was first used in 1983 55 and since then several reports of it have appeared. 56 - 62 The method used is to have the child attached to a multifunction polygraph (recording respiratory raster, ECG, EEG, pulse rate, blood pressure) in a cubicle, thereby ensuring that the infant cannot be moved away from the area of surveillance and thus having physiological criteria that can be compared and timed with images recorded on camera. These children present with an aClJte life-threatening event, 'breath holding attacks' or recurrent convulsions.
--
--
---
-----------------------------------
----
Prevention I
It may be extremely difficult to differentiate between a nat ural phenomenon and abuse phenomenon.63 It can be extremely difficult to find any signs of the external airways obstruction, although a very careful search for bruising on the face, on the inside of the lips and cheeks should be made; in the mucosa l sites bruising may be accomp anied by minor abrasions. Internally, the presence of haemosiderin in the lungs, within both alveo lar cells and in the interalveolar septa, is of great significance. In the acute phase there will have been haemorrhage within alveoli and this may give rise to frothy frank ly blood-stained fluid emerging from the nose and mouth.
ABUSE OF INHALANTS (SOLVENT ABUSE) Inhalants, usually hydrocarbons, are breathable chemical vapours that produce mind-alteri ng effects similar to the effect of alcohol consumption or the use of controlled dnlgs. They are an ingredient in many househoJd products, such as agents used in cleaning, decorating, painting (turpentine, white spirit), sta ins, vamishes, g lu es, air fresheners, hair con ditioners, dry cleaning agents, shoe polish, colour markers, na il polish removers, spot removers and degreasers and deodorants, as well as in fuels such as petrol (gasoline), butane and lighter fuel. Anaesthetic agents, such as ether, chloroform, halothane and nitrous oxide, may be used in a similar manner. The so lvents are taken in to the body by: squirting directly into the mouth from cans and canisters, inha ling from bags or an aerosol can, sn iffin g directly from the lid of the container, breathing in directly (huffing) or after application of the substance in a rag or handkerchi ef. Plas tic bags containing the vapour may be placed directl y into the mouth. There are about 1400 compounds that can act in this manner, including nitlites, alcohols and halo genated compounds, in addition to hydrocarbons. Death may occur in a variety of ways: • direct acute toxicity of the brain, leading to respiratory a nd cardiac arrest; • cardiac arrhythmias brought on by the substance inhal ed, followed by cardiac arrest; this may be brought about by sudden SPUltS of muscular activity and release of a bolus of catecholamines, as if the subject while under the influence is chased and made to run away; • suffocation by the item being used to present the vapo ur to the mouth, for example plastic bag or balloon; • vomiting w hile unconscio us from the effects of the inhalants; • suffi xa tion by displaci ng air from the lungs with an irrespirable vapour; • acute parasympathetic stimulation with inhibition if me substances cause irritation and/or cooling of the upper gullet or n asophary nx;
333
• accidents while und er the influence of these substa nces, due to loss of judgement and impairment of other cognitive functions; • burns produced by inadvertent ignition of the inflammable accelerant used; • long-term toxic effects of the solvent on the marrow, liver, kidneys and brain; • chronic respiratory problems due to chronic bronchitis and recurrent inhal ation al or chemical pneumonitis. Note that in some older children, these substances may be used as sexual aid in the course of autoerotic activities. In cases when such abuse is suspected, there may be lit tle to find at autopsy. There may be a rash around the mouth and blisters where the substance had been app lied and produced a solvent degreasing effect on th e skin. The so lve nt may still be appreciable through its smell at autopsy. One lung should be retai ned in a nylon fibre bag to allow 'a ir' to be sampled from it directly from its airways for gas chromatography assay. The blood , the liver and the kidneys can also be assayed toxicologically. Histology may confirm chronic changes in the internal organs particularly the lungs. No petechial haemorrhages are found either internally or externa lly, except in the few instances of suffocation.
REVERSE SUSPENSION This is perhaps a rarity, which may occur in children dur ing play activity when death results from exhaustion of breathing du e to splinting of the diaphragm 64 ,65 by the upwards displacement of abdominal contents. Death occurs slowly from ex haustion of respiratory effort.
CHEMICAL ASPHYXIA This term is applied to situations when irrespirable gases find their way into the child's environment. For example, carbon monoxide in fires and from exhaust fumes (in closed garages) and from charcoal fires (barbecues) in an enclosed space, chlorine escaping from swimming pools, hydrogen sulphide and methane from refuse tips, old mines and slurry pits are among the gases that can cause death after inha latio n. These events usu ally involve older, active children and, more frequently, boys, in common with many environm ental mishaps.
PREVENTION As incidents involving asphyxia are common, public health offiCials, community doctors, midwives and health visitors have been involved in several programmes aimed at alerting parents to the dangers of asp hyxia. 66 In severely
334 I
Asphyxial deaths in children
disab.led children accidental asphyxial deaths are more common. 57 In the neonate and in young children below the age of 12 months sleeping practices also have to be care fully scrutinized to enquire that they are not potentially dan gerous to the child. 6B - 72 Co-sleeping in particular should be viewed with some discretion, particularly if the parents indulge in smoking and drinking of alcohol. 73 75
REFERENCES
2 3 4 5
6 7
8 9
10 II
12
13 14 15 16 17
18
19 20 21 22 23
Tardieu A. Etude medico -lega le sur les sev ices et ma uva is tra itmen ts exe rces sur des enfants. Ann Hyg Publ tvled Legale 1860; 13 :36 1-98. Luke JL. Conjunctival petechiae. IV Eng/} iVIed 1971 ; 28 4:1101. Rao VJ , Wetli CW The foren sic sig nifica nce of co njun ctiva l petechiae. Am } Forensic ivfed Pathol 1986; 9:32 4. Jaffe FA. Petechial haemorrha ges. Am } Forensic iVIed Patl10l 1994 ; 15:203 -7. Dow nes AJ, Crossland OS , Mell o n AF. Preva lence and distribution of petechiae in well babies. Arch Dis Child 2002; 86:291-2. Hood I, Ryan 0 , Spitz WU. Resuscitation and petec hiae. Am } Forensic iVIed Pat/101 1988; 9:35-7. Beckwith JB. Intrathoracic petechial hemorrhages; a clue to the mechanism of death in SlOS. Ann N Y Acad Sci 1988; 533:37-47. Krous HF, Jordan J. A necro psy study of distribution of petechiae in non-SlOS. Arch Pathol Lab Med 1984; 108:75- 6. Kleeman WJ, Wiechern V, Schuck M, Troger HD . Intrathoracic and subconjunctival petec hi ae in SlOS . Forel1sic Sci lilt 1995; 72:49-52. Bass M, Kravath RE, Gla ss L. Death-scene inves ti ga ti on in sudden infant death. N Engl} Nled 1986; 315 : JOO-5. Iyasu I, Hanzlich R, Rowl ey 0, Willin ger M. Proceedings of 'Workshop on Guid elin es for Scene Investigat ion of SlO S'. } Forensic Sci 1994; 39 :11 26 - 36. CDC. Guid elines for dea th scene investigation of sudde n, unexplained infant deaths: recommenda ti ons of th e interagency pane l on SlOS. MMWR 1996; 45:RR-IO. Haller JA, Donahoo JS. Traumatic asphyxia in ch ildren: pathophysiology and management.} Trauma 1971; 11 :453-7. Campb ell-Hewson G, Egleston CV, Cope AR. Traumatic asphyxia in children. Accidel1t El11 erg iVIed 1997; 14(1):47-9. Sarihan H, Abes M, Akyazici R et a!. Traumatic asphyx ia in children. Cardiovasc Surg 199 7; 38:93 -5. Holland AJA, Liang RWY, Sing h SJ et al. Driveway motor ve hi cle injuri es in children. iVI}A 2000; 173: 192 - 5. Brison RJ, Wicklund K, Mueller BA. Fata l pedestrian injuries to yo un g children: a differe nt patte rn of injury. Am } Public Health 1988; 78 :79 3- 5. Roberts I, Norto n R, J ackson R. Driveway-related c hild pedestrian injuries; a case control study. Pediatrics 1995; 95:405-8. Sumra ll RE, Petty CS, Holman W. Temperature in closed auto mob iles in hot weather. Forensic Sci Gazette 1976; 7:7-8. Surpure J. Heat-rela ted illness a nd the a uto mob il e. Ann Em erg iVIed 1982; It :263-5. McGuirt WF, Ho lmes HD, Feehs R et a l. Tracheo-bron c hial fo reign bodies. Lm)'ngoscope 1988; 98:615- 18. Weiisberg D, Sc hwartz I. Foreign bodies in the tracheo bronchial tree. Chest 190 7 ; 1: 730 - 3. Cleveland RH. Symmetry of bronchia l angles in chi ldren. Radiology 1979; 11 3:89 - 93.
24 Byard RW. Mec han isms of death in in fa nts a nd young children fol.l.o wing fore ign body ingestion. } Forensic Sci 1996; 41 :438-41. 25 Baharloo F, Veyckemans F, Fra nci s C et al. Tracheo-bronchial fore ign bodies: presentation a nd management in children and adults. Chest 1999; 115 : 1357 - 62. 26 Plast ic bags (Ed itoria l). BM} 1959; 1 :1463-4. 27 Knight B, Sa ukko P. Forensic Pathology, 3rd edn. Oxford: Oxford University Press, 2003 . 28 The Bible, 1 J<jngs Ch . 3 v.19. 29 Templeman C. 258 cases of suffocation of children. Edin Med } 1891; 38:22- 6. 30 Nakam ura S, Wind M, Dane ll o MA. Review of hazards associated w ith chi ldren placed in adult beds. Arch Pediatr Adolesc iVIed 1998; 1 53: 10 19- 23. 3 I Drago DA, Dannenberg Ai. Infant mechan ical suffocation dea th s in the United States 1980-1997. Pediatrics 1999; 15 3: 103 -59. 32 O'Hara M, Harruff R, Smialek J , Fow ler D. Sleep location and infa nt s uffocat io n : how good is the evidence? Pediatrics 2000; 10 :9 15-20. 33 Moore L, Byard RW Path o log ical fInd ings in hang ing and wedgin g deaths in infants and yo un g children. Am } Foren sic
ivIed Pa tl101 199 3; 14 :296- 302. 34 35 36
37
38
39
40
41 42
43 44
45 46 47
48 49 50
Collins K. Death by overlaying and wedg ing: a 15-yea r retrospective study. Am} Foren sic iVIed Patho1200 1; 15 : 155-9. Mitchell E, Kraus HF, Byard RW Pathol ogica l findin gs in ove rl aying.} Clin Forensic iVIed 2002; 9:133 - 5. Byard RW, Beal S, Bourne AJ. Potentially dangerous sleep ing environment and accidental asphyxia in infancy and ea rly ch ildhood. Arch Dis Child 1994; 71 :497-500. Kirchner JT. Deaths associated with small children sleeping in adult beds. Am Acad Fam Phy 2000; March I , www.aaf.org/afp. Nakamura S, Wind M, Danello NA. The pediatric forum: should infants sleep with their parents 7 Arch Pediatr Adolesc ivIed 2000; 154:1171-3. Nixon JW, Kemp AM, Levene S, Sibert JR. Suffocation, choking and strangulation in children in Eng land and Wales; epidemiology and prevention. Arch Dis Child 1995; 72:6 - 10. Rauchschwalbe R, Mann NC. Pediatric wind ow cord strangulations in the United States, 1981-95. }AMA 1997; 72: 1696-8. Sabo RA, Hanigan We, Flessner K et a!. Strangulation injuri es in children.} Trauma 1996; 40:68-72. Garros D, J<jng WJ, Brady-Flyer Bet a!. Strangulation with intravenous tubing; a previously und ec id ed adverse even t in children. Pediatrics 2003; III :732-4. Emery JL, Taylor EM, Carp ente r RG, Wa ite AJ. Apno ea monitors and accidental strangulation. BM} 1992; 304 :11 7. Petruk J, Shiels E, Cummings GE, Francescutti LH, Fatal asphyxiations in ch ildren involving drawstrings in cloth in g. Can Med Assoc} 1996 ; 155: 1417- 19. Bya rd RW, Ma rcopo ul os D et al. Early adolescent suicide; a co mparati ve study. } Clin Forensic Med 2000; 7:6-9. Wyatt J, Wyatt PW, Sq uires T, Busuttil A. Hang ing deaths in children . Am } Foren sic Med Pathol 1998 ; 19 :3 43-6. Feldm a n KW, Simms RJ. Stra ng ulatio n in children: epidemiology a nd clinica l co urse. Pediatrics 1980; 65: 1079-85. Cooke CT, Cadden GA, Hilton JMN. Hanging dea ths in children. Am } Forensic Med Pathol 1989; 10:98- 104. Uva JL. Autoero ti c asphyxiation in the United States. } Foren sic Sci 1995 ; 40:574 - 8 1. Bya rd RW, Lipset J. Drowni ng deaths in tod d lers and pre ambulatory children in South Austra lia. Am} Forensic iVIed Path ol 1999; 20:328-3 2.
-
References I
51 Nodlkel JH. Drow ning. N Engl ] Med 1993; 328:253-6. 52 American Academy of Pediatrics Committee on Injury and
53 54
55 56
57
Poison Prevention. Drowning in infants, children and adolescents. Pediatrics 1993; 92 :292-4. Wintemu te GJ. Childhood drowning and near drowning in the United States. Am ] Dis Child 1990; 144:663-9. Byard RH. Covert video surveillance in Munchhausen syndrom e by proxy - ethical comprom ise or esse ntial technique? Med ] Aust 1994; 160 :352-6. Rosen CL. Frost JD, Glaze DG et al. Ch ild abuse and recurrent infant apnea.] Pediatr 1986 ; 109 : 1065- 7. Southall DP, Stebbens VA, Rees SV. Apnoe ic episodes induced by smothering. Two cases identified by covert video surveillance. EM] 1987; 294:1637-41. Samuels MP, McLaughlin W, Jacobson RR. Fo urteen cases of imposed airway obstruction. Arch Dis Child 1992;
67:162-79. 58 Byard RW, Burnell RH. Covel1 video surveillance in
Munchausen syndrome by proxy: ethics compromise or essential technique? Med ] Austr 1994; 160 :352- 4. 59 So uthall DP, Plunkett MCB , Banks MW et al. Covert video recordings of life-threatening child abuse: lessons for child protection. Ped iatrics 1997; 100:73 5-60. 60 Foreman OM, Farsides C. Ethic al use of covert videoing techniques in detecting Muncilausen synd rome by proxy. BM] 1993 ; 307:611-13. 61 Thomas T. Covert video surveill anc e. New Law] 1994; 62
63 64 65
Recurrent apparent life- threatening events and intentional
suffocation (Editorial). Arch Dis Child 1999 ; 81:189.
Friberg TR , Weimreb RN. Ocular manifestations of grav ity
in version. ]AMA 1985; 253:1755-7. Busuttil A, Obafunwa J. Recreational abdominal suspe nsio n: a fatal practice. A case repol1. Am ] Forensic iVIed Pathol 1993 ;
14: 141-4. 66 Towner E, Errington G. The Epidemiology oj Choking in
Childhood and [l17pli catiollsjo r PrelJentiol1. A report prepared for the Department of Health in London, 2002. 67 Ama nuel B, Byard RW. Accidental asphyxia in severely disabled children. ] Paediatr Child Health 2000 ; 36:66- 8. 61l Beal SM, Byard RW. SIDS in South Australia 1968-97. Part 3: Is bed sharing safer in infants? ] Paediatr Child Health 2000 ; 36:552-4. 69 Byard RW, Beal SM. 'V'-shaped pillows and un safe infant sleeping.] Paed ia rr Child Health 1997; 33:171 - 3. 70 Byard RW, Bourne AJ, Beal SM. Mesh side cots: yet another
71
72 73
114:9 6-7.
74
Shabde N, Kraft AW. Covert video surveillance: an important in vestigative tool or a breach of trust? Arch Dis Ch ild 1999;
75
81 :291-4.
335
potentially dangerous infant sleeping environment. Forensic Sci [l1t 1996 ; 83: 105- 9. Moore L, Bourne AJ, Beal S et at. Unexpected infant death in association with suspended rocking cradles. ] Forensic Med Patho11995; 16 : 177-80. Byard RW, Beal SM. Simpson A et al. Accidental infa nt deaths and stroller prams. Med] At/st 1996; 165:140-1. Byard RW. Is co-sleeping in infancy a desi rab le or dangerous practice? ] Paediatr Child Health 1994; 30: 198 - 9. Byard RW. ls breast feeding in bed always a safe practice? ] Paediatr Child Health 1998; 38:498-9. Scragg R, Mitchel l EA, Taylor A et al. Bed sharing, smok ing and alcohol in the SIDS. BM] 1993; 307: 1312-18.
I
CHAPTER 17
I
ACCIDENTAL INJURIES IN CHILDREN
Anthony Busuttil
Overview of paediatric trauma Bicycle helmets Falls Playground injuries Sports injuries on snow and ice Riding injuries
336 338 339 340 340 341
OVERVIEW OF PAEDIATRIC TRAUMA
Tra uma of children has always been a major public health concern in that a significa nt degree of disability, loss of schoo ling and mortality arise as a consequence. Three major sUlveys that looked at these aspects are: (1) the Canadian Hospital Injury Reporting and Prevention Pro gram (CHIRPp),l based on 16 out of the 750 hospitals in Canada; (2) the Australian National Injury Surveillance and Prevention Program;2 and (3) the programme commis sioned by UNICEF Innocenti Research Centre of Florence) (Table 17.1). The annual hospital health-care spend in the USA accrues to abo ut £5.1 million 4 for injuries incurred by children. In 1971, deaths from injuries in countries that belong to the Organisation for Economic Co-operation and Development (OECD) in the 1- to 14-year age group accounted for about 25 per cent of mortality in chi ldh ood (Table 17.2) . In the 1990s, more than one in three deaths in this age group were the result of injury. This has not changed in any appreciable manner since then. In such assessments of paediatric trauma, recorded sur veys vary in the cut-off point in relation to the ages con sidered. It is often the case in many of the published series that persons up to 18 years of age are considered. This may skew the figures, given the increased activity of adoles cents over the age of 14, a nd the high incidence of injury after the age of 12 years. It is unlikely that children under
--
-
--
----
Agricultural injuries Prevention Older children and substance abuse Accidental poisoning Hypersensitivity References
341 341 341 342 342 342
the age of 2 years are involved in many serious accidents, an d any injury below the age of 1 year is sufficiently uncommon that it should always be treated with a mod icum of suspicion .5 A careful match should be attempted in all instances between the findings on medical examination/ investigation and the narrative given by the carers to doc tors as to how the accident occurred. Furthermo re, the for mat and data conventions followed by different countries vary,6 as does the accuracy of data coding.7 In general terms, boys are injured abo ut twice as often as girls,S except in household injuties when girls tend to pre dominate. The difference between sexes increases after infancy and continues through a ll age groups into adoles cence. 9 The largest gro up of injured chi ldren is between the ages of 5 and 9 years, followed by children aged 1-4 years and children aged 10-14 years. Infants under 1 year have the lowest rate. JO The vast majotity of injuries occurred between noon and midnight when children are most likely to be out of the structured schoo l environment. The two most common sites of injuries were the road and home. II Other places included recreation areas, public places, schools 12 - 14 and, less fre quently, farms and workpl aces. Children from deprived backgrounds are more prone to injury. This may be related to a lack of adequate supervision, increased risk-taking and a greater proportion of the ch ild's day spent on the street. 15-18
~
Overview of paediatric trauma I
Table 17.1
337
Mortality (deaths per 700000) from injury: ch ildren aged 7- 74 year53
UK Canada USA New Zealand Australia Mexico Denmark Spain Italy Czech Republi c Portugal Korea
All injuries
Transport
Fire and flames
Falls
Poisoning
Drowning
Firearms
Intentional
Others
6.06 9.68 14.06 13.67 9.53 19.75 8.10 8.12 6.14 11.95 17.76 25.57
2.91 4.33 5.76 6.93 3.37 6.05 3.99 4.02 3.30 4.64 8.65 12.59
0.66 1.01 1.65 0.82 0.68 0.62 0.52 0.30 0.18 0.34 0.6 0.91
0.26 0.20 023 0.58 0.22 1.09 0.13 039 0.51 0.61 0.75 1.18
0.12 0.12 0.17 0.13 0.14 0.55 0.05 0.17 0.12 0.49 0.46 0. 75
0.39 1.26 1.74 1.87 1.97 3.30 1.07 1.12 0.46 2.23 1.34 5.14
0.0 1 0.12 0.4 0.14 0.04 0.35 0.07 0.05 0.03 0.03 0.06 0
0.80 1.45 2.74 1.74 0.85 2.90 0.94 0.33 0.50 1.61 4.23 1.08
0.91 1.19 1.37 1.46 1.26 4.89 1.33 1.74 1.04 2.0 l.67 3.92
From the WHO mortali ty database, with perm ission.
Table 17.2
Mortali ty fram injuries in Organisation for Economic Co-operation and Development (DECO) countries, by age and
gender 1991- 7995, per 700000 af the population
Age (years) 1-4
Boys Girls
21 14
5-9
16 7
10-14
16 7
From the WHO mortality database, with permission.
The use of the term 'accident' to describ e such traumatic episodes is often a misnomer, as few of th ese injuri es are actually due to unpredictable irremedi a bl e ev ents and situ ations; rather they are more often due to interaction with the environment in which children are living and playing, the products that they use and the level of supervisio n t hat they are allocated. Many of these injuries a re preventable and thus the main thrust of the investigation a nd ana lys is of figures from childhood injuries should be in attem pts at prevention. 19-21 There is no simple correlation between mOltality staris tics and non-fatal injury data. 22 FUithermore, the relarive population m ortali ty data are hig her in Eastern Europea n countries. 23 Why this should be so is difficult to determine but, at least in palt, it may be re lated to the different perception of the seriousness of injury a nd a populist fatalistic attitude; there may be a tendency to decry the seriousness of injury combined with Jess willingness to seek medical help for injury until it is too late.
Accident Proneness in Children Intrinsically the behaviour of children is based on curiosity and constant activity. Many paediatricians would class as
abno lmal an inactive, sedentary young child . This combina tion of childhood inquisitiveness, mobility and activity te nds to be impulsive and experimental. This is superimposed on a n obvious backdrop of inex perience and a n inherent imma turi ty of social and intellectual skills to cope wit h adverse situat ions as and w hen they present th emselves. By their very nature, children take risks and are sensation seeking. All of th is compounds their proneness to injury.24 Falls are therefore inevitable. The graspi ng and pulling of objects towards the infant may result in objects being pulled on to the child, and to scalds and burns.25 As the child's territory expands and manual dexterity and p hysical strength improve, play activity becomes more adventuro us and the opportunities for injuries increase. As organized sporting activities are taken up these further raise th e prob ability of injury, be this in organized and team sporting activity - participation in team sports such as rugby, soccer and hockey - or, more simply, in such recreational activiti es as kicking a ball, riding a bicycle or using a sWing. 26 The more v igoro us the sport (e.g. karate, judo, Tae Kwon Do, boxing and skateboarding) the more likely is the occurrence of inju ryY The role of carers in ens uring that the children do not take excessive risks, and indeed to be t here to rescue them from the hazardous situ at ions in w hi ch they find them selves, is essential. When the carers are employed in this capacity by nurseries and schools, there is an ex pectation of a duty of care. Sometimes - indeed , increasingly fre quently - this is called to account in a cou rt of law when this duty is perceived to have been breached. 28 Participa tion in sport also carries with it the possibility and, per haps, in more viol ent sporting activity, a probability, of almost inevitable injury. This willing, wholehearted partic ipation by the child with the sanction of his pa rents, often expressed legally in th e max im volenti non fit iniuria ('to one who is willing, no ha rm is done'), m ay decrease the culpa bility and answerability of the carer when things go wrong
338 I
Accidental injuries in children
but it also bears the expectation and responsibility from them that all that can be done to supervise children's activ ities - to decrease the risk, to provide protective clotting and equipment - has been fulfilled in full exercise of their duty of care. 29
Causes and Mechanisms of Injuries Falls are the leading cause of injuries. Among young chil dren these are often from fu rniture, such as cllanging tables, sofas and couches, cribs, beds or fallin g down stairs, pal1ic ularly in walkers,JO during momentary periods of inattention by their carers or because of the activity of older siblings. As toddlers begin to explore, the oppOl1Unity for falls from heights increases; this includes recorded falls down stairs. Less common are falls from balconies, porches or out of windows;Jl these inciden ts are more frequent in warmer climates when children have more open access . Nearly 20 per cent of the children adm itted after falls have fallen 2.5 metres (8 feet) - the height of one storey or more. The next major mechanism of injury is motor vehicle crashes with children as occupa nts. Most crashes involve collisi ons with other vehicles. Only one-quarter of those chil dren injured have been using some kind of protective restraints, such as child safety seats or safety belts. Many unrestrained children were hurled aro und inside the car dur ing a collision or thrown out of the car through windows or doors. Breast-feeding in the car or holding a child in one's lap is dangerous as the child will be unprotected during a crash. Many other children are injured fatally when they walk, cycle or play on the roadway and are struck by motor veh i cles.J2 Thousands of children are killed annua lly on the roads but most of the deaths are incurred when the child is on foot. There is correlation between the number of child hood fatalit ies on the roads and the size of the popUlation of a country : the bigger the child population and the lower the socioecon omic status, the higher is the likelihood that unsu pelvised children are on the road and thus find their ways into the path of moving traffic. This may also be the reason for the disproportionately higher incidence among children of certain ethnic groups.JJ Many of the deaths occur in the urban location at speeds of less than 30 miles per hour and usually within a Sh0l1 distance away from their homes. Most of these deaths occur in the holiday periods and a t other times during the parts of the day w hen the children are not actually at school. Thus more injuries occur in the summer months. The institution of school crossing patrols manned by retired individual s and the training of children in road safety has produced major dividends in terms of reducing co llision-related injuries and deaths amo ng chil dren; the aims have been to teach children to be vigilant and to improve their skills in quickly integrating and assessing movement, relative velocity, distance and depth in relation to moving motor vehicles, along with the inter pretation of road signs. 34 The setting aside of exclusive
cycle tracks in urban areas has also assisted in the decrease of such incidents. J5 ,36 Children are most frequently hit while walking along the side of the street, during play in the street or while darting into or crossing the street. Children are occasionally hit by motor vehicles while bicycling, but they may also be injured when fallin g off their bicycles after hitting objects such as trees, kerbs or walls. Few of those children admitted to trauma units after bicycling-related injuries, were wearing helmets. 37 Other motor vehicle-related injuries involve motorcycles, all-terrain vehicles and recreational vehicles. Children with hearing impairment from any cause, including chronic ot itis media a nd 'glue ear', are more prone to accidents as they may not be aware of vehicles in their vicinity.
Consequences Many injuries occur while a child is engaged in some kind of activity such as nll1ning, playing, climbing or bicycling. Even when the child is in a more station ary position, such as rid ing in a car, the child is part of an activity involving motion and speed. The combined effects of motion, speed and impact can injure many regions of a child's body. Two or more body regions were affected in about one-half of the children. Chil dren typically had multiple injUlies that resulted in several diagnoses: head inj ury is the most frequent diagnosis among children recorded, followe d by fractures to the bones of extremities and torso. The third most frequen t diagnosis is open wounds, mostly lacerations. The severity of injuries is directly linked to how the chi l dren were injured. Children who are injured in motor vehi cle crashes and as pedestrians are more likely to have four or more functional limitations. The force and speed of col lisi ons are factors determining the severity of injuries. Chil dren injured by falls are less seri ously injured and have fewer functional limitations.
BICYCLE HELMETS Experienced, careful bicycle riders crash every 4500 miles on the average. Head injury causes 75 per cent of our 800 plus an nual deaths from bicycle crashes. Medical research shows that a bicycle helmet can prevent 85 per cent of cyclists' head injuries. 38.39 A helmet reduces the peak energy of a sharp impact. This requires a layer of stiff foam to cushion the blow when crushed. Most bicycle helmets do this with the use of expanded polystyrene (EPS), the white picnic cooler fo am. Once crushed, EPS does not recover. Another foam, expanded polypropylene (EPP), does recover but is much less commonly utilized. A stronger EPS called GECET appeared in 1992 and is widely used now and another synthetic foam, expanded polyurethane (EPU), is used in Taiwan as it has a
Falls I
unifonn cell structure and crushes without rebound; it is heavier and requires a manufacturing process that is not environmentally friendly. The spongy foam placed on the inside a helmet is for comfort and fit, not for impact. The helmet must stay on your head even when the head is hit or bumps on to a hard unyielding surface, more than once in a moving traffic incident: usually there is a bump with a ca r first, and then with the road. It thus needs a strong strap and an equally strong fastener. The helm et should sit level on the head and cover as much as possible of its surface contour. Above all, with the strap fastened, one should not be able to get the helmet off your head by any combination of pulling or twisting. If it comes off or slips enough to leave large areas of the head unprotected, the straps must be readjusted otherwise a better-fitting hel met will be requ ired . The strap must feel comfortably snug when riding the cycle. Most helmets made of EPS foam ha ve a thin plastic outer s hell. The shell helps the helmet skid easily on rough pavement to avoid jerking of the neck during such an impact; it also holds the EPS together after the first impact. Some excellent helmets are made by moulding EPS in the shell rather than adding the shell later. Excessive vents in the helmet mean less actual area con tact with the head, which could concentrate force on one point. 'Aero' helmets are not noticeably faster and in a crash the 'tail' could snag or knock the helmet aside. Skinny straps are also less comfortable. Dark helmets are hard for motorists to see. Rigid visors can snag or shatter in a fall. Helmet standards do not address these problems a nd should be reconsidered. There can be little doubt that bicycle helmets have pre vent some of the more serious head injuries and undoubtedly they have saved the lives of young cyclists. 4o - 42 However, it has been observed that some children, particularly older children, need to be objectively convinced that helmet wear ing is indeed not just useful in preventing injUl)' and that the money spent on their purchase rather than other cycl in g clothes and accessories is money well spent. Above all cycle helmets had to be shown not to make their wearers less 'cool' and trendy but vice versa, and also that their use makes the wearers appear more professional. 43 A variety of strategies have been devised worldwide to encourage the purchase and wearing of cell helmets with varying success levels. 44- 47
FALLS All children trip and fall, indeed from the time that initially they stand up on their own two feet. Falls a re more com mon while the muscle tone is not fully developed but become more frequent once aga in when gait allows chil dren to explore farther and when their innate inquisitive ness takes them further afield. Contusions, haematomas, lacerations and abrasions tend to occur at the site of such fa ll s. OccaSionally, if more
339
severe force is in volved or in falls from a height, fractured long bones a nd closed head injuries may occur 48 The incidence of injury depends to a major extent on the height from which the fall occurs and also on the terrain that the chi ld lands on - the harder and more compact the latter, the higher is the incidence and severity of injury.49 The height of climbing frames, swings, etc. is another factor involved in injury causation and the height of the fa ll is directly related to the severity of the injuries sus tained by children in playgrounds. Injuries are more likely to occur with equipment that is two 2 metres above the ground ,5o so that reducing the height of swings, monkey bars and other equipment improves safety. Reports of children falling out of windows are all too frequent in the summer. Height off the ground and landing surface are two critical factors in the injuries to children that result from falls. Height is directly related to the speed at wh ich a child falls and the force of impact upon landing. Falls are particularly common among children of p re school age who are developing their balance and learning to walk, climb and run ; however, their physical skill s develop sooner than their abilities to anticipate the conse quences and risks of their actions. Amongst falls from heights, about one-quarter of chil dren fell from w indows. Examples of other falls from heights include falls from roofs, out of trees, from ski lifts, off porches or balconies and down flights of stairs. Even escalator falls have been known to cause serious injuries. 51 The incidence of injuries among children h as a distinct seasonal pattern, with an increase in the warmer months of the year. This is even more so for falls from windows as they are more likely to be open in warmer weather. Tod dlers and pre-school children are particularly vulnerable as they lack the judgement needed for risk assessment an d safety precautions. Falls occurred in th e chi ld's hom e in 96 per cent of the cases. Falls from two- to three-storey win dows were most common (72 per cent), followed by fo ur to five storeys (18 per cent). Only 8 per cent of children admitted to trauma centres fell from firs t-sto rey windows; however, 3 per cent fell from six storeys or more. Head injuries, with or without dam age to other body regions, occur among 65 per cent of the children . A small child 's head is at special risk for injury from a fall because it is disproportionately large and heavy in relation to the rest of the child 's body. Height, rate of fall and the force of impact have combined effects that result in injuries to multiple body regions in about one-half of injured children. The greater the height of the fall, the higher the death rate. Among the children who fell from windows, six storeys or more, 8 per cent died. Among those who fell four to five storeys, 5 per cent died; among those who fell two to three storeys, 2 per cent died. Non e died who fell from one storey. Upon hospital discharge, 73 per cent of the children who fell from windows had no identified functional limitations. For the 21 per cent who had one to' t hree limitations, these were most commonly in bathing, dressing, walking and
340 I
Accidental injuries in children
se lF-Feed ing. By contrast, the children with four or more functional limitations at discharge had more difficulties in cogn ition , behaviour speech, vision and hearing, reflecting the severity of head injuries in this group. Among t he chil dren who fell from six or more storeys, 42 per cent had one to three limitatio ns; 8 per cent had four or more limitations. More functional limitations were identified as the height of t he fall increase d; children who fell from four to five storeys had the largest proportion of functional limitations - four or more. Ho wever, because more children died who fell from fi ve storeys or more, there were fewer surv ivors in that group with four or more limitations. Almost one-half (47 per cent) of the falls occurred on playgro unds in recreational areas, one-quarter (25 per cent) occurred at home, and almost one-fifth (19 per cent) occurred on sc hool playgro unds. About 90 per cent of these injuries occurred in falls from playground eq uipment: sw ings, slides, climbers (monkey bars, jun gle gyms or other climbing apparatus) and trampolines. Children fell from dif ferent types of playground equipment depending where the equipment was located. At home, children most often fell from swings or trampolines, whereas at schoo l and in recre ation areas they most often fell from climbing equipment. Children of different ages fell from different types of playground equipment. The youngest children fell most often from slides, whereas older children fell most often from climbing equipment. Seven of the children died in the hospital: six from head injuries and one as a result of a severed sp in al cord. Three of these children fell from slides, two fe ll from sw in gs, o ne fell from monkey bars and one fell from a trampoli ne. Four of the children who died fell from playground equ ip men t at home; the other three fe ll fro m playground equ ip me n t in a recreational area .
PLAYGROUND INJURIES Playground s a re considered as oases from road traffic places w here children can safely enjoy themselves and indeed develop interpersonal, social and team skills. As childhood obesity is bandied around as being of epidem ic proportions, t he impo rtance of play activities cannot be overemphasized. Ho wever, playgrounds - whether at hom e, at schoo l or at rec reational areas - can also be dan gerous. To be safe, all public and private playgrounds must be well designed , we ll mai nta ined a nd used properly. Chil dren need good su pervision , and sho uld be taught how to play safely. Children enjoy experi men tatio n a nd explo ration and will incur risks, w hether calculated or hapha z ard, a nd thus although play areas have to be challenging and inn ovative to keep the ch ildren 's interest, there should also be a backdrop of safety built into playgrounds. Mod ern playgrounds wi ll often h ave eq uipment such as a seesaw, merry- go -round, swings, slide, climbing frame, walking bridge, jungle gym, c hin-up bars, sandbox, spring
rider, monkey bars, overhead ladder, trapeze/trapeze rings, pl ay houses and a maze. Th e scope of many of these items is to assist children to deve lop physical coordination, strength, and flexibility, as well as providing recreation and enjoyment. Common in modern playgrounds are 'play structures' that li nk many different pieces of eq uipment. A little more than one-half of the children injured are boys, with the largest percentage of injured children in the age bracket of between 5 and 9 years 0ld s2 Protective surfaces (suc h as rubber mats, wood mulch, bark or soft sand) of sufficient depth should be installed und er and aro und playground equipment; 53 there is no place for concrete, asphalted s u rface or sto ne slabs. 54 .55 This, togeth er with t he removal of monkey bars from parks, resu lts in a significant reduction of inju ry rate. 56 Equipment must be insta lled properly, with sufficient space a round it. Pl aygro und eq uipment needs to be inspected periodically for structural defects, peeling paint or splinterin g wood a nd these should be repa ired. Protective surfaces under playground equipment should also be inspected periodically, as surfaces such as wood chips or sand lose their protective quality if they get compacted by regul ar usage or are worn awayS7
SPORTS INJURIES ON SNOW AND ICE Areas with cold te mpera tures, snow and ice can make ideal conditions for recreation, but sports on snow and ice create specia l ha za rds for children. Over one-half of the children with t his type of injury are injured while sledging; the next largest groups have injuries caused by skiing, snowboard ing or playing hockey - fewest children were injured while ice-skating. Falling or hitting objects such as trees, rocks or fro ze n ground injured most children. Head injury was the most frequent diagnosis. One out of four children had broken or fract ured bones or had multiple injuries or required sur gery. The average length of hospital stay was 6 days. More than 14 days were spent in the hospital by on ly 10 pe r cent of the children . The combination of speed with a fall or coUision on ice and snow increases the risk of serious injuries . Over one half of the children had head inj uries and few, if any, of these children had been wearing helmets. Data from a study of sk i rental fac iliti es in 1999 looking at 353 rental shops in five European countries (France, Austria, Germany, Sw itzerl and and Ita ly) had shown that 71 per cent failed to meet the sta ndard and abo ut 13 per cent just about reached an acceptable level. In tota l, 70 per cent of rental facilities still set the bindings incorrectly 66 per cent of bindings in children were found to have been set dangerously hi gh. 58 - 60 At the biannual [SSS sy mposium held in Japan in 200 5, injury statistics presented from Scotland, Fra nce and Iran show a rate'expressed as injuries per 1000 part icip ant days of 1-2 per 1000 fo r alpine skiing
Older children and substance abuse I
and 3-4 per 1000 for snowboarding and ski-boarding. More common injuries are knee injuries in alpine skiers, wrist injuries in snowboarders and lower leg injuries in ski-boarders. Children as a group are twice as likely to be injured as adults, especially if using rental gear and on their first day of skiing as were those who had received less than five formal lessons and those skiing with friends. Children were shown to be more frequently injured on lift systems and as a result of bindings failing to release. In relation to the use of protective equipment by skiers the debate about helmet use continues. In Norway, the rates of use amongst those injured were 31 per cent (alpine ski), 26 per cent (snow board) 26 per cent (telemark ski) and 28 per cent (ski-board). In Sweden, where about 58 per cent of all people on the slopes now wear a helmet, a 28 per cent rela tive increase in the risk of a head injury was observed in non-helmet wearers. The main protective effect of headgear is against more minor injmies such as abrasions and bruises. The helmets in use at present simply cannot do much in high-speed impact accidents. Wrist guards were devised to reduce the incidence of wrist fractures amongst snowboard ers but there are currently no international standards for them, so it is velY difficult for a boarder to decide which to use and which is the best buy. Release bindings had led to a reduction in the number of lower leg fractures and may reduce the number of knee injuries that were still occurring. International standards have still to be agreed for them.
341
forbidding children below the age of 13 years to drive a tractor on private ground. Children are also prevented by law from riding as passengers on farming machinery. This applies not only tractors, but also to harvesters and trailers. Quad bikes are quite popular on farms and are often an irresistible attraction to children; crashes with these have resu Ited in injury and deaths.
PREVENTION A major public health issue 71 is the effective prevention of injuries and fatalities from injuries in children and, in most countries, there is a constant active monitoring of the incidence of serious injury and fatalities resulting from accidents; any patterns that emerge from such studies are addressed and this occasionally involves new Jegisla tion. 55 ,55.72 Preventative measures that are adopted are reg ularly audited in terms of their effectiveness and fine tuned as required.73 The availability of guidelines and legal sanction have enabled the commencement of, and the frequent successful completion of, litigation procedures in the civil and, more rarely, the chimerical courts, in addition to public inquests and fatal accident inquiries.
OLDER CHILDREN AND SUBSTANCE ABUSE RIDING INJURIES Each year children are injured and some die as a result of incidents while riding a horse. An adult horse can weigh over 500 kg, gallops at speeds up to 65 km per hour and kicks with the force of 1.8 times its weight. About 65-75 per cent 51 - 54 of children admitted after riding accidents have been riding the horse, with most injuries occurring to the head and in the limbs. Many of the fatalities and injuries can be avoided or have a reduction their severity through a combination of better adult supervision and the use of appropriate protective helmets. 55 .55
AGRICULTURAL INJURIES Children in farming communities often assist their families during peak harvest times and indeed may seek employ ment as fruit and vegetable pickers during vacation peri ods. In the course of holidays on farms, children may also find themselves in trouble with farming equipment. This problem is well recognized in Britain as it is in Australia 57 and in the USA;68 between 1979 and 1988, 167 children died in farming accidents in Britain and about 300 children die annually in the USA.69 Tractor-related injuries are the most common in Britain,l° even although there is legislation in existence
In the older age groups, alcohol use, solvent misuse 74 and drugs may creep into the picture; in addition young adoles cents may carry weapons as part of gang culture or for alleged personal protection from bullying and other preda tory behaviour. 75 In other countries firearms may be used in much the same way and because of easy access to them, injuries and deaths often result from them. More than one person per week dies from volatile sol vent abuse (VSA). The average home has 30 kitchen and bathroom products that can be abused, including: butane gas cigarette lighter refills, liquefied domestic gas, solvent based adhesives, deodorant aerosols, pain relief sprays, aerosol air fresheners, hairspray, other aerosols, correction fl uids, petrol, certain paints, paint thinners and removers, clry cleaning agents, petrol lighter fuel, nail varnish and varnish remover, some shoe and metal polish, and plaster remover. In 2001, VSA was responsible for seven times the number of fatalities than those related to ecstasy (3,4 meth ylenedioxymethamphetamine [MDMA]). The 2003 annual report from the European Union's drug agency warned that VSA is an often overlooked problem, with a big impact on public health of young people; the use of solvents and inhalants is the next common substance misuse after cannabis and alcohol in 15- to 16-year-olds. One in seven 15- and 16-year-olds in the UK sniffs solvents to get high. More than one person dies every week in the UK from the effects of solvent abuse - between 70 and 100 each year.
342 I
Accidental injuries in children
The dru gs charity SOLVE IT says so lven t abuse kills 60-75 youn g people each year. One- thi rd of young people that die from VSA are first-time so lvent abusers; in 1997 in 37 per cent of deaths there was no prior known history of abuse. Most deaths from VSA occurred in the 14-18 age ran ge . Solvent abusers can be male or female, al thou gh there are higher numbers of solvent-related de aths in boys, probab ly due to differences in sniffing behaviour. Sniffing so lvents may cause intox ication similar to the effects of alcoho l,76 so a so lvent sniffer may become drow sy, confu sed, an d aggressive, may take more risks than they would when sober, a nd so on. Accidents, there fore, are quite comm on and so metimes fatal. 77 Over one half of the deaths that have been linked to solvent sniffing appear to resu lt from the direct toxic effects of t he chemi cals that w ere sniffed , but other de aths result from acci dents, choking on vomit or suffocati on. Deat hs are often sudden, and often a mechanism of death invo lv ing cardiac arrest app ears to be the cause . Gas fuels continue to be associated with the majori ty of deaths, accountin g for about two-thirds of VSA deaths. Sniffing the butan e gas in li ghters causes a cardiac arrhythmia. 7s .79 Possession or abuse of a vo latile substance is not a c riminal act in itself, although shopkeepers can be fi ned under the Intoxicating Substa nces (Supply) Act 1985 if they knowingly sell to an abuser who is und er 18 year old. The investigations of such deaths require carefu l sampling; to enabl e later sampling of the air wi thi n the low er airways, it is essential to retain a lun g, with its main bronchus tied up securel y.8o.sl
ACCIDENTAL POISONING The innate inquisitiven ess of children and th eir tendency to sample and taste wha t th ey encoun ter in t heir immediate environm ent may result in accidental poisoning by: con sumption of mushrooms, benies, fruits and vegetation grow ing wild, which are intrinsically poiso no us; and the sampling of liquids held in bottles, including alcoholic bev erages, or other containers of liquids found in the house or in ga rages and outho uses. They are attrac ted by the colour, odour and labels of bottles and contain ers. Poisons range from med icines (such as a nalgesics, contraceptive pills, iron tab lets, vitamins an d antidepressants) to com mon fluids or items foun d in the ho use (cleaning fluids, a lco hol, cigarettes, c ray ons and button batteries) to dangerous garden plants (fox glove and lab urnum seeds). Samplin g by children of medication not intended fo r th em is a recurring prob lem, particularly if the tablets and capsules in questions are coloured a nd resem ble sweets. Poiso ning is the fourth most co mmon ca use of accidental deaths in children. Children under 5 years of age, as well as ado lescents, are pro ne to poisoning but accidental ingestions are most com mon in ch ildren under 5 years old . Abo ut 90 per cent of a ll poisonings occur at ho me, the most common sites being the ki tche n and the bathroom . Unintentional pOisonings
occur most frequ ently whe n routines are disrup ted, for exa mpl e moving and vacations. Chi ld safety caps have help ed decrease the number of p oisonings but they are not 100 per cent effective and should not give a fals e sense of security.82,83 It goes without sayin g that a ll potential poi sons should be prop erly lab ell ed, stored out of reach of children and locked away. In a household where drugs are misused it is po ss ible for a child , even a tod dler, to gain access to liquids, su ch as methadone, and co ntrolled substances of abuse, such as heroin or cocain e powder, which they inges t with the con sequence of serious harm as a direct effect of such inges tion. Passive smoking of cannabis fr om parents smoking close to the child in enclosed restricted environm ents has been known to result in inhal at ion of enough smoke to result in intoxication of the child .
HYPERSENSITIVITY Allergies with the potential for a fatal hypersensitivi ty respo nse have become more freq uent over the years S4 - at least in the public perception. According to the Natio nal Institute of Allergy, food allergy occurs in 6-8 per cent of children und er the age of 4 years, as well as 4 per cent of adults. About 30000 episodes of anaphylaxis and 100-200 deaths per year occur in the USA. At present, the only ways to manage food allergies are to avoid the food s that cause reactions and to treat the sympto ms caused by allergic reac tions. About 20 per cent of th e population worldwide can be considered to have som e form of atopy; this is more common in infants and children und er the age of 3 yea rs. B5 ,S6 Nut allergy is parti cularly important;S7 the inadvertent consump tio n of nuts by children who happen to be sensitive to them may cause deaths unless immediate action is taken. In the UK, 5 out of the 20 fatal a naphylactic re actions reco rded each year are due to food .s8 In addition to peanuts, the foll owi ng fo ods can result in life-threaten ing reac tions on re-exposure: tree nuts, seafood, seeds and cows' milk. 89 - 93 A number of these children always cany with th em an emergen cy pack that contains t he equipment required for an adrenalin injec tion, along with instructions as to how it shou ld be used if they were to develop an acute hypersensitivity respo nse;94 these are carried to school or other extra-domestic sites. 95 Children with bronchial asthma and wh o are kn ow n to respo nd adv ersely to external allergens, su ch as cats, may also develop potentially life-t hreateni ng episodes if exposed acutely to t he antigen .
REFERENCES Macke nzie SG, Pless lB. CH1RPP: Canada 's princ ipal injury surveillance program .il1jul)' Pre/! 1999; 5: 208- 13. 2 Stone DHG , Morrison A. Smith GS. Emergen cy Depart ment inju ry surveill ance system : the best use of resources? inju1Y Prel) 199 9; 5: J66 -7.
-
-
-
--
References I
3 4
5 6
7
8
9
10 II
12
JJ
14
15
16
17 18
19 20
21
22
23
24
25
ECOHOST. Child Injuries in Industrialised Countries. Report commissioned by UNICEF. London: ECOHOST, 2000. Malek M, Chang BH, GalJagher SS, Guyer B. The cost of medical care for injuri es to children. Ann Emerg Med 1991; 20:997-1005. Barber MA, Sibert JR. Diagnosing physical child abuse: the way forward. Postgrad Med ) 2000; 76:743-9. Moyer LA, Boyle CA, PoJJack DA, Validity of death certificates for injury-related causes of death. Am) Epidel11iol 1989 ; 130: 1024-3 2.
Maclntryre CR, Ackland MJ, Chandra raj EJ. Accuracy of injury coding in Victorian hospital morbidity data. Ausr NZ ) PubI Health 1997; 2:779-83. Walton W, Jarvi s SN, Lowe P Jet al. Annual incidence of unintentional injury among 54,000 children. Injury Prell 1996; 2:16- 20. Towner EM, Jarvis SN, Walsh SS, Aynsley-Green A.
Measuring exposure to injury risk in school children aged
11-14. Bi\lIJ 1994; 308:449 - 52.
Levene S. Coron ers' records of accidental deaths. Arch Dis Child 1991; 66:1239-41. Murdock R, Eva J. Home accidents to children under 15 years: a survey of91O cases. BM) 1974; 3:103-6. Lenaway DL, Ambler AG, Beaudoin DE. The epidemiology of school-related injuries: new perspectives. Am) Prep Med 1992 ; 8: 193-8. Latlamme L, Menckel E. School injuries in an occupational health perspective: w hat do we learn from community based epidemiological studies? Injul:V Preu 1997; 3:50-6. Hammarstrom A, Janlert U. Epidemiology of school injuries in the northern part of Sweden. scal1 ) Soc IvIed 1994; 22: 120-6. Alwash R, McCarthy M. Accidents in the home among children under 5: ethnic differences or social disadvantage? BM) 1988; 296: 1450-3. AvelY lG, Vaudin IN , Fletcher JL, Watson 1M. Geographical and social variation s in mortality due to childhood accidents in England and Wales 1975-84. Publ Health 1990 ; 104: 17I -82. Campbell H, O'Dri sco ll S. Th e ep id e miology of leisure accidents in Scotland. Health Bulletin 1995 ; 53:280-3. Hapgood R, Kendrick D, Marsh P. How well do soc iodemogTaphic characteristics explain variation in childhood safety practices? ) Publ Health IvIed 2000; 11 3 :285-9. David RM, Pless LB. BMJ bans 'accidents '. Accidents are not unpredictable. BM)2001; 332:1321-2. Guyer B, Gallagher S, Chang Bet al. Prevention of childhood injuries: evaluation of the Statewide Childhood Injury Prevention Program (SCIPP). Am) Publ Health 1989; 79: 1521-7. Gielen A, Wilson M, McDonald E et al. Randornised trial of enhanced anticipatory guidance for injury pTevention. Arch Pediarr Adolesc Med 200 1; 155:42-9. Botting B, Crawley R. Trends and patterns in childhood mortality and morbidity. In Botting B (ed,) Tile Health oj Our Children: a reuiew in the mid-/990s. Decennial supple ment OPCS DS No. 11,62-81. London: Office of Population Censuses and Surveys, 1995. King A, Wold B, Tudor Smith C, Harel Y. The Health oj Youth; a Cross-national sumey. WHO regional Publications, European Series No. 69, WHO, 1966. Bijur P, Golding J, Haslum M, Kurzon M. Behavioural predictors of injury in school-age children. Am ) Dis Child 1988 ; 142: 1307-12. Waller AI, Marshall sw. Childhood thermal injuries in New Zealand resulting in death and hospita lization. Burns 1993 ; 19:371-6.
343
26 Mott A, Evans R, Rolfe K et al. Patterns of injuries to children on public playgrounds. Arch Dis Child 1994: 71 :3 28 -30. 27 Kujala UM, Taimela S, Antti-Poika I et al. Acute injurie s in soccer, ice hockey, volleyball, basketball, judo and karate: analysis of national registry data. BM) 1995; 311: 1465-8. 28 Swan P. British Schools, Safety Incidents and The Courts. www.articlecity.com/artic les/lega l/ a rticl e_3 8 5.sh tml. 29 Lane K. General Defences Relevant to the Tort of Negligence. www.city.londonmet.ac.uk/-shlane/Generaldefences.htm. 30 Kendrick D, Marsh P. Babywalkers: prevalence of use and relationship with other safety practices. Inju/y Prev 1998; 4:295-8. 31 Spiegel C, Lindaman F. Children can 't tly: a program to prevent childhood morbidity and mortality from window falls. Am) Publ Hralrh 1977; 67:1143-7. 32 Christie N. Social, Economic and Environmental Factors in Child Pedrstrian Accidents: a Research Revirw. Project Report 116 Et 117. Crowthorn e, Berkshire: Transport Research Laboratory, 1995. 33 Lawson SD, Edwards PJ. The involvement of ethnic minoriti es in road accidents: data from three studies of young pedestrian casualties. Traffic Eng COl1trol 1991 ; 32: 12-19. 34 Boxall JA. School crossing patrols: how effecti ve are they? Traffic Eng Control J 938; 29:586. Report CR 192. Crowthorne, Berkshire: Transport and London: Departme nt of TranspOI1. 35 Christie N. Social Economi c and Environmen tal Factors in Child Pedestrian Accidents: A Research Review. Project Report no. 116. Craw thorne, Berkshire: Road Research Laboratory, 1995. 36 Jorgensen E. Bicycle tracks in urban areas in Denmark. Evaluation of the effect on safety. In Biecheler M, Lacombe C, Muhlrad N (eels) Evaluatiol1 85: International Meeting on the Evaluation of Local Traffic Safety Measures, Proceedings of the Paris Conference, 20-23 May 198 5, pp. 755-61. 37 Bicycle Helmet Safety Institute, 199 5. http: //www. bhsi.org/webdocs/hend erson. htm. 38 Cameron M, Heiman L, Neiger D. Evaluation of the Bicycle Helmet Wearing Law in Victoria During its First 12 Months. Report No. 32. Victoria, Australia: Accident Research Centre, Monash Unive rsity, 1992. 39 Vulcan A, Cameron M, Watson W. Mandatory bicycle helmet use: expe rience in Victoria, Australia. World) surg 1992; 16:389-97. 40 Finch C, Newstead S, Cameron M, Vulcan A. Head Injury Reductions in Victoria Two Years After Introduction of MandatolY Bicycle Helmet Use. Report No. 51. Victoria , Australia: Accident Research Centre, Monash University, 1993. 41 Cameron M, Vulcan A, Finch C, Newstead S. Mandatory bicycle helmet use following a decade of helmet promotion in Victoria, Australia - an evaluation. Accident Anal Prey 1994; 26:325- 37. 42 McDermott F. Bicyclist head injulY prevention by hel mets and man datory wearing legislation in Victoria, Australia. Ann R C surg Engl 1995; 77:38-44. 43 Puczynski M, Marshall D. Helmets! All the pros wear them. Am) Dis Child 1992; 146:1465-7. 44 Lee A, Mann N, Takliti R. A hospital - led promotion campaign aimed to increase bicycle helmet-wearing among children aged 11-15 living in West Berkshire 1992-98. Injury Prev 2000; 6:151-3. 45 Britt J, Silver I, Rivara F. Bicycle helm et promotion among low income preschool children. Injurv Prey 1998 ; 4:230-3. 46 Abu larrage J, De Luca A. Abularrage C. Effect of educat ion and legislat ion on bicycle helmet use in a multiracial population. Arch Pediatr Adolesc Med 1997 ; 151 :4 1- 4. 47 Dannenberg A, Gielen A, Beilenson Pet al. Bicycle helmet la ws and educat ional campaigns: an evaluation of strategies
344 I
48 49
50
51 52 53
54
55 56
57
58 59
60
61 62
63 64 65
66
67
68 69 70 71
72
Accidental injuries in children
to increase c hil dren's helmet use. Am} Publ Healt/r 1993; 83:667-74. MacKay M. P layground injuries. Injury Prev 2003; 9: 194 -6. Cha lm ers DJ , Ma rshall SW, Lan g ley JD et al. Heig ht a nd surfacing as risk factors for injury in fa ll s from playground equipment: a case co ntrol study. InjU/)' Prev 1996; 2:98 -1 04. Laforest S, Robaitaille, Lesage D et al. Surfa ce charac teristi cs, equipment height, and the occurrence and severity of playground injuries. Injury Preli 2001; 7:35-40. Platt SL, Fine JS, Foltin GL. Escalator-related injuries in children. Pediatrics 19 97: 100 :E 2. Lilli s KA, Jaffe DM. Playgrou nd injuries in children. Pediatr Emerg Care 1997; 13: 149-53. Laforest S, Robita ill e Y, Dorval D et al. Severity of fall injuri es on sand or grass in playg rounds. } Epidemiol Community Health 2000; 54:475- 7. Ro sevea re CA, Brown JM, Barclay Mcintosh JM, Chalm ers DJ. An intervent ion to red uce playground equipment hazard s. injury Prev 1999; 5: 124-8 . Sosin D, Su rface-specific fall injury rates on Utah school playgrou nd s. Am} Publ Health 1993; 83:733-5. Mott A. Rolfe K, Ja mes R et al. Safety of su rfa ces and eq uipment for c hildre n in playgro und s. Lancet 199 7: 349 : 1874- 6. Sub ve rt JR, Mott A, Rolfe K et a l. Preventing inj uries in public playg round s t hro ug h partnership between health services a nd loca l a uth ority: co mmuni ty interwoven study. BM} 1999; 318:159 5-8. Mayr J , Russel 0, Sp itzer P et al. Playground accidents. Acta Paediatr 1995; 84:573-6. Macar1hur C, Hu X, Wesson DE, Parkin Pc. Risk factors for severe injuries associated with falls from playground equipment. Accident Anal Prev 2000; 32:377-8 2. Langran MA. Summa ry of the research presented at the 16th In ternationa l Sy mposium of the ISSS held in at Mount Arai, Japan, Ap ril 2005. www.ski-injury.com. Barone GW, Rodgers BM. Pediatrics equestrian injuri es : a 14-yea r rev iew.} Trauma 1989; 29:245-7. Bond GR, Christoph RA, Rodgers BM. Pedi atric eq uestrian injuri es; assessing the impact of helmet use. Pediatrics 1995; 95:487-9. Holland AJA, Roy V, Goh V et al. Horse-related injuries in chil dren. Med} Aust 2001; 1756 ;609-11. Ghosh A, Di Scala C, Drew C et al. Horse-rel ated injuries in pediatric patients.} Pediatr Surg 2000 ; 35: 1766 - 9 . Aronso n H, Tough Sc. Horse-rel ated fatalities in the p rovince of Alberta 1975-1 980. Am} ForC'l1sic fVlt:'d Pat/lOl 1993; 14:28-30. Ch itn av is JP, Gibbins CLMH, Hirigoyen M et al. Accidents with horses. What ha s cha nged in 20 yea rs? Inj!lI)' 1996; 27: 103 - 5. Byard RW, Gilbert J, Lipset J , James R. Farm - and tractor related fatalities in South Australia. } PC'diarr Child HC'olrh 1998; 32:139-41. Cogb ill TH, Busch HM, Stiers GR. Fa rm accidents in ch ildren. Pediatrics 1986; 76:562-6. Rivara FP. Fatal and non-fata l farm injuri es in the United States. Pediatrics 1985; 76:5 67 - 73. Cameron D, Bishop C, Sibert JR. Farm accidents in children. EM} 1992; 305:23-5. Millward LM , Morgan A, Kelly MP. NH S Hea lth Development Agency. Health Developm ent Agency. Prevention and Reduction of Accidental Injury in Children and Older People. Evidence briefing, June 2003. Grayling T, Hallam K, Graham D et al. Streets Ahead: SaJe and Liveab/{:, Streets Jor' Children. London: In stitute for Publi c Policy Research, 2002.
-
73 Town er E, Dowswell T, Mackereth C, Jarvis S. What Works in Preventing Uninte ntional Injuri es in Chi ldren and Young Ado lescents 7 An Updated Systemat ic Review. London: Hea lth Develop ment Agency, 2001. 74 Sva nstro m L, Haglun d BJA. Evidence-based Safety Promotion and Injury Prevention - a n Introducti o n. Department of Publ ic Hea lth Sciences, Di vision of Soc ial Medici ne, Karolin ska lnsti tutet, Stockho lm , Sweden, 2000. 7 5 Ano nymous. Solvent abuse: Little progress a fter 20 years (Editori al). 8M} 1990: i:J00. 76 McKeganey N, Norrie J. Association between illegal drugs a nd weapon carry in g in young people in Scotland; schools' survey. EM} 2000: 320:982-4. 77 Joseph DE, Parke r S. inl7alants. Drugs oj Abuse. New York: United States Drug Enforcement Administration , 2005. 78 First Nations and Inuit Health Committee, Canadian Paediatric Soc iety (CPS). Inhalant abuse. Paediatr Child Hea lth 1998; 3:123- 6. 79 Shep herd RT. Mechanism of sudden death associated with volatile substance abuse. Hum Toxicol 1989; 8:28 7-91. 80 Anderson HR, MacNair RS , Ramsey JD. Deaths from abuse of substances : a national epidemiological study. EM} 19 85 ; 290:304-7. 81 www.tox la b.co. uk/vo la tile.htm and www.toxlab.co. uk/dasguide.htm. 82 S ibe rt JR , Clarke AJ, Mitc hell MP. Improvements in child res ista nt co ntai ne rs. A rch Dis Child 1985; 60:1155 - 7. 83 Rodge rs G. The sa fety effects of child-res istant pack aging fo r ora l prescripti on drugs : two decades of experie nce. }AMA 1986 ; 275: 166 1-5. 84 Isolauri E, Huurre A, Salminen S, Impivaa ra O. The all ergy epide mic ex tends beyond th e past fe w decades . Ciil1 Exp A llagy 2004; 34:1007-10. 85 Host A, Halken S. A prospective study of cow milk a llergy in Danish infants during the first 3 years of life. Clinical course in relation to clinical and immun o log ical type of hypersensitivity reaction. Allagy 1990; 45 :5 87 - 96. 86 Sampson HA, Scanlon SM. Natural hi story of food hypersensitivity in children with atop ic dermatiti s. } Pediarr 1989; 115:23-7. 87 Bock SA, Atkins FM. The natural histolY of peanut alJergy. } Allng.l' Ciill illll11ullol 1989; 83:900-4. 88 Pumphrey RS. Lessons for management of anaphy laxis from a study of fata l rea ctions. Clin Exp AI/{:'rgy 2000; 8: 1144-5 0. 89 Pa llares D. Allergy to macadamia nut. A 1111 Allergy Asthma 1111111ullol 2000; 85:385-6. 90 YUll gin ge r .liN, Sweeney KG, Sturn er WQ e t a l. Fata l food induced a naphy laxis. }AMA 1988; 260: 1450-2. 9 1 Sa mpso n HA, Mend e lso n L, Rose n JP. Fatal a nd near-fatal anaphy lac tic reaction s to food in chil dren a nd ado lescents. N ElIgl} MC'd 1992; 32 7:330- 4. 92 Bock SA, Mun oz - Fu rl ong A, Sampson HA. Fata liti es due to anaphylactic react io ns to foods. J Allergy Clin Immunol 200 1 ; 1: 191-3. 93 Yunginger JW, Nelson DR , Squi llace DL et aJ. Laboratory investigation of deaths due to anaphylaxis. ) Forensic Sci 1991; 36:857-65. 94 Simons FE, Chan ES, Gu X, Simons KJ. Epinephrine for the out-of-ho sp ital (first aid] treatment of anaphylaxis in infants: is the ampOUl e, syringe, need le met hod practical? ) Allergy Clin Immunol 200 1 ; 108: 1040-4. 95 Simons FE, Gu X, Silv er NA, Simo ns KJ. EpiPen Jr versus EpiPen in young children weighing 15 to 30 kg at risk for anaphylaxis. ) Allergy Clin Immunol 2002; 109: 171-5.
--=- -
I
CHAPTER 18
I
DROWNING AND NEAR DROWNING John Pearn
345 345
The pathophysiology of drowning
349
e causes of childhood drowning: a perspective
Forensic immersion syndromes
e drowning medium
346
References
351 359
troduction
TRODUCTION
J:o wning remains one of the most common causes of death early chil dhood . In developed and affluent countries, after ?: child has passed th e neonatal period, immersion incidents ~main one of the three principal causes of death.l,2 In many :
child abuse or non- accidenta l injury, homiCide, euthanasia a nd culpable neglect. The demonstration a nd interpretation of secondalY trauma due to attempted cardiopulmonary resuscitation after immersion also depends upon meticu lous autopsy technique. Of all health professionals, pathologists develop the most pragmatic perspective of the serio us threats to children's health, welfare and safety,7 Both the paediatric pathologist and the forensic pathologist have inescapa ble ethical roles in advocacy for the reduction of immersion accidents, a signif icant proportion of which a re preventable.
THE CAUSES OF CHILDHOOD DROWNING: A PERSPECTIVE Childhood drowning ra n ks high as a cause of child death, especially in the pre-school years (Table 18.1). The sites of the drowning incid ent include the family bath tub, private and public toilets, buckets and pails, private, family-owned swimming pools, public swi mmin g pools, fish ponds and ornamental pools, building site tre nches and drains, agri cultural dams and tanks, resi dential and neighbourhood canals, lakes, creeks and riv e rs, boating marinas and the open sears) (Table 18.2). Children who drown in each of these sites have their own s ite-specific precedents a nd forensic concomitants B ,9 The causes, sites, survi val rates and modus operandi of immersion deaths all differ when child victims are compared
346 I
Drowning and near drowning
Table 18.1
Relative rank order (per cent) of traumatic and injury deaths in childhood (1-14 years)
Cause of death
Drowning (accidental) Motor vehicle occupant death s Pedestrian run downs House/ca rava n fires Homicide, child abuse Fa lis, playg round acciden ts, accidenta I strangulati on Other, e.g. poisoning, electrocution Total
Percentage
32 29 17
8 6 6
2 700
Data typical of tropica l and temperate developed nations. Comp il ed from the Austra lian Insti tute of Health and We lfare 6 and other sources.
Table 18.2
Relative rank order (per cent) ofsites of drowning of children aged 0-5 years
Site Private swimm ing pools Family bath tubs Creeks Dams, building trenches, sewers Waterholes, fish ponds
Percentage
64 16
11 5 4
Data typical of tropical and temperate developed nations. Compiled from the Brisbane Drowning Stud y l 7.43 and other sources.
with adu lt subjects. The common patterns of adult immersion incidents, involving alcohol, suicide lO and boating accidents, are rarely encountered in the case of childhood immersion victims. The common situation of difficulty with cadaver id entification often encountered in long-immersed adult vic tims is very rare indeed; the identity of childhood victims is usua lly known and not open to question at the time of extraction from the water. Immersion times of childhood vic tims are measured in minutes rather than hours in more than 99 per cent of repo rted series. 11 The challenge of child immer sion deaths to the forens ic pathologist is that every case is individual and specific. Yet, however rare might be the cir cumstances sunounding an individual traumatic child death, diagnosis as to the cause of death and elucidation and recon struction of its precedents are of the greatest impoliance. In approaching a case of immersion death or near-death in childhood, the pathologist needs to adopt two appro aches. The first is that of the usual detailed history-taking, followed by meticulous autopsy examination of the body with, in some instances, particular attention to post-mOliem radi ographic survey, blood analyses for drugs, chemical analyses of lung tissue, DNA extraction and storage, post-mo rtem photography and, rarely, bone marrow extraction for diatom examination (Table 18.3). The diagnosis of drowning is one of
the most difficult in forensic pathology 12 In the case of child victims this difficulty is compounded when there may be no pathognomonic signs of drowning and where there may be no objective post-mOliem signs of intentional injury, even when such is undoubtedly the case. Alveolar oedema, alveo lar haemorrhages and emphysema aquosum are often the only histological featu res of drowning; even these may not be demonstrable in the case of an infant or child drowning when the duration of immersion is measured in minutes only. The second tool, indispensable to the pathologist, is ref erence to the defined syndromic profiles of child drown ings. Much recent research has built up patterns of quite distinct child immersion syndromes. lJ Such form indispen sable templates not only for the pathologist, but also for the investigating police and presiding coroner (Table 18.4). Childhood immersion fatalities do not form a spectrum of immersion in cidents; rather they constitute subsets of quite distinct forensic syndromes that are unrelated to each other except by the end-point of drowning. Overall, 95 per cent of such cases are accidental (unintentional). In homicide cases, however, drowning is the method of killing in some 10-20 per cent (see Table 18.4). Reference to such immer sion syndromes in the pre-autopsy analysis of any individ ual case is of great help to the pathologist in the choice of ancillary investigations (Table 18.3) and may be of the greatest help to family doctors, coroners and lawyers in subsequent proceedings. Many childhood immersion fata lities are depressingly repetitive as many of these deaths are preventable. Never theless, all of those persons who are involved in the pre mortuary chain - parents, siblings, neighbours, rescuers, bystanders, paramedics and police officers - are inevitably enmeshed in an intense personal and tragic incident. What happens to such individual s often depends on the determi nation of diagnostic truth, and such in turn depends on the patholo gist's skill in recreating the events surrounding such immersion deaths. Because of the well-recognized, and indeed historic, difficulty in confirming death by drown ing, there exists no greater challenge to the attending pathologist.
THE DROWNING MEDIUM In at least 90 per cent of drowning fatalities, death is due to the inhalation of water into the alveoli, with hypoxaemia and subsequent brain hypoxia. The water in which children drown contains a case-specific mixture of inorganic sub stances, gases, dissolved organic matter and particulate mat ter. Particulate matter may include bacteria, fung i, diatoms, zooplankton and phytoplankton. Unlike the situation that is more frequently encountered in the forensic analysis of adult immersion victims, such water composition does not usually have clinical, forensic or legal overtones. Very rarely children drown in vessels t1lled with paint or fertilizer, or in agricultural, industrial or domestic chemicals.
The drowning medium I
Table 18.3
347
Forensic investigotions apposite to immersion victims
Radiology
Total body X-rays essential after immersion death Skeletal survey usually normal in private swimming pool drownings. Exclude cervical spine fracture-dislocations in deaths in public swimming pools, in sea or surf, in diving or suspected diving activity In bath tub drownings, skeletal survey excludes old fractures In unidentified/decomposed cadavers, dental pantograms are useful
Histology: of lung, brain and heart is essential, sampling of other organs advisable Lung: vascular congestion, oedema, alveolar macrophages, alveolar haemorrhage, emphysema aquosum;73 foreign material, including stomach contents in larger airways or bronchioles
Neck muscles: intramuscular haemorrhages 74
Brain: vascular congestion, microhaemorrhages; immunohistochemical changes in hypog lossal nuclei 22 Electran microscopy: not required routinel y Blood alcohol: measure in all immersion victims over the age of 10 years; 60 per cent of adult male drownings have raised blood alcohol levels Antiepileptic drug levels when prescribed Screen for narcotics, psychotropic drugs, amphetamines and other drugs of abuse in all teenage immersion deaths Urine screen: gas chromatography-mass spectrometry (GC-MS); detects prescribed medications, drugs of abuse, 'over-the-counter' drugs, dietary supplements 75 Serum assay: quantitative estimation of anticonvulsants, other prescribed medication and drugs of abuse using high-performance liquid chromatography assay [HPLCF 5 Blood or tissue sampling for DNA identification not routinely required in most paediatric drownings; exceptions
Histology
Alcohol /drug assays
DNA extraction
are suspected neonaticide, bath tub drownings, suspec ted child homicide and when maceration makes visual identification impossible Measurement of electrolytes rarel y helpful
Electrolytes
Serum K may be increased due to haemolysis Serum Na may be elevated due to inhaled or swallowed sea wa ter, or pre- or post-mortem dehydration Vitreous electrolytes are stable for some hours after death and may not reflect agonal serum changes Other biochemical studies
Diatoms
Glucose: not routine; hypoglycaemia may be related to alcoholic intoxication, exhaustion by physical activity or prolonged pre-death immersion in cold water/ 6.n vitreous glucose is more stable Glucose in serum and vitreous should be measured in known diabetics Urinary myoglobin: not routine, normal level < 50 ng/mL; can be very high in drowning but non-specific; elevated levels occur in fatal burns, heat stroke, asphyxia and drowning; may indicate exertional
mu sc le activity or post-mortem change 78
Rarely needed; many different diatom genera and species have siliceous skeleton; femoral bone marrow sampled; tissue acid-digested or macerated, interpretation difficult, diatoms absorbed from the gastrointestinal tract pre-mortem; water from the putative drown ing site essential for interpretation; concordance of diatom genera and species is only 65 per cent 12 Diatom analysis is helpful when: cadaver is badly decomposed; body is found on land but suspicion of prior drowni ng; bod y may have been put into the wa ter after death 12.79
Relative rank order (per cent) of the modus of child homicide in the USA in children under the age of 5 years
Table 18.4
Modus
Percentage
Head injury Drowning
45
Non-drowning asphyxia
13 12 7
Body trauma Other, e.g. neglect, poisoning From ref. 9, w ith permission.
23
Fresh Water Fresh water contains variable amounts of organic material, dissolved salts and free and nascent gases. Most young children who drown do so in chlorinated, freshwater swim ming po ols; or in bath tubs with variable concentrations of dissolved soap or bath salts. 14 The chemical affects of chlorine and soap in fresh water are not of consequence in the pathophysiology of the great majority of fatal immers ions, and are certainly not of any practical significance in the pathophysiology of lung
Drowning and near drowning
348 I
syndromes in more than 95 per cent of survivors. Experi mental animal studies have not demonstrated significant effects on the surface tension of lung surfactant when chlorinated water inhalation is compared with that of pure fresh water. However, any type of fresh water inhalation elevates the minima l surface tension of tracheal and lung surfaces very significantly when compared with the affects that follow sa lt water aspiration.
Sea Water Sea water contains a mixture of inorganic sa lts. Concentra tions vary widely but the relative prop0l1ions of the differ ent types of dissolved salts remain surprisingly constant. The total concentration of sea-water salts may change over rela tively small distances, a phenomenon seen particularly in estuaries and even during tidal flow. II Sa linity is defined as the total mass of disso lved solids (in grams) in one kilogram of water. Such is expressed either as parts per thousand (ppt) or in grams per kilogram (g/kg). In instances where children drown in in land seas, canals, rivers or delta regions (pa11ic ularJy during floods), the salinity of the inha led water may be very low, approach ing that of fresh water. In organic sea water, a typical salt ion profile is: • sodi um , 10.5 g/ kg; • magnesium, 1.3 g/kg; • calcium, 0.4 g/kg; • chloride, 18.9 g/kg; • sulphate, 2.6 g/kg. Oceanic sea water thus typically contains 34.5 g/kg of dissolved salts, of wh ich 29.5 g/kg is sodium chloride; this is equ iva lent to 2.9 per cent sodium ch loride compared with 0.87 per cent sodium chloride that is the concentra tion in human plasma. Irrespective of the salt content or osmolality of the inhaled water, during the drownin g ep isode there is a one-way, mas sive transuda tion of water across the lung membranes into the pu lmonary vasculature. Childhood drowning deaths are exclusively due to cerebral hypoxia fo llowed by acute brain death. Any acute changes in osmo la lity, consequent upon salt concentration differences of the inhaled water, are not of forensic relevance. In survivors, because of rapid compensa tory homeostatic changes - Starling's Law, the Bainbridge Reflex and renal clearance - such diffe rences are not of prac tical relevance in the clinical management of near-drowned victims in the emergency room or intensive care ward. This is not to say that post-mortem electrolytes and osmolality should not be measured, as the issue continues to be a focus of legal debate in subsequent cou11 hearings. How much fluid is needed to drown a child remains unknown. Depth is no guide, as some infants drown in less than 20 cm of water in ornamenta l pools, garden fish ponds, pails or the family bath tub. Drug-intoxicated teenagers, patients with epilepsy and children with spinal or head
---
-
injUlies (sometimes sustained dUling skylarking in the sur rounds of swimming pools) can drown in very shallow water. In experimental mammals, as little as 1.0 mL/kg of fresh water instilled into the trachea causes gross pathophysiolog ical responses in the lungs. Experimental studies, usin g dogs, have shown that drownin g mammals may aspirate more than 22 mi/kg of water. The imp lication is that in the case of a typ ical chi ldhood drowning (with a median age of approxi mately 2 years in most current series) a \ 3-kg child may asp i rate as much as 300 mL of water. Such ac ute increase in blood volume is not greater than the capacity of the nonnal healthy child's heart to compensate for this potential flu id overload.
'Dry' Drowning There exists an historic debate abo ut the proportion of cases of immersion accidents that are due to 'dry' drownings, in which laryngea l spasm is tho ug ht to be the primary cause of fata l hypoxia. Alt hough laryngeal spasm occurs to some degree in every case of fluid inhalati on, water enters the peripheral airways in at least 90 per cent of cases. In child hood, autopsy findin gs of 'dry drownings' are rare. Such evidence is supported also by clinical experi ence in the man agement of near-drowned survivors, where in every case there is evidence of water inhalation. I n adult series, the fre quency of 'd ty drowning' has been va ri ously estimated to occur in 10 per cent to 15 per cent of cases. However, a recent review of the original studies from which such con clusions have been historically drawn has indicated that such may be without firm foundation 15 and that card iac arrest, sustain ed coincidentally while the victim is in the water may be the true cause of death rather than asphyxia. Despite this, there is persuasive experimental evidence to SUpp0l1 the phenomenon of 'dty drowning'. In experimenta l drowning with diatom monitoring, using deliberate immer sion deaths of aquatic mammals such as mink, muskrat and beaver, it has been shown that not all fully submerged ani mals inhale significant amounts of water into their lungs. 16 When small amou n ts of water enter the larynx or trachea, immediate laryngeal spasm results due to an efferent vagal reflex. An immed iate outpouring of thick mucus occurs, probably while the drowning victim is still conscious. Froth develops and in some cases a physical mucous plug forms. When such spasm relaxes preterminally, a significant amount of water is thus prevented from enterin g the trachea and lungs by the foam-froth plug, which acts as a physical barrier. In such cases, loss of consciousness is caused by anoxia or carbon dioxide narcosis. Death fo llows as a result of cerebral anoxia.
Water Temperature Almost all child ren who drown do so in water in a temper ature range of 10-25°(, Water temperatures ab ove 20'C do
----
~
Pathophysiology of drowning I
not influence the brain-protectin g diving reflex, but low temperatures augment it. The hypothermic, brain-sparing effects of near-drowning in very cold water are mediated through physical chilling of the body core and through an augmented diving reflex. In those situ ations when children drown in cold water, phys ical chilling of the body core is the result of not only the conduction and convection effects of cold water on the child's skin and through the involuntary inhalation of cold water, but also from swal lowing relatively large amounts of cold water. All children extracted from the water in immersion incidents have a reduced body temperature, including those who respond to cardiopulmonary resuscitat ion and who may survive the incident. In boating accidents, in accidents involving falls tho ugh ice and in cold water accidents involving older children who can swim, immersion hypothermia may pre cede primary drowning asphyxia. Clouding of conscious ness occurs in a drowning child w hen the core body temperature falls below 36°C; and consciousness is lost ,-,,,hen core body temperature falls below 34°C.
Immersion Asphyxia Occurring Other than in Water Rarely, children drown in non-water media such as paint, liq uid fertilizer or industrial chemica ls. Some children die after falling into wheat or grain silos, or are smothered in falls of earth or sand while playing in 'cubby houses', excavated ·Wendy' houses, or following the collapse of tunnels exca vated into the sides of cliffs at the beach or beside rivers or creeks. In such cases the asphyxiating medium behaves as if it were a fluid. Inhaled paint or sand within the airways, for example, may be a dramatic macroscopic feature at autopsy; the foreign material may be demonstrated histologically (Table 18.3). However, in a ll such cases the mechanisms of anoxia are secondary to foreign substance inhal ation. The mechanisms of oxygen exclusion (first with brain death and then with rapid somatic death) are identical to those encoun tered in cases of fatal water immersions.
THE PATHOPHYSIOLOGY OF DROWNING submerged infant or child holds his or her breath volun tarily until the 'break point' is reached. The 'break point' is determined by both hypercarbic and hypoxic drives. High arterial carbon dioxide levels combined with falling oxy gen concentrations are synergistic. For example, the 'break po int' occurs at PaC0 2 levels of below 55 mmHg if there is asso ciated hypo x ia; however, they may occur at Pa0 2 lev els of below 100 mmHg if the PaC0 2 exceeds 45 mmHg. In the case of older children, particularly those indulging in competitive or peer-demonstrative exhibitionism in swi m ming pools, the practice of pre-immersion hyperventilation is dangerous. Such underwater endurance games or dives .J,.
349
may alter the physiological setting of the breath-h olding 'break point'. If the hyp ercarbic drive to breathe is lost, unconsciousness from cerebral hypoxia may occur before the hypoxia-induced 'break-point' is reached. This is the pathophysiological mechanism leading to drownin g in teenagers or adults who, during snorkelling or free diving, attempt maximum endurance dives. 11 In all cases when the ' break point' is reached and the first involuntary inspiration is made, arterial hypoxaemia has already developed. Tissue hypoxia and acidosis have commenced. Involuntary gasp ing then occurs in a sub merged subject, who may stiU be conscious. Glottal spasm occurs. Even before consciousness is lost, vomiting with aspiration of gastric contents may occur. Consci ousness is lost within 3 minutes of invol untary submersion; and the electroencephalogram (EEG) becomes flat within 4.5 min utes. Breathing movements w ith fluid aspiration , and often the aspiration of sand , mud, g ra vel or vegetable matter, continues after the drowning v ictim has lost consciousness. Respiratory arrest follows. Dysrhythmias may follow and hypoperfusion with hypoxic blood leads to brain death. During the march of intra-immersion hypoxic events, intense autonomic, catecholamine-meditated blood redis tribution occurs. The spleens of drowned victims are rela tively bloodless, as one manifestation of the reflex constriction of sp lanchnic (including splen ic) vessels. The duration of brain hypoxia necessary to cause the death of an otherwise fit and healthy infant or child obvi ously remains unknown. Extensive research from the Bris bane Drowning Study,Il.l7 using a 'bracket method' and involving an ana lysis of hundreds of immersion fatalities and survivors, has indicated that (I) children who are immersed for 3 minutes or less are likely to survive an immersion incid ents and (2) children are unlikely to respond to card iopulmonary resuscitation if the immersion time is lon ger than 10 minutes in the most common water temperatures in which they drown (10-20°C).
The Diving Reflex The brain-sparing diving reflex may be demonstrated in virtually all children, dramatically so in the case of neonates and yo un g infants. The afferent arm of this reflex is the skin surface of the face supplied by the trigeminal nerve. The efferent arc involves t he vagus nerve and the autonomic nerve supply to blood vessels of the skin and all internal body organs except those of the brain and heart. In the case of a submerged child, the diving reflex is manifest by rapid-onset bradycardia and the shunting of blood from cutaneous and splanchnic vascul ar beds to the cerebral and coronary circulations. 18 Blood pressure starts to rise imme diately. The diving reflex is augmen ted by catecholamine release and is probably inhibited by obtundant drugs such as alco hol. It has been estimated t ha t the di ving reflex may provide a drowning infant or young child with an extended
350 I
Drowning and near drowning
'window of salvage' of perhaps 30 seconds of immersion hypoxia.
Lung Pathology As soon as water enters the lungs, a chain of pathophysio logical events occurs, as follows.
that are approximately iso-osmolar (e.g. 0.87 per cent sodium chloride) with plasma, do not denature pulmonary surfactant but may dilute it or wash it out of the alveolar sacs. Fresh water, or inhaled fluid that is significantly hypo-osmolar, causes acute degradation of surfactant activity. Whether the lining surfactant is lost or denatured, the end result is the same and alveoli collapse. FROTH FORMATION
PERIPHERAL AIRWAY RESISTANCE INCREASES Aspiration of even small amounts of water (1.0 mL/kg) is followed by pulmonary vasoconstriction, with immediate development of pulmonary hypertension due to parasympa thetic reflexes. The composition of the inhaled water is important in this context. Vasoconstriction occurs, particu larly following aspiration of fresh water into the mammalian lung; it occurs to a lesser extent in the case of sea water aspi ration, but does not occur with aspiration of amniotic fluid. LARYNGEAL SPASM Laryngeal spasm follows, with an immediate outpouring of thick mucus, followed by froth development. The conse quent degree of airway obstruction probably varies from subject to subject; the larynx relaxes preterminally.
Exudate outpouring from the laryngeal and tracheal mem branes, surfactant washout from the alveolar sacs and dis rupted alveolar membranes result in the formation of froth. The production of froth is characteristic, in variable degree, of a high proportion of the lungs and upper airways of drowned individuals. In some cases, it may be the only forensic mani festation that drowning has occurred. In cases of neonaticide, froth in the air passages is a valuable sign; when interpreted together with changes in the lungs it is indicative that a newly delivered fehls was born alive. The forensic evidence of post-immersion froth in the upper air passages is preserved if freezing has occulTed, and can be demonstrated after warming even following freezing temperatures as low as -18°c. 21 It is important to appreciate that any prolonged freezing of the body of a drowned subject will cause tissue distortion owing to ice crystal formation. This compromises the interpretation of subsequent histological examination.
REFLEX PULMONARY VASCULAR VASOCONSTRICTION This phenomenon leads to immediate pulmonary hyperten sion. Intrapulmonary reflexes then cause shunting of blood through non-ventilated areas of the lungs. Such shunts, com bined with surfactant loss or inactivation and consequent alveolar collapse, reduce mechanical compliance. Normally, intrapulmonary shunting involves no more than 18 per cent of the pulmonary vasculature; but within minutes of fresh water inhalation, some 75 per cent of the lung volume of such drowning victims manifests intrapulmonary shunting. FLUID SHIFTS ACROSS ALVEOLAR MEMBRANES A flux of inhaled water, ilTespective of its osmolality, occurs across the alveolar epithelium, through the basement mem brane and, finally, across the endothelial lining into the cap illary lumen, where haemodilution occurs. This flux causes rapid and severe distortion of pulmonary ultrastructure with damage to both type I and type II pneumonocytes. 19 ,20 Elec tron microscopic studies reveal endothelial changes of cell swelling, microvesical formation, cell detachment from the base of the membrane and cell destruction. SURFACTAI\lT CHAI\lGES Surfactant, primarily produced by type II pneumonocytes, is changed within minutes of water inhalation. Sea water and water containing sodium and chloride concentrations
Hypoxic Cascade In all immersion accidents the cause of death is ultimately irreversible cerebral anoxia. It is the end result of a hypoxic march of events, 11 any step of which may be influenced by the circumstances sUlTounding the immersion incident. This chain of cardiorespiratory events follows an inexorable sequence: • voluntary apnoea; • the diving reflex ensues, pa11icularly in children; • arterial hypoxaemia occurs; • tachycardia and hypertension develop; • tissue hypoxia ensues; • tissue acidosis develops; • inhalation with aspiration of water, followed by glottic spasm leading to a phase of secondary apnoea; consciousness is lost somewhere at this point of the pathophysiological sequence; • involuntary respiratory movements occur, continuing under water until respiratory arrest occurs; the diaphragm may continue to contract after intercostal activity has ceased; • hypotension occurs, with progressive loss of cerebral and coronary perfusion; • dysrhythmias may develop, and in the case of children who drown the preterminal bradycardia mayor may not be followed by ventricular fibrillation, before the inevitable asystole;
- - - - - - - - - ---~----
Forensic immersion syndromes I
• cardiac arrest; • brain death; • somatic death. Brain death after immersion accidents follows a primary state of altered neuronal metabolism. Research studi es of changes in the hypoglossal nucleus (in the brainstem) indi cate that compared with other quicker forms of asphyxia (e.g. hanging, strangulation and choking) pre-mortem damage to proteins such as the c-fos gene protein and the 72-kDa heatshock protein occur in this phase. 22 With increased percentages of the population trained in cardiopu lmonalY resuscitation (CPR) skills and the evolu tion of better ambu lance services and more sophisticated intensive care units, recent years have seen a number of childhood immersion victims maintained on life support prior to their delayed death due to drowning. The forensic pathologist is involved in such cases and coronial autop sies are required in almost all jurisdictions. Exa mination of the brains of these children may demonstrate any or all of the signs of asphyxial brain death. Although cerebral oedema may develop in survivors, it is not a feature of autopsy studies in the drowned victim who dies within minutes of immersion. Besides petechial haemorrhages on the brain surface, there may be features of cerebral oedema, sometimes to the point of hemispheric swelling with resultant mass effect. Such massive oedema has not been reported in 'CPR survivors' who died within 24 hours after rescue. Delayed-death subjects, in whom there exists the legal imperative of coronial autopsy, may also show a wide constellation of lung changes. Such autopsy features range from relatively normal lungs, perihilar pulmonary oedema, generalized pulmonalY oedema, pneumonia, col lapse and consolidation, disrup tion of alveoli, alveolar haemorrhage, foreign inhaled particles in the airways, and signs of unilateral and bilateral pneumothorax. At post mortem , such 'delayed drowning' lun gs may manifest both macroscopic and microscopic features of the adult respira tory distress syndrome (ARDS). 23 In such cases there are often other features of pneumonia, due either to nosoco mial infection or infection with waterborne bacteria or colonization by single-celled organisms. In the case of children who have survived with intensive care or ve ntila tor support for hours or days following extraction from the water, there may be uncommon (indeed, exotic) micro organisms found in pneumonia consolidates or lung abscesses. In the case of salt-water near-drownings, when death has occurred later in hospital, marine Vibrio bacteria or algae may be demonstrable.
FORENSIC IMMERSION SYNDROMES From the perspective of the fore nsic pathologist, the site and circumstances of the immersion incident are of crucial importance in the reconstruction of events as these details
351
will fOlm the substance of the final medicolegal report. Whereas the clinician is concerned reactively in combating the clinical consequences of hypoxia, the pathologist is con cerned with the recreation of the events that led to the pri mary immersion anoxic episode. In more than 95 per cent of childhood immersion incidents, unlike other forms of child hood trauma, the immersion incident is not witnessed. This fact, combined with historic difficulties of confirming death by drowning and the relative absence of specific forens ic signs, places the pathologist in a position of great responsi bility. For these reasons, there is an imperative to approach the pathological investigation against the background of all possible syndromic templates. There are at least 16 such age-site paediatric drowning syndromes, each with specific types of antecedent, different risks for the potential of unlawful child killing, different approaches to investigation and, ultimately, different stratagems for prevention. The importance of such syndromic definition is cognate to the generation of a clinical and forensic differential diagnosis when each new case is being assessed. Review of the child's entire past medical history is essential as a prelude to inter pretation of the forensic autopsy results and indeed may modify the extent of technical investigation (Table 18.3) used in port-mortem analysis. In many nations the crimes of neonaticide and infanti cide are distinct from those of filicide, child homicide or unlawful child killing. Such jurisdictions recognize that the syndromes of neonaticide and infanticide have specific features not only in terms of aetiology, modus operandi and sociocul tural implications, but also in prevention. There exists also the rare problem of false confessions to the drowning of children .24 In these circumstances the recreating of events leading to the child's death is crucial to the legal protection of a parent or other party who may be psychiatrically disturbed. In practice, the forensic examination of childhood immersion victims is straightforward. However, in evelY case judgement is required with respect to the choice of post-mortem ancillary tests and investigations. Most juris dictions operate under conditions of scarce, and sometimes inadequate, resources. This applies particularly to patholo gists or other medical examiners who are called upon to investigate childhood immersion deaths away from sophis ticated fo rensic centres. 25 A small minority, less than 5 per cent of all childhood immersion accidents, have the poten tial for criminal overtones. However, with the increasing tendency for civi l litigation, the pathologist may become enmeshed in highly complex and prolonged civil law suits, in which grieving and aggrieved parents seek compensa tion or damages from parties who are the site owners of water hazards in which children have drowned . For this reason, there is an increasing tendency for forensic pathol ogists to undertake tests a nd investigations that will leave no doubt as to the recreation of factual events that led to death in the water, or after failed attempts at post-rescue resuscitation.
352 I
Drowning and near drowning
Classification of Drowning in Children There are two paradigms by which child drowning may be classified. The first of these uses the primary discriminator of intentionali ty, thus separating accidental drownings from the syndromes of unlawful child killing. The second approach is to classify drownings by site, each with its site-specific syn dromic boundaries. Most medicolegal approaches to child hood immersion incidents use both approaches.
Child-killing Immersion Syndromes NEONATICIDE Neonaticide is defined va riously, in different jurisd ictions, as the killing of a n infa nt in the first 24 hours of life or the killing of an infant in the first 4 weeks of life. In the forensic literature it is usually taken to mea n the unlawful child killing of a subject delivered naturally and of sufficient developmental maturity to have been capable of independent survival, whose killing was perpetrated within the first 24 hours of life. Neonaticide by dro wni ng is not rare, but it is, fOltu nately, decreasing in frequency in many Western soci eties. The syndrome is very specific and has been long recog nized because of its sad sociofamilial overtones. 26 In general, neonaticide is in almost always perpetrated by the mother. In some legal jurisdictions, the crime of neonaticide, with its connotations of diminished responsibili ty, is only recognized as a specific crime if perpetrated by the mother. The mothers are almost always young (95 per cent of cases), often teenagers. They are almost alw ays single. Neonaticide by drowning usually occurs in the context of a concealed pregnancy and in the context of a concealed, solitary labour and delivery. The mothers are often mem bers of ethnic minority groups, often in religious or lan guage isolates living in Western society. Such families typ ically are t hose with religious or tra ditional cultural condemnation of premarital sexua l re lations . These tragic incidents are also sometimes encountered in white or black families, particularly in those of lower socio econom ic sta tus, in which there is a very strict, male-dominated ethos in the microsociety in w hich the parturient mot her is trapped. In these circumstances it is not rare for a Single teenage girl or yo ung woma n to conceal the pregnancy. At the time of the ensuing delivery a nd during labour, the victim will seek a priva te bathroom or public toil et in which to deliver her child secretly.26 The mother adopts the least uncomfort able position for such so li tary delivery. Und er these circum stances, the neonate is often born into a toilet bowl or, less commonly, in to a bath. In Western and oriental societies this usually occurs in a private or public toilet cubicleP There is often doubt about the degree of intent of the deliberate taking of the baby's life, allow ing for the dimin ished responsibility of a dist ressed, usually lonely and always ignorant primigravida. In many such instances, the
young single teenager, ignorant of physiology and without antenata l care, de livers herself of her infant. In the context of fear, pain and ignorance, a proportion of such mothers may not perceive the risk of immersion death. Having said t his, neonaticide is the deliberative taking of the newborn's life and t he related crimes of the deliberate exposure or abandonment of t he newborn have been classic and not uncommon crimes known since antiquity. In this context, drowning is simply the modus of such unlawful child killing, the incidence of which varies, as it is always done with t he sociocul tural mores and the economic circum stances of the parents of the neonate concerned. Forensic autopsy of such infants needs to establish whether spo ntaneous breathing has occurred and whether the airways are patent, and whether asphyxia by water or other agents was the cause of death. As a proportion of such new ly born infants a re found dead in places such as public to ilets, DNA extraction and storage is essential. The tracing, finding and identification of mothers is important in these distressing circumstances, as post-neonaticide identification of the biological father may ha ve important forensic or legal implications. Such may occur if there is assa ult on putative fathers by the girl 's family (particularly by her father or brothers) and in the rare cases when putative or biological fathers may be enjoined in the act of neonaticide itself. INFANTICIDE The crime of infanticide is the unlawful killing of a child under 1 year of age by its mother. The designated crime of infanticide has evolved as the judicial recognitio n that there is a subset of unlawful killings that are the result of diminished responsibility. The deliberate killing of an infant, by a mother often dis abled by psychosis, occurs not at birth but in the weeks or months following birth. Under these circumstances, drowning is, in one sense, a non-specific modus, as the means of ending the child's life. 28 Most such perpetrators are suffering from post-natal depression, with a smaller proportion afflicted with schizophrenia. 29 The intra family dynamics in cases of infanti cide by immersion 3o differ from those encountered in cases of the deliberative, repeated , sub -fa tal U'auma, which is a feature of the crescendo child ab use syndrome that, of course, may ultimately lead to the death of the child concerned. Post- nata l depression is a dangerous syndrome for both infant and mo ther. When depressed mothers kill their infants, the prox imi ty of the family bath tub or of was hing machines, buckets or pails makes this cause of dea th well recognized. The syndrome of infanticide-suicide is well rec ognized by all forensic pat hologists. Sometimes, a mother will kill one, severa l or a ll of her children before taking her own life. 3 1.32 Infanticide by drow ning may reveal, at forensic autopsy, skin features or lesions co nsequent upo n the force exerted by the perpetrator to drown the struggling child. Some parents have attempted to drug their children before immersing them as the fi nal act of killing.
Forensic immersion syndromes I
The forensic pathologist has an important preventative role in infanticide, not of course in realized cases but as an advocate for increased surveillance and help to those mothers afflicted by psychosis (either by endogenous bipo lar disorder or by schizophrenia) when the risk of infanti cide may be hi gh. Such risk remains high in subsequent pregnancies, and the syndrome of sequential infanticide, sometimes by drowning, is well recognized. FILICIDE Filicide is that crime in which the offender is a biological, adoptive or de facto parent. The method of killing is culture specific. In European, Asian, Canadian and Australasian societies the cause of such deaths are head injury, drown ing or suffocation. 33 In the United States, homicidal asphyxia is less common in some regions, where gunshot murder is more frequently employed. 34 Mothers (60 per cent) kill their children more often than do fathers. Eighty per cent of such victims are between 1 and 5 years of age, with a median age of between 2 and 3 years. 35 Particular 'at risk' times for such immersion killings are in the early hours of the evening, particularly during weekends. 36
Site-specific Immersion Syndromes BUCKET OR PAll
A particularly difficult forensic immersion scenario is that which involves a child who has drowned in a bucket or pai1. 37 ,38 The majority of such incidents are undoubtedly accidental, but there remains a subset, of unknown propor tion, in which attempted or realized infanticide or child homicide has occurred. The age range of such victims is 9-20 months. Toddlers are at risk, but infants who are able to pull to stand , but cannot yet wa lk , are also at risk. The infant or toddler may become wedged in the bucket. No complete unselected (fatalities and non-fatalities) series for this type of accident has yet been published. A review of all published papers indicates that the mortality rate approaches 60 per cent and that the risk of post-accident neurological sequelae is high amongst survivors. A related and rare type of immersion involves those toddl ers who climb up beside and fall into washing machines or washing tubs. Buckets and pails are usually ofJ-S gallons (40-70 L) in capacity. The contents of the buckets are almost always water, soiled nappies, dirty mop water, bleach, detergent, soap or antiseptic. 38 These fluids have the potential to cause intense bronchospasm and life-threatening laryngeal spasm , irrespective of the immersion time. In survivors of such immersion incidents, there is always severe pneumo nia and often systemic complications that require the most sophisticated management in the inten sive care ward over subsequent days. In the case of fatalities, analysis of bucket or tub contents and of lung tissue forms an important
353
component of the chemical and toxicological approach that is so important in these accidents. BATH TUB DROWNINGS
Immersion incidents in the family bath tub and in domes tic spa pools have special implications for the forensic pathologist. From both the clinical and the forensic point of view there are eight defined syndromes of bath tub drowning and near-drowning (Table 18.5). Of these specific and definable bath tub or bath spa syndromes, accidental immersion is the most common in childhood. In adult series, suicides figure promin ently.39 A special challenge to both general and forensic patholo gists, and also to clinicians, is the generation of a differen tial diagnosis in respect of causation to encompass the key prese nting feature - the fact that the bath tub is the site of the immersi on incident. If the clinician or path ologist does not consider the various syndromes that constitute such a forensic differential diagnosis (Table 18.5) then it is certain that lawyers in subsequent coronia I, civil action or criminal courts will do so. The fact that perhaps 10 per cent of child hood bath tub drownings mask the tragic fin al result of crescen do child abuse necessitates full post-moltem X-ray skeletal survey in all such cases. Al most all bath tub immersions involve fresh (tap) water, chlorinated to the standard 1 palt per million (ppm). Diatom analysis plays no part in the interpretation of forensic bath tub analyses. By contrast, soap products are inhaled with water in many cases of accidental bath tub immersions involving infants and toddlers, unlike the situation in adult suicides 39 and some homicides 4o for which the bath tub is the site of death. Bath salts are used in many children's baths, especially in the USA and Japan 14 and in many affluent families of all nations. Bath salts contain fluorescein, which is highly sensi tive to detection by high-performance liquid chromatography. Confirmation of fluorescein in lung tissue (and other tissues), in parallel with its demonstration in any residual bath tub water samples, may be helpful and specific. Its demonstration parallels the selected and rare use of diatom marrow demon stration following occasional fatal immersions in open water. Accidental Child Drowning in Bath Tubs
Overall, 80-90 per cent of childhood bath tub drownings are accidental. The usual scenario of such accidents is the vely human and universal situation in which a mother, tired at the end of the day, is attempting to coordinate the feed ing and bathing of a large and vigorous family. As part of this, an infant who is in the bath tub becomes overlooked for several minutes before his or her well-being is checked. The syndrome of accidental infant bath tub immersion is quite specific. Such fatalities and near-fatalities occur only, or virtually only, in working class, labouring and poorer families. 41
354 I
Drowning and near drowning
Table 18.5
Site-syndromes of bath tub and spa pool drownings: forensic and clinical implications
Syndrome Accident
Incidence
Notes
20 per 100000 infants aged 0-"
Poorer and work ing class families Large family size One-th ird are single-paren t families 50 High successful resuscitation rates (60 per cent) Child unsupervised due to sudden unexpected break in routine Parenta l dichotomy of care, each believing the other is caring for the infant 'Vu lnerable periods' such as acute ill ness or marital discord Families often known to ch ild protection agencies Step-parents or de facto relationships inc rease risk Often other types abuse uncovered Some perpetrators 'draw back' after the child loses consciousness Sometimes precipitated by acute stress in a step-parent or de facto partner who is left alone with a toddler at a time of acute interpe rsonal discord Pa rent psychotically dep ressed or with low 10 Child homicide-su icide syndromes fall in this group High rate of realized death Only a small percen tage survive Older children and youn g teenagers Victims may have uncontrolled epilepsy Children with a history of epilepsy who lock themselves in the bath room for ablutions are pa rticula rly at ri sk A proportion of physically disabled toddle rs and yo ung child ren Often not concealed from police, altruistic intent May be a history of cardiac abnorma lities; especially those of bundle of His and right bundle branch block or long 0-Tsyndrome 56 Although very rare, commonly raised as a defence against a charge of homicidal drowning or in civil claims for compensation after drowning
months, annually 8 per 100000 infants aged 12-23 months, annua lly
Child abuse
10-15 per cent of bath tub
immersions estimated to be attempted or realized homicide
Child homicide
Rare - 12 per cent of all child homicides 10
Bath tub epilepsy
Rare
Euthanasia
Rare
Bath tub drowning with cardiac pathology
Very rare; sudden cardiac death (all causes) in apparently normal children has a frequency of 1.2 per 100000 patient-years
These accidents affl ict infants and todd lers in a very defined, age-specific window of 8-18 months; the modal age of such accidents is 9-11 months.42 The mean number of siblings of victims of this type of immersion is greater tha n that of the population-matched average. The victims are usually the you ngest or second-youngest child in the family - often the second-youngest, when there is a new baby claiming a significant part of the mother's attention. This peculiar vulnerability to accidents, to which children in higher birth ranks within the family are especially prone, is known to app ly to other childhood accidents as well, especially accidental poisoning.43 More than one-half of bath tub immersion in cidents occur during a specific 'vulnerable period' when the family routine is suddenl y or unexpectedly broken, such as that which occurs during acute sickness afflicting either the parents or children or in the context of martial strife.41 A typical scenario is that which involves a stressed mother, who is tired at the end of the day, attempting to cope, unaided, with the control, bathing and feeding of several high-spirited or fractious, but always vigorous, yo ung children. The telephone rings, or an appliance breaks, or
-
someone calls unexpectedly at the door. The mother lea ves two or three children, including an infant, in the family bath tu b,44,45 leaving the youngest in the care of an older, but st ill pre-school, child - the older children may hop out of the bath when the mother is gone, leavin g the infant alone. Another factor in some infant bath tub drownin gs is the use of bath seats, wh ich confer inappropriate assurance of safety,45.46 despite clear instruction from the manufacturers that the baby should not be left unsupervised. Unlike other childhood drowning accidents that all show a preponderance of males, accidental bath tub immersio ns show an equal sex ratio. Of those in fan ts and toddlers found unconscious in the family bath tub, some 60 per cent respond to CPR, with residual intact intellect and without neurological sequelae. This reflects the rela tively short immersion times invol ved in such incid ents. Bath Tub Drownings and Child Abuse All reviews of consecutive, unselected series of ch ildhood bath tub immersions with detail ed follow-up have revealed th at some 10-25 per cent of such cases are in fact the result
-.
-
Forensic immersion syndromes I
of non-accidental injury35,42,47 Forensic studies in Ger many have shown that in 5-10 per cent of all child homi cides in the under-5 age group, the modus operandi is drowning, usually in the bath tub. 48 The immersion assault is usually one such in a crescendo series of acts of child abuse perpetrated on the infant or toddler. Step-parents and de facto partners are almost exclusively the perpetrators. Some 'lose their nerve' during the assault and draw back and may summon help for resuscitation. A clue to the specific syndrome of child abuse by bath tub immersion may be that the child is older than the modal age for true accidental drowning in this site (9-11 months); or ou tside the typical 8-18 months age specific window of vulnerability to bath tub immersion homicide. The median age of normal infants to be able to pull to stand is 9 months and the media n age for walking in normal full-term infants is 13 months. Usually, children older than these ages ca n support their heads out of water in a bath tub, even if abandoned or left unattended for short periods of time. There should be a high index of forensic or coronial suspicion in all bath tub immersions involving toddlers older than 11 months, particularly if there is no history of developmental delay or epilepsy. The median depth of water in bath tub immersion incidents is approximately 20 cm. Although deliberate hot water scalding by immersion in the bath tub is not uncommon in reported series of child abuse cases, the pathology of bath tub immersion incidents is primarily that of hypoxia and the two injuries are not reported coincidentally. It may be that the perpetrators do not wish to scald their own hands. Bath Tub Drowning and Child Homicide There exists the separate and distinct syndrome of child homicide, as a 'one-off' event, by bath tub drowning. As a single act, not in the context of ongoing child abuse, this assault is relatively uncommon. Syndromically, it almost always involves a parent afflicted by psychiatric illness or by low intellectual ability. Mothers are usually the perpetra tors, have a psychiatric history and are living in disadvan taged socioeconomic environments. Mothers who kill their children prefer to strangle or drown their victims. Psychol ogists believe that in a large proportion of such cases the psychodynamics are those of surrogate suicide. 49 It is some times the modus operandi of the familiar and tragic 'child homicide-parent suicide' doublet. In this latter incident, tragically familiar to all forensic pathologists, one parent is almost always psychotically depressed, or in unremitting despair, and kills his or her child or several children and then immediately takes their own life. Survival rates for children involved in such incidents are low; less than 10 per cent sUlvive this tragic scenario. Child homicide by drown ing, in the context of sexual assault followed by homicide, always perpetrated by a stranger or non-biological relative, is rare. The mode of killing is almost always by physical
355
violence or strangulation, but the body is sometimes dis posed of in a dam, creek or watercourse. Infant homicide by bath tub drowning approaches the 'per fect crime' in that forensic skills usually cannot distinguish non-accidental submersion from accident. Two series of bath tub immersions have revealed that child homicide is a subset of the bath tub drowning syndrome - known only because of later, unsolicited confessions by the perpetrators. 24 ,50
Childhood Bath Tub Immersion in Epilepsy A small proportion of children and teenagers drown, or almost drown, in the bath tub as a result of an epileptic seizure sl Enquiry about a past his tolY of epilepsy is impor tant in both the clinical and forensic workups of all immer sion accidents. There exist several specific issues relating to this syndrome. It is very rare for children to drown in the sea or swim ming pools as the result of in-water seizures. 52 This applies even to children with uncontrolled epilepsy. The risks are significantly greater if there is a sub-therapeutic serum concentration of anticonvulsant dnlg present. The inci dence of post-seizure bath tub fatality is less than 0.1 per cent of all childhood drownings. There is a higher specific risk to epileptic children, espe cially teenagers with epilepsy, to fatal and near-fatal drown ings in the family bath tub. All parents (and children) in families in whom uncontrolled epilepsy is present are warned not to take private plunge baths in the family bath tub, but rather to shower standing up. Several cases of flan nels or 'washers' occluding the drainage plug, even during such 'stand-up' showering ablutions, have been repOlied in forensic sel1es of fatal bathroom immersion acciden ts related to epilepsy. Teenagers, being what they are, are often insistent on absolute privacy in the bathroom and insist on locking the bathroom door before taking a plunge bath or shower. If a seizure occurs, and the unconscious victim slides beneath the water, the chance of a successful resuscitation is small. Such cases are represented in all forensic series of childhood immersions. It is believed that the (misguided) practice of placing feverish infants into a hot bath - raising a high risk of febrile convulsion - may be the cause of immersion in a small proportion of cases of childhood bath tub fatalities. Water impinging on the face, involving the sensory dis tribution of the trigeminal nerve, is the afferent trigger for the diving reflex. 18 This leads to a complex series of dynamic cardiovascular changes, including' increased vagal tone. There is some evidence that this may be one cause of 'bath epilepsy', initially described in the Indian medical literature but known to occur in all races.
Bath Tub Drowning and Euthanasia All series of unlawful child killings include cases of euthanasia. 53 They are usually perpetrated by a parent but
356 I
Drowning and near drowning
rarely by a grandparent or other close family member. One method of euthanas ia is bath tub immersion. In the major ity of such cases the chi ld has a congenital disability, such as a neural tube defect, mental retardation , cerebral palsy or a chromosome abnorma li ty. In occasional cases, the par ent is psychotic, wrong ly believing that a normal child is disabled or suffering when such in reality is not the case. 49 In instances of euthanasia by drowning, the motive is always (albeit tragically) an altruistic one. The parent believes that the child is suffering and t hat it would be kinder if the child was put out of his or her pain or misery, or that they could be helped from a t ragic black or wicked world. There may be no attempt to conceal the crime. Some perpetrators proactively contact the police to report the incident. Bath Tub Drowning and Organic Pathology A final bath tub drowning syndrome is one in which pre mortem pathology of the heart, recognized or unrecognized, is the cause of loss of consciousness during bathing. This syndrome is rare, but is included both for completeness and because the question commonly arises in the legal defence mounted by those charged w ith homicide by drowning. Sud den cardiac deaths in childhood and during adolescence are we ll known to forensic pathologists. 54 Myocarditis was demonstrated in 5 out of 22 cases of accidental drowning and underlines the need for full investigation of drowning deaths wherever they OCCUr. 55 Importantly, the demonstra tion of pre-existing cardiac abnormalities in a drowned child may protect any innocent person on whom suspicion of cu l pable neglect has faUen. This applies not only to situations where ch ild homicide is being considered in the differential diagnosis, but also to situations such as deaths in public swimming pools where lifeguards may be subject to cliticism of neglect, sometimes of culpable neglect, in subsequent legal proceedings. Congenita l or acquired cardiac abnormalities are also the antecedent to drowning in some cases of older victims who drown in bathtubs,56 swimming pools or open water. Anomalies of the bundle of His, ab normalities of the right cardiac bundle branch and congenital abnormalities of the coronary arteries have all been reported in cases of teenage drowning, including bath tub drowning, in which toxicolog ical tests for alcohol and illegal drugs have proved negative. In rare cases, genetic cardiomyopathies may be present. 57 The mechanism of such sudden and un expected death dur ing swimming or bath tub bathing is conjectural. One pos sibility is that the diving reflex (which follows facial imm ersi on) induces cardiac dysrhythmia with heightened sens itivity, i.e. a priori abnormal hearts. It h as also been suggested that cardiovascular damage may follow reduced venous return due to venous dilatation in hot water immer sions; if this is superimposed on an 'at risk' heart then car diac output may be insufficient to sust ain consciousness. Under normal circumstances, the cardiac output wi ll be sufficient to return brain function to co nsciousness if an
individual temporarily col lapses. However, if the individual is under water, secondary vagal stimulation from laryngeal irritation may compound the hypoxia that is already present. In this co ntext, it is important to note that any drowning episode itself may provide an outpouring of catecholam ines (the 'sympathetic storm'), whi ch, w ith hypox ia , may result in microscopic changes to the myocardium , including myocyte contraction banding and foc al myocyte hypereosinophilia s 8
SWIMMING POOL DROWNINGS In most temperate and tropical countries of the developed world, the principal site of drowning in the childh ood years is the backyard swimming pool. Salt-water imm ersions involving children are now less common than fresh-water immersions, even in littoral regions, for this reason. In the USA alone there are now over 12 million pla st ic wading pools and over 5 million surface swimming pools, of which an estimated 2 million are of the more dangerous, in-ground variety. Proportionate rates may be even higher in other countries, such as Australia and New Zea la nd. In many countri es, pool drownings constitute 70 per cent or more of all consecuti ve unselected series of immersion deaths in ch ildhood. In-ground pools cause 80 per cent of swimmi ng pool fatalities. Motel , hotel , caravan and trailer park pools are a particular hazard in all reported series. 59 Even in countries such as Finland, as many near-drownings take place in swimming pools as occur in lakes, and twice as many in swimming pools as in the sea. The current status of swimming pool drownings has been a feature of forensic series only since the early 1970s. 47 At t hat time, engineering technology and social affluence made private in-ground pools widely accessible. In many commu nities in many nations, one in five homes now has an in- or above-gro und swimming pool filled with water at least throughou t the warmer months of t he year. Even in temper ate climates, where winters may be free zing, pool-house ratios may be as high as 1: 10. In the State of California, some 100 toddler dro wnin g deaths continue to occur annually. The age spectrum of such victims is between 12 and 40 months, with a modal peak between 18 and 24 months. The social class 'risk curve' is U-shaped, with deaths occurring disproportionately w ithin richer families who, most com monly, have in-ground pools, and in poorer families where above-ground pools are often bought as an impulse purchase, installed and then poorly maintained. Above-ground pools are not as dangerous as in-ground pools. However, some children in the 18- to 36-m onth age range will climb a pool ladder, or will place a box or other object beside an elevated pool to gain access. Overall, 70 per cent of toddlers who drown do so in their own pools. Other 'at lisk' pools are those of neighbours, motel, caravan or t rail er park pools and the pools at homes of relatives whom children visit. Toddlers and yo un g children in this age-vulnerable window do not fear water. Many simply crawl into it, or step into such pools if no barrier is present. The Brisbane
Forensic immersion syndromes I
Drowning Study 47 showed that often such vIctIms were attracted to the water by a floating toy, mbber ring or ball , or by an object lying on the bottom of the pool. The follow ing factors lead to higher survival rates following pool-side CPR: clothing (with its buoyancy), the diving reflex (pre served in infants and toddlers) and children found floating (with residual aeration) as opposed to those discovered on the bottom. Toddlers virtually never drown in fenced pools with safety standard-approved self-latching gates with high, hidden locks. Intense advocacy, often supported by pathol ogists, to introduce policed, safety legislation to protect toddlers from such water hazards has been unsuccessful in many communities. Most toddlers who drown in such pools do so within 30 m of their own home. Immersion times are almost always under 20 minutes. Successful resuscitation rates for consecutive, unselected series of all such swimming pool immersion accidents approach 60 per cent and rise to 70 per cent if a trained first aider happens, coinCidentally, to be involved in the attempted resuscita tion. Survivors do well, and some 97 per cent of children who survive this near-drowning syndrome function nor mally. Some 30 per cent of such survivors have wide sub scale disparities on formal psychometric testing but do not manifest clinical neurological signs. Autopsy findings of such children are typically those of a completely fit and well toddler who is alive at one moment and dead 10 minutes later. In such cases there may be mini mal anatomical changes in the lungs. The body often shows post-mortem signs of CPR trauma, including fractured ribs, contused pericardium or injection needle marks from the frantic use of cardiostimulant drugs by ambulance officers, paramedics, firemen (in the USA) or physicians. PUBLIC SWIMMING POOL DROWNINGS
A small proportion of children, almost always in the age range of 3-1B years, drown in public access swimming pools or aquatic parks. One US study draws attention to an excess of deaths in this situation to black, adolescent males 60 Such individuals are typically found on the bot tom of the pool by another casual swimmer. Very com monly, such drownings are followed by major medicol.egal, civil, insurance and regulatory consequences. A high pro portion of such drownings becomes the focus of prolonged court cases, often involving claims for large compensatory sums of money; commercial pool owners, lifeguards, para medics, emergency room physicians and forensic patholo gists may all become enmeshed in complex discussions about pathophysiology and the differential diagnosis of the causes of drowning in general. The majority of such cases are simply due to the fact that individuals, usually poor swimmers, cannot stay afloat and so they inhale and the 'drowning chain' ensues. Swimming lessons and, perhaps, optimistic evaluation of swimming abjJj!y, may engender a faJse sense of security amongst
357
parents. They are no substitutes for parental vigilance and close supervision when children are in the water. 61 •62 A small proportion of such cases involve children who have been skylarking in the pool surrounds, fall and hit their heads or become winded and cannot extricate themselves when they become submerged. Another subgroup is probably second alY to cardiac abnormalities, often congenital in nature but hitherto unrecognized; this group includes those with con genital coronalY artery or valvular abnormalities or those with hitherto unrecognized dysrhythmias, with mechanisms similar to those described in the case of bathroom immer sions. Some rare cases are consequent upon cervical spine injuries following diving accidents. A very small sub-group are consequent upon epileptic seizures in the water. In this latter context, however, it must be said that swimming pool drownings due to proven epileptic seizures are extremely rare. This is not to say that children do not have epileptic seizures in the water but, in practice, most are recognized and the victim is extracted without significant hypoxia and without any subsequent post-hypoxic neurological defects.53 Another subset comprises those young teenagers who have taken alcohol or drugs, and who lose coordination in the water, or sometimes who are illegally or clandestinely swim ming in pools, often as members of group 'larks', sometimes involving 'skinny dipping' (nude swimming) at night. Another subset consists of children performing the danger ously underrated practice of hyperventilating before diving and underwater swimming in such public pools. Because of the extended differential diagnosis that is inescapable in such cases, often dissected in meticulous detail in subsequent court proceedings, it is absolutely essential that the widest array of appropriate post-mortem investigations be undertaken. This includes radiographic studies (especially of the skull, chest and spine), alcohol and drug assays and meticulous forensic autopsy techniques in the examination of heart, lungs and brain (Table IB.3). In practice, it is uncommon for such cases to result in positive findings in any of these investigations, but the careers, rep utation and professional security of many individuals who are involved in the rescue, the attempted resuscitation and the management chain may well be dependent on the pathologist's findings and defended opinion. Perhaps most importantly of all, 'closure' of the grieving process by par ents and loved ones is facilitated by the forensic patholo gist's confident recreation of the events that have led to post-immersion somatic death.
RIVERS, CREEKS AND LAKES
Children occasionally drown in rivers and creeks and in lakes. 63 Those who do are almost always (90 per cent) boys, and the modal age falls in the age window of between B and 12 years. There is often an element of disobedience involved in such cases, when children are swimming far from supervision. The typical scenario is when a boy or
maJe teenage! experiences diff]cuJties in the water and his
358 I _ Drowning and near drowning
friends find that they are unable to rescue him and run for help. Because of the distances involved, resuscitation attempts are unsuccessful in more than 90 per cent of cases. Estim ated immersion times are relatively lon g and often exceed an hour or more before the body is extracted from the water. The forensic autopsy often reveals gravel, sand or aquatic flora in the air passages. DRAINS, TRENCHES AND SEWERS Children who drown in drains, trenches and sewers are almost always boys, and sometimes more than one victi m is drowned in the same episode. The victims are often playing or swimming in forbidd en areas. Often, extraction from the water is difficult and bodies may be wedged in outlet pipes or municipa l watercourses. A series of such drownings always include that subset of children who are playing in fast-movi ng and dangerous waterways during floods. Under such circumstances the novelty of the situation may mask the threat of immersion and the bodies of the victims may never be recovered. Other are found wedged in or entrapped in pipes or outlet drains or under submerged trees. SEA DROWNING Many nations have hi gh shoreline-area ratios, with a pop ulation who are sea-oriented from infancy. In such regions, children grow up in an environment in which respect for the ethos of both water safety and water danger is a natu ral part of culture and folklore. Under these circumstances, sea drownings involving toddlers are relatively uncom mon. 64 Older children are particularly at risk. Child victims of sea drownings are found disproportion ately amo ng immigrants, tourists or other ethnic sub groupS.17,65 The children and their parents may not be aware of the threat of the sea or surf. Occasionally, young teenagers drow n during sailing excursions but, in general , boating and the use of surfboards, in current practice, are lo w drowning threats to children. Childhood drownings in the sea occur almost always duri ng daylight hours and, as a manifestation of the recreational use of the sea, most often occur durin g weekends or holidays. Childhood salt-water immersion rates are unaffected by tidal state. Most sea and surfing beaches are separated by road, dune strip and a beach from residen tial houses. The result of this is that it is very unusual indeed for toddlers or preschool children to drown in the open sea; the modal age for this type of sea drownings is 8 years. The phenomenon of 'secondary drowning' is most often identi fied after salt-w ater drownings. In this latter syndrome an indiv idual may be extracted, pulseless, from the water and may respond well to cardiopulmonary resuscitation, only to manifest life-threatening deterioration hours later due to washout or denaturation changes in pulmon ary surfactant. Children are sometimes envenomed by jellyfish or poi sonous fishes while swimming.66 Under such circumstances, as in the case of adults, there is a very serious risk of
--
secondary drowning consequent upon pain, panic and the ensuing incoordination even amongst proficient swimmers. The victims of stonefish envenomation, which are com monly encountered throughout the Indo-Pacific littoral in both tropical and temperate waters, often become mani aca l with pain . If this occurs during snorkelling, scuba-diving or reef walking, drowning is an ever-present threat. A sea rch for, and identification of, barbs or other foreign material from the venom apparatus is important in such cases. Severe envenomation by cubomedusans can cause death by direct envenomation while the victim is still in the water, or from a combination of envenomation and drowning. Box jelly fishes (sometimes called 'sea wasps') are amongst the world's most venomous creatures and many of the recorded victims of such deaths, a disproportionate number of whom are chil dren, occur before the victim can be extricated from the water. In the forensic examination of such victims, a strip of adhesive tape applied to areas of the skin thought to be affec ted and then examined under the microscope may reveal not only the characteristic nematocysts, but also will enable both genus and species identification to be made. The foren sic autopsy of salt-water drowned victims usu ally shows no specific distinguishing features if the body is retrieved w ithin several hours after the drowning episode. As in adults, the bodies of long-immersed child victims may show secondary changes due to crustacean or piscine post-mortem damage. In such cases forensic identification may be difficult and dental X -rays, post-mortem finger printing or DNA extraction and analysis are essential. AFTERMATH Only a minority of immersion incidents result in fatali ties. 67 ,58 An understanding of the pathophysiology of the immersion sequence leading to somatic death, taken in con junction with the documentation of post-mortem changes, will mean that both more efficient prevention and better car diopulmonary resuscitation will be possible in the future. The documentation and subsequent compilation of child hood immersion statistics is very important from the point of view of future prevention. In this context, case finding of immersion fatalities undertaken for epidemiological research is compromised because of the many different causes and syndromes of drowning in childhood. As many as 1 in 10 cases of childhood immersions may not be retrieved for analysis. For example, some are coded as motor vehicle injuries in those cases when vehicles have crashed into the water.59 Multiple-cause coding is essential if a community's or nation's statistics are to be fully exploited in the context of injury prevention. 70 The pathologist's professional and ethical responsibility is to the dead child and to the determination of the circum stances of that death. The evid ential value of such autopsy findings, however difficult may be their interpretation in the current stage of knowledge, is of the greatest impOltance. Ultimately, the integrity of not only the professions of
~
References I
medicine and of law depends upon such a fo rensic service, but also that of society. The deliberative killing or preve nta ble death of a child defines, in part, the society in which such tragedies occur. From the professional point of view, the pathologist also has both a professional and an ethical responsibility to any living siblin gs of children who have died from immersion, whether accid ental or no t. Much remains to be ach ieved in fos terin g closer coll ab oration between clinici a ns ca rin g for the survivors of near drowning in cidents and the path ologists who investigate fatal imm ersions. 7l •72 Paediatric a nd forensic pathologists have an important role to play in preventative medicine a nd in promoting advocacy for public health stratagems that relate, in particular, to home and water safety in com munities in which drownings are common. A child's death from drowning is, in one sense, from the family's po int of v iew, the beginning of a new era. The role of the pathologist is a crucial one in helpin g parents in t he immediate aftermath of death. To be able to recreate exactly what has happened often brings long-term resolu tion to the aftermath ofa child's drow ning, which is always a t ragedy.
10
11
12
13
14
15
16 17
18 19
ACKNOWLEDGEMENT I thank Dr Terry Sinton, a senior for ensic pathologist of the John Tonge Cent re, Queensla nd Health Pathology Services , Brisbane, Australia, for mu ch encouragement.
REFERENCES
20
21 22
23 24 25
Centers for Disease Contro l and Prevention (C DC). Nonfa tal a nd fata l d row nings in rec reatio nal wa ter settings - United States. 200 1-2002. MMWR 2004; 5 3:447- 52 . 2 Meyer RJ, Theodorou AA, Be rg RA. Chil dhoo d drownin g. Ped iatr Rev 2006; 27:16 3- 9. 3 Fin gerhut LA, Cox CS, Warner M. Internat iona l co mparative ana lyses of injUlY mortality. Findings from the ICE on inj ury sta tistics. International Collaborative Effort on Injury Statistics. Adv Da ta 1998; 303: 1-20. 4 Peden MM , McGee K. Th e epidemiology of drowning wo rld wide. Injury Cont Safety Promotion 2003; 10:1 95- 9. 5 Laraque D, Barlow B, Durkin IVI. P reve ntion of yo uth injuries. J Natl Med Assoc USA 19 99; 9 1 :557-]. 6 Aus tralian Institute of He alth an d Welfare. Australia's Hea lth 2000. The Seven th Biennial Hea lth Report of the A ustra lian In stitute o f Health a nd Welfare. 4.2 Childre n a nd Young People. Canb erra: Aust lnst Health Welfare (AIHW Ca t. No. 19), 2000, pp. 184-92. 7 Byard RW. Accidental childh ood death and the rol e of the pa th o log is t. Pediatr Dev Pat/101 2000; 3 :405-1 O. 8 Pearn JH, Nixon J. A n analys is of the causes of freshwater immersion accid ents involving chiJdren. Accident Anal Prell 1979; 11:173-8. 9 CoUins KA, Nic hols CA. A decade of pediatri c ho mi cid e: a retrospecti ve study a t the Medical Uni versity of So uth Carolina . Am J F01'ensic Med Pathol 19 99; 20:169 -72.
26 27
28 29
30
31 32 33
34
359
Byard RW, Houldsworth G, J a mes RA, Gilbert JD . Cha racte ri stic feat ures of sui ci dal drownings : a 20-y ear st udy. A m J Forensic Med Pathol 2001; 22: 134-8. Pearn JH. Drowning. [n Pearn J, Leditschke JF, Marsha ll V et al. Th e Science of Firs t Aid. Theore tical and Experimental Basis of Modem First Aid Practice. Canberra: St Joh n Amb ulan ce Australia , 1996, pp. 138-47. Ludes B. Coste M, NOlih N et al. Diatom analysis in victim's tissues as an ind ication of the site of drowning. Int J Legal Med 1999 ; 11 2 :163-6. Pearn JH. The manageme nt of near drowning. In Aochi 0, Ama ha A, Takesh ita H red s) Intensive Critica l Care Medicine. Amsterdam: Excerpta Med ica, 1990, pp. 139 -46. Muka ida M, Kimura H, Takada Y. Detect ion of bathsa lts in the lungs o f a baby drowned in a bathtub: a case report. Forensic Sci Int 1998; 93:5 -11. Modell JH, Bellefleur M, Dav is JH. Drownin g withou t asp iration: is this an appropriate diagn osis? J Forens ic Sci 1999; 44:1119- 23 . Gi lbert FF, Gofton N. Te rm ina l di ves in min k, muskrat and beaver. Physiol Beh av 1982; 20:8 35-40. Pea rn JH. Dro wning. In Di cke rmann JD, Lucey JF red s) The Critica lly III Child, 3rd edn. Philad el phia, PA: WB Saunders, 198 5, pp. 129 -5 6. Gooden BA. Mechanism of the human diving response. Integ rative Physio l Beha l! Sci 1994; 29: 6-1 6. Torre C, Varetto L, Tappi E. Scann ing el ectron microscopic ultrastructural a ltera tions of the pulmonary alveolus in ex perimental drowning. J Forensic Sci 1983; 28: 1008-1 2. Gu MN, Xia o JF, Hua ng YR et al. Study of di rect injury by sea water in canine models. Di Yi JUI1 Yi Do X ue Xue Bao 2003; 23:201-5. Tabata N, Morita M, Azu mi J. A frozen newbo m infant: froth in the ai r-passage afte r thawing. Forensic Sci Int 2000; 24:67-74. Gotoh da T, Kubo S, Kitamura 0 et aL HSP70 an d c-Fo s expression of brain stem hyp oglossal nucl eus in drowning. J Med Inl!est 2000; 47:76-9. Sac hd eva RC. Near drowning. CritCare Gin 1999; 15: 28 1-96. Heinemann A, Pus chel K. Discrepancies in homicide statistics by suffo cation. Arch KrimilJo/1 99 6; 197:129-41. Norman MG, Smialek JE, Newman DE, Horemba la EJ. The post- mo rtem exam in ation o n the a buse chil d. Pathological, radiograp hic, a nd legal aspects. Perspect Pediatr Pathol 1984; 8: 313 - 43. Saund ers E. Neonati cides followin g 'secret' pregnanc ies: sev en case reports. Public Hea lth Rep 1989; 104:368-72 . Shiono H, Maya A, Tabata Net aL Med icolega l aspec ts of infant ici de in Hokkaid o District, Japan. Am J Forensic !VIed Path ol 1986; 7: 104- 6. Scott PD. Fatal ba ttered baby cases. IvIed Sci La w 1973; 13:1 97 - 206. Kondo T. Ohshim a T. Re trospective in vestiga tio n of medico legal autopsy cases involv ing mentally handicapped individu als. Nippon Hoiga ku Zasshi 199 5; 49:478-0 3. Geertin ger P, Voigt J. Death in the bath. A survey of bath tub deaths in Copen ha gen, Denmark a nd Go th en burg, Sweden. J Forel1Sic Med 1970; 17:136-47. Bourget D, Gagne P. Matern al fili cide in Quebec~ ) Am Psychia try Law 2002 ; 30 :345-51. Rouge-MaiHart C, Jou sset N, Gaudin A et a l. Women who kill their chil d ren. Am J Forensic Med Parhol 2005 ; 26:320-6. Vanamo T, Kauppi A, Karkola K et aL Intra -familial child homicid e in Finland 1970-1994: incid ence , caus es of death and demographic chara cteristics. Forel1sic Sci Int 200 I; ll7: 199-204. DiMaio VJ. Hom icid a l asphyx ia. Am J Forel1Sic Med Patilo l 2000; 21: 1-4.
360 I
35 36 37 38 39 40 41 42 43 44
45
Drowning and near drowning
Gillenwater JM, Quan L, Feldman KW Inflicted submersion in childhood. Arch Pediatr Adolesc !\lIed 1996; 150:298-303. Schmidt P, Grass H, Madea B. Child homicide in Cologne (1985-94). Forensic Sci Int 1996; 31: 131-44. Scott PH, Eigen H. Immersion accidents involving pails of water in the home. J Pediatr 1980; 96:282 -4. Wa lker S, Midde lkamp IN. Pail immersion accidents. Clill Pediatr 1981; 20:341-3. Copeland AR. Suicide by drowning. Am J Forensic Med Path 1987; 8: 18- 22. Schmidt P, Madea B. Homicid e in the bathtub. Forensic Sci IIlI 1995; 72:135 -4 6. Pearn JH , Brow n H, Wo ng R, Bart R. Bathtub drownings: report of seve n cases. Pediatrics 1979 ; 64:68-70. Kemp AM, Mott AM, Sibert JR. Accidents and child abuse in bathtub submers ions. Arch Dis Child 1994; 70:435-8. Pearn J. Predisposing fact ors leading to child trauma. J Epidemiol COI11I11 Hlth 1978; 32:190-3. Byard RW, deKoning C. Blackbourne B et al. Shared bathing and drowning in infants and you ng children. J Pediatr Child Health 2001; 37:542- 4. Somers GR, Chiasson DA, Smith CR. Pediatric drowning: a 20 year review of autopsied cases: III Bathtub drownings. Am J
63 64
65 66
67 68
69
70
71
Foren sic Med Patho12006; 27:113-16. 46 Sibert J , John N, Jenkins D et al. Drowning of babies in bath seJts: do they pro vid e fal se reassurance? Child Care Health Dev 2005; 3 1 :255-9. 47 Nixon J, Pearn J. Non-accidental immersion in the bath: anoth er extens ion to the syndrome of child abuse and neglect. Child Abuse Neglect 1977; 1 :455- 8. 48 Griest KJ, Zumwalt RE. Child abuse by drowning. Pediatrics 1989; 03:41-6. 49 Ma rl eau JD , Roy R, Laporte L et al. Infanticide committed by the mother. Can J Psychiatry 1995; 40: 142 - 9. 50 Pearn JH, Nixon J. Swimming pool immersi on accidents. An analysis fro m the Brisb ane Drowning Study. Med J Aust 19 77; 1:432- 7. 51 Brenin gsta ll GN. Mortality in pediatric epilepsy. Pediatr
Neuro12001; 25:9- 16. 52 53 54 55
56
57 58 59
60 61
62
Pearn J , Bart R, Yamaoka R. Drow nin g ri sks to epilep tic children. A study from Hawaii. BMJ 1978; 4:1284-5. Wilkey I, Pearn J, Petrie G, Nixon J. Neonaticide, infanticide and chi ld homicide. Med Sci Law 1982; 22:31-5. Wilkinson JL. Sudden cardiac death in childhood and adolesce nce. J Paediatr Child Health 1994; 30:384-5. Somers GR, Smith CR, Wi lso n GJ et al. Association of drown in g and myocarditis in a pediatric popU lation: an autopsy-bas ed study. Arch Pathol Lab J\!led 2005: 129:205-9. Stumpp JW, Schn eider J, Bar W. Drownin g of a girl with anomaly of the bundl e of His and the right bundle branch. A 111 J Forensic !\lIed Pat/wi 1997 ; 18 :208 -10. Schmidt P, Madea B. Death in the bathtub involving chi ldren. Forensic Sci Int 1995; 72:147-55. Lunt DWR, Rose AG. Pathology of the human heart in drowning. Arch Parhol Lab iVIed 1987; 111:939-42. Pearn JH , Brown J, Hsia EY. Swimmi ng pool drownings and near-drownings involving chi ldren. A total population study from Hawaii. Milit Med J980; 145 :15- 18. Brenner RA, Trumble AC, Smith GS et a l. Where children drown, United States, 1995. Pedia trics 200 1; 108:85-9. Brenner RA, Saluj a G, Smith GS. Swimming lessons, swimmin g ability and the risk of drowning. Injury Cont Safety Promotion 2003; 10:211-16. Moran K, Stanley T. Pa renta l perce ption s of toddler water sa fety, swimm in g ability and sw immi ng lessons. lil t J Cont Safety Promotion 2006; 13:139-43.
72 73
74
75
76
77
78
79
Editor. Drowning in a private lake, NOlth Carolina, 1981 -9 0. MMWR 1992; 41:329-31. Patrick M, Bint M, Pearn J. Saltwate r drowning and
near-drowni ng acc idents involving ch ildren. Med J Aust
1979; 1 :61-4.
Schul pen TW. Migration and child hea lth: the Dutch
experience. Eur J Pediatr 1996; 155:35 1-6.
Pearn JH. Th e sea, stingers a nd surgeons: the s urgeon's role in prevention, first aid and management of marine envenomations. J Paediatr Surg 1995; 30: 105-10. Pearn JH. Surv ival rates after se rious immersion acc id ents in childhood. Resuscitation 19 78; 6:271-8. Henderson H, Wilso n RC. Water incident related hospiral activity across England between 1997/8 and 2003/4: a retrospecti ve descripti ve study. Bi\;lC Public Health 2006; 16:2 10. Wintemute GJ, Kraus JF, Teret SP et al. Deaths resu lting from motor vehicle immersions: the nature of the injuries, personal and environmental contribu ting facto rs and potentia l interventi o ns. Am J Publ Healrh 1990; 80:1068-70. Smith GS , Langley JD. Drowning surveilla nce: ho w we ll do E codes identify sub mersion fa ta lities. Injury Prev 1998; 4: 135-9. Start RD , Delargy-Aziz Y, Dorries CP et al. Clinicians and the coroni al system: ability of clinicians to reco gnise reportable deaths. BMJ 1993; 306:1038-41. James DS , Leadbeatter S. Detectin g homici de in hospital. J R Coli PhYSicians Lond 1997; 3 1:296-8. Fornes p, Pepin G, Heud es D, Leco mte D. Diag nosi s of drowning by combined computer-assisted histo morphometlY of lun gs with blood strontium determination. J Fore nsic Sci 1998; 43 :772-6. Puschel K, Schulz F, Darrmann I, Tsokos M. Macromorphology and histology o f intramuscular haemorrhages in cases of drowning. Int J Legal ili/ed 1999; 112:101-6. Hooper WD, Johnson LP. Antiepi leptic drugs: pharmacology and therapeuti cs. In Eadi e MJ, Vajda FJ (cds) Handbook of Experimental Pharma cology. Berlin: Sprin ger- Verlag, 1999. pp. 173-83. Azparren JE , Vallejo G, Reyes E et al. Study of the diagnosti c, chloride, haemoglobi n and dia toms in imm ers io n cases. Forensic Sci Int 1998; 9 1 :123-32. Boles JM, Mabil le S, Scheydecker JL et at. Hyopglycaemia in salt water near-drowni ng v ictims. Intens ilJe Care jvIed 1988; 14:80. Zhu BL, Ishida K, Quan L et al. Post-m orte m urinary
myoglob in levels with refe rence to the causes of death.
Forensic Sci Int 2001; 115 : 183-8.
Pollanen MS. Diato ms and homicide. Forensic Sci Int 1998; 9:29-3 4.
Further Reading DROWNING: PATHOLOGY AND PATHOPHYSIOLOGY Colebatch HJH, Halm agyi DFJ. Reflex pulmonary hypoten s ion of fresh-water aspirati on. J Appl P!JysiolI963; 18:179-8"5. [A classic paper] Good en B. Why some people do not drown. Med J Aust 1992; 157:629-32. Modell JH. The Path ophysiology and Treatment of Dro wning and Near-drowning. Springfie ld, IL: Charl es C. Thomas, 19 71. [A classic book] Pearn JH. Drowning and nea r-drowning. In Black JA (ed.). Paediatric
Emergencies, 2nd edn. London: Butterworths, 1987, pp. 40- 9.
Pearn JH. Pathophysi ology of drowning. iVIed J Aust 1985; 142 :586-8.
References I
361
DROWNING: EPIDEMIOLOGY
DROWNING: FORENSIC CONCOMITANTS
Brenner BA, Smith GS , Overpeck MD. Divergent trends in childhood drowning rates. ]AMA 1994; 271:1606-8. Langley JD, Smeijers J. Injury mortality among children and teenagers in New Zealand co mpared with the United States of America. Injury PreIJ 1997 ; 3: 195-9. Lifesa ving Society (Canad a). Tile National [Canada] Drowning Report, 1997 edn . Ottawa : Lifesav ing Soc. (Canada), 1997. Lindholm p, Steensberg J. Epidemiology of unintentional drownin g and near-drowning in Denmark in 1995. Injury Pm; 2000;
Boles JM, Mabille S, Scheydecker JL, Garo B, Garee M. Hypoglycaemia in sal t water near-drowning victims. Il1tens Care Med 1988 ; 14:80. Cummings P, Qual) L. Trend s in unintentional dro wning: the role of alco hol and medical care. ]AMA 1999 ; 281 :2198- 202. Lunt DW, Rose AG. Pathology of the human heart in drowning. Arch Pathol Lab Med 1987 ; 111:939-42. Orlowski JP. Drowning, near-drowning, and ice-water submersions. Pediatr Ciil1 North Am 1987; 34:75-92. Pearn J. Aquatics for ep il eptic chi ldren. Aust Poediatr] 1982;
6:29-31.
Nixon J, Pearn J, Wilkey [, Corcoran A. 15 years of child drownin g. A 1967-81 analysis of all fata l cases from the Brisbane Drowning Study and an J I-year study of consecutive near-drowning cases. Accident Anal Prev 1986 ; 18:199- 20 3. Pearn JH, Wong RYR, Brow n J et ai. Drowning and near-drowning involving child ren. A five-year tota l population study from the City and County of Honolulu. Am] Publ Hlth 1979 ; 69:450-4. Royal Life Saving Society Australia. Th e National Drowning Report 2001. Sydney: Royal Life Saving Society Austra lia, 2001, pp. 1-7.
18:255-6.
Pearn J. Drowning and alcohol. Med] Aust 1984; 141 :6-7. Pearn J. Bart R, Yamaoka R. Drown ing risks to epileptic children. A study from Hawaii. 8M] 1978 ; 4:1284-5. Wi lkinson JL. Sudden cardiac death in childhood and adolescence. ] Paedintr Child Health 1994 ; 30:384-5.
I
CHAPTER 19
I
SUDDEN DEATH OF CHILDREN IN HOSPITAL Jem Berry
Introduction Definition and frequency Deaths due to natural disease Deaths due to failure to monitor Therapeutic misadventures Deaths due to drug treatment Deaths due to medical devices and procedures Deaths in the dental chair
362 362 363 365 366 366 368 371
INTRODUCTION
Investigation of the sudden an d unex pected death of a baby or child in hos pital is a challenge that requires the combined skills of paediatric and forensic pathology. The death may be due either to the condition for which the patient was bein g treated or to so me other unreco g nized condition. Sometimes death is due to a complication of treatment or failure of a medical device. Occasionally, sud den death is due to an error by professiona l staff. Rarely, death is due to the deliberate act of a carer. Sta ff are often upset and understandably defensive, but it is the pathologi st's role to tease out the facts of what hap pened in a non-judgemental manner so that parents receive a full account of their child's death, and if errors have occurred then the hospital is able to lea rn from them. A pathologist is an advocate for the child and his or her parents, not the insti tution. A post-mortem report indicating medical error seldom results in a coronial findin g of negligence, but the penalties for failing to highlight poor practice may be heavy. The possible causes of sudden unexpected death in hos pital are even more numerous than the many diseases, drugs and procedures th at children en counter th ere. No attempt will be made to describe every eventuality that can lead to sudden death in hospital. The ca uses are broadly grouped in Table 19.1. Deaths in ' the dental chair are not
Sudden death in newborn babies Accidents Suicide Filicide and homicide in hospital
371 372 373 373
Investigation of sudden unexpected death of children in hospital References
375 377
strictly hospital death s, but because they share many fea tures with other anaesthetic de aths they are incl uded in this chapter. The causes of sudden death of newborn babies in hospital are different from those of older babi es a nd chil dren and so will be discussed separately. Sudden hospital deaths pose interesting lega l questions. [s death due to a recognized complication of treatmen t for a lethal disease, natural or unnatural? When is an error negligent? When is an error so serious that the increasingly us ed criminal charge of manslaughter is appropriate? Whatever the answers to these questions, the post-mo rtem report is likel y to be closely scrutinized , and so the investi gation and post-mortem examination must be canied out to the highest stan dards. Studies have shown the value of autopsy examina tion in revealing major unexpected find in gs t hat might have affected outcome in 9 per cent of chi l dren from paediatric inten sive care units (P[CUs), 7 per cent from emergency departments, 2 per cent from gene ral wards,I,2 25 per cent of pa ediatric oncology patients,J and 8.5 per cent of paediatric cardiac patients.4 ,5
DEFINITION AND FREQUENCY The definition of sudden and unexpected death of children in a hospital context is necessarily subjective because these
Deaths: natural diesase I Table 19.1
Causes ofsudden death ofchildren in hospital
Due to natural disease Un recogn ized natural disease Complications of known natural disease Haemorrhage Septic shock Embolism Cerebral oedema Ca rdi ac dysrhythmia Other Failure to monitor
Th erapeutic misadventure Adverse drug events Errors of drug, dose or rou te of administration Anaphylaxis Other adverse reactions Complications of medical devices and procedures Dea ths associated with surgery Surgical dea ths Anaesthetic deaths
Accidents unrelated to medical care Suicide Murder Filicide Homicide by carers and others
children are usually already suffering from sign ificant ill n ess. Buchino et al 6 considered 'sudden' to be death within 1 hour of clinically apparen t distress excluding time spent in resuscitation, and 'u nexpected' to mean that no matter how ill the child, death was no t considered imminent. They esti mated that a children's hospital with 200 beds wo uld have two to four such sudden and unexpected deaths per year. Li fe -threatening collapse is surprisingly common in specialist children's hospitals. Two per cent of more than 6000 children admitted to a I22-bed university children's hospita l during a I2-month period received cardiopul monary resuscitation; a susta ined circulation was restored in two-thirds of these children, 33 per cent were still alive at 24 hours and 15 per cent at 1 year. The most common causes of cardiopulmonary arrest, which was not always unexpected, were respiratory fa ilure and shock. 7
DEATHS DUE TO NATURAL DISEASE Sudden unexpected death in h osp ital may be entirely nat ural. either as a result of a complication of an already diag nosed disease or due to a separate unrecognized disease. In either case, post-mortem exami nation plays a major role in determining the cause of death. There are ma ny possible complicatio ns of natural disease that can cause sudden unexpected death, but a limited number of final common pathways including haemorrhage,
363
septic shock, embolism, cerebra l oedem a and cardiac dys rhythmia. For exa mple, Manto n et al8 described two child ren who died suddenly and unexpectedly in hospital with haemolytic uraemic syndrome complicated respectively by septicaemi a and haemorrhag ic cerebral infarction. Perhaps surprisingly, unexpected hospital deaths from conditions associated with su dden death in the community such as asthm a 9 and epilepsyl° appear to be rare. This is probab ly because sudden death in these two conditions is associated with poor disease control and inadequate management of the acute attack, both of which are more likely to occur out side hospital. Similarly, the 'dead in bed ' syndrome in young persons with type 1 diabetes ge nerally occurs at home and may be due to untreated hypoglycaemia or ketoacidosis. I I - I)
Fatal Haemorrhage Fatal haemorrhage as a complication of an underlying con dition may be overt or if it is internal it may be completely unrecognized prior to autopsy. Haemorrhage can occur in the course of haematological disorders, such as acute leukaemia, or can fo llow erosion of a major vessel by an abs cess, tumour, fore ign bodyl4 or vasculitis. IS Spontaneo us rupture of the spleen with ensuing haemorrhage typically occu rs in infectious mononucleosis, but ca n occur in acute leuk aem ia,16 malaria 17 and other conditions where the spleen is enlarged . Rarely death is due to concealed bleeding into a large tumour. 18 Haemorrhage in hospital is usually treatable but may be ovelwhelming, occasionally, resulting in sudden death.
Septic Shock Bacteraemia and fungaemia may complicate such recognized infections as those of the urinary tract, meningitis, endo card itis, gastroenteritis, pneumonia, osteomyelitis, infective arthritis or they may arise from a n unknown cause. There may be predisposing factors such as chemotherapy, steroid therapy, cyclic neutropen ia, HN infection, extensive burns, eczema or aspleni a . Indwelling vascular catheters and pro longed infusions are other strong associations. The overall mortality in one study of bacteraemia and fungaemia in childhood was 19 per cent, a nd risk factors for death included neutropenia, hospital-acquired infection, polymicrobial infection and ineffective empirical antibiotic treatment. The Single strongest association was septic shock, WHich had a mortality of 60 per cent. 19 Post-mortem diagnosis rests on suspecting the diagnosis, recognizing the source a nd co llecting fresh bacterial cultures from several sites, especia lly a blood culture and a swab of sp leen. Some treatments such as steroids can completely suppress the symptoms of disseminated infection in children and so cultures must be a routine part of the examination of any child who dies suddenly and unexpectedly. 20
364 I
Sudden death in hospital
Toxic shock syndrome occurs in children with burns 21 and after even minor surgical procedures. It has recently been recognized as a complication of varicella infection and is sometimes fatal. 22 25
Fatal Embolism in Childhood Massive pulmonary thromboembolism in childhood is rare, but lesser emboli are more common than is generally recog nized. 2G ,27 A study at the Hospital for Sick children Toronto found only eight cases of massive pulmonary thrombo embolism causing sudden death in 17500 autopsies (0.05 per cent),27 but Buck et al 26 found pulmonary emboli in approx imately 1 per cent of paediatric autopsies and considered that it contributed to death in 31 per cent of these cases. Pulmonary thromboembolism may complicate prolonged immobilization, sepsis, dehydration,28 cardiac surgery, 29 cardiomyopathy, nephrotic syndrome,30-33 haematological disorders with enhanced coagulation, occult malignancy and large vascular malformations. 34 -36 Vascular malforma tions may also be the explanation for reports of pulmonary embolism in conditions such as Proteus syndrome. 37 In chil dren, pulmonary emboli may arise from cardiac, mesenteric or cerebral veins and from the superior and inferior vena cava, as welJ as from pelvic and leg veins, which are the usual sources in adults. Vascular catheters are a potent source of thromboemboli, which may be infected. Inherited thrombophilias such as deficiency of protein C, protein S or the presence of a Iupus anticoagulant are very common in children with thrombotic events, whether or not there is another condition predisposing to the thrombosis. 3B Paradoxical emboli or quite small emboli arising in the left side of the heart, for example in bacterial endocarditis and rheumatic fever,39 can cause sudden death by coronary artery occlusion and will be missed unless the coronary arteries are carefully examined. Fatal coronary artery embolism is also described in newborns. 40 .41 Tumour embolism typically occurs in nephroblastoma, either spontaneously42 or as a complication of surgery,43.44 but may also complicate other childhood renal tumours and tumours with large vein involvement. 45 .4G Fat embolism occurs after trauma , including non accidental injury47 and orthopaedic procedures such as scoliosis surgery.48 Circulating fat is frequently found in venous blood after fractures and bone operations without apparent ill effects, but once cerebral and cardiovascular symptoms have appeared the mortality approaches 90 per cent. 49 Reports describe sudden death due to fat embolism after closed limb lengthening and also after accidental fractures during manipulation of the hips in cerebral palsy, muscular dystrophy and severe dermatomyositis. 49 - 52 Fatal fat embolism appears to be a particular risk for such chil dren who have severely porotic bones and expanded marrow cavities as a result of immobilization or steroid therapy.
Fat embolism occurs in sickle cell disease, and may be a cause of the acute chest syndrome in that condition. 53-55 Gray et al 56 cited a 9-year-old boy with sickle cell disease and acute chest syndrome who was found dead in his hospital bed. A related disorder occurs from lipid overload during intravenous feeding, causing respiratory compromise, hepatosplenomegaly, abnormal liver function and coagu Jopathy.57.58 The condition is generally reversible, but may be fatal. Several studies have demonstrated intravascular lipid in the Iungs post mortem in infants who had been receiving parenteral lipid emulsion, but its significance in the absence of clinical signs is not known. 59- 62 Fatal lipid accumulation in the brain has also been reported. 63 Salmon pink discolouration of the spleen at autopsy is a clue to lipid overload - the fat can be stained using Sudan black in routinely processed paraffin sections. However, this syn drome should be rare with modern preparations of lipid and clinical awareness of the hazard. Gas embolism is mentioned here for completeness, but is usually a therapeutic misadventure resulting from catheter accidents,64 open heart surgery, mechanical ventilation 55 or neurosurgely, especially in the head-up position. Small amounts of venous air embolism are seldom fatal, but much smaller volumes of air are necessary to cause death in infants than in adults and older children. Venous air embolism has been demonstrated in more than 80 per cent of children undergoing craniosynostosis repair, although it resulted in hypotension in only one-third of cases and there were no deaths in this study66 Topical application, acciden tal ingestion or infusion of hydrogen peroxide can result in fatal embolism of oxygen produced by the action of blood and tissue catalase.57 ,58 Buchino et al 69 describe how a nurse mistook a central arterial catheter for a feeding tube and injected a bolus of air with fatal consequences. Cardiovascu lar gas may be seen by computerized tomography (CT) after failed cardiopulmonary resuscitation, and is thought to be introduced during venous catheterization. 7o Diagnosis of fatal gas embolism depends on prompt autopsy before oxygen can be absorbed,?1 radiology, and correct technique including inspection of the large veins for bubbles, aspiration of the heart in situ under water, and examination of the cerebral vessels before removal of the brain.
Cerebral Oedema Cerebral oedema may complicate diffuse hypoxic-':ischaemic brain damage due to any cause, and so may be erroneously assumed to be the consequence of cardiorespiratory arrest rather than its cause. However, it may be a very significant finding in children who have collapsed suddenly and without obvious explanation. Cerebral oedema is a common concomitant of severe head injury, but it also occurs occasionally in young
Deaths: failure to monitor I
children following relatively minor head injuIY and can cause unexpected death after a symptom-free interval. 72 The mechanism of delayed cerebral oedema in children is unknown, but in rare older patients it has been associated with familial hemiplegic migraine. 7J Delayed deterioration after head injury may also occur as a result of the develop ment of subdural or epidural haematoma .74 The effect of space-occupying lesions such as tumours and abscesses may be exacerbated by local cerebral oedema, and they may cause sudden death in hospital from acute hydrocephalus, brain swelling and coning 75,76 Cerebral oedema is a well-recognized but poorly under stood complication of the treatment of ketoacidosis in young children with diabetes in whom it is the most com mon cause of death. 12 ,77-79 Cerebral oedema develops in about 1 per cent of paediatric cases of diabetic ketoacido sis despite what is considered 'best practice' management, and in one study had a mortality of 21 per cent. 79 About one-half of these deaths follow sudden respiratory arrest in hospital. Traditionally, the development of cerebral oedema has been attributed to rapid changes in the blood glucose level and administration of intravenous fluid, causing osmotic swelling of the brain. A case-control study did not confirm these associations and showed hypocarbia, raised serum urea and administration of bicarbonate to be risk factors. 78 Evidence for and against is discussed by Glaser 80 Post-mortem examination occasionally shows that acute deterioration of consciousness and death in childhood dia betic ketoacidosis is due to other cerebral catastrophes, such as subarachnoid haemorrhage, basilar aliery thrombosis or multiple thrombi. 81 Cerebral oedema also occurs in other metabolic disorders such as acute liver failure in which it is a major cause of death,82-84 and in some inherited metabolic diseases such as maple syrup urine disease when it can cause sudden death in hospital. 85 Cerebral oedema should be sought in any hospitalized child who dies after an unexpected deterioration in con scious level. Signs of herniation and coning are a better guide than brain weight, and in babies these signs may dis appear if the brain is put to one side before examination.
Cardiac Dysrhythmia Cardiac arrest is the dramatic common endpoint of numerous processes leading to hypoxia or electrolyte imbalance. Some times it is due to primary disease of the heali. The manage ment of cardiac arrhythmias in children is not always straightfolVVard and adverse events resulting in death or brain damage may become the subject of litigation. 86 Sudden death from cardiac dysrhythmia can occur in myocarditis, cardiomyopathies, myocardial infarction, con genital heart disease, and in abnormalities of the conducting system and electrolyte imbalance. Unrecognized myocarditis can cause sudden death in patients who are hospitalized
-
365
with other illnesses87 and during anaesthesia,88 Myocardial infarction is more common in childhood than is often appre ciated,89.9o occurring in congenital heart disease, anomalous origin of the coronary arteries, cardiomyopathy, as a result of cardiac surgery, as a complication of myocarditis and vas culitis such as Kawasaki's disease and in cystic t1brosis. 91 Sudden death may follow unrecognized cardiac contusion due to trauma. Histiocytoid cardiomyopathy affects infants and young children, especially girls, and, characteristically, causes recurrent arrhythmias or sudden death, sometimes in hospitaJ. 92 .93 Dysrhythmias and cardiac conduction defects are common features of some of the disorders of fatty acid oxidation, but not typically medium-chain acyl Co-A dehy drogenase deficiency.94 Recurrent dysrhythmia as in the long QT syndrome may be mistaken for breath-holding or epilepsy95 and the author has encountered an example of this latter situation causing sudden unexpected death in hos pital. Sudden death in anorexia nervosa may be due, inter alia, to sudden dysrhythmia, and has been associated with acquired prolongation and dispersal of the QT interval. 96 - 98 The cause of cardiac arrest may be immediately apparent at the time of post-mortem examination. If it is not and death appears to have been due to cardiac dysrhythmia then the heart should be retained with the necessary authoriza tion for further detailed examination of the myocardium, conducting system and coronary arteries. Appropriate sam ples should be sent for virology, vitreous electrolyte values, and retained for possible metabolic and toxicological stud ies. The genetic basis of many of the inherited defects of cardiac conduction is now becoming understood, and retained genetic samples may sometimes prove as useful as examination of the heart itself. 99
DEATHS DUE TO FAILURE TO MONITOR It may not always be easy for a pathologist to recognize that there has been a failure to anticipate complications and adequately monitor a sick child, because this some times requires intimate knowledge of standards of clinical care. Failure to monitor is not always due to negligence rather it may be due to lack of facilities or equipment fail ure. As a rule of thumb, when intensive care facilities are adequate an acutely ill child should not die on a general paediatric ward. Failure may involve simple observations such as pulse, temperature, blood pressure, level of consciousness and fluid balance so that a patient's gradual deterioration is not appreciated. Failure may also involve more complex mon itoring, for example of blood gases, electrocardiogram or intracranial pressure. Many routine treatments require monitoring using laboratory tests such as electrolyte values during intravenous fluid therapy. In a retrospective review of case notes, Arieff et al 100
described hyponatraemia due to hypotonic infusions after surgery in 0.34 per cent of children, with a consequent
366 I
Sudden death in hospital
mortality of 8.4 per cent. In a 6-year prospective study they also identified 16 children who were electively hosp i talized for prima ry care and who developed severe sympto matic hyp onatra em ia. Leth argy, headache, nausea and sudden respiratory a rrest occurred from 3 to 120 hours after hypotonic fluid administration, at which time the average serum sod ium level had fallen from 138 to 115 mmol/L. Out of t hese children 10 died, 5 were left in a vegetative state, a nd 1 became mentall y handicapped; all 16 of them had radiological or post-morte m signs of cere bral oedema. When a child suddenly dete riorates and is found to have a low plasma sodium level, ano ther consideration is the syndrome of inappropriate antidiuretic hormone (ADH) secretio n. This can foll ow head injury and other intracranial patholo gy, neurosurgery including spin al surgery, pneumo nia , in tens ive chemotherapy, bone marrow transplantation, anti-epileptic and psychotropic drugs, social use of t he drug 'ecstasy' and many other conditions, and is occasionally fatal. The key to unravelling this and other electro lyte dis orders is expert chart review, supported by measure ments of post -mortem body weight and the osmo lality/el ectrolyte levels of vitreous humour, pl asma and urine. Sometimes, hospital staff do n ot recognize how ill a child has becom e until too late. Exa mples in myex pelience includ e a newb orn baby who died suddenly in hospital of fluid and electro lyte imb alan ce as a result of undiagnosed duod enal atresia, an institu tionalized child w ith epilepsy wh o died un exp ectedly from undi agnosed acute .li ver fail ure att ributable to sodium valproate, an adolescen t thought to have gastroenteritis but who died of undia gnosed Addi son's disease (a similar case has bee n reported by others), 101 and a child being treated for a head injury who died of bleeding from an unreco gn ized splenic rupture. It is now accepted th at so-called delayed splenic rupture is more often due to delayed dia gnosis rather than delayed haem orrhage. 102 ,103 Careful chart review may clarify the sequence of events lead ing to death , but it is importan t not to include con te nti ous judgements best left to others in the post-mortem report.
DEATHS DUE TO DRUG TREATMENT
Death may occur as a result of g iv ing the wro ng dru g, errors in dosage, an inapp ropliate route of admin istrat ion, or adverse reactions to the medicat ion. Adverse reacti ons to drugs may be du e to an excessive therapeutic effect (toxic ity), in to lerance, side-effects, idiosyncrat ic reactions and allergy or drug in teractio ns. Severe adverse reactions within the scop e of this chapter are likely to be due to toxicity, idiosyncratic reactio ns, anaphylaxis or drug interactions.
The Wrong Drug or the Wrong Route of Administration Medication is occasionally given by an inappropri ate route with disastrous consequences. Sudden death h as fo llowed intravenous administration of hy pelio ni c sa line,105 hydro gen peroxide, G8 and connection of an oxygen cylinder to an intravenous camtl a.lO GAlcohol injected in mistake for saline into a line misplaced in the subclavian artery caused the death of a newborn baby. lol Narsinghani et al l08 describe the life - threaten ing effect of infusion of polyethylene glycol electrolyte solution via a misplaced naso gastric t ube into the trachea. Deaths have also followed errors in in travenous adm inistration of potassium chloride in children just as in adul ts. 109 In Phil adelphia, three neonates died when nurses flu shed their vascular access cathe ters wi th dilu te heparin that pharmacy personnel h ad accidentally prepared with bulk potassium chloride sol ution instead of 5 per cent dextrose. llo Multiple incid ents of inadvertent intrathecal injectio n of v incristine or daunorubicin, when mistaken for methotrexate, have been repo rted." I - 1I7 These usually lead to a fatal myel oe ncephalopat hy desp ite intens ive salvage th erapy, but not su dden death. Some incidents have resulted in climinal prosecutions for manslaug hter and so these deaths are of particular forens ic importance." 8,11 9 Cases in which doctors h ave been accused of manslaugh te r involv ing children are sh own in Table 19.2.
Toxicity Due to the Wrong Dosage THERAPEUTIC MISADVENTURES Complications of care are an important cause of morbidity and mortality in hos pitalised patients. A stud y of 1035 consecutive admissions to a PICU showed 2.7 complica tio ns as a resu lt of medical care per 100 PlCU days; overall, 42 per cent were rated as major a nd 36 per cent involved human error. Complications were ventil ator- , drug- and procedure-related, involving infectious or involved inva sive devices, mostly vascular catheters, and affected 8 pe r cent of all patie nts admitted to the PICU . Cardiopulmonary resuscitation was required in six patients and two deaths were directl y due to complications of treatment. 104
Mista kes in dosage are common in paediatrics because of errors in calculating and dispensing the correct dos e fo r body weight in very small babies. 120- 122 This is exacerb ated by the necessity of using adult formulations not packaged in paediatric doses ' off-label' because they are not licensed for use in children. 123 Errors are more likely ' in very ill patients, very small babies, in em ergency department or in tens ive care se ttings, an d by trainee medical staff. 124 In a study of 10 778 medicatio n orders there were errors in 5.7 per cent, and 26 adverse drug events of which five (19 per cent) were co nsidered preventable. 120 In their study Kozer et a l 125 reported 20 errors by a factor of 10 in 22 500 doses in a tertiary children's hospital, potentially resulting in
• -
-
---
-
Deaths: drug treatment I
Table 19.2 Year
367
Doctors occust:d of monsloughtt:r involving childrt:n following drug or onot:stht:tic t:rrors: UK 1970-99
Patient's age Place of death in years (sex)
9 (male)
1974
Hospital
1978
4 (male)
Hosp ital
1991
16 (male)
Hospita l
1994
1 (female)
Paediatric department
1994
9 (male)
GP surgery
1999
14 (male)
Den tist's surgery
1999
12 (male)
Paediatric department
Nature of problem
Outcome of case
Nasotracheal tube inse rted orally
Convicted
and kinked during appendectomy Intravenous dose of methotrexate given intracerebrally Vincristine given intrathecally instead of intravenously Over-correction fo llowing insulin hypog Iycaem ia test Diamorphine given instead of dihydrocodeine during circumcision Nitrous oxide given instead of oxygen postoperatively (wrong connection) Vincristine given intrathecally instead of intraveno usly
Acquitted Two doctors convicted but acqu itted on appeal Case abandoned at committal Pleaded guilty Convic ted No evidence offered
From ref. 119.
death or life-threatening toxic effects in 15 children. Koren et al 121 describe the sudden death of a newborn baby due to a tenfold error in a digoxin injection, and cite serious effects from tenfold overdoses of phenobarbitone, pal1 curonium and salb utamol. Buchino et al 6 quote the case of a 4-month-old infant who received about 10 times the intended dose of ketamine because the verbal order was given in cubic centimetres rather than milligrams. Medica tion errors in paediatrics have been reviewed by Sullivan and Buchino. 126
Adverse Reactions to Drugs Sudden deaths due to correctly administered drugs are uncommon in infancy and childhood. However, infants may react in ways not seen in adults because of differences in drug metabolism. Immature drug metabolism explains the grey baby syndrome due to chloramphenicol in neonates, and probably why children who are under 3 years old are particularly susceptible to sodium valproate-induced hepatotoxicity.) 27 Dysrhythmias can occur during treatment with many drugs including tricyclic antidepressants and as a complica tion of cardiotoxic drugS. 128 ,129 For exam ple, dysrhythmia may follow the intravenous administration of cardiotoxic chemotherapy such as doxorubicin. Although the great majority of dysrhythmias in this situation are benign, sudden deaths have been reported both acutely and as a late effect of such therapy,130-1 33 Tumour lysis syndrome caused by rapid destruction of tumour by chemotherapy may also cause sudden death by hyperkalaemia-induced dysrhythmia,1 34,1J5 Cisapride, widely used for the treatment of gastroesophageal reflux, has recently been implicated in sudden deaths due to dysrhythmias,136 although the relationship has been
chaJlenged, 1)7, 138 Dysrhythmia and even death has occurred in children und ergoing chemical skin peeling using phenol. 139 Respiratory depression caused by sedation used for minor procedures also causes sudden death in children, Adverse outcomes from sedation were more likely in non hospital settings, using multiple agents and when there was no monitoring by pulse oximetry, There was no relationship to the type of sedative or the route of administration, 140 A syndrome of metabolic acidosis, lipaemia, rhabdomyolysis and fatal myocardial failure has been linked to long-term infusion of propofol for sedation of children in intensive care units,141 - 143 It has recentl y been suggested that this syndro me is due to impaired fatty acid oxidation caused by the drug,144,145
Anaphylactic Reactions to Drugs
----------------
Fatal anap hylactic reactions are very uncommon in children. Pumphreyl46 estima ted that there are about 20 anaphylactic deaths of all ages per year in the UK, of which one-half are iatrogenic. A study of anaphylactic reactions in children gave a fatality rate of 2 per cent. 147 Most anaphylactic reac tions in children occur out of hospital and are due to food,148 Anaphylactic reactions in children may be biphasic in 6 per cent of cases, leading to relapse after apparently sllccessful treatment. 147,1 49 Almost any drug given by any route can cause a poten tially fatal anaphylactic or anaphylactoid reaction, even something as apparently innocuous as a bovine gelatine containing chloral hydra te suppository, 150 Other causes include vaccines, 151, 152 anti-venom,1 53 leukocyte infusions,154 skin prick tests, 155-157 cryoprecipitate, 158, 159 gammaglobulin,160 anaesthetic agents, radiological contrast medium,161 ,162 and parenteral alimentation,1 63 Children with chronic illnesses
368 I
Sudden death in hospital
and those requiring multiple procedures including bladder catheterization may develop hypersensitivities. For example, anaphylax is on exposure to latex and ethylene oxide occurs in spina bifida 164,165 and hypersensitivity to ethylene oxide used for sterilization has been reported in renal dialysis patients. 166,167 Children with tumours may become sensitized to chemotherapeutic agents such as asparaginase.'68.169 Patients with cystic fibrosis may become sensitized to anti biotics, pancreatic extract and vascular catheters. 170, 171 Post-mortem findings in anap hylaxis are often non specific, but in children may include oedema of the respi ratory tract and asthma-like changes in t he lungs. 146 .172 , I73 Deaths due to an ap hylaxis are very rare, but failure to diagnose anaphylaxis post mortem is often due to failure to think of the diagnosis and retain the appropriate specimens for measure ment of specific immunoglobulin E (IgEl and mast cell tryptase, altho ugh the latter may give both false positive and false-negative results. 173 - 17 6
unless the device is very close.186- '90 Two-way radios used by emergency crews are more likely to interfere with med ical equipment than cell phones. '91 Addi tio nal hazards of infusion include administration of the w rong fluid 105 or fluid contaminated by bacteria. 192 Culture negative intra venous fluids contaminated with endotoxin caused the death of 36 neonates in an incident in Brazil. I93 Blood transfusion can be complicated by sudden dea t h due to transfusion reactio ns, either as a result of faulty grou ping and cross-matching or administration of blood intended for another patient. Patients who have received multiple transfusions of blood products are particularly at risk of adverse reactions to further tra nsfusions. Frontela et al 194 describe an immunofluorescent technique developed to establish the diagnosis of mismatched blood transfusion using paraffin-embedded tissue from a child at more than 2 years after death . Sudd en death has followed exchange transfusion usin g blood damaged by overheating.
DEATHS DUE TO MEDICAL DEVICES AND PROCEDURES
Nasogastric and Orogastric Tubes
Vascular Catheters and Infusions The insertion of central venous lines is associated with signif icant mortality and morbidity at all ages, despite the use of softer catheters and attempts to educate doctors about the risks. m ,178 Sudden death may occur as a result of cardiac tamponade by blood, intraveno us fluid or parenteral alimen tation fluid, or due to pneumothorax, hydrothorax or haemothorax. I07 Sudden death has also been caused by dys rhythmia provoked by a guide wire 107 and fatal coronary sinus thrombosis du e to a misplaced central venous catheter. 179 Symptoms may occur almost immediately after the catheter is inserted or be delayed. They are so frequently miSinterpreted that tamponade carries a high mortality. 180 The frequency of pericardial effusion or tamponade from long lines in neonates found in a UK-wide survey was 1.8 per 1000 lines with a mor tality rate of 0.7 per 1000 lines. Two-thirds of the fatal cases were not diagnosed until post-mortem examinat ion. 181 In neonates the catheter does not necessarily have to perforate the atria l or ventricular myocardium to produce tamponade, fluid being abl e to permeate the loose cardiac muscle fibres if the tip becomes wedged by thrombus.182-184 The location of catheters should be documented before autopsy by routine radiology, and the identity of unexpected fluid in the pericar dial sac or pleural cavity should be confirmed by chemical analysis. Complications of intravascular catheters in neonatal intensive care have recently been reviewed. 185 Small children are particularly vulnerable to fluid over load if an infusion device fails or is incorrectly set and delivers perhaps a day's fluid in just a few minutes. Mobile phones have been said to interfere with a number of med ical devices including infusion pumps, but the risk for cur rent generation phones has probably been exaggerated
There are few reports of death following attemp ted gastric intub atio n in children despite the potential hazards of acci den ta l tracheal intubation 108 and creating false passages. A nasogastric tube that entered the cranial cavity via a frac ture was probab ly a peri mortem event, rather than the cause of death.'95 Gastric rupture has been described both from gastric intuba tio n and washout l96 and after emesis induced with ipecac. 197 Saline used as an emetic or for gastric lavage can cause fatal salt poisoning.198.1 99 Induced em esis is no longer recommended as a routine treat ment for poisoning in children. 20o
Endotracheal Tubes and Tracheostomies Unexpected death can occur as a result of fail ed tracheal intu bation, accidental placement of the tube in the oesophagus, blockage of the tube with mucus, accidental extubation 201 or rupture of the trachea. 202,2o3 In one stu dy in a PICU the rate of accidental extubation was 1.4 per 100 ventilator days. 104 Tra cheal tubes may become displaced into a tracheo-oesophageal fistula and cause sudden collapse during surgery.69 Tra cheostomy tubes are also prone to blockage and displacement, which very rarely cause death even in hospi ta l.204-2o9
Lumbar Puncture Sudden de ath may occur as a result of cerebellar herniation and brainstem compression if lumbar puncture is carried out in t he presence of raised intracran ial pressure. Tbis was a particular hazard when lumbar puncture was carried out for the diagnosis of bacteria l meningitis, but this is no longer recommended as a routine emergency investigation.210-2 12
-
--
Deaths: medical devices and procedue s I
Conin g in meningitis may sti ll occur despite a noml al CT scan 21J or without lumbar puncture. 214 Herniation may be reversed by vigorous treatment with hyperventilation and m annitol, a nd so it may be hard to confirm post mortem unless there a re irreversible signs, such as necrosis of the cerebella r tonsi ls. 2l1 Coning may also cause sudden death in hospi tal when there is an undiagn osed space occupying intracrani al les ion, such as a tumour or abscess,l5,76 and has been describ ed in childhood acute leukaemia with central nervous system involvement. 21 5
Ventriculoperitoneal and Ventriculoatrial Shunts Shunt failure in treated hydrocephalus has been a significa nt problem and continued to be so in the 1990s.216 Children with shunted hydrocephalus may die sudden ly, sometimes during in vest igation in hospital for a pparently minor symp toms.217 Acute deterioration due to obstruction of the sh u nt is often preceded by a history of headaches, and the child may be found de ad in bed.216 However, symptoms may be non-specific, and shunt failure is not invariably accompa nied by v entricu lar enlargement. 218 Death can also occur due to recurrent pulmonary embolism from vent riculo atrial shunts a nd is usually heralded by respiratory sym ptoms or pulmonary hypertension .217 •219 A ventriculo atria l shunt has caused coronary sinus thrombosis, myocardial infarction and fatal coll apse in hospital. 22o Byard 221 describes sudd en death in a child with a ventriculoperitoneal shunt that had perforated the transverse colon, leading to meningitis. Post-mortem examinatio n should include careful ex am inatio n of the s ite of th e shunt for swellin g - both ends of the sh unt in situ and the va lve for signs of obstruction, examin ation of any intravascular component for t hrombus and careful cultures to rule out infection. The help of a neurosurgeon may be necessary to fully eval uate the func tion of the shunt and valve.
Other Devices Pacemaker failure is a rare cause of sudden death in hospi tal but in all cases when a pacemaker is present it should be saved, with its lead intact for expert ex am in ation, and the site of the tip should be examin ed histologically.222-225 It is important not to let the bare tip of the wire touch the bare body of the dev ice because electrical settings may be lost.
Deaths Associated with Surgery and Anaesthesia Children are affected by the same surgical and anaesthetic accidents as adults but have a higher operative mortality because their sma ll size and immature physiological
-
~
369
responses allow a smaller margin for error. Problems posed by their vulnerability to heat loss, narrow a irway, small extracellular vo lume and age-depe nd ent responses to anaes thetic agents make paediatric a naesthesia a field for special ists. Paediatric surgery often involves children with multiple complex problems, thus addin g to the operative risk. Un expected deaths associated with surgery may be due to the patient's original condition. In that case questions arise about the appro priateness of the decision to operate, and the adequacy of attempts to prepare a nd stabilize the child for surgery. The 1999 report of the UK Nationa l Confidential Enquiry into Perioperative Deaths found that most dea ths within 30 days of sur gery in children were in those with congenital anomalies, necrotising enterocolitis, tumours or trauma. These children gene rally had respira tory or cardiovascu lar problems in addition to their under lying diagnosis prior to surgery. In total, 16 per cent of deaths took place on the day of surgery. 226 Lack of experi ence and inad equate superv ision of junior staff are recu r rent themes in studies of anaesthetic deaths. When unexpected death is due to the surgery itself, the cause is often obvious at post-mortem examination. In con trast, anaesthetic deaths a re usually du e to respiratory failure or cardiac arrest, whic h may leave no trace, and so the ca use often ca nnot be determined by autopsy examination alone. Pathologists cannot have the expertise to assess all aspects of an anaesthetic and should seek independent expert advice.
Deaths Due to Surgery Deaths due to the surgery itself are either due to general haz ards of surgery, such as acute haemorrhage, or specific com plications of pal1icular operations beyond tile scope of this chapter. Out of 289 paediatlic pelioperative cardiac arrests, 48 per cent were considered to be unrelated to tile anaes tIletic; of these, 22 per cent were cardiac surgery patients who could not be weaned off by-pass, anotller 17 per cent suffered ca rdiac arrest as a result of u ncontroll able surgica l haemor rhage, 22 per cent had miscellaneous ca rdiovascular eve nts, 4 per cent had air embolism and 4 per cent had complicatio ns of central venous li nes.227 Complications due to vascular cath eters inserted prior to the commencement of surgery are similar to those already described (see 'Vascula r catheters and infusions', above), but misplaced catheters may have disas trous effects duling surgery if tIley cause intrave nous anaes thetic agents or resuscitative fluids to be delive red into the wrong compartment. Cardiac arrest may follow su rgical traction o~ the viscera (the viscerocardiac reflex) or inte rference with the eye (the occulocardiac reflex). 228 Bietti 229 reported such a case of a child operated on for recess io n of the medial rectus w ho s uf fered a fatal cardiac arrest. Pathologists investigating indi vidual cases should not accept such possible explanations uncri tica lly, and w itllout excluding every other possi ble ca use o f death.
370 I
Sudden death in hospital
Examination of children who die after cardiac surgery requires specialist knowledge of cardiac morphology and the surgical techniques used to treat individual malformations. Cardiac catheterization in infants and children carries a small (about 0.14 per cent overall) risk of death, both from diagnos tic and interventional procedures, as a result of complications such as perforation of cardiac chambers and bradycar dia. 2JO-232 As a general rule, surgical mortality increases with the length of time on cardiopulmonary by-pass, and a long by-pass time is often an indication that problems were encountered during the surgery. Difficulty weaning the patient off the pump usually reflects myocardial damage due to prolonged cardioplegia,233 although if the patient dies intraoperatively or soon postoperatively there may be no morphological evidence of myocardial necrosis. Myocardial necrosis may also be present preoperatively in congenital heart disease. The risks of sutures obstructing branches of the coronary al1eries or impinging on the conducting system are common to several procedures. Allwork 234 gives helpful advice about likely surgical difficulties encountered in spe cific ca rdiac operations and their recognition post mortem.
Complications of Anaesthesia In one US teaching hospital 40 per cent of all anaesthesia related perioperative cardiac arrests were medication related events, 20 per cent due to complications associated with central venous access, 20 per cent were attributed to problems with airway management, and 13 per cent were due to unknown or possibly vagal reaction. The risk of death related to anaesthesia-attributable perioperative car diac arrest was 0.55 per ]0 000 anaesthetics. 235 A study of infants and children found an anaesthetic-related cardiac arrest rate of 1.4 per 10 000 anaesthetics, with 55 per cent of all arrests occurring in babies who were under] year of age and medication and cardiac causes accounting for 69 per cent of all cases. Cardiovascular depression from halothane, alone or in combination with other drugs, was responsible for two-thirds of all medication-related cardiac arrests 22 7 In a study of over 700 000 anaesthetics, Morita et al 236 found the incidence of critical peri operative events such as cardiac arrest and severe hypoxaemia to be greatest in babies who were less than 1 month old, and that these babies also had the highest incidence of critical events related to anaesthetic management. However, although their mortality during and within 7 days of anaesthesia was also the highest of any age group, this was predominantly due to coexisting disease, and no death was attributable to anaesthesia in children under 5 years of age. Causes of anaesthetic deaths in older age groups included anaesthetic overdose, toxic effects of local anaesthetics, improper airway management, incom patible blood transfusion and errors in spinal anaesthesia. Another study found 'major' perioperative events, which included cardiac arrests and death, to be most frequent in babies who were less than 1 month of age, although these
were the sickest group.237 The remarkable safety of paedi atric anaesthesia in a specialist setting is shown by a report of 24165 paediatric anaesthetics in which there was a rela tively high incidence of adverse events in neonates, but no anaesthetic-related deaths in any age group.2JB Cardiac arrest during anaesthesia or surgery is most often due to hypoxia, which has many causes. Induction of anaes thesia may be complicated by broncho- and laryngospasm, failed intubation, or leakage around the endotracheal tube, causing inadequate ventilation. The tube may be kinked, or become blocked by a defective inflatable cuff or mucus. Accidental oesophageal intubation is not necessarily fatal unless spontaneous respiration is prevented by muscle relax ants, and may sometimes be suspected from erosions in the oesophageal mucosa even when the tube has been removed post mortem. Delayed detection of accidental oesophageal intubation is rare, but in one study was more common in infants and in association with emergency surgery.239 Many anaesthetic agents cause respiratory depression, either in therapeutic doses or in over-dosage. Pulmonary aspiration of gastric contents is a much feared complication of general anaesthesia. However, it occurred in only 0.04 per cent of 56] 38 consecutive patients who were less than 18 years old and who under went 63 180 general anaesthetics for procedures performed in all surgical specialities at the Mayo Clinic over a period of 12 years. The risk of aspiration was highest in emer gency procedures, and those involving children who were younger than 3 years of age with bowel obstruction or ileus. Of the 24 patients who aspirated, only three required mechanical ventilation and none died .240 Faulty anaesthetic equipment can cause fatal hypoxia. Incorrect plumbing of piped gases, failure of the oxygen supply, incorrect f10w rates or leaks in the anaesthetic cir cuit, faulty connections and accidental disconnections have all caused unexpected anaesthetic deaths. Cardiac arrest may also result from dysrhythmias due to anaesthetic agents such as halothane, electrolyte abnor malities or vagal inhibition caused by stimulation of the respiratory tract by irritant gases or passage of an endo tracheal tube. Ventricular fibrillation results from release of adrenaline in patients who are inadequately sedated. Collapse or death during anaesthesia and surgery in chil dren may be the first manifestation of an unsuspected inher ited or developmental abnormality. Death has been caused by perioperative fasting in a child with unrecognized very long-chain acyl Co-A dehydrogenase deficiency2 4J Anaes thesia may precipitate cardiac arrest in children with unrecognized congenital myopathy242-244 or cardiomyo pathy.227.245 Cucchiaro and Rhodes 246 reported a 9-year-old boy who was found to have long QT syndrome after he suffered ventricular fibrillation and cardiac arrest following accidental intravascular injection of bupivocaine. Surgery is a potential hazard in children with sickle cell disease.247 McGarry and Duncan 248 described four sudden deaths asso ciated with surgery due to red blood cell sickling in children
Sudden death in the newborn I
with undiagnosed sickle ceLl trait. Acute intermittent por phyria can present for the first time in childhood as a result of an anaesthetic and cause death.249 Cardiac abnormalities have recently been highlighted as a cause of otherwise unex plained sudden deaths during anaesthesia and surgery.250 Malignant hyperthermia is a condition characterized by sudden onset of tachycardia, acidosis, muscle stiffness, and a rapid rise in temperature, which, in susceptible individuals, is precipitated by suxamethonium and certain inhalation anaesthetics, including halothane. It is inherited as an auto somal dominant gene with variable penetrance, and recent work has shown mutations of the RYRl gene in 30-50 per cent of susceptible subjects. 251 The incidence of malignant hyperthermia in children is greater than that in adults, per haps because of the anaesthetic agents used. Dantrolene is a specific treatment. 252 At post-mortem examination, the mus cles may be paJe. Microscopy shows focal muscle necrosis and myoglobin in renal tubules, which may be necrotic. 253 ,254 The fulminant form still occurs, although it represents a fail ure of diagnosis and treatment. 255 Anaphylactoid reactions to neuromuscular blocking agents are quite distinct from malig nant hyperthermia, but may also be fata1. 256 Deaths in the recovery room are uncommon and are likely to be due to the delayed effects of surgery such as haemorrhage or failure to reverse the anaesthetic, resulting in respiratory depression. The airway may become obstructed by mucus, blood clot or a forgotten throat pack. About 1 in 1500 children have an inherited deficiency of cholinesterase, which greatly prolongs the effect of the muscle relaxants succinyl choline and mivacurium, thus delaying the onset of spontaneous respiration after anaes thesia. 257 ,258 Postopera tive cardiac arrests in recovery or the intensive care unit are rarely fata1. 259 Excessive volumes of intravenous fluid cause postoper ative pulmonary oedema, which most commonly presents without warning as cardiorespiratory arrest with normal electrolyte values. Post-mortem examination shows pul monary oedema only, with no other cause of death.260 Postoperative hyponatraemia and cardiorespiratory arrest due to brain swelling is often caused by administration of intravenous fluids that are low in sodium, possibly exacer bated by inappropriate secretion of ADH, and autopsy shows cerebral oedema. 100,261 A meta-analysis confirmed the association with hypotonic solutions,262 and a survey of anaesthetists has shown that children in the UK are still at risk of this complication, especially in non-specialist hospitals. 263 Irrigation of the operative site can lead to sub stantial unrecognized absorption of fluid.
DEATHS IN THE DENTAL CHAIR Most deaths in the dental chair are anaesthetic related. The public expectation of general anaesthesia as an option for minor dentistry and the historical willingness of UK dentists to administer anaesthetics while also carrying out
371
dental procedures resulted in 100 deaths in the UK from 1970 to 1979 264 and 42 from 1980 to 1989,265 one-half of which occurred in children. 266 The number of dental anaesthetics administered in surgeries and those given by operator anaesthetists has declined as a result of a flurry of reports and stricter guidelines. 267 ,268 In the late 1990s approxi mately 300000 patients underwent general anaesthesia for minor dental procedures, with around 2- J deaths per year, usually in fit young adults or children who presented with no particular anaesthetic risk,269 Historically, about one-half of deaths in the dental chair were due to respiratory causes and one-half were due to sudden cardiac arrest. Causes included unobserved respira tory depression due to premedication and multiple anaes thetic agents, sometimes in excessive doses, Cardiac arrest, due to either hypoxia or anaesthetics, such as halothane, is hard to manage in a dentist's surgery. The use of local anaesthetics in dental surgery, as in general surgery, can cause collapse and even death due to rapid absorption, overdose, intravascular injection or ana phylaxis. The frequent simultaneous injection of a vaso constrictor such as adrenaline can also cause adverse effects. Accidental over-dosage with local anaesthetic is particularly easy in children, especially in those who are undergoing multiple procedures on the same occasion. This possibility is exacerbated by the use of multiple agents with names ending in '-caine' requiring different doses. Overdose of local anaesthetic can be fatal in both children and adults and causes nausea, anxiety, excitement, vomit ing, convulsions and cardiorespiratory arrest. 270 Most cases of 'anaphylaxis' that are attributed to local anaesthetic agents are in fact probably due to inadvertent intravascu lar injection. Fortunately, death in the dental chair has become rare since stringent standards have been set for the administra tion of general anaesthetics in dentists' surgeries, but each is particularly tragic because it affects a fit child undergoing relatively trivial procedures. Post-mortem examination should include not only toxicological studies, but also the possibility of unrecognized cardiac, skeletal muscle or meta bolic disorders that were exacerbated by the anaesthetic. Very rarely, oral surgery itself may cause death. Fatal intracranial haemorrhage has followed accidental perfora tion of the base of the skull during surgery on the temporo mandibular joint in a child.271
SUDDEN DEATH IN NEWBORN BABIES A detailed discussion of the causes and investigation of sudden unexpected deaths in the early neonatal period is beyond the scope of this chapter and is covered in standard texts. 272 ,273 As most such deaths will turn out to be due to natural causes they are best dealt with by perinatal pathol ogists, with a forensic pathologist in attendance only if there are forensic concerns.
372 I
Sudden death in hospital
There have been few systematic studies of the incidence and causes of sudden death in the first week of life. Early neonatal sudden hospital death of newborns considered pre viously healthy acco unted for 11 per cent of total neonatal mortality in a Swedish study;274 a more recent study gave the rate of 'early neonatal sudden death syndrome' as 0.14 in 1000 live births, with approximately one-third of infants dying within the first hour after birth.275 Quite similar fig ures were found in a French study that noted that many deaths occurred at night, and that some babies had signs of illness that were not recognized or acted upon. 276 Sudden death of newborn babies, either in special care baby units or maternity wards, is usually due to age group specific causes, or those related to labour, delivery and the transition from intrauterine to extrauterine life. In term babies such likely causes include delayed effects of birth asphyxia or trauma, infection, genetic metabolic disease and malformations, especially ductus-dependent cardiac malformations such as hypoplastic left heart syndrome. 277 In pre-term babies who are receiving intensive care, addi tional consideratio ns are complications of prematurity and adverse effects of procedures and treatment.64.278 Early neonatal death due to frank birth trauma is now rare, thankfully, but minor trauma such as a fractured clavi cle, excess moulding of the skull or minor tearing of the falx may indicate that the delive ry was not quite as simple as the clinical records suggest. Unrecognized birth asphyxia may result in sudden unex pected death in the 'lying-in ' ward. Most fatal malformations are easily recognized at post mortem examination but careful technique is necessary if subtler anom alies, such as those of the upper airway 279 and larynx, coronary arteries or pulmonary venous drainage, are not to be overlooked. Minor atrial septal defects a nd probe patency of the ductus arteriosus are not causes of death in the neonatal period. Infection may be acquired before, during or after birth . The effects of infection are exacerbated by lack of mater nally acquired specific antibodies and immaturity of the immune system in newborns. Bacterial infection is tradi tionally divided into early and late sepsis, the former prin cipally caused by enterovaginal organisms and the latter by nosocomially acquired bacteria. Group B streptococcal infection in newborns is usually an example of the former, and can cause sudden collapse and death with only mini mal signs at post-mortem examination. Viruses such as herpes simplex and the enteroviruses cause devastating disease and sudden death in newborn babies, especially when they have no passively acquired maternal antibodies. Severe infection in small babies is not always accompanied by symptoms obvious to the new mother. Numerous inborn errors of metabolism can cause death in the neonatal period and are listed in appropriate text books. z8o Some, such as galactosaemia, are precipitated by the intro duction of particular comp ounds into the diet. It is worth noting that medium-chain acyl Co-A dehydrogenase deficiency (MCAD), which is familiar as a cause of sudden
death later in infancy, may also cause collapse and death in the early newborn period. Sudden unexpec ted death in hospitalized infan ts with bronchopulmonary dysplasia after prolonged mechanical ven tilation has been reported as a significant cause of late mortality, despite appropriate monitoring and prompt car diopulmonary resuscitation. 281 When a newborn infant is found dead in hospital in bed with its mother then the possibility of accidental asphyxia arises. A familiar scenario is tha t of a mother, perhaps exhausted by childbirth, who falls asleep while breast feeding her new baby and wakes up to find it lifeless. 282 Byard and Burne1l 283 have demonstrated the ease with which the huma n breast can cause accidental asphyxial episodes, but such a possibility should be approached with extreme sensitivity in the context of a neona tal death. Similarly, the possibility of neonaticide has far-reaching consequences, both for the mother and hospital staff, and should be approached with care. Neonaticide is more com mon outside hospital but remains a consideration when there is no medical explan atio n for a newborn's death in hospital , and there are other social and medical pointers such as concealment or denial of pregnancy. Although no explanation may be found in a small pro portion of babies who die suddenly and unexpectedly in the early neonatal period, such unexplained cases are exceptional in the practices of specialist perinatal patholo gists. Careful review of the pregnancy and labour, com bined with a detailed post-mortem examination, including appropriate laboratory investigation and examination of the placenta, will provide an explanation in most cases.
ACCIDENTS Fatal, non-medical accidents are seldom reported in hospi talized children and are probably rare, although patients and visitors are exposed to many of the same hazards as exist at home. For example, a 21-month old child was elec trocuted in hospital by an uncovered lamp switch which he took into his mouth,284 and an infant aspirated the nipple of a makeshift pacifier in a hospital nursery.28S A study of eight hospitals, includin g specialist children's hospitals, over an 18-month period reported 781 non iatrogenic accidents involving patients and visitors under 16 years of age. Overall, 41 per cent of accidents to inpa tients happened when their parents were present. Most involved falls, slips, striking, and scalds; three were seri ous, resulting in fractures or bruising in children with bleeding diatheses 286 There were no fatalities, although one entrapment was potentially lethal. Falls from beds and cots in hospital usually result in only minor injuries even though beds are often higher than those at home, although there are occasional skull fractures.287.288 Toys are second only to beds and cots as a cause of equipment related hospital accidents, and attention has been drawn to
Filicide/homicid e in hospital I
th e dan ger to other children of toys brought into the hos pital by visitors. 284 Finberg et af89 described the death of 6 out of 14 infants in a hospital nursery, who were accidentally poi soned when their formula feed was made up using salt instead of sugar. 289 ,290 The most severely affected survivo r had a serum sodium of 274 mmol/L and made a good recovery. A similar episode occurred in Austria in which five newborn babies died. 291
SUICIDE Of the 30 000 suicides per year of all ages in th e USA, 5-6 per cent occur in hospitals. 292 Suicidal ideation and action s are well recog nized in children and yo ung people in psychiatric wards, a lthough actual suicid es app ear to be uncommon. 293 ,294 Noren et al 295 describe a 15-year-old mental health patient who jumped from the third floor of a hospital and was found to have incidental lymphocytic myocarditis at post-mo rtem exa mina tion. Children with chronic or terminal diseases such as cystic fibrosis may occasionally take their own lives. Repo rts of such deaths in hospital are rare, but the possibili ty should be kep t in mind. Bya rd reported 296 the case of a 14-year-old girl with cystic fibrosis who injected hair conditioner via an intravenous infusion pump intended for systemic antibiotics, although suicidal intent was not proven in this case.
FILICIDE AND HOMICIDE IN HOSPITAL Babi es and children in hospital may be harmed by their paren ts, hospi tal staff or visitors. The motive may be uncl ear, but examples include mercy killing, abuse due to Mun chausen 's syndrom e by proxy, and frank psychosis . Children may be admitted to hospital as a result of deliberate harm that mayor may not have been recognized for what it is, such as physical injury by a parent or simu lated illness. Physical abuse may then continue in hospital, and has been clearly documented by covert video su rveil lance.297 Rarely, a paren t will go to extraordinary lengths to cause real or simulated illness in their child in hospital, using methods such as suffocation , administration of drugs and poisons, tampering with samples, interference with equipm ent and introducing infected material into infusion fluid s.298-301 It has been said that the techniques used are limited only by the intell igence or imagin ation of the per petrator. Some of these methods are capable of resulting in sudden unexpected death, but it is believed th at this is not usually the parent's objective and that such deaths in hos pital are rare. Very rarely a parent is implicated in the 'mercy ki lling' of a severely handicapped child . The scope for filicid e in hosp ital has been increased by the greater availability of rooms in which parents can stay overnight with their child.
373
In recent years it has been recogni zed that rare devian t medical professional s may target vulnerab le pati ents caus ing serial collapses and deaths. Frank discussion of this topic is hampered by the historical difficulty of obtaining sustainable conviction s. How ever, children have fea tured prominently amon g these so-called carer- assoc iated serial killings (CASKs) because, like the elderly and tho se in intensive care units, they are vulnerable and un able to speak for themselves. Some cases invo lv in g children are de ta iled in Table 19.3. In a comprehensive review on which much of this account is based, Forrest 302 esti mated the incidence of CASKs to be approximately 1 per million carers per year. Nurses are the professional group most often involved, per haps because they are the most numerous; however, there are many instances of murderous doctors. 303 Park and Kh an 304 list 13 cases involving nurses from the UK and USA between 1975 and 1998. Perpetrators may be particularly attracted to medical work because of the opportunities it affords, and may be motivated by a desire to control events or to show themselves in a good light in emergency si tua tions. In one adult series the motive appears to have been as trivial as wishing to win an unofficial sweepstake predicting the time when patients would die. These deaths are extremely difficult to recognize because medical professionals often have knowledge of, and access to, methods that are almost undetectable. The victims are pati ents a nd so post-mortem examination shows natural disease that might have caused death and does not ra ise concern. The alarm is usually raised by cli nical staff, but typically not until the number of patients affec ted by col lapses or death has reached double fig ures. The trigger may be the rec ognition of a train of unusual events or increased mortality, or a member of staff may become suspicious after witn essing some strange behaviour by the perpetrator. Yorker3 05 suggested a number of clues to early recognition: • a significant increase in cardiopulmonary arrests, or deaths or both , particularly in patients not tho ught to be in immedi ate danger; • an unusually high rate of successfu l resuscita tion; • multiple events in the same patient; • events occurring more often in a particular shift. It is notable that in so me cases, the staff themselves have id entified a 'death shift' or a particular member of sta ff involved in a large number of events, but their con cerns have not been taken serio usl y by senior personnel. As soon as con cerns are raised they must be shared with senior medical and mana geria l staff, w ho will carry out further investigations and take the decision to inform the authorities. It would be wise to draw on the experience of independent experts and those who have investigated sim il a r cases before at an early stage. Possi ble action s that may be taken when such a death occurs are shown in Table 19.4. The over-riding immediate priority after protecting other children is to secure the
374 I
Sudden death in hospital
Table 19.3
Some well-documen ted corer-associated serial killings involving children Location
Setting
Likely agent(s)
Deaths
Outcome
Prelimin ary hearing 313
Toro nto
Paediatric cardiac ward of children's hos pital
Dig ox in
Uncertain; increased mortality (x 4) on unit over 9 months; nurse charged with four murders
Case against nurse Susan Nelles dismissed No proceedings taken aga inst another known only as nurse 'A' Roya l Commission found that out of 33 deaths 8 were certainly due to digoxin and 15 others were suspicious
Texas v. Genene Jones3 14
San Antonio
Paediatric inten sive ca re unit and paedi at ric pri vate practice
Hep arin, suxa methonium
Su bsta nti a I exc ess mortality in PICU; charged with one murde r an d one assault
R v. AllitP15
Grantham, UK
Paed iat ri c wa rd s of general hospital
Lign oca ine, in sulin , potassium ch loride
Cas.e
Table 19.4 Some immediate actions that may be taken when a child dies suddenly and unexpectedly in hospital in suspicious circumstances to preserve evidence prior to the post -mortem examination Control and limit access to the room Institute log of perso ns entering the room Secure all samples and equipment (includes sharps containers and waste) Ensure that no items on body are removed (vascular lines, end otrach eal tubes, pacemakers, etc.) and postpone 'la st offices' To preserve settings and memory, do not switch off electronic equ ipment Secure clinical records, fluid charts, drug records and records of resuscitation Ask all staff present at the event to make an immediate detailed written record of the event Notify laboratories to save all earlier samp les (haematology, blood bank, chem istry, microb iology, etc.) until notified otherwise Keep any videotape from security cameras and any op erative procedures Inform coron er, medical director, management or police as approp riate Consider taking immedi ate sa mpl es, subject to coroner's permission
scene, sampl es and equipm ent, to prev ent anyo ne remov
ing any tubes or other equipment from the body, and to record who was present and invi te them to write down their recollection of events as so on as possible. The imp or tance of preserving sa mples can not be overemphasized.
Charged with four murders and nine assau lts
Convicted and sentenced to 99 yea rs and 60 years,
respect ive ly, on each cou nt Convicted Clothier Inquiry316
Vital information may be obtained from used syringes, 'empty' ampoules, and discarded samples for bl ood gas analysis, all of wh ich may have to be recovered from the sharps bin, taki ng appropriate precautions against injury. Old samples already in the laborato ry and those from other patients who may have been affected must also be secured against being routinely disca rd ed. Additional samples must be taken at post-mo rtem examination including tissue from any possible injection sites. Toxicological analysis will be informed by circum stances, also remembering that some of the agents used in prev ious cases, such as insulin and heparin, are no t included in a routine toxicological screen . Some cases have involved inj ection of air. Confinnation that serial killin gs have occurred and iden tification of the perpetrator often depend heavily on ep idemi ologica l evidence of increased death rates in a particular part of a unit at particular times when an individual was known to be on du ty (so me drugs with a delayed action, such as digoxin, may be used to produce clinical deterioration and death after the perpetrator has left the wa rd). When finally identified, the perpetrato r is often found to have history of medical and employmellt problems. Such is the difficulty of identifying these individuals that they might never be discov ered if they move around. In retrospect, some are found to have offended before in another centre. Prevention of such rare events as CASKs is 'extremely difficult. The Clothier report into the case of Beverley AJlitt noted that 'a determi ned and secret criminal may defeat the best regulated organization in the pursuit of his or her pur pose '. Nevertheless, sou nd practices such as daily ch ecking of ward drug stocks, ce ntral preparation of drugs adminis tered by intravenous infusi on, exclusion of ampoules of potassium chloride from general wards, the use of two
-
----~ -~ .
-
Investigation of sudden unexpected death in hospital I
witnesses for all injections and the use of swipe cards to limit access to critical areas, not only reduces the omnipresent risk of accidents, but also makes the very rare occurrence of malfiscence more difficu lt. James and Leadbeatter 306 suggest that hospital policies could include: • Detailed recording of incidents of unexpected collapse to include those present at or around the time of collapse. • Taking and storing of a blood sample following such a collapse for possible future analysis. • Regular mortality review including analysis of place of death and persons in atten dance. • Detailed monitoring of potentially harmful drugs on wards (drugs such as insulin, potassium chloride and other arrhythmogenic drugs should be monitored as closely as controlled drugs). Pathologists may contribute to early detection by being scrupulous in assigning causes of death. It was notable in the Allitt case that 'status asthmaticus' and 'status epilepti cus' were used in circumstances in which they did not really fit the facts. In another case sudden infant death syndrome (SmS) was used inappropriately.
INVESTIGATION OF SUDDEN UNEXPECTED DEATH OF CHILDREN IN HOSPITAL When an infant or child dies suddenly and unexpectedly in hospital, parents will be understandably distressed and may be angry. Hospital staff will also be upset and sometimes defensive. In the UK all such deaths should be reported to the coroner or procurator fiscal. Although it is often easier for pathologists to suggest that the post-mortem examination should be carried out by someone from outside the institu tion, an in-house pathologist is more likely to obtain full clinical cooperation and wi ll be better placed to provide feedback to the parents and the hospital. When there is real concern about serious error or negligence then it is wise to seek an independent pathologist. The investigation follows the same pattern as in adults, but with additional considerations due to young age. In par ticular, small children are more likely to suffer accidental over-dosage of drugs, and unexpected deaths may be due to unrecognized congenital or inherited conditions, such as inborn errors of metabolism, myopathy and cardiac conduc tion defects exacerbated by therapy for another condition. The vulnerability of small children occasionally makes it nec essary to think the unthinkable and consider deliberate harm. The scope of the examination varies depending on whether a natural cause for death is very likely, an accident is suspected or there is suspicion of foul play. Deaths clearly due to natural causes can be deal t with in the usual way, but suspicion of accident or other unnatural death should trigger a graded response adopting some or all of the measures in Tab le 19.4, depending on the degree of suspicion. This
375
judgement requires a very detailed clinical history and it is important to speak to several members of staff before accepting the history at face value. The parents' account of events may con tain important details unknown, overlooked or even suppressed by staff. Start and Cross 307 provided general guidance for pathological investigation of deaths following surgery, anaesthesia and medical procedures. Pathologists seldom have the opportunity to carry out a scene investigation with the body still in situ and so the best chance of establishing what went wrong may be lost long before he or she is involved if the hospital does not have procedures in place for securing specimens and equipment after any sudden and unexpected death. The natural urge to clear up the scene and complete 'last offices' should be resisted in favour of a careful appraisal of the scene by a senior member of staff. Where indica ted, used syringes, ampoules, infusion solutions and blood bags should be preserved for possible labora tory analysis. Waste bins and sharps containers may contain useful samples, but should be searched with great care to avo id injury. Infusion pumps, temporary pacemakers, monitors, ventilators, anaest hetic machines and other equipment should be set as ide for testing, but e lectroni c equipment should not be switched off or unplugged until any settings or memory have been checked by an expert . All tubes and catheters should be left in the body, the reasons for so doing care fully explained to the parents. Staff should be asked to write down their recollection of events, who was present, t he position of the body as it was found and other details. Th e post-mortem examination itself should only take place when there is a full history and, if the cause of death is truly unknown must be a hybrid of the best of paediatric and forensic practice (Table 19.5). It is essential that the pat holog ist is fully inform ed about the child's illness and any procedures undergone. In the past, coro ner's autopsies have been carried out by pathologists who seemed unaware of the complexities surrounding sudden death in the par ticular disease concerned. 56 A clinician who looked after the child in life should be present to explain treatment, procedures and any special con ce rns. In surgical deaths the anaesthetist should be consulted and the surgeon should attend in order to explain the procedure and ass ist (but not carry out) the examination of the surgica l site. Radiographs obtained soon afte r death are necessary to document a ny gas in vessels, and are useful as an objective re cord of t ubes and catheters. The value of routine photo gra phs to supplement a full externa l examination and description may only become apparent in retrospect. How ever, photographs are not a substitute for a cohtempo rane ous written and drawn record on a pre-printed body chart (those used in clinical departments for recording injuries to children are suitable). All punctures and injection sites should be recorded with an estimate of their age. Standard measurements ane! body weight should be measured accurately. When compared with the chi ld 's we ight on admission to hospital, the post-mortem body weight is
376 I
Sudden death in hospital
Table 19.5
A checklist for post-mortem examination of a child who has died suddenly and unexpectedly in hospital
Before the event Check that the institution has a suitable protocol to be followed by staff in the event of sudden unexpected death approved by the coroner/procurator fiscal.
Before the post-mortem examination Ensure that relevant equipment and ante-m ortem samples have been secured Obtain a full clinical history Decide whether the exam ination wou ld be better carried out by or with another pathologist (forensic, sub - specia list, independent, double doctor. etc.) Refresh own knowledge about clinical aspects of the child's disease, its patho logy, complications and treatment Discuss the case with relevant clinicians (paediatrician, surgeon, anaestheti st, etc.) Obtain parents' account if present before or at the time of death Discuss relevant samples with , for example, the clinical biochemist or toxicologist in ad vance Consider the need to take immediate samples Obtain radiographs within 12 hours of death if gas embolism is a possibility
During the post-mortem examination Document all tubes, lines and catheters (whole body radiograph s, photograph s, drawings, text) Document all puncture sites and incisions Document any injuries Record the body weight and standard measurements Full post-mortem with particular care to exclude and record: Pneumothora x
Gas embolism
Thromboembolism
Cerebral oedema and herniation
Exa mine any operation site with the surgeon present Save any implanted devices Save appropriate samples: Histo logy of all major organs, any lesion and operative site
Vitreous humour for electrolytes, and glucose if hyperglycaemia relevant
Culture blood, spleen, lung, cerebrospinal fluid, any infective lesion, thrombus on catheter tips, etc. for bacteria
Consider virolog ica l cu ltures
Frozen tissue samples, e.g. skeletal muscle, kidney, myocardium, liver [possible enzyme, muscle or toxicological studies)
Fibroblast culture or other source of DNA
Frozen lung if fat embolism possible
Serum, urine, stomach contents, bile, injection sites, etc., for possible toxicology as advised by local toxicolog ist
Serum for mast cell tryptase if anaphylaxis possible
Consider saving heart and/or brain for specialist examination
After the post-mortem examination Discuss findings with coroner/procu rator fi scal and clinicians Give cause of death when possible Inform relatives of tissue, organs and samples retained, and any progress towards finding the cause of death, perhaps via the legal authority Attend or instigate an institutional death review meeting (Have there been simi lar deaths, and what is the mortality rate of that particular clinical unit7) Ensure that any adverse event has been appropriately documented and reported If the cause of death remains unknown, then seek an independent detailed clinical review or take other appropriate action to exclude adverse event, error, negligence or malfiscence
sometimes a useful clue to excess or inadequate adminis tration of intravenous fluids. The body should be opened with care, noting the posi tion and condition of the tips of all tubes and catheters. The
examin ation will include all body cavities (any excess fluid should be measured an d saved for analysis and comparison with infusion fluids) an d all major organs. Special atten tion should be taken during dissect ion to exclude gas
•
-
References I
embolism, pneumothorax and cerebral oedema. Trauma, such as chest wall contusion, retroperitoneal haematoma, stomach perforation and even rib fractures may sometimes be attributable to vigorous resuscitation. 308 ,309 Subdural haemorrhage may be enco untered at post-mortem exami nation if there has been unsu ccessful open heart massage during card iac surgery. In addition to standard histological sampling, it is nec essalY to sample any lesion and any operative site. Snap frozen samples of urine, serum, li ver, myocardium, skeletal muscle an d kidney will prove valu ab le if a n inherited metabolic disorder or myopathy becomes a consid eration late r, bu t when metabolic disease is suspec ted it is adv is ab le to con sul t a clinical biochemist over sa mplin g in advance and to save a source of DNA. Bacteriological samples should include blood and cere brospinal fluid as a minimum. Vitreous humour biochem istry may point to an unexpected electrolyte disturbance. Samples must be saved for toxicology in case a drug error is subsequently suspected. Sub-optimal specimens hampered one study of anaesthetic-related deaths an d so a standard set of samples that will include blood, urine and gastric contents should be agreed in advance with the lo cal toxicologist. Samples taken in life an d already sto red in the labo ra tory may prove invaluabl e for metabolic, toxicological and serological examination . In problematic cases the patholo gist must take steps to secure these from routine disp osal. If the death remains unexplained after the gross post mortem examination and was consistent with sudden car diac arrest then the heart should be saved with authorization or, at the very least, extensively sampled, with the conduct ing system saved for possible serial sectioning and samples saved for 'molecular autopsy'.99 Any implanted pacemaker should be submitted for testing with its leads, although in a series of 38 surgically induced cases of complete heart block in children treated by permanent pacemaker implantation there were no hospital deaths.223 If a cerebral cause of sud den death is possible from the history then the fresh brain should be examined closely for fla tten ing of gyri, herniation and coning and authorization sought for fixation and formal neuropathological examination. Ifthe cause of death is still undetermin ed when all inves tigations are complete then it is often appropriate to seek an independent clinicopathological review within t he institu tion in the first instance. It is not acceptable or safe to ascribe death to the patient's original illness if it does not fit the facts. The question may arise whether SID S can occur in a hospitalized patient. Peterson et a1310 calculated that if the risk of SIDS is the same for hospitalized in fants as fo r those in the community then SIDS would be rare even in a busy children 's hospital. However, they suggested that the rate might be higher in sick infants, citing cases that we might find hard to accept as SIDS today. 'Undetermined' is not usually a popular conclusion, but the truth is preferable to obfuscation a nd fabrication. Inappropriate use of the term SIDS may lead later to suspicion of a 'cover up'.
377
In England an d Wales a co roner's pathologist is often asked to determine w hether the death was natural or unnatural. There is unfort una tely no definition of 'na tural causes'. Deaths after surgery or medical treatment often fall into the g rey area between natural and unnatural causes of death. 311 Coroners themselves do not agree on how to use these terms, and so it is as well to discuss borderline cases individually. A majority of coro ners will accept natural causes for deaths following appropriate med ica l interve n tion for natural disease from which the patient was likely to die, provided tha t there was no element of negligence and the relatives do not object.312 Great care should be taken in assigning the cause of death and in any accompa nying comment not to go beyond the pathologist's expert ise. A si gnificant propOJ1ion of medical negligence claims are attributable to inaccurate post-mortem diagnoses and speculative comments in post-mortem reports.
ACKNOWLEDGEMENT [ would like to acknowledge the assistance of Professor ARW Forrest with the section on carer-associated serial killings.
REFERENCES
2
3
4 5 6 7
8
9 10 II
12
Go ld stein B, Metlay L, Cox C, Rubenstein JS. Association of pre morte m di agno sis and autopsy findin gs in pediatric intensive care unit versus emergency department versus ward patients. Crit Care iVIed 1996 : 24:683-6. Stambouly JJ, Kahn E, Boxer RA. Correlation be tween clinical diagnoses and autopsy fin dings in critica ll y iLl children. Pediatrics 199 3; 92 :248-51. Koszyca B, Moore L, Toogoo d I, Byard RW Is postmortem
examinat ion usefu l in pediatric oncology? Pediatr Pat/wi
1993; 13 :709- J 5.
Gatzoulis MA, Sheppard MN, Ho SY. Value and impact of
necropsy in paediatric card iology. Heart 1996; 75:626-31.
Russell GA, Berry P J. Postmo rtem audit in a paediatric
card iol ogy unit. J Clin Path 01 1989; 42:9 J 2-18.
Buchino JJ, Corey TS, Montgo mery V. Sudden unexpected
death in hospitalized children. J Pediatr 2002; 140:46 1-5.
Reis AG, Nadkarni V, Perondi MB et a!. A prospect ive investigation into the epidemiology of in-hospital pediatric ca rdiopulmonary resuscitation usin g the internation al Utstein reporting style. Pediatrics 2002 ; 109:200-9. Manton N, Sm ith NM, Byard RW Unexpected childhood death due to hem olyti C uremic syndrom e. Am J Forensic iVIed Pathol 2000; 21 :90-2. Champ CS, Byard RW Sudden death in asthma in,childhoocl. Forensic Sci Int 1994; 66: 117-27. Breningstall GN. Mortality in pediatric epilepsy. Pediatr Neurol 2001; 25 :9-16. Dahlquist G, Kall en B. Mo rta lity in childho od -onset type I diabetes: a population-based study. Diobetes Care 2005; 28:2384 - 87. Ed ge JA, Ford-Adams ME, Dunger DB. Cau ses of death in chi ldren with insulin-dependent diabetes 1990-96. Arch Dis Child 19 99; 81:318-23.
378 I
Sudden death in hospital
13 Thordarso n H, Sovik O. Dead in bed sy ndrome in youn g diabeti c patients in Norw ay. Diab et Med 1995; J 2 :782-7. 14 Grey TC, Mittleman RE, Wetli CV, Horowitz S. Ao rtoes ophageal fistul a and sudden dea th . A rep olt of two cases and review of th e literature. Am J Fore nsic Med Fath ol 1988 ; 9:19-22. 15 Cohle SO, Colby T. Fatal hemoptysis from Behcet's disease in a child . Cardi ollClSC Pathol 2002 ; 11 :29 6-9. 16 Lash ley PM , Ramesh J , Ferdinand S. Spontaneous rupture of th e spleen. West Indian Med J 1998; 47:172-3. 17 Vidyashankar C, Basu A. Kulk arn i AR, Chou dhury RK. Spontaneous rupture of sp leen in falc iparum malaria. India'i J
Gastroentero/2003; 22:10 1- 2. 18 Koszyca B, Moore l, Byard RW. lethal manifestations of neurofibrom atos is type I in ch ildhood . Pediatr PatllOl 199 3; 13:573-81. 19 Ashkenazi S, leibovici l, Samra Z et al. Risk factors for morta lity due to bac te remia and fungemia in child hood . Clin I'~fect Dis 1992; 14:949 -51. 20 lascari AD , Pearce JM, Swanso n H. Sudden death due to disseminated cly ptococcos is in a child w ith leukem ia in remission. South Med J 1997; 90: 1253-4. 21 Brown AP, Khan K, Sinclair S. Bacterial toxico sis/toxic sho ck synd rome as a contributor to morbidity in children with burn injuries. Bums 2003; 29:733-8. 22 Doctor A, Harper MB, Fleisher GR. Group A beta- hemolytic st reptococcal bacteremia: hi sto rical overview, changing incidence, and recent association with varicell a. Pediatrics 1995; 96:4 28 -33. 23 Hidalgo-Carballal A, Suarez - Mier MP. Sudden unexpected death in a chi ld with varicell a caused by necrotizing fasc iitis and strepto co ccal toxic sho ck syndrome. Am J Forensic Med PatllOl 2006; 27 :93-6. 24 Tseng HW, Liu Cc, Wang SM et al. Complications of varicella in children: emphasis on skin and cent ral nervous system disorders. J Microbiol Immunol Infect 2000; 33:248-52. 25 Yang YJ, Liu CC, Wang SM et al. Streptococcal toxic shoc k synd rom e complicating varicella in children. J Formos Med A ssoc 1997; 96:749-53. 26 Buc k JR, Co nnors RH, Coon WW et al. Pulmonary embolism in children. J Pediotr Surg 1981; 16:385 -91. 27 Byard RW, Cutz E. Sudden a nd unexpected death in infancy and childhood due to pulm ona lY thromboembolism. An autopsy study. Arch Pathol Lab Med 1990; 114:142-4. 28 de la Grandmaison Gl, Durigon M. Pulmonary embolism: a rare cause of sudden infant death. Am J Forensic ivIed Pathol 2002; 23 :257 -9. 29 Baudet EM, Puel V, McBride JT et al. long- term resu lts of valve replacement with the St. Jude Medical pro sth esis. J Thorac Cardiovasc Surg 1995; 10 9:8 58-70. 30 Citak A, Emre S, Sairin A et al. Hemostatic problems a nd thromboembolic com pli cat ions in nephrotic ch ildren. Pediatl'
Nephrol2000; 14:138-42. 31
Jones CL, Hebert D. Pulmo nalY thrombo-embolism in the nephrotic syndrome. Pediatr Nephro! 199); 5:56-8. 32 Mehls 0, Andrassy K, Koderisch J et al. Hemostasis an d thromboembolism in children with nephrotic syndrome: differences from adults. J Pediatr 1987; 110:862-7. 33 Zimmerman Rl, Novek S, Chen JT, Roggli Y. Pulmona ry thrombosis in a lO-yea r-old child w ith minimal change disease and nephrotic syndrome. A clinical, radiologic, and pathologic correlation with literature review. Am J Ciin Patho! 1994; 101:230- 6. 34 Baskerville PA. Thromboembolic disease and congenital venous abnormalities. Phlebologie 1987; 40:531 - 6.
35 Bohm N, Bre ndlein F, Niederhoff H, Sutterle H. Diffuse hemangiomatosis of the sm all pelvis. Death from lung embo lism in an ll-yea r-old girl. Med Welt 1978; 29:839-43. 36 Machin GA, Kent S. PulmonalY thromboembolism from a large hemangioma in a 4-week-old infant. Pedjatr Pathol 198 9; 9:7 3-8. 37 Slavo tinek MI, Vacha SJ , Peters KF, Biesecker LG. Sudden death caused by pulmonary thromboembolism in Proteus syn drome. Clin Genet 2000; 58:386-9. 38 Nuss R, Hays T, Manco-Johnson M. Childhood thrombosis. Pediatrics 1995; 96:291-4. 39 Sta hl J, Santos lD, Byard RW. Coron ary artery thromboembo lism a nd un expec ted death in childhood and adolescence. J Forensic Sci 1995; 40:599-601. 40 Muraskas J, Besinge r R, Bell T et al. Perinatal myocardial infarction in a newborn with a structurally normal heart. Am J Perinatal 1997; 14:93- 7. 41 Tillett A, Hartley B, Simpson J. Paradoxical embolism causing fatal myocardial infarction in a newborn in fant. Arch Dis Child Fetal Neol1atal Ed 2001; 85:F 137 -8. 42 Zakowski MF, Edwards RH, McDonough ET. Wilms' tumor presenting as sudden death due to tumor embolism. Arch Pathol Lab Med 1990; 114:605-8. 43 Moore L. Byard RW. Fatal paradoxical emb ol ism to the left carotid artery during partial resection of Wilms' tumor. Pediatr Parhol 1992; 12:551 -6. 44 Shurin SB, Gaud erer MW, Dahms BB, Conrad WG. Fatal intraoperative pulmonary embolization of Wilms tumor. J Pediatr 198 2; 101 :559-62. 45 Arthur OS, Stephe ns CA, Brummitt WM et al. Fatal tumor embolism during examination under anesthesia. Surgery 1973 ; 74:466-8. 46 Chan KW, Fry er CJ, Fra ser GC, Dimmick JE. Sudden cerebral death in malignant presacral teratom a. Med Pediatr Oncol 1985; 13:395-7. 47 Nich ols GR, Corey TS, Davis GJ. Nonfracture-associated fatal fat embol ism in a case of child abuse. J Forel1sic Sci 1990; 35:493-9. 48 Shapiro G, Green OW, Fatica NS, Boachie-Adjei O. Medical compli ca tions in scoliosis surgery. Curl' Opin Pedi atr 2001; 13 :36-41. 49 Shulm an ST, Grossman BJ. Fat embolism in childhood. Review with report of a fatal case related to physica l therapy in a ch ild with dermatomyositis. Am J Dis Child 1970 ; 120:480-4. 50 Edw ards KJ, Cummings RJ. Fat embolism as a comp li cation of clo sed femoral shortenin g. J Pediatr Orthop 1993; 12:542-3. 51 Ganel A, Israeli A, Horoszowski H. Fatal compli cation of femoral elongation in an achondroplastic dwarf. A case repo rt. Clin Orthop 1984 ; 185:69-71. 52 Pender ES, Pollack CV Jr, Evans OB. Fat embolism syndrome in a chi ld with muscular dystrophy. J Emerg JvIed 1992; 10:705-11. 53 Horton DP, Ferriero DM, Mentze r We. Nontraumatic fat embolism sy ndrom e in s ickl e cell anem ia. Pediatr Neurol 1995; 12:77-80. 54 Vi chinsky E, William s R, Das M et al. Pulmonary fat embolism: a distinct cause of severe acute ches t syndrome in sick le cell anemia. Blood 1994; 83:3107-12. 55 Vichinsky EP, Neum ay r lD, Ea rles AN et al. Causes and outcomes of the acute chest syndrome in sickle cell disease. National Acute Chest Syndrome Study Group. N Eng! J Med 2000; 342:1855-6 5. 56 Gray A, An ionw u EN, Davies S(, Brozovic M. Patterns of mortality in sickle cell disease in the United Kingdom. J Clin Pathol 19 91; 44 :459-63. 57 Haber lM, Hawkins EP, Seilheimer OK, Saleem A. Fat overload syndrome. An autopsy study with eva luation of the coagulop athy. Am J Clin Patho! 1988; 90:22 3-7.
References I
58 Landry BA, Melhem RE. Pulmonary nodules seco ndary to total parenteral alimentation. Pediatr Radial 1989 ; 19:456-7. 59 Leve ne MI, Batisti 0, Wigglesworth JS et al. A prospective study of intrapul mon ary fat accumulation in the newborn lun g following intra lipid infusion. Acta Paediatr Scand 1984; 73 :454-60. 60 Puntis JW, Rushton DJ. Pulmonary intravasc ular lipid in neonatal necropsy specim ens. Arch Dis Child 199 1; 66: 26-8. 61 Schroder H, Paust H, Schmidt R. Pulmonary fat embo lism after intralipid therapy: a post-mortem artefact? Light and electron microscopic invest igations in low-birth-weight infants. Acta Paediatr Scand 1984; 73:461- 4. 62 Shulman RJ , Langston C, Schanler RJ. Pulmonary vascu lar lipid deposition after ad ministra tion of intravenous fat to infa nts. Pediatrics 1987 ; 79:99-102. 63 Schulz PE, Weiner SP, Haber LIvI et al. Neurological complicatio ns from fat emulsion therapy. Ann Neural 1994; 35:628-30. 64 Vald es -D ape na M. Iatrogeni c disease in th e perinatal period as see n by the pathologist. Monog r Pathol 1981 ; 22:382-418. 65 Lau KY, Lam PK. Systemic air embolism: a comp lication of ventilator therapy in hyalin e membrane disease. Ciin Radial 1991; 43:16-18. 66 Fab erowski LW, Black S, Mi ckl e JP. Inciden ce of venous air embolism during craniectomy for craniosynostos is repair. Anesthes iology 2000; 92:20-3. 67 Cina SJ , Do wns JC, Conradi SE. Hydrogen peroxide: a so urce of lethal oxygen embolism . Case report and review of the literature. Am) Forensic Med Pathol 1994; 15:44-50. 68 Lubec B, Hayn M, Denk W, Bauer G. Brain lipid peroxidation and hydroxy rad ical attack fo ll owi ng the intravenous infusion of hydrogen peroxide in an in fa nt. Free Radical Bioi Med 1996 ; 21 :2 19- 23 . 69 Buchino JJ, Kee nan WJ, Pi etsch JB et at. Malpos itioning of the endotrac heal tube in infants with tracheo esophageal fistula. ) Pediatr 1986; 109:5 24-5. 70 Shiotani S, Kohno M, Ohashi N et al. Cardiovasc ular gas on non-traumatic postmortem co mputed tomography (PMCT): the influence of card iopulm onary resuscitation. Radiat Med 2005; 23 :22 5-9. 71 Rudd PT, Wiggl esw orth JS. Oxygen embo lus during mechanical vent ilation with di sappearance of signs afte r death. Arch Dis Child 1982; 57:237-9. 72 Denton S, Mileusnic D. Delayed sudden death in an infant followin g an accide ntal fall: a case report with rev iew of the literature. Am) Forensic Med Pa thol 2003 ; 24:371-6. 73 Cox C. When ca n nurses be charged with manslaughter? Nurs Times 2001; 97 :21. 74 Hahn YS, McLone DG. Risk fact ors in the outcome of children with minor head injury. Pedia tr Neurasurg 1993 ; 19: I 35-42. 75 Ch un CH, Johnso n JD , Hofste tter M, Ra ff MJ. Brain abscess. A study of 45 co nsecuti ve cases. Medici ll e (Baltimore) 1986; 65:4 15-31. 76 Shemie S, Jay V, Rutka J, Armstrong D. Acute obstru ctiv e hydrocephalus and sudden death in children. Ann Em erg Med 1997; 29:524- 8. 77 Edge JA. Cereb ral oedema durin g treatment of diab eti c ketoacidosis: a re we any nearer finding a ca use? Diabetes Metab Res Re/J 2000 ; 16:3 16- 24. 78 Glase r N, Barnett P, McCaslin I et al. Risk factors for cerebra l edema in children with diabetic ketoacidos is. The Pedi atric Emergency Medicine Co llaborative Researc h Committee of th e American Academy of Pediatrics. N El1gl ) Med 2001; 344:264-9. 79 Marcin JP, Gl ase r N, Barnett P et al. Facto rs associated with adv erse outcomes in children with dia betic ketoacidosis related cerebral edema. ) Pediatr 2002; 141 :793-7.
379
80 Glase r N. New perspec tives on the pathogenes is of cerebra l edema co mplicati ng diabetic ketoacidosis in children. Pediatr E17do crinal Re/J 2006 ; 3:3 79-86. 81 Rose nbloom Al. Intracerebral crises during treatment of
di abetic ketoacid osis. Diabetes Care 1990; 13:22-33.
82 Alper G, Jarjour IT, Reyes JD et al. Outco me of children with cerebral edema ca used by fulminant hepat ic failure. Per/ia tr Ne ll 1'01 1998; 18 :299 -304. 83 Ee LC, Shepherd RW, Cleghorn GJ et al. Acute liver failure in chi ldre n: a regional experience.) Pa ediatr Child Health 2003; 39:107-10. 84 Liu GT, Urion DK, Volp JJ. Cerebral ede ma in acu te h epatic fai lure: clinicopathologic co rrelati on. Pediall' Neural 1993; 9:224- 6. 85 Riviel lo JJ Jr, Rezvan i I, DiGeorge AM, Foley CM. Cerebral
edema causing death in children with map le syrup urine
di sease. ) Pediatr 199 1; 119:42-5.
86 Garso n A Jr. Medicolegal problems in the management
of cardiac arrhythm ias in children. Pediatrics 1987;
79 :84-8.
87 Sm ith NM, Bourne AJ, Clapton WK, Byard RW. The spectrum of prese ntation at autopsy of myocarditis in infancy and childh ood. Pathology 1992; 24: 129- 31. 88 Critchley LA. Yet ano ther report of anesthetic death due to
unsus pe cted myo carditis.) Ciin Anesth 1997; 9:676-7.
89 Celermajer DS, Sholler GF, Howman-Giles R, Celermajer JM. Myoca rdi al infarction in childhood: clinical ana lysis of 17 cases and medium term follow up of survivors. Br Heart) 1991; 65:332- 6. 90 John srude CL, Towbin JA, Cecc hin F, Perry Jc. Postinfarction ventricular arrhythmias in children. Am Heart) 1995; 129: 1171 -7 . 91 Zebrak J, Skuza B, PogorzelskJ A et al. Partial CFIR
genotypin g an d characterisation of cystic fibro sis patients
with myo ca rdial fibrosis and necrosis. C/in Genet 2000;
57 :56-60.
92 Sheh ata 8M, Patterson K, Thomas JE et al. Histiocytoid
ca rdi omyopathy: three new cases and a review of the
literature. Pediatr De v Pat/JOI 1998; 1:56-69.
93 Stahl J, Couper RT, Byard RW. Oncocytic cardiomyopathy: a rare cause of unexpected early childhood death associated with fittin g. Med Sci Law 199 7; 37:84-7. 94 Bonnet D, Martin D, Pascale DL et a1. Arrhythmi as and
conduction defects as present in g sy mptoms of fatty acid
oxidation disord ers in children. Circulation 1999 ;
100:2 248-53.
95 Allan WC, Gosp e SM Jr. Seizures, sy nco pe, or breath-h olding presenting to th e pediatric neurol og ist - when is the etio logy a life-threatening arrhythmia? Semil1 Pediatr Neural 2005; 12:2-9. 96 Cooke RA. Chambers JB. Anorexia nervosa and the hea rt. Br ) Hasp Med 1995; 54:313 -17. 97 Swen ne I, Larsson PI. Heart risk associated with weight loss in anorexia nervosa and eating diso rders: risk factors for QTc interval prolongation and di spersion. Acta Paedia tr 1999; 88:304-9. 98 Va nn acc i A, Baronti R, Masini E et al. Anorex ia nervosa and the ri sk of sudde n death. Am ) Med 2002; 112:3 27-8. 99 Teste r DJ, Ackerm an MJ. The role of molecular a~topsy in
unexplained sudden ca rdiac death. Curr Opin Cardiol 2006;
2 J: 166-72.
100 Arieff AI, Ayus JC, Fraser CL. Hyponatraemia and death or permanent brain dama ge in healthy children. BM) 1992; 304: 1218- 22. 101 al Sab ri AM, Smith N, Busu ttil A. Sudd en death due to auto immune Addison's disease in a 12-yea r-old girl. Int) Legal Med 1997; I JO :278- 80.
380 I
Sudden death in hospital
102 Berlatzky Y, Shiloni E, AnneI' H, Weiss Y. Delayed rupture of the spleen or de layed diagnosis of the splenic injury? lsI') lVled Sci 1980; 16:659-64. 103 Olsen WR, Polley TZ Jr. A second look at delayed splenic rupture. Arch Surg 1977 ; 112:42 2- 5. 104 Sta mbouly JJ, McLaughlin LL, Mandel FS, Boxer RA. Co mplications of ca re in a ped iatric intensive ca re unit: a prospective stud y. Intensive Ca re Nled 1996; 22: 1098- 104. 10 5 Yo ung RS , Tru ax BT. Hypernatremic hemo rrh agic enceph alopathy. Ann Neurol 1979; 5:588-91. 106 Rouse DA, Hargrove R. An unusual case of gas embo lism. Am) Forensic lVled Parhol 1992; 13 :268-70. 107 Bagwell CE, Salzberg AM, So nnin o RE, Hay nes JH. Potentially letha l comp lications of central ve nous catheter placement.) Pediar Surg 2000; 35:709- 13. 108 Na rsingha ni U, Chadha M, Farrar HC, Ana nd KS. Life threatening respiratory fa ilure followin g accidental infusion of polyethylene glycol elec tro lyte so luti on into the lu ng. ) Toxicol Ciin Toxicol 2001; 39 : 105-7. 109 We therton AR, Corey TS, Buchino JJ , Burrows AM. Fata l intravenous injection of potassiu m in hosp italized patients. Am) Forensic lVled Patho12003; 24: 120 - 31. 110 Cohen MR. Potassium chloride injection mix- up. Am ) Hasp Pharm 1990; 47:2457-8. III AJ Ferayan A, Russell NA, AJ Wohaibi ]\II et al. Cerebrospinal t1uid lavage in the treatment of in advertent intrathecal vincristine injection. Childs Nerv Sysr 1999; 15 :87 - 9. 112 Alcaraz A, Rey C, Concha A, ]\Iledina A. Intrathecal vincristine: fatal myeloencephalopathy despite cerebrospi nal t1uid perfusion.) Toxicol Ciin Toxicol 2002; 40:557- 61. 113 Dettmeyer R, Dri eve r F, Becker A et al. Fatal myeloencephalop athy due to acc id ental intratheca l vincristine administration: a report of two cases. Forensic Sci Int 2001; 122 :60-4. 114 Fernandez CV, Esau R, Ham il ton D et al. Intrathecal vincristin e: an analysis of reaso ns for recurrent fata l chemotherapeutic error with reco mm end ations for prevention. ) Pediatr Hemarol Oncol 1998; 20 :587 - 90. 115 Kwack EK, Kim DJ, Park TI et al. Neural toxicity induced by accidental intrathecal vincristine administration . ) Korea n Med Sci 1999; 14 :688-9 2. 116 Seale JR. Not again! Erroneous intratheca l. inj ect ion res ul ts from a problem with protocols. BM) 200 1; 322:548. 117 Walter AW. Errors involving pediatric patients rece iving chemotherapy: a literature review.1Vfed Pediatr Oncol 1997; 29: 155-6. 118 Dyer C. Doctors face trial for manslaughter as criminal charges against doctors continue to rise. BM) 2002; 325:63. 119 Ferner RE. Med ication errors that have led to manslaughter charges. BlVl)2000; 32 1:1212 - 16. 120 Kausha l R, Bates DW, Landrigan C et al. Medication errors and adverse drug events in pediatric in patients. )AMA 2001; 205:2 11 4- 20 121 Ko ren G, Barzil ay Z, Gree nwald M. Tenfo ld errors in adm inistratio n of drug doses : a neg lec ted iatroge nic disease in pediatri cs. Pediatrics 1986; 77:848- 9. 122 Koren G, Haslam RH . Ped iatri c med ication errors: predic tin g and preventing tenfo ld disas ters. ) Ciin Pharmacal 1994; 34: 1043-5. 123 Conroy S. Unlicensed and off- label dllJg use: issues and recommendations. Paediatl' Drugs 2002; 4:353-9. 124 Kozel' E, Scolni k D, Macpherson A et al. Variables assoc iated with medication errors in pediatric emergency medicine. Pediatrics 2002; 11 0:737 - 42. 125 Kozer E, Scolnik D, Keays T et al. Large errors in the dosing of medications for chi ldren. N Engl ) Med 2002; 346:1175-6.
--------
-
- ---
-
.
-
126 Sullivan J E, Buchino JJ. Med ic ation errors in pediatrics: the octopus evading defeat.) Surg 011co12004; 88:182 - 8. 127 Johnson TN. The development of drug metabolising enzymes and their influence on the susceptib ility to adverse drug reactions in children. Toxicology 2003; 192:37- 48. 128 Swanson JR, Jones GR, Krasselt \IV et al. Death of two subjects due to imipramine and deSipramine metabolite accumulation during chronic therapy: a review of the li terature a nd possib le mechanisms.) Forellsic Sci 1997; 42 :335 - 9. 129 Varley CK. Sudden death of a child treated with imipramine. Case study. ) Child Adolesc PsychopiJarmacol2000; 10 :32 1-5. 130 Gup ta M, Thaler HT, Steinherz L. Presence of prolonged dispers ion of QT in terva ls in late surv ivo rs of childhood anthracycl ine the rapy. Pediatr Hematol Ollcol 2002; 19 :533 -42. 13 1 Krischer JP, Epste in S, Cuthbertso n DD et al. Clinica l card iotoxicity fo ll ow in g anthracycline treatment for chil dhood cancer: the Pediatric Oncology Group experience. ) Clill Ol1col 1997 ; 15: 1544-52. 132 Ma ssin MM, Dresse MF, Sc hmi tz V et al. Ac ute arrhyth mogen icity of fir st-dose chemo therap euti c age nts in chil dren. Med Pediatr 011co/2002; 39:93-8. 133 Wortman JE, Lucas VS Jr, Schuster E et al. Sudden death dUJi ng doxorub icin administration. Cal1cer 1979; 44:1588-9 1. 134 Arseneau ]C , Bagley CM, Anderson T, Canellos GP. Hyperkalaemia, a sequel to chemothera py of Burkitt 's lymphoma . Lancet 1973; 1: 10-14. 13 5 Chasty RC. Liu-Yin JA. Acute tumour lysis syndrome. Br) Hasp Med 199 3; 49 :488-92. 136 Anon. Severe cardiac arrhythmia on cisapride. Prescrire JJ1f 2000; 9: 144-5. 137 Levy J, Hayes C, Kern J et al. Does cisapride influence card ia c rhythm? Results of a United States multicenter, double-blind, placebo-controlled pediatric study. ) Pediatr Gastroellterol NutI' 2001; 32:458-63. 138 Ramirez-Mayans J, Garrido-Garcia LM, Huerta-Tecanhuey A et al. Cisapride and QTc interval in children. Pedia trics 2000; 106: 1028-30. 139 Unlu RE, Alagoz IVIS, Uysal AC et al. Phenol intoxication in a child.) CmJ1ioJac Surg 2004; 15:1010-13. 140 Cote CJ, Karl HW, Notterman DA et al. Adverse sedation events in pediatrics: analysis of medications used for sedation. Pediatrics 2000; 106 :633-44. 141 Cannon ML, Glazier SS, Bauman LA. Metabo li c ac idosis, rhabdomyolysis, and cardiovascular collapse after pro longed propofol infusion. ) NeuroSL/rg 2001; 95: 1053 - 6. 142 Parke TJ, Stevens JE, Rice AS et al. Metabolic acidosis and fatal myocardial failure after propofol infusion in children: five case reports. BM) 199 2; 305:613 - 16. 143 Vasile B, Rasulo F, Candiani A, Latronico N. The pathophysiology of propofol infusion syndrome : a simple name for a complex syndrome. Intensive Care Med 2003; 29: 1417-25. 144 Wo lf A, We ir P, Segar P et al. Impaired fatty acid oxidation in propofo l infu sion sy ndrome. Lal1cet 200 1; 357:606-7. 145 Wo lf AR, Potter F. Propofo l infusio n in ch il dren: when do es an anesthetic too l become an intensive care liability? Paediatr Al1aestil 2004; 14 :435- 0. 146 Pumphrey RS. Lesso ns for manageme nt of a nap hylax is from a study of fata l reac tions. Clin Erp Allergy 2000; 30: 11 44-50. 14 7 Lee JM. Greenes DS. Bip hasic anaphylactiC reactions in pediatrics. Pediarrics 2000; 106:762- 6. 148 Novembre E, Cianferoni A, Bernard ini R et al. Anap hy laxis in ch il dren: clinical and all ergologic features. Pediatrics 1998 ; 101 :EO.
-
"""'""
References I
149 Novembre E, Calogero C, Mori F et al. Biphasic anaphylactic reaction to ketorolac tromethamine. Int J Immul1opathol Pharmacal 2006; 19:449-50. 150 Sakaguchi M, Inouye S. Anaphylaxis to gelatin-containing rectal suppositories. J Allergy Ciin Immunol 2001; 108: 1033-4. 151 Bohlke K, Davis RL, Marcy SM et al. Risk of anaphylaxis after vaccination of children and adolescents. Pediatrics 2003; 11 2:815-20. 152 Ponvert C, Ardelean-Jaby D, Colin-Gorski AM et al. Anaphylaxis to the 23-valent pneumococcal vaccine in child: a case-control study based on immediate responses in skin tests and specific IgE determination. Vaccine 2001; 19:4588-91. 153 Singh A, Biswal N, Nalini Pet al. Acute pulmonary edema as a complication of anti-snake venom therapy. Il1dian J Pediatr 2001; 68:81-2. 154 Zylberberg R, Schott RJ, Fort J et al. Sudden death following white cell transfusion in a premature infant. J Perinatal 1987; 7:90-2. 155 Devenney I, Faith-Magnusson K. Skin prick tests may give generalized allergic reactions in infants. Ann Allergy Asthma Immunol 2000; 85:457-60. 156 Lockey RF, Benedict LM, Iurkeltaub PC, Bukantz Sc. Fatalities from immunotherapy (IT) and skin testing (SI). J Allergy Ciin Immunol 1987; 79:660-77. 157 Reid MJ, Lockey RF, Iurkeltaub PC, Platts-Mills IA. Survey of fatalities from skin testing and immunotherapy
1985-1989. J Allergy Ciin Immullol1993; 92:6-15. 158
Burman D, Hodson AK, Wood CB, Bmeton NF. Acute anaphylaxis, pulmonal)' oedema, and intravascular haemolysis due to cl)'oprecipitate. Arch Dis Child 1973; 48:483-5. 159 Harper JL, Gill ]C, Hopp RJ et al. Induction of immune tolerance in a 7-year-old hemophiliac with an anaphylactoid inhibitor. Thromb Haemost 1995; 74: 1039-41. 160 Shemin ER. Anaphylactic reaction to gamma-globulin. JAMA 1968; 203 :59. 161 Lasser EC, Lyon SG, Berl)' Cc. Reports on contrast media reactions: analysis of data from repolts to the U.S. Food and Drug Administration. Radiology 1997; 203 :605-10. 162 Shehadi WH. Death following intravascular administration of contrast media. Acta Radio/ Diagn (Stockh) 1985; 26:457-61. 163 Pomeranz S, Gimmon Z, Ben Zvi A, Katz S. Parenteral nutrition-induced anaphylaxis. J Parenteral Enteral NutI' 1987; 11:314-15. 164 Bimlingham PK, Suresh S. Latex allergy in children: diagnosis and management. Il1dial1 J Pediatr 1999; 66:717-24. 165 Moneret-Vautrin DA, Laxenaire MC, Bavoux F. AJlergic shock to latex and ethylene oxide during surgel)' for spinal bifida. Anesthesiology 1990; 7J :556-8. 166 Bommer J, Ritz E. Ethylene oxide (EIO) as a major cause of anaphylactoid reactions in dialysis (a review). Artif Organs 1987; 11:111-17. 167 Hurley RM. Anaphylaxis during hemodialysis. Int J Pediatr
Nephro11984; 5:53-4. 168 Dellinger CI, Miale ID. Comparison of anaphylactic reactions to asparaginase derived from Escherichia coli and from Erwil1ia cultures. Cancer 1976; 38:1843-6. 169 Zanotti KM, Markman M. Prevention and management of antineoplastic-induced hypersensitivity reactions. Drug Safety 2001; 24:767 -79. 170 Ganier M, Lieberman P. IgE mediated hypersensitivity to pancreatic extract (PE) in parents of cystic fibrosis (CF) children. Clil1 Allergy 1979; 9: 125-32. 171 Haworth CS, Niven RM, Moorcroft AJ et a!. Acute anaphylaxis following midline catheterisation in a patient with cystic fibrosis. Thorax 1999; 54:747.
172
381
Pumphrey RS. Fatal anaphylaxis in the UK, 1992-2001.
Nouartis Found Symp 2004; 257:116-28. 173 174
Pumphrey RS, Roberts IS. Postmortem findings after fatal anaphylactic reactions. J Clin Patho/ 2000; 53:273-6. Caughey GH. Il)'ptase genetics and anaphylaxis. J Allergy
Clin Immunol2006; 117:1411-14. 175 Horn KD, Halsey JF, Zumwalt RE. Utilization of serum tryptase and immunoglobulin E assay in the postmoltem diagnosis of anaphylaxis. Am J Forensic Med Pathol 2004; 25:37-43. 176 Yunginger JW, Nelson DR, Squillace DL et al. Laboratol)' investigation of deaths due to anaphylaxis. J Forensic Sci 1991; 36:857-65. 177 Co.llier PE, Blocker SH, Graff DM, Doyle P. Cardiac tamponade from central venous catheters. Am J Surg 1998; 176:212-14. 178 Collier PE, Goodman GB. Cardiac tamponade caused by central venous catheter perforation of the heart: a preventable complication. J Am Call Surg 1995; 181 :459-63. 179 Suarez-Penaranda JM, Rico-Boquete R, Munoz JI et al. Unexpected sudden death from coronal)' sinus thrombosis. An unusual complication of central venous catheterization. J Forensic Sci 2000; 45:920-2. 180 Pezzati M, Filippi L, Chiti G et a!. Central venous catheters and cardiac tamponade in preterm infants. Intensive Care Med 2004; 30:2253-6. 181 Beardsall K, White DK, Pinto EM, Kelsall AW. Pericardial effusion and cardiac tamponade as complications of neonatal long lines: are they really a problem? Arch Dis Child Fetal Neonatal Ed 2003; 88:F292-5. 182 Byard RW, Bourne AJ, Moore L, Little KE. Sudden death in early infancy due to delayed cardiac tamponade complicating central venous line insertion and cardiac catheterization. Arch Patitol Lab Med 1992; 116:654-6. 183 Leibovitz E, Ashkenazi A, Levin S, Nissim F. Fatal pericardial tamponade complicating total parenteral nutrition via a silastic central vein catheter. J Pediatr Gastroel1terol NutI' 1988; 7 :306-7. 184 Rogers BB, Berns SD, Maynard EC, Hansen IW. Pericarclial tamponade secondal)' to central venous catheterization and hyperalimentation in a vel)' low birthweight infant. Pediatr Pathol 1990; 10:819-23. 185 Hermansen MC, Hermansen MG. Intravascular catheter complications in the neonatal intensive care unit. Ciin
Perinatoi2005; 32:141-56. 186 Anon. Radiofrequency interference with medical devices. A technical information statement. IEEE En9 iVIed Biol Mag 1998; 17:111-14. 187 Betts IR, Simpson lAo Inhibition of temporary pacing by a mobile phone. Heart 2002; 87:130. 188 Robinson MP, Flintoft lD, Marvin AC. Interference to medical equipment from mobile phones. J Med Eng Technol 1997; 21 :141-6. 189 Iat FH, Wah KC, Hung YH. A follow-up study of electromagnetic interference of cellular phones on electronic medical equipment in the emergency department. Emerg Med (Fremantle) 2002; 14:315-19. 190 Iri JL, Severson RP, Firl AR et al. Cellular telephone interference with medical equipment. Mayo Ciin Proc 2005; 80: 1286-90. 191 Boyle J. Wireless technologies and pati,ent safety in hospitals. Telemed J E Health 2006; 12:373-82. 192 Ihong ML, Iay LK. Septicaemia from prolonged intravenous infusions. Arch Dis Child 1975; 50:886-9. 193 Garrett DO, McDonald LC, Wanderley A et aJ. An outbreak of neonatal deaths in Brazil associated with contaminated intravenous fluids. J Infect Dis 2002; 186:81-6.
References I
239 Chinachoti T, Surase ranivongse S, Pengpol W, Valairucha S. Delayed detecti on of esophageal intubation: Thai Anest hesia Incidents Study (THAI Study) d atabase of 163,403 cas es. J Merl Assoc Thai 2005; 88(Suppl. 7):S69-75. 240 Warner MA, Warner ME , Warner DO et a1. Periopera tive pulmon ary aspiration in infants and chi ldren. Anesthesiology 1999; 90:66-71. 241 Roe CR, Wiltse HE, Sweetman L, Alvarado LL. De a th caused by perioperative fasting and sedat ion in a child with unrecognized very lo ng chain acy l- coenzyme A dehydrogen ase deficiency. J Perlia tr 2000; 136 :397-9. 242 Breucking E, Rei mnitz P, Schara U, Mortier W. Anesthetic co mplication s. The incidence of severe anesth etic com plication s in patients and families with prog ressive muscular dystrophy of the Duchenne and Becker types. Anaesthesist 2000 ; 49 : 18 7-95. 243 Larach MG, Rose nberg H, Gron e rt GA, Allen Gc. Did anesthetics tr igge r cardiac arrests in patients with occult myopathies? Anesthesiology 2001; 94:933- 5. 244 Oudesluys-Murphy AM, van Berke l JH, van den Bran d HM. Unexplain ed sudden deatb und e r genera l anaesthesi a. Lancet 1985 ; 1 :696-7. 245 Schmidt GN , Burmeister MA, Lilje C et a l. Acute b ea rt failure during spin a l s urgery in a boy with Duchenn e muscul ar dystrophy. Br J Anaesth 2003 ; 90:800-4. 246 Cuccbiaro G, Rbodes LA. Unu s ual presenta tion of long QT sy ndrome . Br J A naesth 2003; 90:804 -7. 247 Lagard e MC, Tun e ll WP. Surgery in patients w ith hemoglobin-S disease. J Pediatr Surg 1978; 13 :605-7. 248 McGarry P, Dunca n C. Anesthetic risks in sickle ce ll trait. Pediatrics 1973; 51:507 - 12 . 249 Dover SB, P le nderleith L. Moore MR, McC oll KE. Safety of ge ne ra l anaesthes ia a nd s urgery in acute he pat ic porphy ria. Gut 1994; 35: 111 2-15. 250 Tabib A, Lo ire R, i\Iliras A et a l. Unsuspected card iac lesion s associated with sudden un ex pec ted periop e rative death. Eur J Anaesthes iol 2000; 17:230- 5. 25 t Galli L, Orrico A, Loren z ini Set al. Frequen cy a nd localization o f mutation s in the 106 exo ns o f the RYR J gene in 50 indiv idu als w ith malignant hyperthermia. Hum Mulat 2006; 27:830. 252 Hallid ay NJ. Malignant hyperthermia. J Cran iojac Surg 200 3; 14 :800-2. 253 Karger B, Te ige K. Fatal m a lig nant hyperthe rmia: delayed onset and aty pical course. Forensic Sci lilt 2002 ; 129:1 87 - 90. 254 Prescott RJ, Roberts SP, Wi.lIiams G. Malig nant hyp erpyrexia: a rare cause of postoperative death. J Clin
Path o11 992; 45:361-3. 25 5 Ellis FR. Malignant hype rpyrex ia . Arch Dis Child 1984 ; 59: 1013-1 5. 256 Watkin s 1. Adverse rea ct ion to neuromusc ular blockers : frequ e n cy, investiga tion, a nd epidemi ology. Acta Anaesthes iol Scand 1994; 102(Suppl.):6-10. 25 7 Fox MH, Hunt Pc. Prolonged neuromusc ular block associated with mi vacurium. Br J Anaesth 1995; 74: 237-8. 258 Pasqu ariello CA, Schwartz RE. Plasm a ch olinesterase defici ency in a neon ate. Ca n J Ana esth 1993; 40:529- 3 1. 259 Hoffman TM, Wern ovsky G, Wieand TS et al. The in cid ence of a rrhythmias in a pedia tric ca rdiac inte nsive care unit. Pedia tr Cordial 2002; 2 3:598-604. 260 Arieff AJ. Fatal pos toperative pulmonary edema: pathogenesis and litera ture revi ew. Chest 1999; 115:1 371-7. 261 Brouh Y, Paut 0, Tsim a ratos M, Camb o ulives J. Postoperative hyponatremia in chi ld re n: pathophysiology, diagnosis an d treatme nt. Ann Fr Anesth Reanim 2004 ; 23 :39-49.
383
262 Choong K, Kho ME, Menon K, Bohn D. Hypo to nic v ersus isoto nic saline in hospitalised children: a systematic revi ew. Arch Dis Chi ld 2006; 91 :828-35. 263 Way C, Dhamrait R, Wade A, Walker 1. Perioperat ive fluid the rapy in c hildren: a s urvey of cu rrent presc ribing pract ice. Br J Anaesti1es ioi2006; 97:37 1-9. 264 Copl ans MP, Curson I. Dental anaesthesia a nd mOliality. Br Med J (Ciin ResEd) 1982; 284:661 -2. 265 Copl a ns MP, Curs on I. Dea ths associate d w ith dentistry an d d en ta l di sease 1980-19 89. Anaesthesia 1993; 48:43 5- 8. 266 Worthin g ton LM, Flynn PJ, Struni n L. Dea th in the de nta l chair: a n avo idab le ca tastrophe? Br J Anaesth 1993; 80: 13 1-2. 267 De nta l Anaesthesia Comm ittee of the Royal College of Anaesthetists. Standards and Guideli nesjor General Anaesthesia jar Del/tistry. London: The Roya l College of A na esthe tis ts, 1999. 268 Poswillo D. General Anaes thesia Sedation a nd Resusc itat io n in dentist ry: Repon of a n Expert Working Party for the Standing Dental Advisory Commi ttee. London: Dep artment of Health, 1990. 269 Cartvl'right DP. Dea th in the dental ch a ir. A naesthesia 1999; 54: 105-7. 270 Malamed SF. Morb idity, mortality and local anaesthesia. Primary Dent Care 1999; 6:11-1 5. 271 Lifsc hultz BD, Kenn ey JP, Sturgi s CD , Donoghue ER. Fatal intrac ranial hem orrhage follo win g ped iat ric oral s urgica l proced ure. J Fo rensic Sci 199 5; 40: 13 1-3. 272 Keelin g JW. Fetal and Neonatal Pa thology, 4th edn. Lo ndon: Springer, 2007. WiggleswOlih JS. Perina tal Path ology, 2nd edn. Lond o n: 273 WB Saund ers, 1996. 274 Po lberger S, Svenningsen NW. Ea rly neonatal sudden infant dea th a nd near death of full-term infa nts in mate rni ty wards. Acta Pa ediatr Scand 198 5; 74:861-6. 275 Rodriguez-Alarcon J, Melchor Jc. Linares A et a1. Early neo natal sudden death or near death syndrome. An epidemiolo g~cal study of 29 cases. Acta Paediatr 1994; 83:704-8. 2 76 Deh a n M, Imbert MC, Benisvy C et a l. [Sudden newbo rn infa nt deat h in mate rnity. Anatomo-c1inica l study of 3 1 casesJ. Arch Fr Ped iatr 1992; 49: 609- 16. 277 Pina r H. Postm onem findings in term neonates. Sel11in Neonato l 2004 ; 9 :289-302. 278 Keeling JW. Iatrogenic disease in the newborn. Vi rchoUis
Arch A Pathol A nat Histo1 198 1; 394: 1-29. 279
Byard RW, Kennedy JD. Dia g nost ic difficulties in cases of s udden death in infants with mandibular hyp op la sia. Am J Forensic Med Pathol 1996; 17 :255-9. 280 Wigglesworth JS. Performan ce o f the perin atal a utopsy. In Perinatal Pathology, 2nd edn . Lo nd on: Saund ers, J 996, p. 36. 28 1 A bman SH, Burc he ll MF, Schaffer MS, Rosenberg AA. Late s udden unexpected deaths in hospitalized infants with bronchopulmonary dyspl asia. A m J Dis Child 1989; 143:815- 19. 282 Espagne S, Hamo n I. Thiebau geo rges 0, Hascoet JM. [Sudd e n death of n eo n ates in the delive ry roo mJ . Arch Pediatr 2004 ; 11 :436-9. 283 Byard RW, Burn ell RH. Appare nt life-threatening events and infant h olding practices. A rch Dis Child 199 5; 73 :502-4. 284 Frank DJ, Drobish NL. Toy safety in hospi tals: or beware of parents bea rin g gifts. Ciin Pediatr (Phila) 1975; 14:400-2. 28 5 M illun ch ic k EW, MeAno r RD. Fatal a sp ira tion o f a makesh ift pacifier. Pediatrics 1936; 77 :369-70. 286 Levene S, Bo nfield G. Accidents on hospital wards. Arch Dis Child 199 1; 66: 1047-9. 287 Lyons TJ, Oates RK. Falling out of bed: a relatively benig n occurrence. Pediatrics 1993; 92:125 -7 .
384 I
Sudden death in hospital
288 Nimityongskul P, Anderson LD. The likelihood of injuries
when children fall out of bed. ) Pediatr Orthop
302 Forrest ARW The Investigation and Prosecution of Health
304
Care Workers Who Systematically Harm Their Patients. LL.M thesis, Uni versity of Wales, 1992. Kinnell HG. Serial homicide by doc to rs: Shipman in perspective. BM) 2000; 321: 1594-7. Park GR, Khan SN. Murder and the ICU. Eur) Anaesthesiol
305
Yorker Be. Nurses accused of murder. Am ) Nurs
198 7 ;
7: 184-6. 289 290
Finberg L, Kiley J, Luttrell CN. Mass accidenta l sa lt poisoning in infancy. )AlI!LA J963; 184:187-90. Finberg L. Unfo rgettable patients.) Pediatr 1992;
303
2002; 19:621-3.
121 :323-4. 291
292 293
294
295
Walter GF, Maresch W [Acc id ental sa lin e poisoning in newborn infants. Morphologic findings and pathogenetic discussio n]. Klin Padiatr 1987; 199 :269-73. Busch KA, Fawcett J, Jacobs DG. Clinical correlates of inpatient suicide. ) Ciin Psychiatry 2003; 64: 14-19. Pfeffer CR, Plutchik R, Mizruchi MS, Lipkins R. Suicidal behavior in child psychiatric inpati ents and outpatients and in nonpatients. Am) Psychiatl)' 1986; 143:733-8 . Wo lfersdorf M. [Patient suicide in the psychiatric hospita l: selected results of the Clinic Suicide Wo rking Group Study 1/11 1970-1992 of the 'Su icidal Behavior and the Psychiatric Hospi ta l' Study Group]. Psychiatr Pra_r 1996; 23 :84-9 . Noren GR, Staley NA, Bandt CM, Kaplan EL. Occurrence of myocarditis in sudden death in children. ) Forensic Sci
James DS, Leadbeatter S. Detecting homicide in hospital. ) R Call Physicia11S LOlldol1 1997 ; 31:296-8. 307 Start RD , Cross SS. AC P. Best practice no 155. Pathological investigatio n of deaths following surgery, anaesthesia, and medical procedures.) C/il1 Patlwl 1999; 52:640-52. 308 Bush CM, Jones JS, Cohle SD, Johnson H. Pedia tric injuries from cardiopulmonary resuscitation. Anl1 Emerg Med 1996; 306
28 :40-4. 309 Maguire S, Mann M, John Net al. Does cardiopulmonary
resuscitation cause rib fractures in children? A systematic review. Child Abuse Negl 2006; 30:739-5 1. 310 Peterson DB, Beckwith JB, Benson EA. The sudden infant death syndrome in hospitalized babies. Pediatrics 1974;
1977; 22:188-96. 296 Knight DM, James RA, Sims DN et at. Sudden death due to
54: 644 - 6.
intravenous infusion of hair conditioner. Am ) Forensic Merl Pathol 1998; 19:252-4. 297 Southall DP, Plunkett MC, Banks M et at. Covert video recordings of life-threatening child abuse: lessons for child protection. Pediatrics 1997; 100:735-60. 298 Feldman KW, Hickman RD. The central venous catheter as a source of medical chaos in Munchausen syndrome by proxy. ) Perliatr Surg 1998; 33:623-7. 299 Goldfarb J, Lawry KW, Steffen R, Sabella e. Infectious diseases presentations of Munchausen syndrome by proxy: case report and review of the literature. Clin Pediatr (Philo)
3 II
Gill JR, Goldfeder LB, Hirsch CS. Use of 'therapeutic complication' as a manner of death. ) Forensic Sci 2006;
3 12
Roberts lSD, Gorodkin LM, Benbow EW What is a natural cause of death? A survey of how coroners in England and Wales approach borderline cases.) Clin Pathol 2000;
51 :1127-33.
53 :367 - 73 .
Buehler JW, Smith LF, Wallace EM et al. Unexplained deaths in a children'S hospital. An epidemiologic assessment. N El1gl) Med 1985; 313:211 -1 6. 314 Istre GR, Gustafson TL, Baron RC et al. A mysterious cluster of deaths and cardiopulmonary arrests in a pediatric intensive care unit. N El1gl) Med 1985; 313:205-11. 315 Hunt J, Goldring J. The case of Beverley AJlitt. IVIed Sci Law 313
1998; 37:179-85. 300 Kamerling LB, Black XA, Fiser RT. Munchausen syndrome by
301
proxy in the pediatric intensive care unit: an unusual mechanism. Perliarr Crit Care Med 2002; 3:305-7. Meadow R. Munchausen syndrome by proxy. Arch Dis Child
1997; 37:189-97 . 316
Clothier e. The AliitT InquilY. London: HMSO, 1994.
1982; 57:92-8.
~~
--------_.
1988 ;
88: 13 27-8, 1332.
-
------ -
~
-
I
CHAPTER 20
I
ROAD TRAFFIC ACCIDENTS IN CHILDREN Anthony Busuttil
Road traffic fatalities Investigation of a fatal road traffic collision Vehicular collisions Other supervening problems in collisions Pedestrian injuries Child cyclists
385 387 390 391 392 392
ROAD TRAFFIC FATALITIES Road traffic deaths are still common, although rates among countries vary considerably (Table 20.1). They deserve the same level of investigation as any other unexpected and violent death; however, this does not always occur because of prioritization of other fatal ities. In the thorough investi gation of any road traffic death , particularly one that involves a child, a number of sources of inform ation must be considered and integrated to enable a useful and faith ful reconstruction of the individual incident and, perhaps more importantly, the development of appropriate preven ta tive measures. The detailed information obtained about the collision from each potential source must be carefully collated to ensure that the incident can be reconstructed in detail - not only to assist the bereaved in coming to terms with the death, but also for court inquiry (criminal and civil) and insurance purposes. 1, 2 If the death can be shown to have been preventable then steps should be taken to ensure that similar deaths in t hat particular location or manner could be avoided in the future. In the UK nearly two out of three road accidents happen when children are walking or playing and almost two thirds of child accident victims are boys. As a child gets old er, the risk of a road accident increases. A child from a low-income family is five times more likely than a child
Diffuse axonal injury Whiplash injuries Injuries to children in utero Other vehicular accidents References
393 393 393 393 394
Table 20.1 International comparisons: road accident fatality rates per 70 6 papUlation year - 2000 (aged 0-14 yearsj4s Iceland Sweden Japan United Kingdom Italy Germany Netherlands Austria Finland Scotland Norway Switzerl and Denmark Hungary
0 12 13
15 17 19 19 20 21 21 22 23 26 26
Ireland Australia Belgium Canada Spain Czech Republic France Greece Luxembourg Poland USA New Zealand Portuga l Republic of Korea
27
29 29 30 31 32 33 33 37 37 40 47 52 58
from a high-income family to be killed ,on the road . Chil dren from ethnic minorities are involved in up to twice as many accidents as the national average while walking or playing. Also, the risk of being involved in a road accident when walking or pl ay ing is more than 10 times greater for a child with hearing difficulties. Child pedestrian deaths
Fatal road traffic co llis ions I
others are in use and no restraint is available. Children must not sit in the front seat withou t the right child seat. • In France, children under 10 years of age are not allowed to travel in the front seat of a car unless the child concerned is an infant of 9 months old or younger and weighs less than 9 kg; however, the infant must be placed in an appropriate baby seat (rear facing) unless there is an air bag fitted, in which case the infant is not allowed on the front seat at all. In the rear seats, a ll children under 10 years must be in the right restraint; if the child's weight is between 9 and 15 kg then he or she must use a child seat and if over 15 kg then a booster type seat should be used. • In Spain, children aged 3 or under must use a suitable child seat, and child ren up to 12 should use one if ava il able. If not they can use ad ult seats. • In Italy, it is illegal for children under 3 years to sit in the front or rear seats without the proper child seat. Children aged betvveen 4 and 12 years cannot be a passenger unl ess they are using a su itable safety restraint seat or an adaptor for a seat belt. If no seat is available for children aged between 4 and 12 years then they can use the ad ult seat belts in the rear if accompanied, in the rear, by another passenger aged over 16 years. The design of cars and the restrain ts within them has exercised engineers and car designers for severa l decades, with the emp hasis on design ing safer vehicles. Data col lected from actual incidents have assisted in ensuring that proposed designs are effective 4 6
INVESTIGATION OF A FATAL ROAD TRAFFIC COLLISION The inves tigation of a fatal road traffic collision has a vari ety of face ts, described as foll ows.
The Scene of the Collision Data collection should always start at the actual scene of the fate" road in cident. The deceased child may have been a passenger in a veh icle, a pedestrian or a pedal cyclist; more rarely, the deceased child may have been a pillion rider on a motorcycl e, a passenger on a pedal bicycle or the occupant of a sidecar of a motorcycle. Establishing this information should be the point of departure of the inves tigation of the particular death. AJthough it may not be fe as ible for a forensic pathologist to attend at evelY such scene of a fata l road traffic collision , if the accident appears to be likely to resu lt in an eventual criminal prosecution then it is imp ortant in such cases for the pathologist to actually attend the scene. When the acci dent is a 'hit and run inc ident' it is a lmost imperative that the pathologist attends, together wit h t he rest of th e inves tigative team. In each case, evelY effort sho uld be made to
387
docum ent the scene fully by sti ll photographs and on video to ensure that the findin gs at the site of the incident are availabl e for futu re reference, briefing and reconstmction. It is often the case that the deceased has not been pho tographed in situ , as the attendance of ambulance person nel and paramedi cs usually results in the child being taken away to hospital for at least atte mpted resuscitation. Pub lic access to the locus and ready visibility of the scene in the interest of decency may also require the early removal of the body from the scene of the collision . Specialists in road traffic collision investigations are nowadays available7 in most police forces. The scope of their activity is to document as much as possible at the scene, tak ing measurements of such items as skid and tyre marks,8 oil and other fluid leaks and any other data that can be g leaned. They will log the location of di scarded items of clothing and shoes on the roadway, bicycles, broken portions and debris fro m a vehicl e (for example paint fra gments,9 side mirrors, lamps,1O glass fragme nts and pools of blood) in an attem pt at a full reconstmction of the incident. They wi ll eventua lly produce a sketch of the incident and, if the incident warrants it, a computerized reconstmction may be attempted. Valu able information about such ma tters as the condition of the road, street lighting and prevailin g weather conditions will be collected and later included in a detailed report. An important part of the eventua l investigation, is the determination of the speed at which the vehicle{s) inlJollJed in the incident was trauel/ing. lI It has to be kept in mind that the injuries on injured or deceased individuals invo lved in t he collision are of little or no help in determining this . One of the main pointers to impact speed are the markings left on the roadway by a vehicle attem pting emergency braking and the movement and transport of objects a long the road. Any contact with walls, street furnitu re (such as lamp- posts or signposts) and the defo rmation caused to such items is more useful in the estimation of imp act speed. Similarly, any deformation in the actua l framework of vehi cles involved can , after an assessment by a mechanical engineer and appropriate calculations, assist in answering t his importa nt question. The pathologist or clinician should be wary about being led into giving definitive answers to the question of speed estimates, i.e. responses that cannot be later substa ntiated scientifica lly on cross-examination. These investigators would also collect any trace evidence that many be useful to identify the vehicle involved, particu larly in 'hit and mn' accidents. Even a small fragment of pain t from the offending vehicle can be analysed to enable such detailed information as to the make and year of manufacture of the vehicle, any recent res prays of the bodywork, etc. 12
Eyewitness Accounts
The second source of data are the accounts given by eye witnesses to the elJent. Th ey will be interviewed by police officers as early as possible after the incident and while
386 I
Road accidents and ch il dren
constitute the most common accidental cause of child fata li ties in the UK. The speed at which the vehicle is travelling has an essen tial bearing on the number of pedestrian fatal ities and speed reduction measures that have been used on roads (e.g. speed bumps across the road and limiting the speed limit) have all been shown to benefit from a diminution of child pedestrian fatalities. An impact speed below 20 mph results in signifi cantly fewer fatalities. The faster the traffic, the greater the 11sk of death and serious injury. When children are hit by cars travelling at 20mph, 5 out of 100 are killed; most suffer only minor injuries and about 30 out of 100 suffer no injury at all. At 30 mph nearly half of all children are kjJled and many are seriously injured. At 40 mph 85 out of 100 children are killed. Children make unpredictable pedestrians. Traffic-coping skill s are complex and children do not have the ability to judge speed, distance and danger accurately until around the age of I I years. Even children above this age are easily distracted and may not always behave as drivers expect. Young children are also much smaller than adults and so are difficult for drivers to see. Children within vehicles who are sitting in the correct child seat for their size and one that had been fitted properly usually suffer only minor injuries in a car crash, but arou nd one-half of all child seats are still not properly fitted. Rear facing infa nt seats reduce the risk of fatal injury in a crash by more than 70 per cent, fon-vard-facing toddler seats by more than 50 per cent and safety belts by 45 per cent. At one time in the USA, only 10 per cent of children under the age of five travelled unrestrained, but they accounted fo r more than one-half of child deaths in cars. It is important that child restraints are used properly if they are to save young lives. A study by the Automobile Association (AA) and Devon County Council found that 24 per cent of child car seats were too loose, 21 per cent of the harn esses were too loose and 21 per cent of the buckles were not fitted properly. The law app lying to child car seats changed in September 2006 in the UK, meaning that all children under 14 years should have some form of extra restraint as appropriate to their weight and age while travelling in a car. Every year around 30 children aged up to I I years are killed during car journeys, 400 are seriously injured and over 7000 more are injured; many of these deaths and iqjuries could have been prevented if the children had been strapped in correctly. Children who are not secured with the appropriate child restraint, such as a chi ld seat, may be seriously injured in a car accident. A ch ild should never be carried on an adult's lap. Also, an adult sho uld never try to put an ad ult seat belt around both him- or herself and a child; in the event of an accident the child would be crushed by the adulfs weight. As from 18 September 2006 it became law that a child under the age of 3 years must have an appropri ate child restraint when travelling in the front or rear seat. The only exception to this is travel in a taxi - by law children aged under 3 years may travel unrestrained in a taxi because the ordinary ('adult') seat belt in the car is not suitable for them.
A chi ld aged from 3 to 12 years or under 135 cm (about 4 feet 5 inches tall) must use the appropriate child restraint in the front seat and, if there is one availab le, in the rear. Children over 3 years old may wear an ordinary ca r seat belt in a taxi . If there are no seat belts in the rear of the car, by law a child over 3 years old can travel unrestrained. Chil dren who are aged 12-13 years old, or younger children who are over 135 cm tall, must use the ordinary car seat belt, if available. Children aged 14 and older a re classed as adult passengers. They must use the ordinary car seat belt and it is their own responsibility to do so. Car seats a re so ld in stages appropriate to the chi ld 's age, and they range from rear-facing carders for bab ies up to the weight of 13 kg to forward-facing seats that are suit able from 9 months to 4 years. After this age, booster seats can be used until the child reaches the height of 135 cm. The new law states that it is illegal to install a rear-facing child seat in the front of any car when the air bag feature has not been disabled. Instead, it is recommended that instead any rear-facing infant seat should be insta lled behind the driver on the rea r seat. Frontal ai r bags and side air bags are fitted as sta ndard practice in new cars within the UK and are designed to dou ble up on the effectiveness of the traditiona l three-point seat belts, whic h have been comp ulsory in new cars since 1967. Containing 35 L of propellant, air bags inflate to full capac ity within 25 milJiseconds, at a speed of 160 mph, which goes some way to explain why it is imperative to deactivate any air bag if the child is to be carried in the front. In a crash, the air bag infl ates very quickly. It could hit anything close to the dashboard with enough force to cause severe injuries or even death. Because the back of a rear-facing child seat sits very close to the dashboard , the seat could be struck with enough force to cause serious, or even fatal , injuries to a baby. Even older children (who have outgrown child seats) are at risk from a deploying air bag if they are not properly restrained with a lap/shou ld er belt. Reports of eight deaths of child passengers in crashes involving air bag deployment are of special concern because they involved low-speed crashes that the children otl1ervvise might have survived. 3 Up to the age of 3 years it should always be observed that no child should travel unrestrained unless in the exception of traveLling in a taxi or licensed car hire vehicle in which no restraint is availab le. The use of the co rrect car seat for the child's weight is essential, although it is a lso velY important to make sure that yo ur child 's seat is in sta ll ed properly and is the right design for the particular car. It is important to note, how ever, that the law varies in other co untries in rela t ion to child restraints: • In Genl1any, for example, children under 12 years of age and under 1.5 m in height must use an appropriate child restraint (e.g. a child seat or booster seat while in the front or rear of a car). Rear-facing seats cannot be used in front seats that are equipped with air bags. Children may travel in the rear seats without restraints if all the
-
.....
388 I
Road accide nts and children
events are still fresh in their minds; they may be inter viewed again at a later stage to enable the police to go over their initial statement with them once again, in more detail and in the light of results of other investigations. It has to be remembered than these incidents are always the cause of major emotional trauma to the onlookers, and thus it may often be useful to wait until they have had a time to compose themselves before such statements are taken; at the scene all that needs to be done is to obtain names and addresses. Great sensitivity has to be exercised by the inves tigating police officers in the timing of taking statements, particularly from next-of-kin eyewitnesses. On the other hand, delayed attempted rationalization of the perhaps jum bled recollections of the rapid incident may result in prob lems in distinguishing fact from rationalization. In this aspect of the data collection , it is always useful to interview the paramedics , police officers and sometim es also fire officers who are called to the scene. The informa tion that they may have picked up in the course of their scene atte ndance, i.e. their involvemen t afte r the collision, the appearances of the locus of the collision and the posi tion of the body when found, is particul arly useful. They may have had to move vehicles or the body of the child or to extricate the child 's body from within or under the vehi cle. It is then most important to know exa ctly the position and appearances of the body prior to its movement. In this respect, if any resuscitation has been carried out then it is also essential to interview the ambulance crew, paramedics and the members of the nursing and medical staff who have been directly involved. This will enable them to give information as to the methods of resuscitation used, any drugs administered, and any other changes effected by them.
Clothing of the Decedent The clothing of the injured parties may also have useful trace evidence 13 on it, as well as features that indicate that certain matters had taken place - for example, crushing under a wheel, contact with the greasy undercarriage of the vehicle, etc. If the child was taken to hospital, it is often the case that clothing is cut off with scissors and sometimes has been completely disposed of before the police have had a chance to claim it. It is essential that such items of clothing be retrieved and that those responsible for clothing removal are instructed to be aware of its potential value as evidence.
Inspection of the Vehicles Involved Another accessible source of data in fatal road incidents is the vehicle (or veh icles) involved in the incident. If a mecha nically propelled vehicl e is involved it is essential that exp erts care fully check its roadworthiness, including the brakes. The vehicle is usually towed or taken away to a police-manned garage where it is fully inspected and examined in terms of
problems with its components, seat belts, tyres, steering mechanism and other features relevant to safe vehicle opera tion. A history of the vehicle may be usefully be sought through maintenance garages. The deformation and denting on the metallic body sur faces of a vehicle that have occurred in the course of the fatal incident can be extremely usefu l in determining, with some degree of accuracy, the speed at impact. Mechanical engineers are able to examin e the vehicle and determine by mathematical calculation, often assisted by computer aided programs , the speed at impact. The vehicle will also bear in it dents and breakages at sites of direct impact with the human body. If a child pedes trian has been knocked down, an inspection of the vehicle may well indicate where the initial contact with the child's body was; this co uld be the front bumper or bonnet of the car. This inspection may be able to determine whether the child has been elevated after the initial impact and pro pelled upwards to strike the car still moving in his direction or wheth er the child had fallen beneath the car. Impacts with pedal cyclists often result in the child being thrown up in the air if the combined speed of impact is over 16-20 mph (25-32 km per hour). The body will often land on the vehicle with second imp act: for example a primalY impact may be with the windscree n and with the secondary impact with the A-pillars or windscreen of the car. At such sites of impact, there may have been transfer of hair, blood and tissu e from the cyclist. When the deceased child has been an occupant of a vehicle, it is important to determine where he or she was placed inside the vehicle, whether the child h ad been prop erly restrained and wheth er the restraints available in the vehicle were appropriate, properly fitted and functional, and whether there were any other occupants or other items inside the vehicle that might have resulted in fuliher impact and injury.
Information from Others Involved FUliher useful data sources are the narratives from other persons invoJved in the co llision, not least the driver of any vehicle involved. In such instances, in addition to docu men ting any injuries which they may have sustained, it is also impoliant to determine the driver's state of health and whether or not they were incapacitated in any way, or intoxicated, at the time th at the collision took place. In all road traffic collisions the drivers of any vehicle will invariably be breatha lysed for alcohol at the roadside and a general observation made of them by police officers present as to their state of orientation\ coordination, pupil Jary size and reaction. If the situation warrants it, statute provides for their medical examination a nd the collection of further samples for the estimation of alcohol and drugs in body fluids. More detailed medical examinations and the collections of specimens for drug testing wi II be carried out
_. - - - - - - - - -------
-
Fatal road traffic collisions I
at the police station rather than at the scene, or in hospital if the driver has been injured.
The Decedent The main source of evidence accessible to the pathologist is the body of the deceased. If the child has not been unclothed, it would be essential that the clothing is removed carefully to avoid any further damage to it and to avoid any loss of evidence from it as well as any cross-contamination. The clothing, including the shoes, should be separately bagged into brown paper bags and given to the police for further examination as required. The child's external injuries, no matter how trivial, should be carefully measured and documented. It is useful to photograph all aspects of the body. This may not always be feasible and, in such instances, the importance of a care ful description of these injuries is crucial. Injuries with pat terns to them must be documented fully photographically. This may have to be followed, depending on local legal custom, by the formal identification of the decedent by members of his family and by police officers who were actu ally present at the scene. Even if there is no legal require ment, it is essential than the parents are aJlowed early access to their child after they have been fully briefed about the appearances of their child by those who have been trained to so do, for example police family liaison officers. It is essential that the body of the child is presented to them in an aesthetic way, perhaps with partial covering or bandaging to obscure injuries or even following some clean ing and reconstruction of the body. The identification process is fraught with emotional overlay for those who are doing this in an official capacity, and indeed much more poignantly so to the members of the family of the deceased. The latter should be given enough time with the body and not officiously rushed. They should not be discouraged from touching the body and if, indeed, the family asks for mementoes, such a locks of hair or handprints, then these requests should be met, with the consent of the appropriate legal personnel. Some families may wish to obtain a Polaroid photograph of the face of the deceased child. Religious ritual may also be requested as part of the acute bereavement pro cedure. All reasonable requests made by the family should be heard and, if at all possible, agreed. It is extremely useful to document bony injuries using radiology before the autopsy commences. This ensures a much fuller documentation of such injuries, particularly of the ribs and vertebrae and in younger children in whom ossification may be incomplete. Another useful investigative adjunct, which is now avail able to most police forces, is the availability of a portable laser source of light that enables the child's unclothed body to be scanned in a dark room; this shows any marks on the body that have been caused by tyre imprints or patterned injuries that have resulted from contact with specific parts of the vehicle.
--- -.
- ~ -
~ ~
389
A full autopsy must be carried out and all internal injuries documented fully being careful to examine closely the vetiebral column, the rib cage in its entirety and the scapulae - fractures at these sites are often missed. If physical abnormalities are present then these should be carefully recorded. It is also important to check that the child that has not suffered from any previous significant medical histolY by look.ing through the child's general practitioner records; if the family doctor is not known then this should be checked through the police. If such a history exists, it is important to give due weight to any relevant information. If the child suffered from learning difficulties, epilepsy, visual impairment or hearing problems then these may all be relevant; with regard to a history of hearing problems the middle ears have to be opened and conditions such as otitis media and 'glue ear' have to be excluded. The autopsy should enable a full documentation of all injuries, externally and internally. If any injuries show any patterns in them (e.g. tyre mark imprints) then these have to be documented fully by photographs and accurate measurements. Any paint or glass fragments or any other trace evi dence should be collected, particularly if there is some dubiety as to the vehicle involved . Even in autopsies on children, particularly in the late teens, it is important to collect the following samples for further analysis: 1. Blood for DNA profiling, particularly in 'hit and run' accidents. 2. In appropriate cases, it may be useful to exclude acute inhalation of solvents prior to the collision. In addition to the collection of blood, this may require the retention of a lung, with its main bronchus tied off securely and placed inside a nylon fibre bag to avoid loss of intrapulmonary air content. This enables assay of the solvent content within the bronchial air. 3. In some cases, it may be necessary to test for alcohol and other drugs. For these assays, appropriate specimens of blood, urine, vitreous fluid and liver will have to be retained on the off chance that these become necessaty. In the older teens, it is always wise to consider intoxication as a potential complicating factor in road traffic collisions. 4. (Control) samples of plucked scalp hair may also come in useful for matching with any hairs found at the scene and to exclude any previous exposure to controlled drugs. Recent problems with organ retention from necropsies may make it difficult for the brain to be retained for neuro pathological examination. In cases whe,n the death is very acute, retention of the brain has only limited value in terms of documenting injury. In delayed deaths, and in cases where there is known to be a pre-existing cerebral problem (e.g. epilepsy, hyperactivity syndromes, autism, etc.) it may be necessary to suspend and fix the brain for an appropriate
390 I
Road accidents and children
period in form alin and to examine it as soon as possible afterwards. Diffuse axonal injury does occur in babies and children. If the child has surv ived long enough after the acci dent, this pathological damage is discernible with appropri ate special staining techniques an d microscopy. In deaths in the intens ive therapy unit, bacteriological samples may be required to exclude nosocomial infections that may have contributed to the death. This may be of importance if civil litigation is in progress in terms of a novus actus interveniens and the apportioning of blame for the death. In some deaths, with the consent of the family and of the legal authorities, organs have been harvested for tra nsplanta tion purposes. The current dearth of organs compared with the demands of patients, combined with the current major successes of transplanta tion, has made this phenomenon more common than it once was. The contlibution of the pathologist in such cases is more limited given that it is assumed that all the organs removed had been intact, healthy and functional. In such instances to complete the autopsy examination, it is essential to speak with and to obtain a report from tbe transplant coordinator as to by whom, wben and wh ich organs were removed and whether these have functioned adequately in their seco nd hosts. If it is considered that some organs or tissues may be required for other pur poses it is essential that fully informed consent is obtained from the family and that tbis matter is approached in as sen sitive and measured a manner as is possible.
VEHICULAR COLLISIONS The data in Table 20.2 show that, by and large there has been a major welcome decrease in fatal and serious road incidents involving children. Unfortunately, in relation to less selious non-fatal transportation injuries there is serious underreport ing of such collision-induced injuJies, particularly if the injuJies are minor. One estimate placed this at 25 per cent of all child pedestrian casualties l4 and another at 20 per cent with injuJies in boys being less frequently reported than in Table 20.2
girJ. 1s However, such incidents still occur and they should warrant the same amount of investigation that any other sudden suspicious or accidental death is given in the medico legal context. Unfortunately, this is not always the case and important information of potential preventative and safety improvement importance may remain uncollected. Car occupants are the largest category of road users who sustain injulies if they becom e involved in collisions, although fatalities are more common amongst pedestrians (Table 20.3). This has been the reason why large amounts of research, time, money and effort have been spent in ensur ing that these injuries are prevented by improvement in car design, stre ngthening the fra mework of vehicles, the intro duction of internal safety devices - seat belts, air bags - and through legislation improv ing road worth iness of cars and general safety of cars. Car collisions cannot be rationalized simplistically from an isolated assessment of the forces invol ved; in what appear to be entirely simil ar incidents, one car occupant may survive with little injury, whereas the other may experience fatality. The physical size, injury tolerance and the age of the occupant are all important in the outcome of the collision. Thus the different physical build of children is an importan t consideration in predicting the injuries that they will sustain in road traffic collisions (Table 20.4). The bead of a youn g child forms a higher proportion of the overall body mass than that of an older child or an adu lt. The neck of a child cannot support the weigh t of the head as adequately as in an adult. The lib cage is smaller and more flexible a nd pliable, and leaves a larger unprotected area of the abdomen; the child's pelvis is not as resilient as that of an adult and its cur vature is different. All of these factors will have a bearing on the kinematics in the collision situation and the possible types of injury that are sustained and their outcome. Most collisions are frontal, with a tendency for the offside fron t of the vehicle to become involved. In side impacts, t he door of the vehicle on the side of the collision remains sup ported by the striking object dUling the impact phase and acts as an unyi elding barrier that is pushed inwards into the vehicl e's compartment and against which the occupant may
Child fatalities and casualties in Scotland, England and Wales J997-2005
Scotland Child casualties
-
England and Wales Fatal
Fatal and serious
All severities
Fatal
Fatal and serious
1997
26
745
379 8
229
5708
40751
1998
32
698
3536
174
5382
3991 4
All severities
1999
25
625
3196
196
5073
38872
2000
21
561
2999
170
4641
367 15
2001
20
543
2956
199
4447
35361
2002
14
527
2747 247 8
2003
17
431
2004
12
383
2394
2005
11
368
2166
-- ---
---
. _ -- - --
-
---
-
Supervening problems in collisions I Table 20.3
Average number
of road traffic fatalities per year for
1997-2001 in Scotland
Deaths
Pedestrians Driver or rider Passenger vehicle! pillion
Age (years)
Total
0-4
5-11
12-15
2 (0.01) 1 (0.00) 1 (0.00)
8 (0.02) 1 (0.00) 3 (0.01)
6 (0.02)
1 (0.00) 3 (0.01)
16 3 7
Figures in parentheses indicate rates per 1000 of the population.
Table 20.4 Site affatal injuries in child car occupants by age Site of injury (no.) 0-9 months (4) 10 months to 6 years (29) 7-14 years (33)
Total (66)
Head
Neck
Chest
Abdomen
Other
4 19
0 2
0 4
0
0 3
20 43
4
4
6
8
2
7
4
'Most of the children were unrestrained and about one-third of them were ejected from the ve hicle. Modified from Andersson and Jonansson 4 6
sustain injuries. Collisions with trees, lamp-posts and other street furniture, and walls produce a more concentrated load, with the possibility of an increased intrusion into the passenger compartment and thus a higher risk of serious and even fatal injuries being sustained, particularly trunk and pelvic injuries in front-seat passengers as well as head and neck injuries. In addition to the initial direction of the principal force at the time of the collision, there will also be post-impact vehicle rebound, depending of the type and magnitude of the force of the collision and the part of the car that has been struck. This also depends on the size, resilience and safety features of the vehicle involved. These phenom ena will introduce a sideways and lateral component to the impact in the immediate post-impact phase, which, in tum, will have a further direct bearing on both the configuration of the impact and the contact between the body and other parts of the inside of the vehicle as a seco ndary impact phe nomenon. Such an understanding of the kinetics of a colli sion has led to improvement in the inner padding in cars and the removal from the inside of the vehicle of any pointed surfaces and edges. Hyperextension of the neck, secondary to forehead and facial impacts, may result in serious neck and brainstem injuries that could be instantaneously fatal; 16, 17 in severe intrathoracic injuries in particular there may be traumatic rupture of the aorta 18 and compressive injuries to the heart. 19
391
Obviously, the mass ratio between the two vehicles involved and the velocity change that occurs on impact are further critical factors, and these are crucial elements that have to be considered in assessing the force of the impact and the injuries caused therein. If the struck object is station ary and unyielding then the velocity change in terms of acute deceleration will take place in the impact vehicle only. In terms of injury, it is self-evident that the reduction and restriction of movement of the occupants of a vehicle would tend to decrease or remove propulsion and thus secondary impact against the interior of the vehicle. The rational e of a seat belt, which is fitted properly and functioning well when required , is to enable the wearer to 'ride down' the impact when the rapid deceleration occurs; thereby the energy of the impa ct is extended over a slightly longer timescale; the longer this time frame, the greater the dissipation of the forces at the time of the impact. The webbing stretch of the belt enables this deceleration rate and interval to be length ened over a somewhat longer period , thus allowing the energy of the impact on transportation casualties to dissipate plior to contact of the body with an unyielding surface. An evaluation of the impact of front seat belt legislation in Britain showed a marked reduction in injuries to children aged II to J 4 years who were restrained in the front seats of vehicles when collisions took place. 20 If the child has been restrained inside a vehicle, there may be injuries on the body that indicate that this restraint has been in position and functi oned: friction-induced abrasions, bruising and even fractures - of the clavicle, sternum and ribs 21 - may follow the line along which a seat belt would have been applied with force at the time of impact. Internal damage to the body such as intra-abdominal injuries 12,22,23 (including ruptures of the liver and mesentery and major blood vessels 24 ) and intrathoracic injuri es (such as pul monary and cardiac injuries) may indeed be associated with such external injuries; fatalities due to seat belt use have been recorded in children involved in transpoltation injuries. In telms of these considerations, given the relative mass of pedestrian to a vehicle, particularly a high-sided vehicle, the relatively greater speed of the vehicle than of the pedestrian, the decreased resilience of the human body as contrasted with metal, and the Jack of protection in the pedestrian are all factors that render such collisions very traum atic, even at low speed. 2s
OTHER SUPERVENING PROBLEMS IN COLLISIONS Occasionally, the occupant of the car who had been poorly restrained inside it is thrown out of the ,car and injuries are sustained on external impact. If the car catches fire with the occupants still inside then secondary fire-induced injuries and inhalation of smoke will occur. These changes will result in a variety of other injuries to the child involved in such an incident.
392 I
Road accidents and children
PEDESTRIAN INJURIES
If the child is a pedestrian then the primary impact is usually by the front of the car with one or other side of the body that happens to be closer to the vehicle, particul arly if the child is in the process of walking or mnning in front of the vehicle. The distribution of the casualties and fatalities in these inci dents are examined further in Tables 20.5 and 20.6. Depending on the child's height above the ground, dif ferent varieties of impact are present. This will determine whether the prim ary impact with the vehicle is above or below the centre of gravity of the child's body and in this the kinetics of the child's body after the collision takes place, vary. By and large, the umbilicus is at about the level for the centre of gravity. A collision producing a primary point of con ta ct above the centre of gravity will cause the
Table 20.5 Casualties by speed limit at site of incident, mode of transport and severity, 7997-2007 (Scotland) Speed limit (mph) 30
Fatal Pedestrian Pedal cycle Motorcycle Car Bus/coach Other Serious Pedestrian Peda I cycle Motorcycle Car Bus/coach Other
Table 20.6
53 4 6 22
0 2 893 141 114 482 54 52
40
50
5
2
0 7 0
0
37 6 10
8
73
2 9
0 0
7 23 0 5
60
70
Other and unspecified
15 3 30 146 0 12
9 2 2 21 0 3
0 0 0 0 1 0
49 29 190 1187
12 2 2 175 1 33
2
102
child to fall into the path of the oncoming veh icle, and an impact below the centre of gravity would cause the child to be lifted up into the air, then, as the vehicle proceeds, on to the vehicle, and after impact with the vehicle to fall behind the vehicle. If the child is on a bicycle then the centre of gravity at the time of the collision has been so displaced further up (and this is so the larger the bicycle) that on impact the child will be thrown over the vehicle. From an examin ation of that body and the veh icle, the points of pri mary and secondary impact may be worked OUt. 26 The size of the vehicle involved , the age and build of the child, projecting parts of the front part of the veh icle and the speed at impact are all factors that have a bearing on the severity of the injuries sustained by the child. Even minor impacts at low speed may produce severe or fatal head and neck injuries in a young childY Most pedestrian childhood fa tali ties (about 80 per cent) occur in built-up areas with in a I-km radius of the child's home ; a number of such fatalities occur on the way to or from school so that most deaths occur at the times of school opening and closure during weekdays. Injuries are more prevalent during the summer months when children are more likely to be outside their homes. The psychological capabilities and developmental level of the child have a major bearing on their abili ty to cope with road conditions; they have generally less well-developed ski lls relating the perception of movement and velocity, the ability to judge distance and depth, the abi li ty to OIientate themselves in relation to the source of sound and the ability to concentrate for a sufficient length of time, putting aside play activities and concentrating while croSSing or using roads. They are also not equipped with the ability to inte grate all of this information swiftly and efficiently, and come to the appropriate evasive conclusions.
CHILD CYCLISTS 1 0 0
These accoun t for about 17 per cent of deaths on the roads. Arguments have raged for the last decade or so as to the usefulness of cycle helmets for pedal cyclists and the design
Child pedestrian (0-15 years) casualties in single-vehicle accident: mean per annum, 1997- 2007 (Scotland)
CrOSSing road not concealed by vehic le Crossing road concealed by ve hicle Sta nding/wa Iking Other/unknown Total
On pedestrian crossing
Within zigzag area of a pedestrian crossing
Within SOm of a pedestrian crossing
68
12
58
Other
Total
804
84
1026
22
360
~1
423
0 0
0
0 7
99 46
99 54
13
81
260
1602
9 0 78
Crossi ng road in other areas
1
1171
Other vehicular accidents I
of such protective head gear. 28 There can be little doubt that most pedal cyclists die because of head injuries 29 a nd that helmets will afford some degree of protection, particularly in collisions where little force has been expended. In this vein, a number of individual states and countries have suggested that legislation may be required on this matter in concert with legislation concerning the use of seat belts. In Victoria, Australia, following on from a massive 10-year-long educational campaign, legislation was intro duced ensuring that all cyclists wore helmets. 30 ,31 The helmet-wearing rate tripled and a 70 per cent reduction in head injury admissions occurred during the periods 1989-90 and 1991-92. There was also a corresponding decrease in non-head injuries of 23 per cent in the first year and 28 per cent over the first 2 years. From a comparison of statistics between countries where cycling is very popular (The Netherlands, Belgium) and those where it is a less frequent activity and more of a leisure pursuit, the incidence of injuries is lower in the former. This has been extrapolated to the fact that in the Low Countlies cyclists and motorists are segregated and cyclists have their own cycle paths and do not mix with other vehicles on busy roads, thereby decreasing the likelihood of collisions.
DIFFUSE AXONAL INJURY Brain trauma produces a diverse spectrum of injuries in which diffuse axonal injury (OAl) is a well-recognized major contributing factor to long-term disability. Magnetic resona nce imaging and computerized tomography have long been thought to cause underestimation of the extent of OAl because of their lack of sensitivity and therefore newer and more sensitive imaging techniques have been developed. Using these techniques with magnetic reso nance spectroscopic imaging, an increased rate of detec tion was shown in children in whom at one stage it was thought that OAI did not OCCUr. 32 ,33
WHIPLASH INJURIES 34 Generally, very young people and elderly people are those less at risk of sustaining whiplash injury from a car acci dent. Research suggests that 17 per cent of under-21-year olds and just 2 per cent of over-60-year-olds are less likely to be involved in an accident resulting in whiplash than in any other type of motoring accident. These results have come to light despite the fact that younger people, espe cially small children, have far smaller and less well developed neck muscles. The action of a car coming to a sudden halt usually causes the whiplash motion to occur. The impact of a rear-end collision (when one vehicle hits the rear of the vehicle in front) shunts the car forwards, forcing the body forward with it while the h ead and neck are thrown back. The head then tilts downwards towards the steering wheel and the neck extends forwards. The
393
extended head and neck are thrown backwards until they hit the headrest with the consequence that the soft tissues of the neck are stretched and torn. This is also known as hyperextension followed by hyperflexion, and it occurs in a frontal collision when the neck and head continue to move forward while the body remains relatively still. 35 Rear-end collisions can result from any number of sce narios. For example, the driver in front may suddenly brake to avoid a cyclist that pulls out in front of him or the car behind may accelerate more than the leading car at a round about. The impact of hitting another car is the equivalent of hitting a firm surface, such as a wall, at half the speed. Th at is, rear-ending a stationary car while travelling at 30 mph has the equivalent force of hitting a wall at 15 mph. Overall, 43 per cent of people who sustain whiplash injuries from car accidents are involved in rear-end collisions. Side collisions involve the side of one or more cars being impacted. These are common in car parks and at T-junctions. Personal injuries suffered as a result of side collisions are determined by the part of the car that is struck, the speed of the vehicles involved, the presence of safety fea tures such as side-impact air bags, and the weight and con struction of the object that strikes. In total, 35 per cent of people who have been involved in car accidents and sus tained a whiplash injury were involved in side collisions. A head-on collision involves the front ends of two cars crashing into one another. This type of collision may result from a dIiver coming round a bend on a narrow country lane too quickly or a motOlist swelving into oncoming traffic on a B-road to avoid an animal. Head-on collisions are often fatal but only 32 per cent of people who receive whiplash injuries from car accidents sustain them from head-on collisions.
INJURIES TO CHILDREN IN UTERO Road incidents may involve pregnant mothers who are passengers, drivers and, less frequently, pedestrians. 36 ,3 7 Injuries to the pelvis and abdomen may result in severe damage to the uterus, rupture of the uterus and abruption of the placental site with dire, often fatal, consequences to the unborn child. Litigation may focus on whether or not the road incident has directly caused the death of the fetus or unborn infant. 38 ,39 The potential contribution of the seat belt to these injuries may have to be considered 40 - 42 and indeed pregnant mothers are legally exempt from wearing a seat belt for this reason. This topic is considered in more detail in Chapter 10.
OTHER VEHICULAR ACCIDENTS Childhood accidents on the farm,4J,44 often involving vehicles or machinery, are not uncommon, particularly in rural envir onments; they may require intervention from governmental agencies involved in sentimental and work-reared health and safety issues, as well as requiring public inquest.
394 I
Road accidents and children
REFERENCES
2
3
4
5
6
7 8
9
\0 11 12 13 14
15
16
17
18
19
Basham DJ. TraJfic Account i\lIanagement. Springfield, IL:
Charles C Thom as, J 979.
Clark WE. Traffic Management and Collision Investiga tio n. Englewood Cliffs, NJ: Prentice Hall, 1982. Centers for Disease Control and Prevention (CDC). Air- bag assoc iated fatal injuries to infants and children riding in front passenger seats: United States. MMWR J995; 44:845-7. Ashton SJ, Hayes HRM, Mackay GM. Child pedestrian injuries. Proceedings of the Internationa l Meeting on Biomechanics of Trauma to Children. Lyon: lRCOBl, September 1974, pp. 83-100 and 159-70. Roy AP, Mackay GM, GIOYllS PF. So me observations on the modelling of children in car collisions based on field accident investigations. Proceedings of the Fifth International Conference. Birmingham: tRCOBl, September J980, pp. 286-304. Mackay GM. Children in cars and pick-up trucks. Saudi Med) 1987; 8:123-7. (Also: Seat Belt Loading. Proc. A.A.A.M. 36th Annual Conference, Portland , OR, October J992.) MacKay GH. The role of the accident in ves tigator. ) Foren sic Sci 1970; 10:245- 54. Hamm ED. Locating an area on a susp ect tire for compa rative examination to a questioned track. Int ) Forensic Identification 1988; 38/40: 143-5 I. Dabdoub G. Dixon AC, Watson AD et al. The identifica tion of domestic and foreign automobile manufacturers through body primer characterisation. ) Forensic Sci 1989; 34: 1395-404. Dolan DN. Vehicle lights and their use as evid ence. ) Forensic Sci Soc 1971; 11 :69-82. Ambourn RF. The calculation of motor car speeds froll1 curved tire mark s.) Forensic Sci 1989; 29:371-86. Ryland SG, Kopec RJ. The evidential va lue of automobile paint chips.) FOre11s ic Sc i 1979; 241(1):140-7. Monahan DL, Harding HWJ. Damage to clothin g: cuts and tears. ] Forensic Sci 1990; 35:901-13. Tun bridge RJ, Everest JT. Wild BR. Johnstone RA. An In Depth Study of Road Accident Casualties and their Injury Pattern s. Report RR. Crowthorne: Transport and Road Research Labo ratory, 1988, pp. 136-7. Ward H, Norrie J. Sang A et al. Urban Safety Project: the Sheffield Scheme. Contractors report CR J 34. Crowthorne: Departm ent of Transport, Transport and Ro ad Research Laborato ry, J 989. Kondo T, Sa ito K, Nishigami J. Ohshima T. Fata l injurers of the brain stem and/or upper cervical spinal co rd in traffic cases; nine autopsy cases. Sci Justice 1995; 3 5:197-201. Gunji H, Mitusaawa I, Hiraiwa K. The mecha nism underlying the occurrence of traumatic brainstem lesi ons in victims of traffi c accide nts. Legal Med 2002; 4:84-9. Shkrum MJ, McCafferty KJ. Green RN et a l. Mechanisms of aOltic injuri es in fatalities accming in moto r vehicle collisi ons. ) Forensic Sci 1999; 44:44-60. Feezko JD, Lynch L, Pl ess JE et aJ. An autopsy review of 142 [non-pen etrating] blunt injuries to the aorta. ] Trauma 1992' 33:846-9. .
20 Lo wne R, Robelts A, Roy P et al. The effect of UK Seat Belt Legi s lation on Restraint Usage by Children. Technical Paper Seri es 840526. New York: Society of Automotive Engineers (SAI), 1984. 21 And erso n PA, Rivara FP. Maier RV, Drake C. The epidemiology of seatbelt-associated injuries. ] Trauma 199J; 3:60-7. 22 May AK, Chan B, Da niel TM , Young JS. Ante rior lung hern iat ion: another aspect of the seat-belt syndrome. ) Trauma 1995; 38:587-9. 23 Reid AB. Letts RM, Black GB. Pediatric chance fractures: association with intra- abdo minal injuries a nd seat belt use. ) Trauma 1990; 30:384-91.
24 Riches KR. James RA, Gilbert JD , Byard RW. Fatal childhood vascular injuries associated with seat belt usage. Am) Forensic Med Pat/lOi200]; 22: 193 - 5. 25 Beattie YT, Ri chard s D, Belton N et al. Injuries in the adolescent population in Scotland: patterns and types of injury sustained. Heath Bulletin 1999; 57: 165-74. 26 Hitosuigi H, Takatsu A, Sigeta A. Injuries of motorcyc lists and bicyclists examined at autop sy. Am ) Forensic Med Pathoi J999; 20:251 -5. 27 Wilson MH, Baxter SP, Tenet SP et al. Saving ChildTeI1. A Guide to Injury PreIJelltion. Oxford: Oxford University Press, J991. 28 Hillman M. Cycle Helmets: The Case For and Agail1st. London: Policy Studies Institute, 1993. 29 Thomas S. Acton C. Nixon J et al. Effectiveness of bicycle helmets in preventing head injury in children - a case control study. 8M) 1994; 308:J73-6. 30 Cameron M, Vulcan A, Finch C et al. Mandatory bicycle helmet uses following a decade of helmet promotion in Victoria. Australia: an evaluation. Accidel1t Anal Pre/} 1994; 26:325-37. 3 1 McDerm ott FT. Bicyclist head injury preventi on by helmets and mandatory we aring leg islation in Victoria Australia. Ann R Coli Surgeons 1995 ; 7 :38-44. 32 Tong KA. Ashwal S, Holshouser BA et aJ. Hemorrhagic shearin g lesions in children a nd adolescents with posttraumatic diffuse axonal injury: improved detection and initial results. Radiology 2003; 227:3 32-9. 33 Holshouser BA. Tong KA, Ashwa l S. Proton MR spectroscopic imagin g depicts diffuse axonal inju ry in children wi th traumatic brain injury. Am) Neuroradioi 2005; 26: 1276-85. 34 Freem an MD. Croft AC, Rossignol AM. Whiplash-associated disord ers: redefining whiplash and its manage ment. Quebec Task Fo rce. A critical eval uat ion. Spine 1998; 23: 1043-9. 35 Minton R. Galasko CSB. Murray PA, Pitcher M. Whiplash injuries, causati ve studies: final report 1998. Report of the Transport Research Laborato ries ISBN: 1-84608-256-0. 36 Patterson RM. Trauma in pregnancy. Ciin Oustet GYl1eco/ 1984; 27:32-8. 37 Corsi PRo Rassian S, Bechelli de Oliviera L et al. Trauma in pregnant women; analysis of maternal and fetal mortality. Trauma 1999 ; 30:239-43. 38 Stafford PS, Biddinger PW, Zunwaite RE. Lethal intrauterine fet al trauma . Am) Ousret GYlleco/ 1988; 159:484-9. 39 Weinberg L. Wyatt JP, Busuttil A. Traum atic intrauterine fetal spina l fracture following seat belt use : a case report. ] Trauma 2001; 51: 1195-6. 40 Mamhews CD. In correctly used seat belts associated with uterine rupture following v ehicular collisio n. Am ] Oustet
GYl1eco /1975; 121:111 5-16. 41
Schoenfelkd A. Ziv E, Stein Let al. Seat belts in pregnancy and the obstetrici an. Obstet Gynecol Surg 1987; 42: 275 -82. 42 Griffiths M. Hillm an G. Usherwood M. Seatbelt hazards in pregna ncy (Case report). Br ] Obste/ Gynaecoi 1991; 98:320-1. 43 Byard RW. Gilbert J, Lipsett J. James RA. Farm and tractor related fatalities in children in South Australia. ) Paediatr Child Health 1998; 34 :139-41. 44 Byard RW, Gilbert J, James RA, Lipsett J. Farm and tractor related fatalit ies in children. ] Forensic Med Pachol 1999' 20:73-7. . 45 Road Accidents - Scotland 2006. Edinburgh: A Scottish Executive National Statistics Publication. 2007. 46 Andersson A, J onasso n K. Proceedings oj the Si.xth
Il1ternatiol1al COIlJerenceJor Accidellt and-TraJfic MedicinE. Melbourne. 1977 ; 279.
I
CHAPTER 21
I
FORENSIC DNA PROFILING IN CASES INVOLVING CHILDREN Alex M Graham and David J Harrison
Introduction Inheritance of genetic materia l Forensic DNA analysis: history and techniques Sample collection and processing DNA evidence and child sexual offence Y chromosome short tandem repeat typing Mixed samples Additional sample problems and solutions
395 395 398 402 403 405 406 407
INTRODUCTION
For centuries philosophers have deliberated which came first, the chicken or the egg. In the context of forensic paediatrics, the egg - or rather t he fertilized egg - bein g the precursor and genesis of the livin g child, is the appropriate place to start off the discussion on the role of genetics. It is, after all, the fertili zed egg that develops into all 100 trillion cells that make up the human being, and it is the genetic information that comes together from mother and father at the point of fertili za tion that provides the instructions for the creation and development of the child.] While the result is obviously of biological relevance, it is the analysis of the individuality of the inh erited complement of genetic mate rial from person to person that has proved of significan t value in a forensic context. In a forensic investigati on, genetic evidence may be used to confirm or refute a known individual as being the likely so urce of trace material or to identify a body or body remains. This chapter will focus on the above use of DNA in paternity testing. 2
INHERITANCE OF GENETIC MATERIAL Humans inherit half of their genetic material from each bio logical parent in the fo nn of deoxyribonu cleic acid (DNA) molecules neatly packaged up into chromosomes. The mother's egg contrib utes 23 chromosomes, 22 autosomes
Mitochondrial DNA Paternity testing Identification of body remains and missing persons Identification of the 'abandoned baby' or fetal materi al and avenues for identifying the source of an unknown profile DNA databases References
407 408 411
413 414 414
numbered 1-22 in decreasing order of size, and one sex chromosome, in this case an X chromosome. The father's spenn contributes another 23 autosomes, again numbered 1-22, plus either an X or a Y chro mosome, which determines the sex of the offspring.] When an ovum fuses with a spenn car!)'ing an X chromosom e, resulting in two X chromosomes in the fertilized egg, the conceived child is female, denoted XX. When an egg ce ll fuses with a sperm carrying a Y chro mosome the offspring is male, denoted XY.4 The chromoso mal complement of the fertilized egg and of the offspring is, therefore, 46 (Fig. 21. 1), which makes up the individual's genome. Figure 21. 1 shows these chromosomes as a human male ka!)'otyp e in which the 22 pairs of autosomes are neatly aligned in descending order of size, followed by the X and smaller Y sex chromosomes. In a femal e ka!)'otype, the Y chromosome is replaced by a second X chromosome. Because it is the male that carries the Y chromosome in humans, the presence of Y chromosome DNA in a genetic sample indicates that a male has contributed to the sample, and thus, this is a useful way of determining the sex of ori gin of a sample. At the point of fertilization it is not possible to predict whether the sperm cell that fuses with an egg cell carries an X or a Y chromosome and, hence, which sex the offspring will be; likewise, one cannot predict which one of the par ents' two copies of each of the other 22 chromosome pairs will end up in the egg and sperm cells (also known as gametes or germ cells) that combine. The actual combination
396 I
Forensic
DNA
profiling
2
14
5
4
8
6
13
3
9
10
15
12
11
16
17
18
I
19
20
21
22
x
y
Figure 21.1 Karyotype of a normal human male showing chromosomes in a condensed state. Chromosomes were stained, photographed, then cut out, paired with their partner and the pairs aligned in descending order of size. The sex chromosomes, the Xand smaller Y chromosome, are positioned at the end of the pairs of autosomes. (Courtesy of RMurray, South East Scotland Cytogenetics Service.)
is the result of the parental chromosome pairs randomly and independently segregating during gamete cell production by the process of meiosis.4 It is this randomness, in part, that gives rise to the diversity of genetic material from individual to individual. Because which of the two possibl e copies of each of the 23 chromosomes of the parent that enters the egg or sperm cell is random, there are 223 possible combinations of the mother's chromosomes in a human egg cell and, like wise, 223 possible combinations of the father's chromosom es in a hum an sperm cell. Therefore, when a human egg cell and sperm cell randomly fuse, the result is one of 223 x 223 possible different combinations of the parents' original chro mosomes, which equates to one of over 713 possible combi nations of the parental chromosomes in the resultant offspring. This helps to explain why siblings, despite being deJived from the same parents and inheJiting half of each of their parents' genetic mateJial, have a different combi nation of chromosomes and a different genetic make-up. In actual fact there is a further level of genetic variation. This arises because a process aki n to chromosomal shuf fling takes place between pairs of chromosomes, known as homologous chromosomes, such that the members of a pair swap segments of DNA to give rise to novel sequences along the length of the chromosome compared with the original chromosomes. In this way, one can appreciate the biological mechanisms that contribute to the interindividual genetic variation, the result of which we can see all around
us in terms of the diversity of features and characteristics that we all possess. It gives us an understanding of why we feel justified in saying that, with the exception of identical (monozygotic) twins, it is highly probable, thoug h not real istically feasible, to prove definitively that every individual has a unique combination of chromosomes and a unique genetic sequence. Monozygotic twins are the exception, because they are derived from a single fertilization event when the embryo splits to form two individuals at an early stage of develop ment. Thus, monozygotic twins have the same chromoso mal and DNA complement. This precludes existing forensic DNA analysis techniques from discriminating between their DNA, and is a factor that should not be overlooked when claims of the power of DNA evidence are made. Conven tional forensic fingerprint evidence is increasingly coming under attack for its alleged lack of scientific basis and cred ibility an d question ab le reliability, as exemplified by recent high-profile cases of misidentification, such as the Madrid bombing and the Shirley McKie case. 5 - 11 One feature of conventional fingerprinting that is often cited in its favour is its abi lity to differentiate between identical twins. How ever, such cases are equally, if not more, susceptible to the assumptions and un certainties of the technique. 12 Nearly all body cells of a human possess the full comple ment of 46 chromosomes and, as such, are said to be in the diploid state. The chromosomes are comp artmentalized into
Inheritance of genetic material an orga nelle of the cell known as the nucleus. Any cells with a nucleus are, therefore, a potential source of this nuclear DNA, which can be used in an investigation. Notabl e excep tions include the egg and sperm germ cells previously men tioned, which only have half the number of chromosomes and are said to exist in the haploid state, and mature red blood cells (erythrocytes), which are anucleate and lack nuclear DNA. However, an ucl eated cells are not totally devoid of DNA because, li ke nucleated cells, they have a much smaller DNA component which is stored in the cells' mi tochondria and is known as mitochondrial DNA (mtDNA). Mitochondrial DNA has its own characteristic inh eri ta nce patterns and features which make its analysis a valu able forensic tool in its own righ t in some situations. These will be conside red later in this chapter.
I
397
Vt=:::ILC:;;:;~_ Sugar phosphate backbone
~~~~~~'- Adenine
~=:TI:t=~~~- Cytosine Thymine
-*=::::::Jli=~~
Guanine --h~3:tt=;;>
Deoxyribonucleic Acid While morphologically the members of a homologous chromosome pair in any individu al look the same (see Fig. 21.1) and possess a high degree of similarity across their length, they are not identical. This is due to the specific make-up of the DNA that is transmitted by way of the chromosomes. DNA lies at the heart of genetic analysis, and in order to appreciate the techniques employed and the evidence derived from forensic DNA analysis a basic level of understanding of the molecule is required. DNA is made up of three components: a sugar part (deoxyrib ose), a phosphate part and a nitrogenous base (Fig. 21.2). The sugar and phosphate components are joined by phosphodiester bonds to form the sugar-p hosphate backbone of the molecule, but it is the bases that link two of these strands together, in what has been likened to rungs of a lad der bridging two strands of linked sugar-phosphate units, that are of interest in terms of conferring attributes and individuality on living beings. There are four bases, adenine, cytosin e, guanine and thymine, commo nly referred to as A, C, G and T. These bases hold the two strands of DNA together in a very spe cific way: A can pair only with T (and vice versa) and C can pair only with G (and vice versa). When correctly paired , the two complementary strands can spi ral into the recog nizable double-helix form that is so familiar (Fig. 21.2). It is the specific sequence of the bases along the length of the chromosomes that may vary between members of a homologous chromosome pair in an individual and between chromosomes from different individuals. It is this sequence variation that differentiates us at the genetic level and which may be used as a means of tracing a DNA sample to its source.
Genes. Alleles and Genetic Variation Some of these vari ations may confer phenotypic effects, i.e. characteristics that can be observed. These lie in regions of
~=TIt:::~~i\- Nitrogenous base
L....l_ _ _ _
Base pair
Figure 21.2 Structure of deoxyribonucleic acid (DNA). This figure represents the double-stranded he lica l structure of DNA. The molecule comprises two strands of linked sugar and phosphate molecules. The sugar-phosp hate backbone is held together by nitrogenous bases forming pairs between adenine and thymine or cytosine and thymine. These linked strands then twist to form a helical spiral as shown. (Courtesy of Talking Glossary of Genetic Terms, National Human Genome Research Institute, National Institute of Health, USA: www.genome.gov/10002096.)
the genome called genes, which provide the instructions to make specific proteins that build and run our body, for example the pigments that give hair its natural colour, the enzymes that digest our food and metabolize ingested drugs, the ion channels found in cardiac cells, or compo nents of the immune system. The last two examples may be of paliicular interest in the investigation of sudden infant death syndrome (SmS) as particular variants have been implicated as putative contributors to a subset of SIDS deaths.13 Di fferent variations of a paliicul ar gene's sequence, known as alleles, may typically take the form of a change in one or more base to an alternative base(s), the insertion of extra bases or deletion of sections of DNA. These differences mayor may not alter the fun ction or production of the pro tein product. For example, the different colours of eyes of different people result from different alleles of the genes that determine eye colo ur. 14 Whether a child suffers from the genetically inherited condition cystic fibrosis depends on which combination of allelic variants of the gene encoding the cystic fibrosis transmembrane conductance regulator
398 I
Forensic DNA profiling
(CFTR) protein are inherited. 15 Sufferers from this condition inherit two faulty, or mutated, versions of the gene, which affect the production or function of the gene product and give rise to the observed clinical syndrome. 16 Different alleles of particular genes may be of impor tance in investigating the cause of death in a child or infant if they affect the viability of the offspring, but in the major ity of conventional forensic work the focus is on regions of the genome that Jie outside the genes because this is where maximum interindividual variation in genetic sequences is believed to exist and, hence, where the power of discrimina tion is greatest. Interestingly, recent findings suggest that the level of variation between individuals is much higher than previ ously thought. Large segments of DNA within and outside genes are present in the genome of different individuals a different number of times. This phenomenon, called copy number variation (CNY), means that variation between indi viduals is 5- to lO-fold greater than previously believed. 17.18 That said, it is amazing how similar our genomes are. Prior to this study it was generally accepted that over 99 per cent of the genome is the same in every individual, not actually much of a difference, especially when you also consider also that there is only a couple of per cent difference between humans and chimpanzees! 19,20
-- - - -
--
--
-
Figure 21.3 Deoxyribonucleic acid (DNA) samples analysed by restriction fragment length polymorphism (RFLP) in a paternity case. Each column represents a sample: M, mother; F, father; C, child. It is clear that each individual is different. However, closer examination of the children's samples reveals that whereas every band in the profile of child 5 (C s) is seen in one of the parents, as would be expected, the profile of child 7 (C 7) includes bands that are not seen in either parent, indicating that this child is not their biological offspring. (Courtesy of Cellmark Forensic Services, Abingdon, Oxfordshire.)
FORENSIC DNA ANALYSIS: HISTORY AND TECHNIQUES Restriction Fragment Length Polymorphisms From a historical perspective, the advent of forensic DNA analysis as a means to identify, visualize and trace the interindividuality of genetic material, enabling a sample to be matched to its source, was somewhat serendipitous. Alec Jeffreys, dubbed the 'father of DNA analysis', was not aiming to develop a forensic investigative tool when he discovered a potentially groundbreaking way of visualiz ing individual genetic specificity in the 1980s; in fact, he was looking for a human version of the myoglobin gene in seal meat. 21 Jeffreys used a technique called restriction fragment length polymorphism (RFLP) analysis in which genomic DNA is cut or digested in a sequence-specific way by restriction enzymes; this results in a mass of fragments of DNA of differing lengths. These fragments can then be sep arated on the basis of their size and electric charge by a process called gel electrophoresis; specific fragments that share similarity in sequence are visualized by binding a labelled DNA probe of complementary sequence to the sought-after sequence, which is shared by a subset of the fragments. 22 ,23 The result is somewhat akin to a 'barcode', with DNA from different individuals forming different-sized fragments that are represented by bands positioned at a dif ferent location along the length of the developed 'barcode'
(Fig. 21.3). Because of the analogy of comparing bands on an electrophoresis DNA gel with points of comparison of conventional fingerprints, Jeffreys named the technique 'DNA fingerprinting'. He claims it was coining this term that attracted attention to the potential forensic application of the technique. Using RFLP analysis, it is possible to compare DNA extracted from biological samples from a crime scene with those of a suspect; if the DNA patterns match, the sam ples most probably came from the same individual. If the DNA patterns do not show the same pattern of bands, the suspect can be eliminated as a likely source of the crime scene sample. This was, in fact, the outcome of the first application of the technique in a forensic case. Two 15-year-old girls in Narborough, Leicestershire, UK, were sexually assaulted and murdered. RFLP analysis showed that DNA extracted from semen from both crime scenes came from a common source, and thus a single man was deemed to have been responsible for both crimes. How ever, the crime scene DNA did not match DNA from a blood sample taken from the prime suspect. Subsequently, the first mass DNA intelligence screen (DNA man hunt) was launched. However, in an interesting twist in this investiga tion, the DNA screen failed to identify a match to the crime scene profiles. The offender, Colin Pitchfork, was eventually identified and convicted after it was overhead in a pub that a colleague had been persuaded to give a DNA sample on
Forensic DNA ana lysis I
his behalf. Subsequently, Pitchfork's DNA was found to share the same RFLP pattern as the crime scene samples. Pitchfork was convicted of the murders and sentenced to life imprisonment in 1988. 24 ,25 This case provides the first example of a suspect being exonerated on the basis of DNA evidence, an outcome of many subsequent analyses that should not be overlooked. Although DNA evidence has often been hailed as being responsible for Pitchfork's con viction, some argue that it was, in fact, 'an old-fashioned tip-off from an informanf 26 that resulted in the case being solved. However, there is little doubt that the probability that two non-related individuals will show the same RFLP pattern by chance is extremely low; this has led to many convictions when samples from a crime scene and suspect matched.
Short Tandem Repeat Profiling PRINCIPLES
Although RFLP analysis is highly effective and shows a very high level of discrimination, practical issues, namely the requirement for large amounts of high-quality DNA (which is often a luxury in forensic investigations), the need for radioactive materials in RFLP analysis initially, the com plexity of analysing and interpreting the results (Fig. 21.3) and the development of new techniques to analyse DNA variants, mean that the majority of forensic DNA analysis today focuses on a specific type of variation called micro satellite repeat polymorph isms, rather than the longer mini satellite repetitive DNA on which RFLP was based. A polymorphism simply refers to the occurrence of more than one allele or variant at a particular region or locus in the genome within a population. Strictly speaking, to count as a polymorphism the allele must be present in the popu lation at a frequency greater than 1 per cent; rarer variants or de novo events giving rise to new sequences are termed mutationsY The important point to be aware of is that dif ferences exist and, in terms of differentiating between indi viduals, any difference is of relevance. Interspersed througho ut the genome are stretches of microsatellite DNA in which short core sequences (4-6 bases) of DNA are found to be repeated in tandem, multiple times, the exact number of repeats varying from chromo some to chromosome and from individual to individual. A number of these so-called short tandem repeats (SIRs) have been found to be highly polymorphic with lots of different alleles, corresponding to the number of times the core sequence is repeated. STR polymorphisms have proven to be an ideal tool to use to compare DNA samples and look for similarities and differences between individuals. The greater the level of variation in repeat number at an SIR locus in t he population, the greater the power of dis crimination. By analysing the number of times a particular selection of STRs is repeated on each homologous chromo some in an individual, it is possible to build up a profile of the
399
individual's DNA. An individual who has an SIR, for exam ple THOI, which resides on chromosome II and has the core sequence AATG, repeated six times on one copy of chromo some 11 and nine times on the other chromosome 11 would be denoted as having a genotype 6, 9 for that SIR. Using esti mates of the frequency at which each known allele ex ists in the population, one can determine the probability that the particular allelic combination in question would arise in a person chosen at random from that population. For example, using data from the US Caucasian population, one can esti mate that this allelic combination at THOI would be found at a frequency of 0.053 (or 5.3 per cent) in the population. 2s If a crime scene sample sho ws the genotype 6, 9 at THOI, it wiJl be appreciated that looking at this single locus or region alone will not be particularly helpful in identify ing the individual who left the DNA because many people in the population are likely to share this genotype. This is similar to the situation that may arise using conventional blood group typing as an investigative tool: many people have the same blood group and thus the discriminatory power of blood group typing is limited. 29 However, if one also looks at other STR loci on different chromosomes, then the more loci that are examined, the less chance there is that any two unrelated individuals wi ll have exactly the same combination of a ll eles at every STR examined. TECHNIQUE
The result, in terms of STR repeat numbers, of analysing a selection of SIR loci is known as an SIR profile. This can be illustrated graphically, with the different alleles being repre sented as peaks at different positions along a horizontal axis, an example of which can be seen in the electrophoretogram shown in Fig. 21.4. These profiles are obtained by extracting DNA from biological samples from a wide range of possible sources, the proviso being that the sample contains cells that contain a nucleus and that these have DNA of sufficient quantity and quality. Although the erythrocytes in blood lack the requisite nuclear DNA, the white blood cell compo nent makes blood a particularly good source of DNA for forensic analysis. Blood is frequently analysed in a range of investigations, for example, child murder or assault, or for diagnostic purposes in relation to genetically inherited con ditions. Semen, saliva or fingernail scrapings may be exam ined in the case of alleged child sexual abuse; buccal swabs may be taken from subjects in a paternity dispute or from suspects apprehended by legal autho rities. Bones, skeletal remains or teeth are particularly useful in cases in which body remains are badly decomposed, have suffered fire dam age or have been exhumed, because of the increased resist ance of DNA from these sources to degradation. Muscle tissue, skin, fetal material, placental tissue, urine or faecal material and inanimate objects QJ1 which DNA traces may have been deposited (such as cigarette butts, postage stamps, chewing gum and toothbrushes) are also suitable for DNA ana lysis (see Butleilo for further references). It is of note that
Forensic DNA analysis I
his behalf. Subsequently, Pitchfork's DNA was found to share the same RFLP pattern as the crime scene samples. Pitchfork was convicted of the murders and sentenced to life imprisonment in 1988. 24 ,2 5 This case provides the first example of a suspect being exonerated on the basis of DNA evidence, an outcome of many subsequent analyses that should not be overlooked. Although DNA evidence has often been hailed as being responsible for Pitchfork's con viction, some argue that it was, in fact, 'an old-fashioned tip-off from an informanf 26 that resulted in the case being solved. However, there is little doubt that the probability that two non-related individuals will show the same RFLP pattern by chance is extremely low ; this has led to many convictions when samples from a crime scene and suspect matched.
Short Tandem Repeat Profiling PRINCIPLES Although RFLP analysis is highly effective and shows a very high level of discrimination, practical issues, namely the requirement for large amounts of high-quality DNA (which is often a lux UlY in forensic investigations), the need for radioactive materials in RFLP analysis initially, the com plexity of analysing and interpreting the results (Fig. 21.3) and the development of new techniques to analyse DNA variants, mean that the majority of forensic DNA ana lysis today focuses on a specific type of variation called micro satellite repeat polymorphisms, rather than the longer mini satellite repetitive DNA on which RFLP was based. A polymorphism simply refers to the occurrence of more than one allele or variant at a particular region or locu s in the genome within a population. Strictly speaking, to count as a polymorphism the allele must be present in the popu lation at a frequency greater than 1 per cent; rarer variants or de novo events giving rise to new sequences are termed mutations. 27 The impoliant point to be aware of is that dif ferences exist and, in terms of differentiating between indi viduals, any difference is of relevance. Interspersed throughout the genome are stretches of microsatellite DNA in which short core sequences (4-6 bases) of DNA are found to be repeated in tandem, multiple times, the exact number of repeats varying from chromo some to chromosome and from individual to individual. A number of these so-called short tandem repeats (STRs) have been found to be highly polymorphic with lots of different alleles, corresponding to the number of times the core sequence is repeated. STR polymorphisms have proven to be an ideal tool to use to compare DNA samples and look for similarities and differences between individuals. The greater the level of variation in repeat number at an STR locus in the population, the greater the power of dis crimination. By analysing the number of times a particular selection of SIRs is repeated on each homologous chromo some in an individual, it is possible to build up a profile of the
399
individual's DNA. An individual who has an STR, for exam ple TH01, which resides on chromosome 11 and has the core sequence AATG, repeated six times on one copy of chromo some 11 and nine times on the other chromosome 11 would be denoted as having a genotype 6, 9 for that STR. Using esti mates of the frequency at which each known allele exists in the population, one can determine the probability that the particular allelic combination in question would arise in a person chosen at random from that population. For example, using data from the US Caucasian population, one can esti mate that this allelic combination at TH01 would be found at a frequency of 0.053 (or 5.3 per cent) in the popuJation. 2o If a crime scene sample shows the genotype 6, 9 at THO]' it will be appreciated that looking at this single locus or region alone will not be particularly helpful in identify ing the individual who left the DNA because many people in the population are likely to share this genotype. This is similar to the situation that may arise using conventional blood group typing as an investigative tool: many people have the same blood group and thus the discriminatory power of blood group typ ing is limited. 29 However, if one also looks at other STR loci on different chromosomes, then the more loci that are examined, the less chance there is that any two unrelated individuals will have exactly the same combination of all eles at every STR examined. TECHNIQUE The result, in terms of STR repeat numbers, of analysing a selection of STR loci is known as an STR profile. This can be illustrated graphically, with the different alleles being repre sented as peaks at different positions along a horizontal axis, an example of which can be seen in the electrophoretogram shown in Fig. 21.4. These profiles are obtained by extracting DNA from biological samples from a wide range of possible sources, the proviso being that the sample contains cells that contain a nucl eus and that these have DNA of sufficient quantity and quality. Although the erythrocytes in blood lack the requisite nuclear DNA, the white blood cell compo nent makes blood a paliicularly good source of DNA for forensic analysis. Blood is freque ntly analysed in a range of investigations, for example, child murder or assault, or for diagnostic purposes in relation to genetically inherited con ditions. Semen, saliva or fingernail scrapings may be exam ined in the case of alleged child sexual abuse; buccal swabs may be taken from subjects in a paternity dispute or from suspects apprehended by legal authorities. Bones, skeletal remains or teeth are particularly useful in cases in which body remains are badly decomposed, have suffered fire dam age or have been exhumed, because of the increased resist ance of DNA from these sources to degradation. Muscle tissue, skin, fetal material, placental tissue, urine or faecal material and inanimate objects 0)1 which DNA traces may have been deposited (such as cigarette butts, postage stamps, chewing gum and toothbrushes) are also suitable for DNA analysis (see Butler.30 for fUliher references). It is of note tha t
400 I
Forensic DNA profiling
_
03S1358
vWA
_
t
• A
1500
--~~ImUJ;---------------/}fE~----------~~----.L[ ~~~O
j
-
021S11
08S1179
,.------~
m
j
>----_
_
J
----~
~ m
jm
III 05S818
018S51
;---_~
0138317
A
2000
1500 1000 500
Ilil
07S820
~H~
m----------------~~m~------------'m m
Figure 21.4 A short tandem repeat (STR) profi le. An electrophoretogram show ing the deoxyribonucleic acid (DNA) profile of a sample analysed at nine STR loci (three on each horizontal row) and a sex-determining region, ame logenin (labelled AMEL on the second row), using the comm ercial Profiler Plus® kit (Applied Biosystems, Foster City, CAl. Each peak represents a different allele and the number in green below the peaks denotes the number of times the particular STR sequence is repeated. Where two peaks are seen at a particular STR locus, for example, 14 and 15 for D3S 1358 on chromosome 3, the individual has a diffe rent number of repeats on their homologous chromosome 3s; where a singl e peak is seen at a particular STR locus, for examp le, 13 for D8S1179 on chromosome 8, the individual has the same number of repeats on both copies of their homologous chromosome 8s. In this example a single peak representing the X chromosome is seen at the amelogenin locus, indicating that the sample was from a female. (Courtesy of Dr JR Gilder, Forensic Biomatics, Inc., Fai rborn, OH, USA.)
with the advent of lo w-copy-number (LeN) DNA analysis, it is possible to obtain a DNA profile from as little starting material as a single cell or the few epidermal cells left behind in a fingerprint, and trace quantities of DNA that may sub sequently be analysed may be deposited simply by touching articles or surfaces. 31-35 However, the evidential value should be treated cautiously and the serious technical and interpre tative issues relating to such evidence should be taken into account. Once a DNA sample has been extracted, it is amplified by a process called the polymerase chain reaction (peR) in order to generate a sufficient quantity of DNA for detec tion. 36)7 peR is a technique that revolutionized molecular biology w hen it was developed in the 1980s, and its intro duction enabled the shift from RFLP to STR profiling. It provides a simple and readily automated way to replicate specific preselected regions of the genome so that, from a small number of initial copies of the specific DNA sequence of interest, millions of copies of that sequence can be gen erated in a cyclical process. During each cycle of the reac tion, the number of copies of the targeted templ ate sequence of interest is doubled, resulting in an exponential in crease in the tota l amount obtained after multiple cycles. In the case of STR profiling, the target sequences within the genome are a selection of STRs distributed on different chromosomes of the genome. With the use of commercial kits, it is possible to amplify as many as 10 SIRs in a single reaction tube . These kits app ly a fluorescent label or tag to the different STRs to enable them to be detected and
differentiated from each other. The kits have been designed and validated for specific conditions and quantities of tem plate DNA, conditions not met by the aforementioned LeN analysis. In order to compensate for the trace levels of DNA that would not otherwise be detected, LeN analysis increases the number of rounds of peR amplification, but the cost is an increase in potential artefacts. Spurious alleles can be detected and genuine alleles may fail to show Up,34 casting severe doubts on the acceptability of the results as being of a sufficient evidential weight for a court of law. Amplification of selected STRs of the sample DNA results in a mixture of amplified STR fragments of different length that need to be separated and visualized. This is done by a process of gel electrophoresis or, more recently, capillary electrophoresis, which separates the labelled fragments on the basis of the electric charge and size of the fragments. 38 The fluorescent dye-l abelJed fragmen ts are passed through a reten tive material (either a polyacryla mide gel or viscous polymer), wh ich retards their migration through the material under the action of a potential difference. Shorter fragments pass through the material more quickly and are detected by a laser detector and registered as a peak on the electrophore togram before the longer fragments. In this way profiles from different samples can be generated and compared. In addition to the SIR loci that are analysed, a gene on the sex chromosomes called amelogenin is also typed in order to identify the sex of the source of the sample. Although much of the X an d Y chromosomes differ from each other and are non-homologous, there are some stretches of the chromosomes
-
Forensic DNA analysis I
that show homology, and the amelogenin gene is one of these. There is a difference in size between the X amelogenin gene and the Y amelogenin gene, which can be differentiated on the basis of the size of the PCR products when the amelo genin gene is amplified in the same way as the SIR loci described above. When only a single peak corresponding to the X version of the gene is detected, the sample is taken to be from a female; when two peaks corresponding to the X and the Y copies of the gene are detected the sample can be considered to be from a male. Considering any locus that is typed, we cannot be absolutely confident in our interpretation in every case. Genetics is no exception to the rule that biological systems have their anomalies, and there are times when a sample shows only an X amelogenin signal when the true source is in fact male. As for all the STR loci, it is possible for DNA to be present at too Iowa concentration or to be too badly degraded to be detected, or for the reaction to fail; thus, a negative result does not necessarily mean there is no sam ple present. Potentially more interesting and problematic, in terms of forensic investigations, are cases in which a mutation may exist in an individual, resulting in anomalous results. For example, the Y amelogenin gene has been found to be deleted in a number of individuals, including specific Indian populations ; conventional sex typing would suggest that these males were female, with considerable consequences for forensic casework. J9 - 42 Such events are not restricted to the amelogenin locus; mutations giving rise to anomalous results can occur at any locus.
INTERPRETATION
If two samples analysed by STR profiling give the same result at all the loci examined, the samples are said to 'match' (Fig. 21.5). A match tells us that the two samples could have originated from the same source but is not defin itive in identifying the source of the sample. Who knows if one continued to compare additional regions of the genome whether a difference between the 'matched' samples might in fact be identified? One could tell this for certain only by sequencing the entire genome of the two samples, which is an unrealistic expectation given current technology. So it is important to remember that failure to show a difference should not be regarded as confirmation of identity - it is all down to statistical probability, which will be discussed later. On the other hand, if two samples do not match, i.e. they do not share the same combination of repeat numbers ,at every STR analysed, they would be deemed to have come from different individuals because there has been a failure to illustrate identity. Alternatively, the result may be incon clusive. This may be because the DNA is of poor quality or present in too Iowa quantity to produce sufficient data, or the data that are produced may be ambiguous or one may be unable to interpret the results, for example, in the case of mixed samples.
M,Xa.,.,:_
401
•••••••••••••••
BI:J\....___ -:1=-_____-::::u'--____.J. .P:~ IIil IIil m m
m
::JJ m
u
u
n~~
~--~IIil------~mm~-------L
m
~____~J~------~u~------~ti~~
IIil
IIil
m
m
m
r
Suspect 4
__~___-.:!J
m
m.--------..l..t
!-----.J,
mL---~m
4000 2000
III Figure 21.5 Comparison of short tandem repeat (STR) profiles at three STR loci (D3S 1358, vWA and FGA) from different sources. The profiles of samples ta ken from four suspects were compared with that of a blood stain found at a crime scene. The comparison shows that only suspect 3 'matches' the blood sample at these loci, i.e. shares every allele at each locus in common. The rema ining suspects would be excluded as a source of the blood stain. (Courtesy of Dr JR Gilder, Forensic Biomatics, Inc., Fairborn, OH, USA.)
It is worth noting that, in an analysis of DNA by STR pro filing, scientists are not able to tell if the sequence of As, Cs, Gs and Ts of two matched samples identified as having the same-sized SIR PCR product and concluded to share a par ticular number of repeats at the SIR locus is actually the same over the whole of the SIR region. What is being looked at are differences in lengths of sections of the genome rather than the actual sequence across the entire length. It is possi ble for two amplified STR products to show the same length. indicative of being the same, while actually differing in sequence when they are examined in greater detail. The investigative value of a match is typically presented in terms of the 'random match probability' (RMP). which expresses the likelihood that the match occurs by chance in the population. The RMP is the result of a calculation using statistics derived from population studies and assumptions that originate from the study of population genetics, in
402 I
Forensic DNA profiling
particular the Hardy-Weinberg equil ibrium. 43 ,44 Analysis has shown that in a stable, randomly mating popul ation, devoid of selective pressures that would alter the frequency at which existing alleles are found in the population , it is possible to predict the frequency of the possible genotypic combinations of these alleles in the population. 45 ,4G Using statistics, one can get an idea of the likelihood th at the given profile would arise by chance in the population from which the frequency estimates of the different a lle les have been derived and, hence, the weight of the evidence. One should not lose sight of the fact that the data derived from the population databases are based on predictions established from relatively small samples of individuals in the popUlation; the results genera ted are based on probabilities. The RMP is calculated by mult iplying together each of the predicted fr equen cies of occurrence of the profiled genotype for the STRs that have been typed in the sample using t he product rule to determine the probability that the given genotype would occur by chanceY The product rule can be used because each STR typed is situated on a differ ent chromoso me and, as describ ed at the start of this chap ter, c hromosomes segregate to the ga metes randomly and are inherited independently of each other. Th e number of STR loci analysed depends on whichjuris diction one is in and, thus, the pow er of discrimin ation ofthe resultant RMP. In the UK, 10 STR loci are routinely exam ined at present, whereas in the USA 13 loci are looked at, the latter number providing a greater power of discrimination. Despite the fact that such analyses are typically prese nted in terms of extremely sma ll RMP chance matches, in the order of one in a billion in the UK COUlts 48 and smaller in the USA, it is of note that Jeffreys has argued that we should be analysing as many as 16 STR loci to increase the power of discrimination in light of the increasing number of individ uals entered on the national DNA databases aga inst which the crime scene profiles are compared. 49 This is because the greater the number of profil es on the DNA databases with which unknown samples are compared, in order to identify a match, and thus a source of the sample, the greater t he probabi lity of a chance match. One only needs to look aro und at differe nt ethnic and racial populati ons to see that interpopulat ion genetic dif fere nces exist. As such, it is important to generate popul a tion databases for different race groups because certain alleles will be found to ex ist at a higher frequency in some populations than in others, for exa mple, the frequ ency of the nine-repea t allele at the D1 3S317 STR lo cus is 4.5 times higher in the US Hispanic popul ation than in the US African-American popul ation (Hisp anic frequen cy = 0.15357; African-American = 0.0 32955) (data taken fro m Butler2B ) . If one were to calculate the RMP for a sample based on data from a popul ation other th an the true source of that sampl e, one may end up with an RMP that misrep rese nts the true likelihood of a match. Whether the differ e nces would be statistically significant w he n dealing w ith
---
.
-~~~~~~~~
-
RMPs of the order of one in billions or less is debata ble, but for a calculation based on an incomplete profile in which not all the STR loci tested have generated a result, thus resulting in hig her probabilities fo r the RMPs being calcu lated, discrepancies may be more significant. The issue is ge nerally addressed by calcul ating an RMP based on the population da tabase that generates the most co nservative results, i.e. those most favourable to the defence in cases in which a suspect's DNA is found to match a DNA sample fro m a crime scene. Further issues regarding population structure may be take n in to account when deriving the predicted genotype fr equencies for population databases when one is dealing with populations with a hig h level of inbreed ing. It will be apparent that the closer the relationship of two indiv idu als, the more DNA they will have in commo n and, thus, the more likely it is for individuals from inb red populations to share a particular STR allele than would be the case for two non-related individuals in a randomly ma tin g population , i.e. the conditions of random mating fo r Hardy-Weinb erg equilibrium are not satisfied. If a susp ect a nd the true per petrator of a crime are blood relatives then their genotype frequencies will not be independent and a correction fa ctor or para meter measuring the popul ation differentia tion/ substructure, denoted as or FsT , is required. 47 ,48.50,51 An alterna tive to the RMP for the presentation of DNA evi dence is likelihood ratios. A likelihood ratio expresses the ratio of two alternative probabilities of the same evidence under different hypotheses. In the context of DNA evidence, when a suspect's DNA profile matches a crime scen e sample profile, there are two possible hypotheses : either th e sus pect's profile matches because his biological sample was found a t the Clime scene or the sample came from some unrelated individual who has not been arrested and who happens to match the suspect by cha nce. The like lihood ratio is usually calculated by dividing the likelihood of the prosecu to r's hypoth esis (the first above hypothesis) by the likelihood of the defender'S hypothesis (the second above hypothesis).47, 5! In its Simplest form, the likelihood ratio is the inverse of the RMP. However, things become more com plicated when alternative scenarios for the evid ence that is seen are presented, or if the DNA sample is a mix ture from more th a n one contributor. To give a feel for the numbers, a likel ihood ratio of 1000 or greater provides velY strong sup port for the prosecution's hypothesis, a likelihood ratio of 100-1000 provides strong supp ort and a likelihood ratio of 10-100 provides moderate supportY
e
SAMPLE COLLECTION AND PROCESSING The ge nera l principles underly ing DNA evidence as applied to a crim inal case involving a child, e.g. the identification of a n unknown child's body or remains or other forensi c investigations, are no different from those of an adult con text. Typically, a DNA sample believed to be central to the
---~~~~~~~~~~~~--
-
----
---
DNA evidence/child sexual offence I
crime or investigation - be it a blood sample, a sample of saliva swabbed from a bite mark or, in the case of alleged child sexual abuse, semen extracted from underwear, bed clothing or an intimate sample - is analysed to produce an STR profile, which is used to link a suspect to a crime scene. The profile is compared with a reference profile taken from a suspect, identified on the basis of additional evidence in the case or, if no leads are originally present to identify a suspect, with profiles lodged on a national DNA database. A profile on its own is of no investigative value unless there is a profile with which to compare it. Samples on such DNA databases may be associated with a known and named indi vidual or they may be from anonymous samples collected from previous unsolved crime scenes, which enables sus pects to be matched with crime scene profiles or crime scenes to be associated via matches. The legislation governing the power to collect DNA sam ples and retain profiles on the databases varies from nation to nation. In the UK such powers are wide-ranging, and DNA can be collected from anyone suspected of a record able offence and profiles retained indefinitely, regardless of conviction or acquittal of the suspect (Police and Criminal Evidence Act 1984 [amended]), but in other jurisdictions the power of the criminal investigative authorities may be much less. From the point of identification of a crime scene or sub ject from which a sample for DNA analysis is collected to the point at which a profile is generated and interpreted, strict adherence to protocols, an unbroken chain of custody of the evidence and careful sample handling must be adhered to. 53-55 The risk of contamination of a sample and the need for correct handling to avoid it cannot be overemphasized; pro fessional bodies have described procedures and guidelines to help ensure quality control and quality assurance in these processes, for example, the Federal Bureau of Investigation (FBIl-appointed Technical Working Group on DNA Analysis Methods (TWGDAM), now known as the Scientific Working Group on DNA Analysis Methods (SWGDAM), and the DNA Advisory Board (DAB).53,55,57 A case may be lost if doubt can be cast on whether samples have been correctly handled risk of contamination, compromise of a sample or poor handling - the most noteworthy of which is the trial of O.J. Simpson, accused of the murder of Nicole Brown Simpson in 1994;30,58 false convictions may be secured as a result of bad practice. 59 DNA is ubiquitous: we are all continuously shedding DNA and leaving a trace of our movements in this way. In fact, we need not even have been at a location ourselves in order for our DNA to be found there. It may be deposited by secondary or tertiary transfer, being carried and subsequently deposited elsewhere by an animate or inanimate object. 5O - 55 Individuals who pass through a crime scene after the event could unwit tingly leave their DNA, implicating themselves in the process. Likewise, it is possible for DNA to be present at a scene prior to the event and collected when samples were taken,55 or,
403
depending on the nature of the sample, a perpetrator may be able to suggest a plausible reason for the presence or his or her DNA at a Clime scene, leading to termination of an inves tigation. 57 It is crucial, therefore, that there is evidence to link samples directly to the event; that all persons entering the crime scene and crime scene officers, scientists or other indi viduals collecting samples take precautionary steps to avoid contaminating evidential samples with their own DNA, be this from sneezing, coughing, shedding skin or the like on to the sample; that clean equipment and gloves are used when handling each sample; that protective clothing is worn to limit contamination from the sampler; and that the samples are securely sealed and appropriately transported and stored to prevent degradation of the sample. 53 During processing of the samples, it is imperative that case samples and reference samples are handled separately and that samples that have not been amplified by PCR are kept away from those that have been amplified. It takes only a minute amount of ampli fied contaminant DNA to swamp any DNA present in a case sample, resulting in only the contaminant amplified DNA being detected, at the expense of the genuine sample DNA, when the two are simultaneously subjected to PCR. Plas ticware and consumables used in the analytical steps are another potential source of contamination as they may be contaminated when they enter the laboratory,55,68 and the problems of contamination are particularly apparent when LCN analysis is conducted. 34 ,54,59 Forensic laboratories have sbict standard operating procedures to which they adhere, such as the DAB Quality Assurance Standards of recommen dations for forensic laboratories, in order to standardize and validate procedures. 70 Despite all these efforts, errors will arise, be they the result of human factors, the nature or quality of the sample or technical or biochemical artefacts, and this should not be overlooked. 71 - 73 There is debate as to whether the exceed ingly small RMP figures that are presented in court are unfairly prejudicial when they are not accompanied by esti mates of these errors, particularly when the RMPs presented are smaller than the probability of laboratory error. How ever, some argue that a meaningful statistic for an error in a particular case cannot be provided. 74
DNA EVIDENCE AND CHILD SEXUAL OFFENCE The major problem facing the forensic investigation in cases of alleged child sexual abuse, and which is not unique to DNA evidence per se, is that the majority of cases arise ret rospectively, which presents several difficulties in terms of gathering forensic evidence (see Chapter 1). A number of other issues that relate to DNA analysis in such cases are also of relevance to DNA evidence in a more general sense, for example mixed samples, DNA quantity considerations and Y chromosome STR typing; discussion and introduction of these in the context of sexual abuse investigation seems appropriate.
404 I
Forensic DNA profiling
In the typical case of sexual abuse that comes to attention only months or years after the event, the opportunity to obtain DNA samples has passed by the time investigation begins. However, other indications, such as evidence of pregnancy or sexually transmitted infections, wh ich may be investigated using DNA techniques, can be highly probative should the opportunity for collection of immediate post coital samples not present itself. The occurrence of sexually transmitted infection in a child is unlikely in the event that sexual contact had not occurred, though vertical transmis sion is a possibility, and one also needs to be aware that con sensual sex with an infected partner may have taken place, rather than abuse. In one study, genital human papillo mavinls (HPV) infection, detected by the presence of HPV DNA from perineum and vaginal samples, was found in some of the sexually abused preadolescent girls examined but not in children in whom sexual abuse was excluded, the majority of the HPV DNA-positive cases being subclinical infections.75 It is noteworthy, however, that the incidence of sexually transmitted infections in victims of child abuse is low, a reflection of the generally low incidence of infection in perpetrators. In a recent study in Edinburgh, Scotland, sexually transmitted infection was seen in less than 10 per cent of referrals. 76 The appropJiateness of DNA testing for sexually trans mitted infectious agents using nucleic acid amplification test (NAATs)77 in suspected cases of child abuse has been ques tioned because of t he potential for false-positive results, with a possible consequence of false convictions. 78 .79 However, it is paramount that the well-being of the child is prioritized in any investigation and that the child's health is not compro mised at the expense of gathering forensic evidence. While NAATs may not be as specific as traditional culture tests, they show increased sensitivity. From a medical perspective, the downside of a false- positive result is limited to the cost and side-effects of superfluous antibio tic treatment com pared with the long-term potential serious health conse quences for the individual of untreated infection. 8o From a legal perspective, given that culture tests are sti ll the only test currently accepted by the UK courts 81 and some labora tories no longer conduct culture tests, having implemented NAATs alone,82 there are issues relating to obtaining suitable forensic evidence. The recommended approach is to take tvvo swabs so that a n initial NAAT-positive result can be con firmed by culture,81 though this involves additional distress and discomfort to the child. If pregnancy follows sexual abuse, it may be possible to gather evidence as to the likely father by way of DNA paternity testing of offspring or aborted fetal material ,83,84 a technique that is discussed later in this chapter. Although child sexual abuse includes a spectrum of activities from intercourse to physically less intrusive sexual ab use of a minor, whether samples taken for DNA analysis are informative depends on many factors. These include the nature of samples in relation to the offence; the time since t he offence; whether a male perpetrator ejaculated during
the abuse, was azoosperrnic, oligos permic or used barrier protection; and whether the victim subsequently washed, changed his or her clothes, ate, rinsed his or her mouth, uri nated or defecated, all of which can diminish the quality or quanti ty of detectable DNA, or even remove all traces of detectable DNA. In the event that a child does present in an acute case, conflicting opinions have been expressed about the value of ta king intimate fore nsic samples up to 72 hours after the event in prepuberta l victims 85 as recommended by the American Academy of Pediatrics. 86 It is important to weigh the distress and discomfort that may be inflicted on the victim against the likelihood of evidence being found. Christian et al 85 argue that, if one is seeking evidence of semen or sperm in a vaginal swab sample to provide objec tive evidence of sexual contact with a child who does not have a previous sexual histoty, data detived from adult cases is inappropriate to prepuberta l situations. While the litera ture reveals marked differences in survival time of sperm cells in different body orifices, the generally accepted maxi mum times for observin g sperm heads range from 24 hours in the mouth to 7 days in the vaginal cavity, and in the region of 2-3 days in the anus and rectum .87,88 This argu ment 55 may be va lid in the case of cytological observation of sperm cells or the detection of semen by, for example, acid phosphatase activity or the presence of phosphate-specific antigen (PSA), which have a shorter period of detection BG ,89 However, g iven the sensitiv ity of DNA analysis and its abil ity to detect trace material refractoty to other detection methods, particularly using LCN techniques,31-34 one should not discount the possibility of obtaining DNA evidence beyond the time periods suggested by Christian et al. 85 In one study, evidence of male DNA, detected by Y-STR ana lysis, was found in over 90 per cent of vaginal or ana l swabs taken from child sexual assault victims up to 72 hours after the incident,9o supporti ng the fll1dings of others that Y-STR DNA amplification can produce positive resu lts in swabs cytologically negative for spelm cells. Such negative results wo uld usu ally have led to tennination of biological investigation,91 illustrating the va lue of conducting the DNA analysis. The presence of Y-chromosome DNA in a high vaginal swab from a fe male would be interpreted as tanta mount to evidence of intercourse, and in the case of a minor, where consensual sex is not a defence aga inst statutoty rape, is highly probative evidence of abuse. There are practical considerations, in the case of a sma ll child, in ensuring that internal samples genuinely represent material exclusively of an internal nature. The anatomy of a young girl does not lend itself easi ly to collection of such samples, and care has to be taken to avoid contact with external regions of the child 's body. There may be perfectly acceptab le re asons for male DNA to be present on a child's body, particul arly if male relatives have frequent contact with a child, and sampling of t his material could lead to inappropriate conclusions being drawn, incrimin ating an innocent person. In this respect the use of the colposcope with the magnification and illumination that it affo rds has
-
.
Y chromosome STR typing I
assisted in the collection of non-contaminated intimate samples.
Y CHROMOSOME SHORT TANDEM REPEAT TYPING As mentioned above, if one can detect the presence of Y-chromosome DNA in a sample then, barring syndromes with abnormal chromosome complement or structure, one can conclude that male DNA is present. Such evidence is particularly useful in the case of all eged sexual abuse by a male against a female, although typing the Y chromosome of a sample has add itional valuable applications, including paternity testing and the identification of missing persons. If an internal vaginal sample is taken from a girl under the legal age for sexual intercourse and Y-chromosome DNA is detected, this is convincing evidence of abuse. In theory, one could test for male DNA using the amelo genin locus by conventional SIR typing, the X and Y amelo genin alleles being of different size. The problem with this approach is the frequently encountered imbalance in the rel ative quantity of any male DNA that would be present in sam ples routinely taken from the female victim in an alleged case of sexual abuse, i.e. vaginal, anal or oral swabs. Such samples typically comprise a vast excess of female DNA, usually from host epithelial or inflammatory cells, with only traces of male DNA, typically in the form of sperm cells, although epithelial and inflammatory cells may also be present, the last two being morphologicaJly indistinguishabl e from the female equivalents. Given the imbalance in the amount of male and female template DNA added to the PCR, there would be pref erential amplification of the female DNA to the extent that male DNA may go undetected. This is analogous to the situa tion described earlier of a DNA sample to be typed being con taminated with PCR-amplified DNA, due to poor laboratory practice, and swa mping the sma ll amount of DNA present in the sample, leading to preferential re-amplification and detec tion of the contaminant DNA. As ever, absence of detectable DNA is not evidence of absence. One way to get round this problem is to type a series of SIR loci which are specific to the Y chromosome (Y-SIR typing),92,93 so that female DNA is not amplified in parallel and a male-specific profile or Y-SIR haplotype results. There are hundreds of STRs distributed on the Y chromosome,94 and by typing a chosen set of these one can identify the par ticular Y haplotype of an individual male, i.e. the closely linked alleles of that individual at the chosen loci. One gets more information about the source of the samp le using Y-SIR profiling than if one was just to type the amelogenin locus to determine maleness because of the variation that exists in the Y chromosome between non-related males and between different ethnic, racial and geographic groups. However, the power of discrimination of Y-STR typing in no way approaches that of conventional autosomal STR typing. This is because the Y chromosome is passed from father to
405
son, essentially unchanged in sequence, so all male offspring through the generations descended from one particular male will share the same Y chromosome and, hence, the same Y-chromosome haplotype. Thus, given a Y-chromosome haplotype alone, it would not be possible to conclude that the man who was the origin of a DNA sample was its source, or whether it came from a blood relative such as his brother, son, father, grandfather, grandson, or uncle on his father's side of the family. A number of internet-accessible Y-SIR databases have been established,92 which vary in the number of SIR loci for which informatio n is avail ab le. For example, the frequ ently used and largest Y-SIR haplotype reference database (YHRD) was established in 2000 with the objectives of generating ' reliable Y-SIR haplotype frequen cy estimates for minimal and extended Y-STR haplotypes to be used in the quantitative assess ment of matches in forensi c and genealogical case work' and assessing the 'ma le population stratification among world-wide populations as far as reflected by Y-SIR haplotype frequency distributions'.95 The database can be searched for frequency data and geographical distribution data of a particular haplotype, both of which may aid in a forensic investigation ; however, the haplotypes entered cur rently vary in the number of SIR loci for which data are available. In 1997, a 'minimal haplotype' was recommended that typed seven 10ci 96 for which data are ava ilable for all entries on the YHRD SIR Database. Since then, European labo rato ries have moved towards typing nine loci , and SWGDAM recommends typing 11 loci. 97 Kits for these and additional loci are commercially ava ilable, and current prac tice is moving towards 17 loci being analysed routinely, further increasing the discriminatory power of the analysis. It is likely that the foc us in the future will move towards supplementing the information derived from Y-SIRs with single-base sequence variations, called single nucleotide polymorphisms (SNPs). Various ways are used to try to enri ch a fema le/male mixed sampl e for the male cell component in order to assist in the analysis and maximize the chance of getting an inter pretable Y-SIR haplotype data. For example, differential DNA extractio n techniques may be employed, which take advantage of the increased resistance of sperm cells to par ticular reagents routinely used to digest cells during the DNA extraction step of ana lysis. Epithelial or other non-sperm cells are initially lysed and separated from residual sperm ceJls, which can then be processed separately?O,98 Antibodies specific to sperm cell antigens coupled to magnetic beads 30 or laser microdissection (LM) techniques 30 ,99.lOo may be used to directly capture sperm cells and, more recently, fluores cence in situ hybridization in conjunction with LM has been developed as an effective way to isolate not only sperm cells, but also male epithelial and inflammatory cells. 101 , I02 The latter technique presents a solution to the problem of deposited semen samples that are devoid of sperm cells as a result of a medical conditions or vasectomy, and offers a way to include not only sperm cells, the subject of previous
406 I
Forensic DNA profiling
analysis, but also additional male cells, which will boost the quantity of extractable male DNA in small samples. As with conventional STR typing, the result of Y-STR typing can lead to an individual being excluded as the source of a sample, if the two Y haplotypes do not match. The results may be inconclusive, for the same reasons as dis cussed previously, or they may lead to an individual being included as a possible source of the sample, when the sam ples show sufficient similarity that they could have origi nated from the same individual. In this case, in a court setting, the findings are typically presented as being unable to exclude the suspect (the individual found to match the haplotype) as the donor of the crime sample, but also unable to exclude all patrilineally related male relatives, as well as an unknown number of unrelated males, as potential donors.92 Courts often seek some kind of statistical meaning to this conservative statement. The RMP calculation described above is not appropriate because, unlike the auto somal STR loci that are typed for conventional STR profiling, which are located on separate chromosomes and, hence, inherited independently of each other, the Y-STR loci are inherited together as a group of linked loci on all but rare occasions. Instead, an indication of the probability of a coin cidental match may be evaluated using the counting method or a Bayesian-approach haplotype surveying method. 92 • lo] The counting method is the simpler and expresses the num ber of times the given haplotype is found in the database that has been searched and upper-bound confidence limits that can be placed on the haplotype's frequency. The signif icance of inferences made from these calculations is highly dependent on the size of the database used. The 20th release of the aforementioned YHRD STR Database included 46831 seven-loci 'minimal haplotypes'. Of these, 44974 were typed for the nine European recommended loci and 17935 for 11 SWGDAM recommended loci. 95 The US National Center for Forensic Science is currently in the process of creating a National US Y-STR Haplotype Reference Database, which will contain data on ll-loci haplotypes. The database is to be created by compiling and consolidating existing govern ment, commercial and academic Y-STR databases and will allow continuous updating of haplotypes entered on the database by retyping existing samples as new STR loci are introduced to create extended haplotypes.lOJ.l04 Making allowances for advancing technology and new typing tech niques in this way shows foresight, but raises some of the highly contentious issues that exist in relation to whether only profiles of samples or actual DNA samples of DNA database entries should be retained once an entry is lodged on a DNA database. In the context ofY-STR typing and forensic applications, interesting data are coming forth suggesting a way in which the common biological attribute of the Y chromosome shared by male relatives might be exploited in conjunction with another commonly shared feature of many male rela tives, namely their surname, in societies in which the custom is for the family name to pass through the male lineage. 105.106
Sharing a surname in such societies significantly elevates the probability of sharing a Y-chromosome haplotype, leading to the proposal that the establishment of a large surname-Y-chromosome haplotype database might be of intelligence value in a case devoid of other leads and when male DNA central to the case has been identified, for exam ple, in a rape case. By comparing the Y-chromosome haplo type of such a sample against the database, a potential surname of the source might be identified and lead to iden tification of a pool of potential suspects. Because many indi viduals share a Y-chromosome haplotype, existing databases do not attempt to attach identity to an entry. There are unde niably many barriers to the success of this proposed approach, including non-paternity events, surname chang ing, adoption, the existence of more than one founder of a particular surname, whether the theory holds true for com mon surnames and mutations. Civil liberty and privacy issues in establishing and using such a database need to be addressed. The method would not be able to conclusively exclude or include any individual as a perpetrator, and more research is required if the proposed approach is to become a realistic option, but it is interesting to be aware of new opportunities to exploit genetic information that may come to fruition.
MIXED SAMPLES Y chromosome-specific short tandem repeat typing is of value when more than one male perpetrator may have con tributed to a sample because it reduces the number of possi ble allelic combinations that one is trying to analyse and is useful in the event of a mixed male and female sample. Where more than one individual contributes to a sample. then multiple peaks representing multiple alleles may be seen on the electrophoretogram of a conventional STR analysis. If two individuals contribute to a sample then up to four different alleles could be detected at anyone locus if both individuals were heterozygous at the locus; in the case of three contributors there could be up to six different alle les. Such scenarios introduce additional practical and inter pretative problems over conventional sample analysis. 107 - 11l How does one know which of the possible allelic combina tions to attribute to the different contJibutors? One approach is to attempt to allocate alleles on the basis of the strength of the signal (peak height or peak area on the electrophore togram), working on the assumption that if one of the con tributors' DNA was in excess of the other at the start of the PCR, then the alleles initially present in greater quantity would produce larger peaks after amplification. This may not hold true in every instance, however, and in cases where near-equivalent quantities of DNA from more than one indi vidual are present, all signals detected being of similar inten sity, the approach will not be possible. If one considers the scenario of five or more peaks, things get even more compli cated in trying to detem1ine how many different individuals
406 I
Forensic DNA profiling
analysis, but also additional male cells, which will boost the quantity of extractable male DNA in small samples. As with conventional STR typing, the result of Y-STR typing can lead to an individual being excluded as the source of a sample, if the two Y haplotypes do not match. The results may be inconclusive, for the same reasons as dis cussed previously, or they may lead to an individual being included as a possible source of the sample, when the sam ples show sufficient similarity that they could have origi nated from the same individual. In this case, in a court setting, the findings are typically presented as being unable to exclude the suspect (the individual found to match the haplotype) as the donor of the crime sample, but also unable to exclude all patrilineally related male relatives, as well as an unknown number of unrelated males, as potential donors. 92 Courts often seek some kind of statistical meaning to this conservative statement. The RlVIP calculation described above is not appropriate because, unlike the auto somal STR loci that are typed for conventional STR profiling, which are located on separate chromosomes and, hence, inherited independently of each other, the Y-STR loci are inherited together as a group of linked loci on all but rare occasions. Instead, an indication of the probability of a coin cidental match may be evaluated using the counting method or a Bayesian-approach haplotype surveying method. 92 , IOJ The counting method is the simpler and expresses the num ber of times the given haplotype is found in the database that has been searched and upper-bound confidence limits that can be placed on the haplotype's frequency. TI1e signif icance of inferences made from these calculations is highly dependent on the size of the database used. The 20th release of the aforementioned YHRD STR Database included 46831 seven-loci 'minimal haplotypes'. Of these, 44974 were typed for the nine European recommended loci and 17935 for 11 SWGDAM recommended loci. 9s The US National Center for Forensic Science is currently in the process of creating a National US Y-STR Haplotype Reference Database, which will contain data on ll-Ioci haplotypes. The database is to be created by compiling and consolidating existing govern ment, commercial and academic Y-STR databases and will allow continuous updating of haplotypes entered on the database by retyping existing samples as new STR loci are introduced to create extended haplotypes. 103 ,104 Making allowances for advancing technology and new typing tech niques in this way shows foresight, but raises some of the highly contentious issues that exist in relation to whether only profiles of samples or actual DNA samples of DNA database entries should be retained once an entry is lodged on a DNA database. In the context of Y-SIR typing and forensic applications, interesting data are coming forth suggesting a way in which the common biological attribute of the Y chromosome shared by male relatives might be exploited in conjunction with another commonly shared feature of many male rela tives, namely their surname, in societies in which the custom is for the family name to pass through the male lineage. lOS, 106
Sharing a surname in such societies significantly elevates the probability of sharing a Y-chromosome haplotype, leading to the proposal that the establishment of a large surname-Y-chromosome haplotype database might be of intelligence value in a case devoid of other leads and when male DNA central to the case has been identified, for exam ple, in a rape case. By comparing the Y-chromosome haplo type of such a sample against the database, a potential surname of the source might be identified and lead to iden tification of a pool of potential suspects. Because many indi viduals share a Y-chromosome haplotype, existing databases do not attempt to attach identity to an entry. There are unde niably many barriers to the success of this proposed approach, including non-paternity events, surname chang ing, adoption, the existence of more than one founder of a particular surname, whether the theory holds true for com mon surnames and mutations. Civil liberty and privacy issues in establishing and using such a database need to be addressed. The method would not be able to conclusively exclude or include any individual as a perpetrator, and more research is required if the proposed approach is to become a realistic option, but it is interesting to be aware of new opportunities to exploit genetic information that may come to fruition.
MIXED SAMPLES Y chromosome-specific short tandem repeat typing is of value when more than one male perpetrator may have con tributed to a sample because it reduces the number of possi ble allelic combinations that one is trying to analyse and is useful in the event of a mixed male and female sample. Where more than one individual contributes to a sample, then multiple peaks representing multiple alleles may be seen on the electrophoretogram of a conventional STR analysis. If two individuals contribute to a sample then up to four different alleles could be detected at anyone locus if both individuals were heterozygous at the locus; in the case of three contributors there could be up to six different alle les. Such scenarios introduce additional practical and inter pretative problems over conventional sample analysis.107-lll How does one know which of the possible allelic combina tions to attribute to the different contlibutors? One approach is to attempt to allocate alleles on the basis of the strength of the signal (peak height or peak area on the electrophore togram), working on the assumption that if one of the con tributors' DNA was in excess of the other at the start of the PCR, then the alleles initially present in greater quantity would produce larger peaks after amplification. This may not hold true in evelY instance, however, and in cases where near-equivalent quantities of DNA from more than one indi vidual are present, all signals detected being of similar inten sity, the approach will not be possible. If one considers the scenario of five or more peaks, things get even more compli cated in tlying to determine how many different individuals
--
Mitochondrial DNA I
might have contributed to the sample. There are numerous scenarios depending on whether the contributors are het erozygous for alleles or homozygous for alleles and whether individuals share alleles in common . Mixture analysis is complex and should not be considered definitive. loa Sophis ticated software is available to assist in the interpretation of these profiles,112 but one is justified in being wary of over reliance on algorithms, just as one is justified in being wary of human error and misinterpretation or evaluation of such complex scenarios.
ADDITIONAL SAMPLE PROBLEMS AND SOLUTIONS Clearly, there are many practical and technical hurdles to be overcome when analysing DNA in the case of alleged child sexual abuse. The problems of failing to obtain a sample of suitable quality or quantity will determine whether an interpretable profile or Y haplotype is obtained, be it mixed or otherwise. However, if one spreads the net wider and looks at the ge neric situation where the identity of the source of a DNA sample is being sought (this could be samples collected in connection with abuse, murder, kidnap, the identification of the remains of a body follow ing a natural or man-made disaster, or any other investiga tion one may wish to consider), then other pitfalls may hinder successful analysis, over and above the potential for contamination of DNA irrelevant to the case in hand,109 which has been touched on ea rlier (see p. 403). Amplification of a sample by PCR may be a complete or partial failure due to partial or complete degradation of the DNA or as a result of the presence of PCR inhibitors in the sample that have not been removed during extraction of the DNA. Degradation may be the result of enzymatic, chemical or environmental factors,30, 113 including warm humid con ditions, contamination with soil bacteria or gastric contents, or exposure to the air, fire or chemicals. Fai lure to amplify DNA may be caused by exposure to ultraviolet (UV) light or by a variety of contaminating substances depending on the sample, e.g. humic compounds in soil, haem in blood, poly saccharides and bile salts in faeces, melanin in hair or tiss ue, urea, certain lubricants used on proctoscopes or specula when collecting sampl es or by leather and textile dyes such as indigo, used to dye denim, to name a few. 30,I09,114 Clearly, not only the collection, handling and storage of the samples, but also the environment and treat ment experienced by the sampl e since its deposition have to , be considered. Deliberate attempts may be made by a perpetrator to eliminate incriminating evidence, for example, setting fire to remains, which can adversely affect DNA analysis. STR typing of cremated remains is considered unreliab le and by some to be unsuitable for forensic purposes,115 although others have been able to get successful STR profiling results even from badly charred remains. 116,11 7
407
MITOCHONDRIAL DNA
In some cases there may be no way around the problems of poor quality or quantity of a sample, but it is worth bearing in mind that an alternative to nuclear DNA as an investiga tive tool can be found in the form of mitochondrial DNA (mtDNA). STR and Y-chromosome typing use nuclear DNA as a template and length of SIRs as the basis of discrimina tion and source attribution of a biological sample. However, our cells also have a DNA component contained in their mitochondria, the organelles that are the site of energy pro duction in the celJ. 118 The mitochondrial genome is much smaller than the DNA content of the nuclei, being approxi mately 16.5 kilo bases compared with over 6.4 billion bases in the DNA of the 46 nuclear chromosomes. 119 However, although our cells contain only one nucleus, with one copy of each maternally and paternally inherited allele that are STR typed, there may be hundreds of mitochondria per cell, with each mitochondrion having an average of four or five copies of mtDNA. It has been estimated that on average there are 500 mtDNA molecules per cell, 120, 121 which is highly advantageous if one is faced with very sma ll biological sam ples with nuclear DNA at sub-SIR detection levels because the higher copy number of mtDNA per cell increases the likelihood of recovery of detectable DNA. Another advantage of mtDNA is its increased resistance to degradation, so that in samples in which nuclear DNA can no longer be analysed it may be possible to get genetic informa tion by analysing the residual mtDNA component. For exam ple, mtDNA from human skeletal remains from the Vietnam War which had been exposed to extreme environmental con clitions, including heat and humidity, for at least 17 years has been successfu lly extracted and analysed despite unsuccess ful attempts to type nuclear DNA.122 Mitochondrial DNA has also been sequenced from ancient remains including the Ice man, also known as Otzi, found in the Alps in 1991 , who lived more than 5000 years ago,i23 as well as from Nean derthal fossilsl 24 and from the mammoth. 125 It is worth bear ing in mind, however, that the environment experienced by these remains may have been conducive to survival of DNA, for example, the frozen state of the Iceman. Successful extraction of mtDNA, far less nuclear DNA, is in no way guaranteed, particularly when the sample has been subjected to adverse environmental conditions. While mtDNA may have its advantages, there are down sides too, the main one being the low power of discrimina tion. In a somewhat analogous way to inheritance of the Y chromosome, mtDNA has a characteristic inheritance pattern, being inherited only through the maternal line. It is only the mother of any offspring who passes on mtDNA to her sons and daughters. Mitochondrial DNA that is pres ent in the sperm cell does not enter or survive in the fertil ized egg. 121 There have been reports suggesting rare instances of paternally inhelited mtDNA, but others have questioned the authenticity of these findings as being truly paternally derived mtDNA sequences. 126 It is generally
408 I
Forensic DNA profiling
accepted that, with the exception of mutation events, all siblings and maternal relatives have the same mtDNA, so it is not possible to identify the source of a particular mtDNA sample to the degree of specificity with which nuclear DNA may be attributed. For forensic purposes mtDNA is analysed by sequencing two specific sections of the mitochondrial genome that show the most interindividual variation, namely hypervariable regions I and II. The sequence results of an analysis are reported in terms of differences relative to a reference sequence l27 of the hypervariable regions. 121 When an unknown sample and a known sample are compared and they share the same sequence, it is not possible to exclude the known sample as being the source of the unknown sam ple, but, analogous to the case ofY-STR typing, nor can one exclude all individuals originating from the same maternal lineage. If more than two differences in nucleotide sequence are seen, the known sample is excluded as the source of the unknown sample. If there is one difference in the sequences, the result is said to be inconclusive to account for mutations that have been seen to occur from mother to child. 128 Mito chondrial DNA has a higher mutation rate than nuclear DNA, and tracking the conservation of a particular sequence through generations and the emergence of mutations has been a useful tool in genealogy studies, acting like a molec ular clock, and has provided support for the 'Out of Africa' hypothesis of human genesis from a woman who lived 200000 years ago. 129. 130 A consequence of mtDNA's susceptibility to mutation can give rise to a situation of heteroplasmy whereby an individ ual can possess mtDNA molecules with different sequences. As there are multiple mitochondlia and multiple mtDNA molecules, if a viable mutation alises in a subset of these molecules they may be passed on as the cells divide, giving rise to mixed populations of mtDNA molecules in an indi vidua1.121.131 Whether more than one mtDNA sequence is detected on sequencing a sample from an individual exhibit ing heteroplasmy will depend on the relative proportions of the different populations of mtDNA molecules in the sample analysed, which may vary from cell to cell and from tissue to tissue. Heteroplasmy appears to be a more common phe nomenon than previously thought 13l and the consequences for forensic analysis should be taken in to account when interpreting results. This appears to be particularly relevant in relation to hair samples, which are often found at crime scenes and used as evidence in a case. Because the root of the hair is needed for nuclear DNA analysis and is lacking in many hairs from crime scenes, mtDNA is more frequently performed and heteroplasmy has been found to be more common in hairs than in other tissues. 132 Probably the most famous example in which DNA evidence was found to exhibit heteroplasmy was in skeletal remains identified as Tsar Nicholas II from exhumed remains believed to belong to members of the Russian Royal family murdered at Ekaterin burg in 1918. The investigation concluded that both Tsar Nicholas II and his brother exhibited heteroplasmy.1 33,134
As for Y haplotypes, a statistical estimate of the signifi cance of a match is needed when mtDNA sequencing fails to exclude an individual as being the source of an unknown sample. The process is similar to the Y haplotype evidence, being a reflection of the frequency at which the given mtDNA sequence occurs in established mtDNA databases. 121
PATERNITY TESTING Inheritance of mtDNA and the Y chromosome provides a means of tracking blood relationship through the genera tions, but often one wishes to have a more definitive means of detelmining the biological parentage of a child. Proof of paternity or maternity may be needed to settle inheritance disputes, child maintenance responsibilities or immigration applications, not to mention the psychological and relation ship issues and problems that go along with doubts in trust and fidelity of relationships. The principles behind paternity testing may also be employed to assist in the identification of the remains of an unknown when no reference sample of the individual is available, if there is evidence to support parentage by individuals who can provide samples. DNA paternity testing uses STR typing and the princi ples behind the manner by which DNA is inherited from our parents. It will be appreciated that, because any indi vidual inherits half of his or her genetic material from each biological parent (one copy of each of the chromosome pairs), the presence of genetic material that is not consis tent with this mode of transmission, i.e. of SIR alleles not present in the alleged parents, is strong evidence that the alleged parents are not in fact the biological parents. In most cases, the biological mother is known and one is seeking to determine whether an alleged father could be the biological father, though it may be the father who is known and it is maternity that is being investigated. Alternatively, one may be faced with the more difficult situation that only a parent and alleged child are available for analysis 135 or it is necessary to determine, in the case that the mother is known, whether two children have the same or a different father. 136 In the first of these scenarios in which the mother is known, if one SIR types the mother, the alleged father and the child, and compares the mother's alleles with those of the child at each SIR locus, it may be possible to determine the allele that the father must have passed to the child, the 'obli gate allele'. For example, if the mother has alleles 15, 16 at a particular SIR locus and the child has alleles 13, 15 then the child must have inherited allele 15 from the mother and, hence, must have inherited allele 13 from the father. If the father has alleles 13, 14 then he could not be ruled out as the father on the basis of DNA evidence, but nor could any other man carrying allele 13 at this locus. If, on the other hand, the mother was 13 , 15 and the child was 13, 15, then it would not be possible to nalTOW the father'S contribution to a sin gle allele because the mother could have passed on either allele 13 or allele 15. A man with either a 13 or a 15 allele
Paternity testing I
could have been the father, i.e. there would be more possible men who cou ld have fathered the child. Obviously, one looks to more than a singl e STR locus when conducting this analy sis to determine if the alleged father has alleles consistent with fatherhood at each locus typed; the greater the number of loci examined, the greater the evidential weight. Tf a n alleged father cannot be excluded as the father, on the bas is of his genotype being inconsistent with those required to fi t the genealogy, then one needs to have a sta tistical measure of the weight of this evidence as indicative of paternity. This is most commonly presented in terms of the combined paternity index (CPl), though the probability of paternity, which is based on a Bayesian approach, is also w idely used. 121, IJ7 - 139 The paterni ty index is a likelihood ratio w here the proba bility of obtaining the observed genotype in the child at a particular locus, given that the alleged fa ther is t he biologi cal father, is di vided by the probability of obtaining the observed genotype, g iven that a random man is the father. Tn order to determine the probability of obtaining the observed genotype, given that a random man is the father, one uses the allele frequencies of the obligate allele from population databases of the race of the alleged father. Tn order to obtain the probability of obtaining the observed genotype, given that the alleged father is the biological father, one looks at his genotype. If he is homozygous for the obligate allele the probability would be 1 because there is no alternative allele that he could have passed on to a child; ifhe is heterozygous for the allele the probability would be 0.5, because there is an equal chance that he passed on either of his altern ative alleles. 121, IJ8 Having calculated the paternity index for each locus, a CPI is calculated by multiplying the paternity indexes for each locus analysed. A cpr of 100 is accepted to establish 99 per cent probability of paternity and a CPI of greater than 1000 indicates that the probability of paternity is greater than 99.9 per cent. A CPI of 100 is the genera lly accepted minimum standard for inclusion,140 although some laboratories use a hi ghe r cpr as their minimum.141 In the context of paediatric cases, paternity testing is not just applied to the situation where one wishes to iden tify the father of the subj ect of the case; t he subject may be the mother, for instance in the case of statutory rape. By DNA profiling feta l material from a borted pregnancies and from the child (mot her) one can determine the likelihood that a given suspect was t he father an d, hence, guil ty of statutory rape. 8J It is important to appreciate that no CUlTent DNA test is definitive and there a re a numbe r of issues relating to pater nity testing that need to be considered. As stated earlier, relatives share a highe r proportion of their genome than random men, so problems may arise if a relative of the real father, for example, father, brother or son, is being consid ered as the biological father. One study found that, in over 31 per cent of cases, an un cle could not be excluded as the father, with five cases showing no mismatches. These five cases would have resul ted in paternity by the uncle being
409
identified had he been the only putative father tested. 142 Other studies also highlight the problem of relatives being mistaken as a biological father; 14J a way to determine more definitively if a given allele is inherited from a given indi vidual would be of great value in preventing potentially mis taken conclusions. Determining the parental origin of alleles would also be of help in cases in which the mother is not available for test ing and DNA is availabl e only from the child and alleged father, in the case that t he mot her and child share the same heterozygous genotype and when multiple male relatives are suspected of incest. A novel test has been proposed that migh t be app licable to these cases. It would look at genomic imprinting that resul ts fro m epigenetic modifications, which change the chemical structure of the DNA without altering the underlying DNA sequence.144-146 Certain genes have been shown to exhibit this phenomenon, commonly by way of DNA methy lation, with the changes being parent specific. It is possible to detect these changes; thus, there may be a way of determining the parental origin of a particular allele. The approach has been proposed for forensic investigations and paternity testing, 147, 148 and it will be interesting to see if it becomes a routine part of forensic DNA analysis, add ing strength to support the RMP, where not only the same alle les are shown to match a suspect but a common geno mic imprinti ng pa ttern is observed. On a related vein, analys is of epigenetic differences has also been proposed as a potential way to differentiate between monozygotic tw ins. As dis cussed above, monozygotic twins inherit the same ge ne tic sequence, and it has been shown that they are epigenetically indistinguishable at birth and in the early years of life ; how ever, older twins exhibit detectable differences in epigenetic modifications of particular sequences t hat were acqu ired during life. This may be one way to differentiate between identical twins.149 While the increase in shared alleles between relatives may lead to problems in identifying the real father in a paternity case, it is worth mentioning that, because relat ives have more alleles in common, the effect of th is in incestu ous relationships gives rise to an increased level of homo zygosity in the offsprin g, which may be used as an ind icator of incest. 150 CPT or alternative statistics may be reported when an alleged man is not excl uded as the biological father. But on the basis of ge nea logy, how different do the alleles need to be from the alleles we would expect the father to have in order to lead to an exclusion? DNA paternity testing works on the same basis as blood group testing, proposed in the 19205, 151 and late r refined, as a means of investigating paterni ty disputes, but may add a greater degree of discrim ination. Rather than looking at the phenotypic effect, i.e. the observed effect of our genetic material, namely what blood group an individual is, 29 DNA typ ing looks at the genetic material itself, which provides more information. Despite frequently being overl ooked today, blood group testing was, and still is, an effective way of, relatively quickly and
410 I
Forensic DNA profiling
cheaply, excluding an individual as being the source of a samp le or the biological father of a child. If the blood groups did not match a suspect when trying to identify the source of a sample, then a suspect could be excluded or, in the case of a paternity dispute, if the offspring's blood group was not consistent with the possible combinations that could have arisen from the mother and alleged father, then the individ ual in question is excluded . Despite this appa rentl y clear-cut log ic, things may not be quite as simple as far as DNA profiling is co ncerned, for reasons to be explained below, and, from histo rical cases, court decisions have not always followed the scientific theory and evidence for blood group testing. For instance, in the famou s paternity dispute in 1946 involving the actor and director Charlie Chaplin, the apparent clear impossibil ity of fatherhood, on the basis of blood group testing, was not viewed as bei ng conclusive, and Chaplain was ordered to pay child support to the child of Ms Joan Berry.152.153 In 1983, Shaw l52 stated 'Motherhood has always been a biological certainty; now fatherhood will be as well. We will have come one step closer to equality of sexes.' Little did he anticipate the consequences of errors that have been, and no doubt will continue to be, made in conducting in vitro fertili zation (NFl. A growing number of cases have come to light in which embryos have been transferred to the wrong . woman, so that she has been impl anted with another woma n's eggs; the wrong spelID has been used to inseminate eggs; or both errors have occurred. 154 ,155 It has been admit ted that these mix-ups are a regular occurrence in fertility clinics in the UK;156 similar cases have arisen in other coun tries. 157 ,150 Th e majority of cases t hat have come to light involve white parents having a black child, or vice versa. This is the most readily apparent way of detecting any mis takes. How many other cases go undetected is open to ques tion. Paternity testing is one way that ca n bring the errors to the fore . The ramifications of such events for the parties involved cannot be underestim ated, encompassing psycho logical trauma, legal parenthood and custody issues as well as compensation considerations. 155 ,158-1 60
Mutation and Anomalous Genotypes While the decision in the Charlie Chaplin paternity case has been cliticized, given what appeared to be an impossibility of blood group inheritance, using DNA analysis, failure to exclude an individual as a parent, even when alleles of the alleged father do not appear t o fit those expected, do es have scientific reasoning. The fact is that SIR loci, like any region of the genome, are susceptible to mutation. 161 In fact, these regions of the genome are particularly susceptible to muta tion and show a high rate of mutation during gamete forma tion, when the cells undergo meiosis, which can lead to allel ic change from one generation to the next. 162-165 This can result in an allele in the offspring being different from the parental allele from which it was derived, and without further testing
might lead to a conclusion of non-paternity being dra wn. 166- 169 Mutation rates vary across the SIR loci and some may be as high as 0.3 per cent of meiotic events. 165 Evidence of two mutations occuning between generations in several cases has led to a minimum of three independent allel ic inconsistencies being required for a putative father to be excluded as the biological father. 167 ,168 When there is doubt whether inconsistent alleles are due to mutatio n or a genuine excl usion of the alleged father, typing the Y chromo some SIRs of the child, if male, may provide further evidence to show a genuine exclusion or not, as the case may beY6,1 70 The mutations discussed above relate to ge rmline events, which are passed from parent to child and affect all the cells of the offspring, though somatic mutations may also arise in cells other than the parental gametes. If a mutation arises at an STR immediat ely post fertilization and that STR is subject to forensic ana lys is, then the resulting cells of the offspling will have a profile that differs from that expected on the basis of genetic inheritance. All the child 's cells will carry the mutated allele and not the parental one. This has consequences for pat ernity investigations. It is not just in paternity investigations that mutations can affect the outcome of t he investigation. Somatic cell mutations in SIRs can lead to anomalous profiles with more than two alleles being seen per locus 171, 172 and discrepan cies in results of comparison of profiles, even if the profiled sa mples originated fro m t he same individual. If a mutation a rises in a subset of cells that are profiled , and the mutated cells are at a level above the threshold of detection, a com parison between pro files of cells of a different tiss ue, or a sample of cells lacking a sufficient proporti on of mutated cells for detection, can lead to the possibility of a common source of the samples, and thus a match, being missed. This may be of relevance if, for example, blood cells sampled at a crime scene have a different profile from those of buccal cells taken as a reference sample from a suspect. Individuals who possess cells with different genotypes that have arisen from a single zygote by way of mutation are known as mosaics. 173, 174 An individual may also possess cells of differen t genotypes because of ch imerism resulting from natural anomalous events in which more t han one zygote fuses to create a single individual; 175-1 78 when cells derived from a different zygote are present in a single indi vidual (e.g. if dizygotic twi ns exchange blood across a pla centa l79 ,lool or as a result of medical intervention, be it bone marrow or organ donation.181-184 In the case of bone mar row transplantation, for example, it is possible for false identification to ar ise when the donor may, in fact, be the guilty party, but because the recipient harbours cells shown to match the donor's cells left at a crime scene the recipient may be falsely accused or convicted. 185 It will be apparent that chimerism and mosaicism can lead to inappropliate exclusion of the true source of a sample, when looking for a match, and to the exclusion of a biologi cal parent as being so if the possibility of the phenomenon is not taken into account during an investigation. 175 ,176, 186
Identification: body remains/missing persons I
The problem of the perception of the infallibility of DNA evi dence was highlighted in the media-grabbing case where a mother was accused of committing benefit fraud because it was alleged that the she could not have mothered the children in question and risked them being taken in to care. It was only after her third child's birth was witnessed, and DNA evidence also showed that baby appeared not have been her child, that it was discovered that she was a chimera. The cells giving rise to her offspring were of different genotype to those typed during the maternity investigation. 177
parent. The best it can do is provide evidence in terms of sta tistical probability and it is up to a jury or those concerned to determine if it fulfils their criteria of proof.
IDENTIFICATION OF BODY REMAINS AND MISSING PERSONS
The Use of Mitochondrial DNA and the Y Chromosome in Paternity Cases As mentioned above, the Y chromosome may be used to pro vide additional information in cases where there may be uncertainty over whether a mutation has arisen from father to child. There may be occasions when paternity is in dispute and samples of the putative parent are not available, which causes practical problems. Sometimes, material may have been taken for medical investigations and, if it has been stored, it may be accessed and used as a source of DNA.166 Obviously, this is not always possible, and one approach that might be adopted in the case of a male offspring is analysing the Y chromosome haplotype of male relatives of the puta tive father, given that the Y chromosome is passed from father to so n and, barring mutation, aU these family relatives would share a common Y chromosome. This was put in to practice in the attempt to address the long-standing histori cal controversy over allegations that the US President Thomas Jefferson fathered the sons of one of his slaves, Sally Hemmings. The first source that on e would look to in order to type a deceased's Y chromosome would be known male offspring of the alleged father, but Jefferson had no surviving sons. Instead data were gathered from male-line descendants of two sons of the president's paternal uncle, male-line descendants of the two men Jefferson was alleged to have fathered (Thomas Woodson and Eston Hemmings) and male line descendants of the sons of Jefferson's sister, also suspected of fathering one of Hemming's children. Analysis of the Y chromosome haplotypes excluded Jeffer son of fathering Thomas Woodson, but the younger son, Eston Hemmings, sha red the same Y chromosome haplotype as Jefferson 's relatives, providing support that he may have been Jefferson's illegitimate son. 187 However, the very fact that Jefferso n's male relatives had a common Y haplotype also means that any of Jefferson's male relatives alive and capable of fatherhood at the time co uld not be ruled out as the father and the president was only one of several candi dates, despite the incriminating publicity the results of th e study attracted. 188 The lack of conclusive evidence has meant the descendants of Jefferson's daughters have not accepted the relationsh ip,1 89 highlighting the fact that DNA evidence cannot conclusively identify an individual as being the source of a sample or an individual as being a biological
411
-
The loss of a child is devastating for any parent, and one can not overemphasize the importance of closure for families in the event of such tragedies by way of positive identification that their child has actually died. If the remains of a deceased person can be identified, this can put an end to uncertainties or hopes that a child may have sUlvived and can enable return of the body or remains to the next of kin for appropri ate burial, events that may be of deep psychological and reli gious significance and comfort at a time of distress. DNA analysis is one way that can assist in the identifi cation of human remains, but the approach is not without problems. Not only does one have to address the state in which the remains are found , which is often far from ideal because of degradation or decomposition, but one needs to acquire a reference profile with which to compare that of the remains. There is also the issue that following some incidents, for example, explosions, body parts may be frag mented and widely dispersed, adding to the difficulty of reunify in g an identified whole body. If the remains show few signs of decomposition, it will generally be possible to obtain DNA from blood samples or internal soft tissue which has been largely unaffected. As the degree of decomposition increases, blood becomes less of an option and superficial soft tissue may exhibit signs of putre faction, in which case tissue or muscle deep er within the body and/or bone marrow may yield DNA suitable for STR profiling. When remains are at an advanced state of decom position, bone marrow may still be able to provide material for a DNA profile but, as decomposition progresses, the extraction process increasingly faces problems of cell debris, decomposition products and contaminants that can inhibit the PCR reaction, and thus need to be removed. If bone mar row does not yield a profile then hair may be STR typed, if the root rema ins, or mtDNA from the shaft may be sequenced. Alternatively, DNA could be extracted from skeletal struc tures, which is also the option when remains are fully skele tonized. The resistance of structures such as bone and teeth offers protection to DNA contai.ned in the bone matrix or tooth pulp; however, this, in turn, requires special proce dures to extract and purify it. The level of nuclear DNA that can yield an STR profile from these sources is low; sequenc ing the more resistant mtDNA may be the only option avail able. On rare occasions one may be faced with naturally mummified or desiccated remains, which present their own challenges, particularly if there is a degree of decomposition. Extracting DNA from soft tissue of such remains has given little, or marginal, and inconsistent success in STR profiling (Jason Eshl eman, personal communication; Carlos Morales,
412 I
Forensic DNA profiling
personal communication), though extracting and typi ng DNA from bones and teeth of such remains does not appear to offer significant problems over other sources of bone and teeth (Tim Clayton, personal communication; Jason Eshle man, personal commu nication) . DNA has been extracted and sequenced from dried superfici al parts of Egyptian mummi fied remains, suggesti ng that it is possible to get genetic information from such material, but the DNA in this case was in a fra gmented state, making its suitability for STR analysis questionable. It was only one of many samples studied that indicated that recoverable DNA might be obtained. 190 It is possible tha t natural and artificial mummi fication may affect the success of extraction. Despite these issues and potential problems, DNA analy sis has been of Sign ificant value to investigating teams in major incidents or mass disasters over the years. Although techniques such as visual identifi cation, odontology, anthro pology, fingerprint analysis, radiology and facial reconstluc tio n can go a long way in helping to identify remains, they may not always be appropriate or possible. DNA analysis has succeeded in many investigatio ns, in conjunction with these tech niques, or in cases in which other techniques have been unsuccessful. For example, in 1998 a tragic fire in Manila resulted in the death of 23 children between the age of six months and eight years. The bodies were initially buried unidentified, but three mon ths later the burial site was exhumed in an attempt to identify the remains. 191 Exhumation recovered 22 bodies, and a combination of autosomal and Y chromo so me typin g was carried out on 21 of the bodies that had not been otherwise identified. The identification process was assisted because th e identity of the children present at the tragedy was known, though this may not always be the case in such investigations. It was also possible to deter mine the approx imate age range of the re mains, which was of further benefit. Ten of th e exhumed bodies were believed to be sibling pairs and 11 were unrel ated to each oth er. 191 Despite the bodies hav ing been burnt, buried an d exhumed, DNA analysis succeeded in matching 18 of the remains with a known ch ild. In two of the cases mothers were able to provide reference mate ria l in the form of stored umbilical tissue from their child, which had been kept in accordance with Philippine custom an d which provided a positive identification . 192 Iden tificatio ns of the other children, for whom ante mortem reference sa mples were not ava ilable, was achieved by patern ity-type ana lysis using autosomal STR profiling, as well as by comparison ofY haplotypes, by analysing DNA reference samples from 10 mothers, three fathers and a paternal grandfather of two children for whom a paternal sample was unava ilab le. Y haplotypes were valuable in iden tifying not only father/grandfather to son/grandson rela tionships but also those of sibling brothers. 191 In the Manila fire tragedy, the remains had been subject to extreme conditions so it was no t su rprising that a full profile was not obtained for evelY sample typed. Where
alleles cannot be identified at evelY 10CllS, th e evidential weight is lower but va luable information can still be obtained, as illustrated by this case. Having an ante-mortem reference sample of a mlss1l1g individual with which to compare DNA profiles of remains goes a long way to aiding the process of identification, and over the years such samp les have been taken from tooth brushes, hairbrushes, clothes, towels and leftover food with bite marks.30.193 It is important to realize, however, that sllch items are mobile and may have been used by someone other than the owner and, hence, might cany DNA of someone other than the individual in ques tion. Archived medica l samples are a very useful source to turn to for ref erence material if they are available. The paternity-type approach, using pedigree analys is to identify remains when reference material is ava ilable from the pa rents of a ch ild suspected to have di ed, is commonly used. Knowing each parent's alleles, one can predict the range of possible genotypes that their child could have and eliminate remai ns that do not cOITespond to these. If a match is identified, the li kel ihood that the remains are the biologi cal child of the couple rather than the child of an unrelated co uple can be determined statistically. Admittedly there can be a velY large number of possi ble combinations (for a 10 loclls profile this cou ld be as high as 1048576), but the approach has been used successfully in many mass disasters. For example, after the Waco, Texas, incident of 1993 - when the remains of 61 individua ls were recovered in varying states of preservation, some in a n advanced sta te of putre faction and others badly charred - 26 positive identifications were made in this way.194 It is also possible to gather suf ficient inform ation when a single paren t is available as a reference sample, 195 though the statistical weight of the evi dence will be less, and comparison with a sibling's genotype may also be helpful. In aeroplane traged ies, families are often present together on the same flight and the common ality of alleles across the generation can be used in identify ing family members. The 9/11 disaster of 2001, when 2792 people were killed in a terrorist attack on the World Trade Center, must surely have been one of the most challenging investigations, and those w ith the task of recoveling and identifying the remains faced lo gistical, technical, practical, scien tific and medical difficulties. Although direct matches with referen ce samples were used where possible to identify the remains, other DNA techniques, including the paternity approach, were also employed. 196 It will be appreciated that in paternity analysis there is always the chance of a coincidental match. A particular all ele at a locus is shared not only by a parent and child, but also by approximately 1 in 1000 random pairs of indi viduals, so an attempt was made to minimize the number of false-positive matches obtained during the identification process. 197 The number of false-positive results is related to the size of the reference list and the number of victims, and therefore is a major consideration in investigations such as
Identification of an unknown profile I
the 9111 tragedy, but it should also not be overl ooked in investigations invol ving smaller numbers of victims. As men tioned previ ous ly, the state of the remains can affect the amount of information that can be obta ined from STR analysis. Poor-qu ality DNA may not provide results for the alleles present at all loci typed, or it may resu lt in iden tifica tion of only a sin gle allele in a hete rozygous individual when there really should be two alleles; one always needs to be aware of the possibility of this so- ca lled allelic dropout. In many cases, the reason for poor SIR resu lts is that the STR fragment in the sa mple is no t intact along its whole length, as it should be. As DNA degrades it fragments into smaller pieces a nd breaks may occur within the SIR; thus, the true alleles are not detected. It may be possible to design PCRs for STR profiling that look at smaller sections of DNA, 198 and commercial kits adopting this approac h are no w available ; the sma ller t he targeted DNA being identified, the greater the chance it remains intact. Where there are deficiencies in SIR typing, mtDNA may be sequenced, as discussed ab ove, or the in format ion gathered from partial STR profiles may be sup plemented by, or rep laced, with SNP data. 199 SNPs are single base sequence variations between indiv iduals at a particular positio n in the gen ome. These sites are ab undant (estimated 10 million) throughout the genome, but because they are mostly biallelic they lack the polymorphic variation across the population that SIRs exhibit. 2oo However, if one geno types a DNA sample at many SNPs, one can approach the discriminatory power of SIR profiling. Estimates vary as to how many are needed, but it is likely that between 50 and 100 may be required,199 consistent with the 70 used in rela tion to the 9/11 investigation. 196
IDENTIFICATION OF THE 'ABANDONED BABY' OR FETAL MATERIAL AND AVENUES FOR IDENTIFYING THE SOURCE OF AN UNKNOWN PROFILE The approac hes described above, which are used to identify the remains of a ch ild , ha ve been geared towards identify ing a previously know n livi ng individual, either by directly matc hing a profil e with a reference sam ple from the indi v idual or through pedigree analys is. However, one may be faced with the scenario of an aba ndoned newborn or fetus requirin g investigation. In such a scenario t here will be an endeavour to iden tify the mother. Whethe r maternal DNA can be recovered from the scene to ass ist will depend on the post-partum circum stan ces and the conditi on in which the newborn is found. Ma ternal DNA has successfully been recovered and pro filed from placental material found abandoned with a wrapped dead newborn, as well as from blood found on the wrappin g material. Using the profile, it was possible to determine the mothe r of the baby in this case. 83 Obviously, this outcome requires that a reference sample is available fro m the mother fo r comparison. For exam ple, entries on a
--.
~;
413
nationa l DNA database can be searc hed agai nst the profile in question or from a woman suspected of being the mo ther during the investigation. Placental material is an ideal source of maternal DNA in such circumstances, but it may not always be available. In another repolied case, a live newborn was fou nd abandoned in a box but no pl ace ntal materia l was found with the baby. It was, however, possible to obtain a profile of the mother by extracting maternal DNA from vernix caseosa on the ha ir and body of the newborn, which conta ined maternal blood. 83 One should be aware of the potential for such mate rial being relevant and useful in similar investigations. A more problematic scenario is the case of the abandoned baby fou nd long after deli vely. The rem ains may have decomposed, with the concomitant difficulties of obtaining DNA profiles discussed previ ously, or there may be circum stances in which the deceased has been concealed or the environment has been conduci ve to mummification. As men tioned earlier, DNA profiling of desicca ted material is prob lematic. [n the case of a mummified baby, there is the addition al issue that bone and teeth, a commonly used source of DNA in the case of remains, will be unavailable if the neonate's bone is cartilagi nous and teeth have not erup ted. Despite the difficulty in obtaining suitable genetic informa tion from desiccated tissue, profiles have been obtain ed from mummified tissue from babies, though in these cases there was no accompa nying decomposition of the bodies, which may have contri buted to the successful results. [f no maternal DNA is found with the fet al remains but a DNA profile of the baby can be obtained, genetic analysis resorts to the pedigree type of analysis: comparing the baby's profile with those of suspected parents in a paternity testing approach to look for consistent transmission of alleles. As stated above, the statistical weight of such an analysis is less than if information is available from at least one parent. 195 The problem is obtaining DNA from a suspected mother when there may be no lead to the parents. However, this is true of any investigation centring on DNA. An SIR profile alone provides no meaningful information other tha n the sex of the individual. What fo llows desc ribes possible means of identifying a candid ate so urce when there is no match with existing profiles on national DNA datab ases or with suspects in the case, these being the first two options to consider. Typ ing the less specifi c mtDNA or Y chromosome might be of assistance in spread ing the net wider: identifying a match might go so me way in identifying a possible pedigree that may lead to a parent, but this type of investigation may also lead to nothing. A similar technique, so-called 'familial searching', is also increasin gly being used. In this case, rather than searching existing databases for a profile that matches all alleles at all loci, a less stringen t search is employed, allow in g profiles with some allelic mismatches to come to light. The logic beh ind this approach is that such profiles might belong to relatives of the true source, who would be expected to share more all eles in common than a random individual,
414 I
Foren sic DNA profiling
and it may be possible to home in on the true source of the sample through a family member. The techn ique has been used successfully20' but raises many ethical issues and remains highly controversiaI.202-204 As with many genetic based investigations, there is the potential to uncover other wise unknown non-blood relatio nshi ps in individuals believed to be true relatives, with ramifications for family relationships that shoul d not be considered lightly. A number of studies have pointed to the potential to obtain an indication about the physical charactelistics of a source of DNA in a forens ic investigation in which no matching reference sample is found, based on the fact that the physically recognizabl e characteristics of an individual are, at least in part, genetically determined.14.2os-207 Acquir ing information about the race of the source of the sample has also been proposed, based on inter-racial genetic differ ences. However, moving from looking at physically incon sequential as pects of our genetics in the form of rou tinely analysed STRs, currently believed to have no influence on our characteristics or health, to more personal and sensitive information is something that wiJi have social implications that need careful consideration, particularly when such information is unlikely to be conclusive and may be of no investigative va lue, given the com plex play of genetic and environmental facto rs t hat make us who we are.20B-2IQ In the specific case of the abandoned baby or fetus, one may come across rare cases in wh ich there are indications in the deceased of a genetically inherited condition, for exam ple the charactelistic features of Down's syndrome. 173 In such cases prenatal genetic testing of the fetus may have been carri ed out, and it may be possible to co mpare the pro file of the baby with stored clinical samples. A match wi th a clinical sample could lead to the parents through medical records. With increased use of electronic medical records, th is approach could become a more realistic prospect in the future. Admittedly these are tenuous possibilities, requiring significant investigative power and resources, and one has to accept that DNA evidence may not provide constructive evidence in an investigation. There may be times when an investigator wants to know the age of the source of a sample, for example if a newborn is illegally removed from hospital, or to determine if fet al blood is present in putative products of conception in crim inal abortion investigations. Physical chara cteristics can be of some help in some cases, but inte resting research has identified a method that can differentiate between blood fro m a newborn less than a day old , blood from an infant less than four months old and blood from infants over four months of age by assaying the expression levels of partic- ular haemoglob in genes, which exhibit a highly age specific expression pattern .211 Implementation of this ap pro ach in forensic investigations may provide va luable inform ation for future investigations. Rath er than looking at the DNA itself, one is looki ng at the translated product of DNA, ribonucleic aci d (RNA), which acts as a intermedi ary in translating genes into an end product.
DNA DATABASES
In this chapter, mention has frequent ly been made of DNA databases. Currently, national DNA databases are the norm, com ing under national legislation and management, wi th the nations' authorities having varying powers to collect and analyse samples from individuals depending on the severity of the suspected crime. In the UK, authorities have the power to take and analyse a DNA sample from anyone suspected of having committed a recordable offence. These samples are entered on th e UK National DNA Database, the largest data base in the worl d, where they are cross-checked against exist ing entries on the database to look for a match. Profiles may be from individuals convicted of a crime, fro m previous arrestees or from samples taken from unsolved crime scenes; matches connecting crimes with named known individuals or linking crimes with other crimes can be made. Under English law samples taken from a suspect remain on the UK National DNA Database indefinitely, regardless of whethe r the suspect is co nvic ted, acquitted or never brought to trial. Authorities in other countries or states have lesser powers to take or retain samples, and there is significant valiation throughout the world. The question is whether full harmonization of laws and intercountry procedural standardization can ever be established. The Combined DNA Index System (CODIS) oper ates in the USA, and it enables laboratories to exchange and compare DNA profiles at a state and national level. The exis tence of an Interpol DNA database and DNA Gateway enables the sharing of data across nations, so there are moves towards global cooperation. One has to wait to see whether a truly international database is esta blished containing every one's profile with the a im of assisting in solving Clime at an internationa l level. This is something aspired to by some, but, unsurprisingly, a view not shared by many, who feel that such a move constitutes a serious violation of civil liberties. DNA has had a major impact on the field of forensic sci ence and investigations. Without DNA we would not be here as living beings; without DNA a nalysis, undeniably, many investigations would have gone unsolved, but we should be aware that the techniques have their limits. DNA provides only one piece of evidence in any case, and any resul t and statistical likelihood of a chance match needs to be cons id ered in the context of the case as a whole.
REFERENCES Albe rts B, Johnson A, Lewis J et al. Development of multicellular organisms. In Molecular Biology of the Celi, 4th edn. New York: Garland Scien ce Taylor and Francis Group, 2002. 2 Jobling MA, Gill P. Encoded evide nce: DNA in forensic analysis. Nat Rev Gen 2004; 5:739-51. 3 Hartl DL, Jones EW. The chromosomal basis of heredity. In Essential Genetics: a Genomics Perspective, 3rd edn . Sudbury, MA : Jones and Bal11 ett Publishers, 2002. 4 Alberts B, Johnson A, Lewis J et al. Germ cell s and fertilization. In Molecular Biology of the Cell, 4th edn.
Referen ces I
5
6
7
8 9
10
II
12
13 14
15 16
17 18 19 20
21
22 23 24
25 26
27
28
New York: Garland Science Taylor and Francis Group,
2002 .
Rafferty S. Fingerprint evidence doubts clear policewoman of perjury. Scotsman, Saturday 15 May 1999, p. 4. McKie IAJ. There's nane ever fear'd that the truth shoul d be heard but they whom the truth would indite. Sci Ju sriu 2003; 43: 161-5. Cole SA. More than zero: accounting for error in latent fingerprint identificati on . J Crim Law Criminol 2005; 95:985 -107 8. Webb J. The myth of fingerprints. New Scienrisl 2005;
2517:3 .
Saks MJ, Koehler JJ. The coming paradigm sh ift in forensic id ent ification sc ien ce. Sciellce 2005; 309:892- 5. Broeders APA. Of earprints, fingerprints, scent dogs, cot deat hs and cog nitive contaminati on - a brief look at the present state of play in the forensic arena. Forensic Sci 1m 2006; 159: 148-57. Dror IE, Charlton D, peron AE. Contextual information renders experts vulnerable to making erroneous id entificati ons. Forensic Sci Inr 2006; 156:7 4-8 . Jain AK, Prabhakar S, Pankanti S. On the similarity of id entical twin fingerprints. Pattem Recognition 2002; 35:2653-63. Opdal SH, Rognum MD. The sudden infant death syndrome gen e: does it ex ist? Pediatrics 2004; 114:506 - 12. Duffy DL. Montgomery GW, Chen Wet al. A three-single nucleotide polymorphism ha plo type in intron I o f OCA2 explains mo st human eye-co lor variation. Am J Hum Gener 2007; 80:241-52. SudbelY P. Single-gene disorders. In Huma n Molecular Generics, 2nd edn. Harl ow, UK: Pearson Edu cation , 2002. Kum ar V, Abb a AK, Fausto N. Robbin (lnd Corran's Pathologic Basis of Disease. Philadelphia: Elsev ier Saunders, 2005, pp. 489-95. Red on R, Ishik awa S, Fitch KR et a!. Global variation in co py number in the human genom e. Narure 2006; 444:444-54. Shianna KV, Willard HF. In sea rch of no rmality. Naru re 200 6; 444:4 28 - 9. Li W-H, Saunders MA. The chimp anzee and us. Nature 2005; 437:50-1. The Chimpanzee Seq uen cing and Analysis Consortium. Initial se quence of the chimpanzee gen ome and compariso n wi th the hum an genome. Naru re 200 5; 437:69-87. Newton G. Discovering DNA fingerprint ing. Wellcol11e Trusr: The Human Geno me Web sire. 2004 (accessed April 2007). http://geno me.wellcome.ac,uk/d oc_wtd020877.h tm I Jeffreys AJ, Wilson V, Th ein SL. Hypervariabl e 'millisatellite' region s in humon DNA Narure 1985; 314:67- 73. Rudin N, Inman K. Forensic DNA Allalysis. Boca Raton, USA: CRC Press, 2002, pp. 71-6. Forensic Science Service Casefi les: Colin Pitch fork - first murd er co nvicti on on DNA ev idence also clears the prim e suspect. Th e Forensic Science SeflJice Website. www.forensic. gov.ukl fo rensic_ t/i nsidel news/! isCcasefi les.p hp?case = I (accessed April 2007). Butler JM. Overview and histoJY of DNA typing. In Forensic DNA Typillg, 2nd ed n. Amsterdam: Elsev ier, 2005, p. 3. Aronson JD. DNA fingerprinting on trial : the dramatic early history of a new forensic technique. Elldeauour 200 5; 29 : 126-31. McKonkey E. Lecture 3 - genet ic diseases: the conseq uences of mutat ion. In How the Human Genom e Work s. SudbuIY, 1\rJA: Jones and Bartlett Publishers, 2004, p. 43. Butler JM . Appe ndi x II: U.S. population data - STR allele frequencies. In Forensic DNA TVpiHg, 2nd edn. Amsterdam: Elsevi er, 2005.
415
29 Spalding RP. Id entificati on and characterization of blood and bloodstains. In Jam es SH, Nordby JJ (ed s) Forensic Science: all In troduction ro Sciellrijic Investigative Techniques. Boca Raton, FL: CRC Press, 2003. 30 Butler JlVl. Sam ple collection, DNA extracti on an d DN A quantita tion. In Forensic DNA Typin g, 2nd edn. Amsterdam: El sev ier, 2005. 31 Van Hoofstat DEO, Deforce DLD, Brochez V et al. DN A typing of fing erp rints and skin debris: sensitivity of capillary electrophoresis in forensic applicati ons using multiplex PCR. Proceed in gs from the Second European Symposium of Hum an Identifi catio n, 1988, pp. 131-7. 32 Findlay I. Tayl or A, Quirke P et al. DNA fingerprinting from single cells. Nature 1997; 389: 555-6. 33 Vall Oorschot RAH, Jones MK. DNA fing erprin ts from fing erprints. Na ture 1997 ; 38 7:767. 34 Gill P, Whitaker J, Flaxman C et al. An investigation of the rigor of interpretation nIles for STRs derived from less than 100 pg of DNA Forensic Sci lilt 2000; 112:17- 40. 35 Wickenheiser RA Trace DNA: a review, discussion of theory, an d application of the transfer of trace quantities of DNA through skin contact. J Foren sic Sci 2002; 47:442-50. 36 Hartl DL, Jones EW The chemical structu re, repli cation, and manipulation of DNA Essential Gen erics: A Gellomics Perspectiue, 3rd edn. SudbulY, lVlA: Jon es and Baltlett Pu blishers, 2002. 37 Butler JM. The polymerase chain rea ction (DNA amplification). In Fore nsic DNA Typin g, 2nd edn. Am sterdam: Elsevier, 2005. 38 Butler JM. DNA separat ion methods : slab ge l and capill ary electroph oresis. In Forensic DNA Tvping, 2nd edn. Amsterd am: Elsevi er, 2005. 39 Thangaraj K, Reddy AG, Singh L. Is the amelogenin gene reliab le for ge nd er identification in forensic casework and prenatal diagnosis? Inr J Leg !VIed 2002; 116:121- 3. 40 Chang YI\rJ, Burgoyne LA, Both K. Hi gher failures of amelogeni n sex test in an Indi an population group. J Forellsic Sci 2003: 48:1309-1 3. 41 Mitchell RJ, Kreskas M, Baxter E et al. Amelogen in Y negative males: multiple origins. Illr Congress Ser 2006; 1288 :274-6. 42 Jobling lVlA, Lo IC, Turner DJ et al. Structura l variation on the short arm of the hum an Y chrom osome: recurrent multigene deletions encompassing amelogenin Y. Hum Mol Gen 2007; 16:307-16. 43 Sudbe ry P. Hu man population geneti cs and evolution. In Humal/ Molecular Generics, 2nd ed n. Amsterdam: Pearson Ed ucation, 200 2. 44 Butler JM. STR popul ation database analysis. In Forensic DNA Typing, 2nd edn. Amsterdam: Elsevier, 2005. 45 Hartl DL, Jones EW Molecular mechanisms of gene regulation. In Essential Gel/etics: A Genomics Perspective, 3rd edn. Sudbury, MA: Jones and Bartlett Publishers, 2002. 46 Winter PC, Hickey Gl, Fletcher HL. Section D popul at ion genetics and evolution. In Ham es BD (ed.) Genetics: Insrant Notes Series, 2nd edn. Ox ford, UK: BIOS Scien tifi c Publishers Ltd, 2002. 47 Butler JfVl. Profile frequency estimates, likelihood ratios, and so urce ottributioll. In Forensic DNA Typing, 2nd edn. Amste rdam : Elsevier, 2005 . 48 Foreman LA, Evett IW Statistical analysi s to support fore nsic interpreta ti on for a new ten-locus STR profiling syste m. Inr J Leg Med 200 1; 114 : 147-55. 49 Jha A DNA fingerprinting 'no longe r foolproo f. Guardian, Thursday 9 September 2004.
416 I
Forensic DNA profiling
50 Lu cy D. DNA I. In Introduction to Sra listics for Forensic Scientists. Chichester: John Wiley and Sons, 2005. 51 Lucy D. Eva luation of evidence in practice. In Introduction fO Statistics fo r Forens ic Sci entists. Chichester: John Wiley and So ns, 2005. 52 Evert IW. Towards a uniform fram ework fo r reporting opini ons in forensic science casework. Sci Justice 1998; 38: 198-202. 53 Forensic Science Service, Association of Police Surgeo ns and Association of Chief Police. Officers' Guidelines for the Collection of Specimens. London: Forensic Science Service, 2000. 54 Inman K, Rudin N. Good field practice - processing a crime scene. In Principles and Practice of Criminalistics: the Profession of Forensic Science. Boca Ra ton, FL: CRC Press LLC, 2001. 55 National Criminal Justice .Reference Servi ce. DNA evidence: what law en forcement officers should know. NCJ 200908. Nat In st Justice J 2003; 249: 10- 15. also Available online: www.ncjrs.gov/pdffilesl/jr000249c.pdf (accessed April 2007). 56 Kaye DH, Sensabaugh Jr GF. Reference guide on DNA evidence. In Reference Manual on Scienfific Evidence, 2nd edn. Washington, DC: Federal Judicial Center, 2000, pp. 485-57 6. 57 Fa igman DL, Kaye DH , Saks MJ , Sanders J. Modem Sc ientific Evidence: Th e Law and Science of Expert Testimony. 25 -2.5 DNA Typing; The scientific statu s af DNA typing: Laboratory performance. St Paul , MN: West Publishing Co., 2002, pp. 257-63. 58 Lynch M. The discursiv e production of uncertainty: the OJ Simpso n 'Drea m Team ' and the sociology of knowledge machin e, Soc Stud Sci 1998; 28 :829-68. 59 Scientifi c Testimo ny, an onlin e journal: DNA tes ting problems. Avail abl e online: www.scientific.org/ DNAProblems/DNA-Problems.htm (acc essed April 2007) . 60 Toledano T, Quarino L, Leung S et al. An assessment of DNA contamination risks in New York City medical ex amin er facilities. J Forensic Sci 1997 ; 42:721-4. 61 Rutty GN. Human DNA contamination of mortuaries: does it matter? J Palh al 2000; 190:410-1 I. 62 Rutty GN, Watson S, Davison 1. DNA con tamin atio n of mOltuary instruments and work surfaces: a significant problem in for ensic practice? In l J Leg Med 2000; 114:56-60. 63 Lowe A, Murray C, Whitaker J et at. The propensity of individuals to deposit DNA and second ary transfer of low-level DNA from individuals to inert surfaces. Forensic Sci lilt 2002; 129:25- 34. 64 Rutty GN. An investi gati on into the transference and survi vabili ty of human DNA follo win g s imul ated manual strangulation with consideration of the problem of third party con taminat ion. lnt J Leg Med 2002; 1J 6:1 70-3. 65 Sullivan K, Johnson P, Rowlands D, Allen H. New devel op ments and challenges in the use of the UK DNA database: addressing the issue of contaminated consumables. Forensic Sci Int 2004; 146:5175-6. 66 Peel C, Gill P. Attribution of DNA profiles to body fluid stains. lnl Congress Ser 2004; 1261 :53-5. 67 Bond JW. Value of DNA evid ence in detectin g crime. J Forensic Sci 2007; 52:128-36. 68 Gill P, Kirkham A. Development of a simulation mod el to ass ess the im pa ct of co ntamination in casework using STRs. J Forensic Sci 2004; 49:485-91. 69 Rutty GN, Hopwood A. Tucker V The effectiveness of protective clothing in the reduction of potential DNA contaminati on of the sce ne of Clime. ln t J Leg Med 2003; II 7: 170-4. 70 Butler JM. Appendix N: DNA Advisory Board Quality Assuran ce Standards. In Forensic DNA Typing, 2nd edn. Ams(erdam: Elsevier. 2005.
71
72
73
74
75
76 77 78
79
80
81
82
83
84
85
86
87
Thom so n WC, Ford S, Doom T et aJ. Evaluating foren sic DNA evidence, Part 2: Essential elements of a co mpe tent defense review. Champ ion Magazi ne, 24-28 May 2003, www.nacdl.org/_852 570 IC00653 9F8.nsf/0/9 C5F 7C6D724A4 7518 52 56E540074C156?Open (accessed April 2007). Teichroeb R. .Rare look inside state crime labs reveals recurring DNA test problems. Seattle PosHntelligencer, Thursday 22 July 2004, http://seattlepi.n wsource.com/ local/183007 _crimel ab22 .html (accessed April 2007). Pompa non F, Bonin A, Bellemain E, Taberlet P. Gen otyp ing errors: causes, consequences and so lution s. Na t ReIJ Genet 200 5; 6:847- 59. Faigman DL, Kaye DH, Saks MJ, Sand ers J. Modem Scientific E,Jidence: The Law and Science of Expert Testimony. 25-1.2 DNA Typing; Th e legal rele/Jance af DNA tests: Objections ro DNA eIJidence. St Paul , MN: West Pu blishi ng Co., 2002 , pp. 210-36. Stevens -Simon C, Nelligan D, Breese P et al. The prevalence of geni tal human papillomavims infections in abused and non ab used preadolescent girls. Pediatrics 2000 ; 106:645-9. Thompson C The GUiltl/Paediarric il/ le /joc e. Presentation at 22nd lusn Europe Meeting. Versaill es, Abstract 0.048, 2006. Kl aus ner JD. The NAAT is out of th e bag. Clin Infect Dis 2004 ; 38:820-1. Hammerschlag MR. Use of nucleic acid amplification test in inves tigating child sexual ab use. Se.r Tran sll1 Infect 2001; 77: 153-7. Hammerschlag MR. Appropriate use of nonculture tests for the detection of sexually transmitted disea ses in children and adolescents. Semin Pediatr Infect Dis 2003; 14:54-9. Kellogg ND, Baill argeon J, Lukefahr JL et al. Comparison of nucleic acid amplification tests a nd culture techniques in the detect ion of Neisseria gonorriloea e and Chlamydia trachoma tis in victims of suspected child sexual abuse. J Pediatr Ado/esc Gynecol2004; 17 :33 1-9. Tho mas A, Forster G, Robinson A, .Rogstad K. National Guidelines on th e Management of Suspected Sexually Transmitted Infectio ns in Children and Young people. Clinical Effectiven ess Group (Assoc iation of Genitourinary Medicine and th e Medical Society for the Study of Ven ereal Diseases), 2002. Availabl e on lin e: www.bashh.org/guidelin es/2002/ ado lescencfi nal_0903.pdf (accessed April 2007). 2001 Nati onal Guidelines on the Management of Adult Victims of Sexual Assault. Clinical Effectiveness Group (Ass ociation of Genitourinary Medicine and (he Medica l Society for the Study of Venereal Diseases). Avail able online: www.bas hh .o rg/gu idelines/2002/sexassauIC0601.pd f (accessed April 2007). Csete K, Beer Z, Varga T. Prenatal and newborn paternity testing with DNA analysis. Forensic Sci Int 2005 ; 147:S57-60. Robino C, Barilaro MR, Gino Set al. Inces tuous paternity detected by STR-typing of chorionic villi iso lated from archival formalin-fixed paraffin-embedd ed abortion mate rial using laser microdissection. J Forensic Sci 2006; 51 :90-2. Christian CW, Lavelle JM, De Jong AR et a!. Forensic evidence findin gs in prepubertal victims of sexua l assault. Pediatrics 2000; 106: 100-4. American Academy of Pediatrics (Committee on Child Abuse and Neglect). Guidelines for the eva lu ation of sex ual ab use of children: subj ect revi ew. Pediatrics 1999; 103:186-91. Allard JE. The collecti on of data from findings in cases of sexual assau lt and the sig nificance of spermatozoa on va gina l, an al and oral swabs. Sci Ju sti ce 1997; 37:99 - 108.
---
418 I
131
132
133
134
135
13 6
137 138
139
140
141
14 2
143
144
145
146 147
148
149
150
Forensic DNA profiling
Tully LA, Parso ns TJ, Steighner RJ et aJ. A sensitive denaturing gradi ent-gel electro phoresis assay reveals a high frequ ency of heteroplasmy in hyp erva liabl e region I of the human mtDNA co ntrol region. Am j Hum Gel1et 2000; 67:432-43. Paneto GG, Martins JA, Lon go LVG et a!. Heteroplasmy in hair: differences am ong hair and bl oo d From the same individuals are still a matter of debate. Forel1s ic Sci lilt 2007; 173: lJ 7-21. Gill P, Ivanov PL, Kimpton C et aJ. Id en tifi cation of th e remains of the Romanov family by DNA analysis. Nat Gel1et 1994; 6:130- 5. Ivan ov PL, Wa dham s MJ, Roby RK et al. Mitochondrial DNA sequence hetero plasmy in the Grand Duke of Russia Georgij Romanov establishes the authenticity of the remains of Tsar Nich olas II. Nat Gellet 1996; 12:417-20. C6li ca MY, Rodriguez Card ozo MB , Abovi ch MA et aJ. Paterni ty cases when the alleged father is missing. Int Co ngress Ser 2004; 1261 :520- 2. Mayor LR, Bald ing DJ. Discrim in ation of halF-sib lings when maternal genotypes are known. Forel1sic Sci lilt 2006; 159: 141-7. Tra cey M. Sho l1 tandem repeat-based identification of indi vid ual s an d parents. Croatian Med j 2001; 42 :23 3-8. Linacre A, Goo dwin W. Role of and techni ques in fo rensic hem atogentics - paterni ty testing. In Payne-James J, Busutti l A, Smock W (eds) Fo rensic Medicille: Clinical al1d Pathological Aspects. London: Greenwich Med ica l Media Ltd, 2003, pp. 677-88. Fun g WK, Hu Y-Q , Ch un g Y-K. On statistica l analysis of forens ic DNA: th eory, methods and co mputer programs. Forel1s ic Sci lilt 2006 ; 16 2:17-23. Coleman H, Swenso n E. DNA ill the Courtroom: a Tmil Watcher's Gu ide. Washin gto n, DC: GeneLex Press Seattle, 2000. Avai lab le on line: www.genelex.com/ pate rnitytesting/pate rnityboo k.ht ml (accessed Ap ri l 2007). Relationship Testin g Program Unit 2004. Annual Report Summary for Testi ng in 2004. Ava il abl e online: www.aabb. 0 rglD ocu men ts/ Accred ita tio n/Pa ren tage_Testin g_ Accreditation]rogram/rtannrpt04.pdf (accessed April 2007). vo n Wurmb- Schwa rk N, M
151 \ 52 153
154
155
156
157 158 159
J 60
161
162
163
164
165
166
167
168
169
170
171
Ottenberg R. Medicoleg al application of human bl oo d grouping.] Am Med Assoc 192 1; 77:682-3. Shaw MW. Paterni ty determination 1921 to 1983 and bey ond . jAm Med Assoc 1983; 250:2 53 6-7. Albrecht K, Sch ulth eiss D. Proof of paternity: historical refl ect ions on an andrological-forensic ch al lenge. Andrologia 2004 ; 36:3 1-7. BBe. Embryo mix-up at IVF hospital. Monday 28 October 2002. http://ne ws. bbc. co.u k/ l/hi/h ea lth/2 36770 5.st m (accessed Ap ril 2007). BBe. $1111 pay-outJol' IVF embryo mix-up. Thursday 5 August 2004. Available onl ine: ht tp://news.bbc.co.uk/ l/hi/he alth /353 7950.stm (accessed Apri l 2007). BBe. NF mix-ups 'occur regularly'. Wedn esday 24 July 2002. Ava il ab le online: http://news.bbc.co.uk/ l/hi/health/21484 23 . still (accessed April 2007). Dy er O. Black twins are born to white parents after infertility treat ment. BM] 2002; 325:64. Spriggs M. lVF mixu p: white couple have bla ck ba bi es, ] Med Ethics 2003; 29:65. BBe. ORA: Legal minefi eld of IVF 'm istake '. Monday 8 July 2002 . Avai lable on line: http://news.bbc.co.uk/ I/hi /health/ 2116589.stl11 (accessed April 2007). BBe. Court rules on fath er of IVF mix- up twins. Wedn esday 26 Febru ary 2003. Avail ab le onlin e: http:// news.bbc.co. uk/ I/hi/ hea lth /2797J69.stlll (a ccessed Apri l 2007 ). McKonkey E. Lectu re 2 - the mechanisms of mutation: how the hum an ge nome cha nges. In How tiJe Hum an Genome Works. Sud bury, MA: Jones and Bartl ett Publishers, 2004, pp. 19-33. Brin km ann B, KJintscher M, Neuhuber F et al. Mutati on rate in human microsatellites: influence of the structure and length of the tand em repea t. Am j Hum Gel1 1998; 62:1408-15. Saja ntil a A. Lukka M, Syvanen A- e. Experimentally observed germ line mutatio ns at hum an micro- and minisateJlite Jo cl. Eur] Hum Genet 1999 ; 7:2 63 -6. Kayser M, Roewe r L, Hed man M et al. Characteristic and frequency of germline mutations at microsatellite lo ci from the human Y chromosome, as revealed by direct observation in father/son pairs . Alii j Hu m Genet 2000; 66:1580- 8. STR DNA Interne t Data base Mutation Rates for Common Loci. Ava ilabl e onlin e: www.cstl.n ist.gov/biotec h/strbase/ mutati on.htm (accessed Apri l 2007). Yamamoto T, Tam aki K, Hu ang XL et al. The application of minisatellite variant repea t mapping of PCR (iVIVR-PCR) in a paternity case showin g fals e exclu sion du e to STR mutation. ] Forel1sic Sci 2001; 46:374-8. Brandt- Casad eva ll C, Gene M, Piqu e E et al. Presence of two mutations between fathe r and child in two cases of paternity tes ting. 1111 Congress Ser 2003; 12 39:S 15-1 8. Nutini Ai, Mariottini A, Giunti L et al. Double incompa tibili ty at human alpha fibrinogen and penta E loc i in pa te rnity testing. Croatian iVIed] 2003; 44:34 2- 6. Thangaraj K, Reddy AG, Si ngh L. Mu ta tion in th e STR locus D2 1S 11 of father caus ing all ele mi smatch in the child. ] Forel1s ic Sci 2004; 49:99-103. Jun ge A, Brinkmann B, Fimmers R, Madea B. Mutations or exclusion: an unusu al case in paterni ty testing. Int] Leg Med 2006; 120:360-3. Crouse CA, Rogers S, Ami ott E et a1. Analysis and interpretation of sh ort tand em rep eat microvariants and three-ba nded allele patterns usin g multiple allele detection systems. j Fore nsic Sc i 19 99 ; 44 :87 -94.
References I
172 Clayto n TM , Guest JL, Urq uhart AJ, Gill PD. A ge netic bas is for ano malous band patterns enco untered during DNA STR profiling. J Forensic Sci 2004; 49: 1207-14. 173 Sudbery P. Hu man genetic disease. In Hum lln Molecu lllr Genetics, 2nd edn. Harlo w, UK: Pea rso n Ed ucat ion. 2002 . 174 Dauber EM, Wenda S, Glock Bet aJ. Mosa icism as a poss ibl e reason for poor amp lification of ame logenin-Y in three hum an male indiviciua ls. fl1l Congress SCI' 2004: 1261 :508-10. 175 Farber CM. Georges M, De Bock G et aJ. Dem ons tration of spo ntaneous XX/XY chimeri sm by DN A fin ge rprinting. Hum Genet 1989; 82:19 7-8. 176 Yu N, Krus kall MS, Yun is JJ et al. Disp uted maternity leading to identificatio n of tetra ga metic chimerism. N Eilgi J Med 2002; 346:1545. 177 Mayne E, Borland S. The moth er with three chi ldren who don' t share her DNA. Ma il all SUlidav. 5 March 2006. 17 8 Mild e A, KJ1I1-Burm eiste r R. Ritz-Timme S. Ka atsch HJ. DN A typin g in cases of blood chimeri sm. flit J Leg Med 1999; 11 2:333 - 5. 179 Rub ock i RJ, McCue BJ, Kelly MT et aJ. Natural DNA mixtu res ge nerated in fraternal twin s in utero. J ForellSic Sci 200 1; 46: 120-5. 180 Dau ber EM, Fat' i, Stad lbache r S et al. STR typ in g in a pair of chimeric tw ins. In( Congress Ser 2003: 1239 :569 - 71. 181 Dau ber EM, Dorner G, Mitte rbau er Met al. Discrepant results of samples taken from diffe rent ti ss ues of a singl e ind ividual. jilt COllgress Ser 2004; 1261 :48- 9. 182 Dauber EM, Muller CJ. Schii niger-H ekele M et al. Artificial blood chimerism and grafl-versus-host disease after liver transpl anta tion. ll1 i Congress Su 2006; 1288:840- 2. 183 Pop e S, Chap man I-I, Lam bert J. The effec t of bon e marrow transp lants on DN A profi les : a case exam ple. Sci Justice 2007; 46:23 1-7. 184 von Wurl1lb-Schwark N. Bos inski H. Ritz-Timme S. What do the X and Y chromosomes tell us about sex and gende r in foren sic case analysis? J ForeJIsic Legal Mer! 2007: 14 :27-30. 185 Aldhous P. Bone marrow donors risk DNA id enti ty mix-up. New Scientist 2005; 2523: 11. 186 Rodri guez Cardozo MB , Co lica MV, Abov ich fVIA et aJ. Maternity testing in a ch imerical child . jJIt COll(Jress SCI' 2004; 126 1 :502-4. 18 7 Foster EA. Job ling MA. Tay lor PG et aJ. Jefferson fathered slave's child. Nature 1998; 39 6:27-8. 188 Marshall E. Wh ich Jefferso n was th e father? ScieJIce 1999; 283: 153- 5. 189 Check E. Jefferso n's desce nd ants continue to deny slave link. Nature 2002; 41 7: 213. 190 Pa abo S. Molecular cl on in g of anci ent Egyptian mumm y DNA. Naiure 1985; 314 :644- 5. 19 1 Ca laca l G, De lfin F, T'111 M et aJ. Id entificatio n of ex hum ed rema in s of fire tragedy victims using conve ntio na l method s and autosomallY-chromosomal short tandem rep eat DNA pro filin g. Am J Forensi c ivied Patlw12 005; 26:285- 9 1. 192 Ca lacal G, De Ungria MC, Delfin F et al. Iden titlca tion of two fire vict im s by co mp arat ive nuclear DNA ty ping of skeleta l remains and stored umbili ca l tissues. AI1I J Forensic Med Patilol 2003 ; 24 : 148- 52.
419
193 Meye r HJ. The Kaprun cab le ca r fire disaster - aspecrs of forensic organ isation t'ollowing a ma ss fata lity with 155 victims. Fore nsic Sci fJI t 2003 ; 138: 1- 7. 194 Clayton TM. Whitaker JP, Maguire CN. Id en tificati on of bodies from the sce ne of a ma ss disaster lI sin g DN A amplification of sholt ta ndem repeat (STR) loci. Forel1sic Sci fl1t J995: 76:7 -15 195 Hsu CM, Hu ang NE, Tsai LC et al. id entificatio n of victims of the 1998 Taoyuan Airbus cras h accid ent using DN A analys is. JJIt J Leglll Mer! 1999; 11 3:43 - 6. J96 Pres iden t's DNA initiative. Lessons learn ed from 9/ll: DNA identification in mass fatality inciden ts, 200 6. Avail able online: www.ncj rs.gov/pdftil es I/110/214 78 1.pd f (acc essed Ap ril 2007). 197 Bre nner CH. We ir BS. Issues a nd strateg ies in the DN A identification of World Trade Ce nter victims. Theor Popul Bio i 2003; 63:173 -8. 198 Coble 1\110, Butler J1\I1. Ch'lracterization of new miniSTR loci to aid ana lysis of deg raded DNA. J Foreusic Sei 2005; 50:43 -53. 199 Butler J iVI. Single nucleotid e polymorphis\11s ami oth er bi -allelic mark ers. In Forellsic DNA T)'pin g, 2nd edn. Amste rdam: Elsevier, 2005. 200 Sob rin o B, Brion lVI, Carracedo A. SNPs ill fore nsic genetics: a review on SNP typ ing methodologies. Forens ic Sci fni 2005 : 154:1 8[ - 94. 201 Fore nsic Science Se rvice Casefiles : Craig Harman - Family DNA link offers crime breakthrough. Tile Forensic Science Sen'ice Website, www. forensic.gov. uk / forensic_t/inside/ news/lis c casefiles.php?case=24 (accessed April 2007). 202 Beiber FR. Guilt by associa tion. Ne/U Scientist 2004; 2470:20. 203 Gree ly HT, Riordan DP. Garrison NA , Mou ntain JL. Family ties: the use of offen der da ta bases to catch offen der's kill. J Law M ed Erliics 2006; 34: 248- 62 . 204 Haimes E. So cia l and ethical issues in the USe of familial searchi ng in forensic investigations: insi ght from famil y and kin ship stu dies. J Law Med Ethics 2006; 34: 263 - 76. 205 Grimes EA, Noake PJ, Dixon L. UrlJ uart A. Sequ~nce polymorphism in the human l11 elanocoltin I rece ptor geJl e as an indicator of the red hair phenotype. Forensic Sci flit 200 1; 122: 124- 9. 206 Rees JL. Geneti cs of hair and sk in co lor. AIIII Re/J Genet 2003; 37 :67 -90. 207 Uil11aso n RL, MohicJeen MPK, Mes t JR et al. SLC24A5, a pu tative cation excha nger, affects pigmentation in zebra fi sh and huma ns. Sciellce 2005: 310: 17 82-6. 208 Cho MK. Sa nk a r P. Fore nsic genet ics and ethica l legal and social implicati ons beyon d th e clinic. Nat Genet 2004; 36:S8 - J 2. 209 Ossorio PN. About face : fo ren sic genetic test ing for race and v isible trai ts. J Law Med Ethics 2006; 34:277-92. 210 Will iams R. Johnson P. Inclusi veness, eITectiveness and intrusiveness: issues in the develop ing uses of DNA profilin g in supp ort of cri minal investiga tions. J Law /Iiled Eth ics 2006; 34:234-47. Alva rez M, Ballantyne J. Th e ide ntification of newbo rn s 211 usin g messenger RNA profiling analys is. Anlll Biochem 2006; 357:21-34.
I
CHAPTER 22
I
THE DENTIST'S ROLE IN CHILD ABUSE AND NEGLECT David Whittaker
Introduction Dental neglect Facial and oral pathology
420 422 422
INTRODUCTION
Damage to a child, other than by accidental means, has been recognized for centuries I but app ears to have firs t been described in a formal manner in 1962. 2 The authors used the term 'batte red child syndrome' but this forms part of what is more commonly called 'child abuse'. Ch ild abuse may include non-accidental injury (NAIl but also emotional ab use, sexual abuse and neglect. It has been increasingly recognized that the denta l surgeon may be the first profes sional to suspect physical abuse, especially in relation to injuries involving the orofacial structures. 3 A child is consid ered to have been abused if he or she is treated in a way that is unacceptable in a given culture at a given time. 4 For the purposes of this review, only physical abuse in the form of NAI and neglect will be considered . Neglect occurs when an adult knowingly allows a child to endure pain or suffering or fails to provide the basic requisites for proper maturation. Since the syndrome was specifically named it has become possible to add fu rther injuries to the pattern of injuries described. These further injuries include bite marks,s which may be found anywhere on a child's b;dy. Bite marks in children rep resent child ab use until proven otherwise. They are rarely acci denta l and are good indica t ors of genuine child abuse. 6 There have been ap proxi mately 50 articles in the international dental li terature relating to child abuse and bite marks. The dentist's rol e in diagnosis and management was described in 1986. 7 How ever, the problem is still not well appreciated by the profes sion at large. Apart from the specific injury of bite marks,
The dentist accused of child abuse Bite marks References
425 425 432
it is not the responsibility of general dental practitioners (GDPsl to diagnose child abuse. B Dental surgeons are, how ever, in a good position to diagnose NAI to the face and oral structures and they should be aware of the possibility of abuse and have some knowledge of the key factors involved in its diagnOSis. The diagnosis of NAI is difficu lt to make with certainty. and va ri ous pathological or post mortem changes have been mistakenly identified as child abuse.9 Conversely, child abuse or neglect has gone unde tected in some unexp ected deaths in children 3 or in re po rts of facial or intraoral injuries. 3 The incidence of NAJ to the head and neck region is high, v arying between 11 and 50 per cent of cases. 10,11 Therefore, the dental profession com prises an important part of health professionals involved in identifying and reporting child abuse or neglect. 12,13 A number of key indicators of NAt have been identified . 14 The more of these that are present the more likely it becomes a defini tive dia gn osis: • delay in seeking help; • a vague history or a history that v aries between interviewees; • a history incompatible with the injury seen; • abnorma l behaviour of the parents; • the child's app earance, e.g . withdrawn, frightened or sad ; • other signs of abuse, such as neglect or deprivation; • dis closure of an incident by the child. The diagnosis of NAI is discussed in greater detail in Chapter I.
Introduction I
The prevalence of child abuse is reported differently in different countries, presumably reflecting social and cultura l differences. 15 The true extent in a given population is diffi cult to determine, but in 1967 in London it was believed that at least 300 cases occurred per year. 16 In terms of mortality, in the 1980s at least 4 children in Britain 17 and 80 children in the USA 18 died weekly as a result of abuse or neglect. It was reported in 1988 in Britain that at least one child in every thousand under 4 years of age suffers severe physical abuse each year. This might include fractures, brain haemor rhages, severe internal injUli es or mutilation. 19 In 1991, an average of 4.2 children out of every thousand in the UK were known to have been physically abused , the highest rate being seen in children less than 1 year of age. 20 It is thought that one in 10 000 children in Britain will die as a result of abuse, although many workers consider mortality rates to be considerably higher. It is generally believed that proven cases of child abuse may be four or fi ve times as frequent as t hey were some 10 years ago. 4 In 1974, 97 loca l authorities in the UK were approached by questionnaire in a survey covering 90 per cent of the population. Some 5700 cases came to light in the last 9 months of that year with a mortality of 0.7 per cent and a significant risk of re-injury.21 In the USA, it is mandatory for health-care workers to report suspected cases of child abuse. The second National Incidence a nd Prevalence Study of Child Abuse and Neglect22 reported that 4.95 children in every thousand were physically abused. Girls were more li kely to be sexually abused than boys, although there was no gender difference for physical abuse. Low family incom e is significant, and children of Afro-Caribbean origins were more likely to be physically abused than other racial groups. Sexual abuse is more common in Caucasian children. Whether or not the true occurrence is increasi ng, it is clear that increased aware ness results in many more cases being reported and, on the basis of current statistics, an individu al dental surgeon may be expected to encounter a case at least once every 3 years. In spite of this, British dentists have not formally been requested to develop an awareness of NAls to children, and many health authority child abuse procedure documents make no mention of the possible role of the dentist. 23 Since the Children Act (1989) and Working Together (1991), child welfare services in the UK are required to have procedural guidelines to deal with suspected child abuse cases. These app ly to aU persons working with children, and general denta l practitioners are includ ed in this group. National guidelines do not exist for GDPS,8 and the document Working Together (1991) gives very little information about the role of the GDP. The British Dental Association in its booklet Ethics in Dentistl)! does produce guidelines for child protection 24 but it suggests that each individual GDP should meet the guidelines laid down through their local district health authority. It is important that GDPs keep adequate and accurate up-to-d ate dental records of all their patients; in the case of child abuse, these may need to be referred to in case conferences or even court procedures. The guidelines point
421
out that the relationship between parent and child, the child's reaction to other people, the child's reaction to a dental or medical examination or the general demeanour of the child may have a bearing on possible child abuse. It is important that the dental practi tioner without direct extensive experience in referral techniques refer cases to the local consultant in paediatric dentistry, the local con sultant in dental public health or an appropriate consultant paediatrician. It may be necessary and wise to discuss any potential referral with the general medical practitioner of the child concerned. Under the 1989 Children Act and Working Together (1991) the local authority has a legal duty 'to make or cause to be made' such enquiries (by another agency, such as the National Society for the Pre vention of Cruelty to Children [NSPCC]) as t hey consider necessary, to enable them to decide whether t hey should take any further action to safeguard the child 's welfare.25 ,26 There are a number of reasons why general dental practi tioners or their staff may fail to report potential child abuse. Therefore, each local area should have a simple system, preferably referrin g t he case through the normal referral sys tem that dentists might use, such as to their local paediatric dentist or paediatric consultant. Failure to refer a child may, of course, have serious consequences. Referral rates by GDPs are not known, but it is known that healt h professionals in general refer 17 per cent of suspected abuse cases, whereas teachers and school nurses refer 23 per cent of cases,u Reporting cases of child abuse is mandatory in the USA, where 10 subtypes of child abuse and neglec t have been classified 28 The most frequent of these is physical abuse, with an incidence of a lmost 32 per cent, and this is closely followed by educational neglect (28 per cent) and byemo tional a buse (26 per cent). Suggestions as to how t he dentist should examine a child thought to be suffeling from child abuse or child neglect have been made. 29 These include the relationships between the child and his or her parents; the cleanliness, stature and nutritional status of the child; the quality of his or her cloth ing; and the presence of physical problems, such as a limp or difficulty in climbing into the dental chair. It is recommended that the body be systematically examined for traumatic injury, bruises, previous scars, burn marks or injury caused by foreign bodies. If the dentist's suspicions are aroused, the par ents should be informed that a n injury has been observed; the dentist must then contact the approp riate child abuse or child neglect authority. In the UK, it has been said that it is not the responsibility of GDPs to make the diagnosis of child ab use, nor should they ever do SO.8 However, the same authors state t hat no GDP should feel gui lty abou t referring children with suspected child abuse. This does not mean that they are accusing either parent or carer but they are simply seeking a second opinion on a condit ion that is outside their expertise. Failure to refer a child in need may result in more severe injuri es in t he future. The failure to follow up a suspicion of child abuse by a professional health-care worker may be construed as a form of professional negligence. Accidentally
422 I
Dentist's role in child abuse and neglect
acquired bruises or pseudobntises must be differentiated from inflicted injuries. It is self-evident that infa nts who ca nnot crawl cannot cause a self-inflicted accident. It follows that severe bruising or fractures in a child below th e ages of 6-9 months are almost always inflicted non- accidentally by a second party. Alth ough not directly involved, general dental practition ers should be aware of the problem of child sexua l abuse, although this requires specialist and expert intervention from those health-care workers trained specifIca lly in this area. 30 There is certainly in adequ ate in for mation both in training and in th e dental literature to assist the paediatri c dentist in examining these patients. Any physical examina ti on of a child should normally be confmed to those parts of the body considered to be within the expertise of the dentist. Suspicion of abuse lying in the purely medical fi eld should be referred to a consultant paediatrician for complete examination.
DENTAL NEGLECT Neglect of a child 's dental treatment needs could be con strued as a form of abuse. In this insta nce it is best described as an act of omission rather than an act of commission. Gross dental caries, exacerbated by poor diet, use of car cinogenic drinks and poo r ora l hygiene procedures, clearly lie within the resp onsibility of the child's parents. However, the rate of progression and extent of dental ca ries may also be genetically influenced and may dep end to som e ex tent on the 'q ua li ty' of the tooth ena mel, on the microbiological flora of the oral cavity an d on the flow and buffering cap ac ity of sa li va. These are clearly not the responsibility of the parents. Th e issu e is therefore co mp lex an d sho uld be app ro ached with care. In the USA, state laws have defined chi ld neglect as negligent tre atm ent, including a failure to provide ade quate care, su pport, nutrition or med ica l or surg ica l care so as to threaten a child's health or welfare. In Ca lifornia, den tal neglect is listed as a component of physical neglect. 28 In 1998, 15000 denta l surgeons in California were ques tioned concerning their knowledge, attitudes and practices regarding child ab use and neglect. 31 Seventy-one per cent had never seen a case of child abuse or neglect. Twen ty-eight per cent received no forlllal training in the recogn ition of child abuse or neglect whilst at dental schoo l. However, 84 per cent had read postgraduate literature on the subject s~nce qualifYing. Sixty-four per ce nt were aware that they were required by la w to report ab use and neglect, but alm ost the sa me percentage did not kn ow whether failure to report was a felony. Approximately the same percentage were aware that fa ilure of parents or gu ardians to foll ow through in dental treatment once they had been infoillled about a chi ld 's ram pant caries could be considered as child neglect. AJ.most hal f of those questioned were aware that there is a strong con'ela tion between dental neglect and the presence of physical
neglect. 32 Betvveen 1991 and 1992, 2274 paediatric dentists and orthodon tists were surveyed in relation to reporting child abuse. Forty-two per cent of pa ediatri c dentists reported hav ing seen a case of suspected child abuse in the previous year, whilst only 15 per cent of the orthodontists did. More signif icantly, only 11 per cent of the olthodontists actually reported their suspected cases, whereas 48 per cent of the paediatric dentists did. Among those paediatric dentists, denta l neglect was a frequent reason for suspicion?1 Dental neglect, a com mon form of child maltreatment, should be suspected if ram pant caries and oral infection, bleeding and trauma persist despite the elimination of financial and transportation obsta cles. Questioning the child and parents separately may help un cover an obvious discrepancy ben,yeen the clinical findings and the histo ry33 There are no published repolis of child abuse or criminal neglect in th e UK relating to advice or con trol of gross caries. Guidelines from the USA 34 indicate that dentists an d dental personnel have an obligation to repoli occurrences of dental neglect. Dental neglect is 'wilful failure of parent or guardian to seek and follow through with treat ment necess ary to ensure a level of oral health essential for adequate function and freedom from pain and infection'. Dentists and dental staff should be trained in recognizing and reporting this type of ab use and neglect. There is evidence that dental neglect is misunderstood an d under-reported. Certain fami lies appear to be more at risk of such neglect than others 35 Dentists, particularly paediatric dentists, are in a unique position to eva luate not only the dental needs of chil dren but also famil y situations and circumstances. Greater collaboration between paediatricians and dentists in the eval uation and management of ab use an d neglect is in the best interests of the child 36 It is clear that further research is required to develop effective methods, both for eva lu ation and for the education of at-risk individu als.
FACIAL AND ORAL PATHOLOGY The facial region is said to be a common site of NAls in chi l dren, and it has been claimed that almost 50 per cent of chil dren suffer injUly to this part of the body.5 Of 260 victims of ch ild ab use surveyed in Boston, 49 per cent had sustained orofa cia l trauma 3 ? and a to ta l of 65 per cent sustained head and facial tra um a. In fact, head or facial trauma was the prin cipal reason for ad mission to the hospital in 45 per cent of all cases. Despite this, few of the 537 dentists surveyed at that time knew of their legal responsibility to repOli these injuries. Eleven per cent of them had seen orofacia l trauma cases of a suspicio us nature. Ho wever, only 22 confirmed cases of child abuse had been noted by th e den tist ; of these, only four were repolted to the social agencies. Other in vestigato rs have reported an incidence of 60 per cent of cases involving injury to the head and neck region 38 The incidence in the UK of fa ci al injuries related to child abuse has been cited as 43 per ce nt. 39 Whilst it may be open to discussion as to whether dentists should become involved in the diagnosis of child
Facial and oral pathology I
abuse, 15 it seems clear that he or she would have an imp or tant role to play in the recognition, diagnosis and treatment of injuries to the face and oral cavity. The largest survey ava il able appears to be one from 1992 which investigated the denta l aspects of 1248 cases of child abuse in a major Amer ican hospital over a 4-year period. 12 There were an equa l number of girls and boys, except where sexual abuse was involved. Fifty-three per cent of children were in the 0- to 4 year age group, and when all cases were considered together 38 per cent presented with injuries to the head, face and mouth. In a smaller group surveyed the same yea r, 56 per cent of children had abnOlmalities within the head and neck region. In a larger studlOover a 5-yea r period, 4340 patients were studied. Sixty-eight per cent of these were victims of sex ual abuse and 32 per cent of physical abuse. Of these physically abused children, 49 per cent had evidence of injury to the head and neck region, whereas in the sexually a bused children only one per cent had injuries in this area. There were 11 deaths in this particular series.
Extraoral Injuries The head and facial region is a common site for NAI not only in t he West, 12 but also in t he Indian continent. 41 It is not known why the face often seems to be the focus ofvio lent attack in children, but it may be because it represents the persona of the individu al. The types of injury have been classified J4 and include contusions and ecchymoses (66 per cent), abrasions and lac erations (28 per cent), burns and bites (four per cent) and fractures (two per cent). Bruises, lace rations, burns, cuts, scars and black eyes will suggest NAI.4 2 Extensive bruising of the face with a history of limite d trau ma, especially if both old a nd n ew bruises are present , is almo st diagnostic. High velocity slaps or lashes may cause an unbnlised, negative image of the object that caused the injury. 2o The margin of the injury may often be demarcated by petechial haemor rhage from capiliaries. 4J This type of injury is sometimes known as the tattoo bruise. Bruising to the soft tissues of the fac e a nd head may occur in many different ways, but acci dental falls rarely produce bruises in the areas of the cheek; they are more likely to involve skin overlying bony promi nences, such as the forehead or zygomatic bone. Bruises change colour as they age,44 but the rate of change from red to blue, to purple, to green, to yellow cannot be used as a reliabl e indicator of the age of the injury. Many injuri es to the head and neck are caused by the adult human hand . This can leave press ure bruises in the form of grab marks by fingers and thumb, and these are velY commonly seen follo w ing pinching or pulling of the ear. In these circumstances, there wiH usually be a match ing bruise on the posterior surface of the ear. They are seen more commonly in babies and infants and may be indica tors of potenti ally more serious injuries, such as subdura l haemorrhage from shaking, and should always be taken
423
seriously. Grip marks on the cheeks or chin can arise dur ing restraint during feeding. Hand marks may also be in the form of a slap or a punch. Slaps may leave an imprint of the abuser's hand, whereas punches to th e forehea d, nose and eye region may result in unilateral or bilateral peri orbital ecchymosis. Other reasons should be eliminated , such as infective, ren a l or haematological conditions but, if the dentist has any doubts concerning unexplained or sus picious lesions on the face, the child should be referred to a consultant in paedi at ric dentistry or in paediatrics.2,27 Approximately 10 per cent of NAIs to the child involve thermal injUly 4 5 These may be burns (73 per cent) or sca lds (27 per cent). Thirteen per cent of non-accidental thermal injuries occur on the face , head and neck;4Gthe rol e of the dental surgeon is clear in recognizing, diagnosing and reporting such injuries. Between 2 and 12 per cent of all inpatient burns are NAls infl icted with the purpose of pun ishing or controlling the child's behaviour. Burns to t he face have been found to be the most frequent injury in so me studies,J9 and those from hot, solid objects a re the easiest to diagnose, and are usually of second degree with out blister formation. The shape of the burn may resemble the hot object producing it. For example, cigarette burns produce a circular, punched-out lesion of uniform size, but these are relatively rarely seen on the face. However, they are pathognomonic for child abuse. The use of hot hair dryers to inflict facial injury has been reported. 47 Friction burns have been described , and these present with broken blisters over bony prominences, including those of the face, and can be caused by being dragged across a carpet. Scalds can be on any part of the body but, in so far as the face and head are concerned, the most common injUly is due to hot liquid being thrown at a child. 2o Non-accidentally produced fractures of the facial bones are uncommon in children and feature in only two per cent of cases. 42 Several patterns of skull fracture that are indicative of NAI have been described. 48 These are multiple fractures with complex configurations - fractures that a re depressed, wide or more than 5 cm long ; involvem ent of more than one cra nia l bone and non-parietal fractures ; or fractures associated with intrac ranial haemorrhage. Multiple injury and inade quate explanation in the history are the best indicators of a skull fracture being of non-accidental origin. Most fractures in physically abused children occur up to the age of three. 49 Accid ental fractures occur more commonly in older children of schoo l age. The most common facial fracture is nasal (45 per cent), mandibular (32 per cent), and zygomatic and orbit complex (20 per cent).50 Specialized treatment is required for facial fractures and their definitive management has been dis cussed in the oral surgery literature. J7
Intraoral Injuries A review of 1248 cases of child abuse 12 desc ribes five tooth fractures , three tongue and tongue frenulum lacerations,
424 I
Dentist's role in child abuse and neglect
two lip frenulum lacerations, 11 injuries to the ora l mucosa , eight palatal lesions, loose a nd missing teeth a nd ev idence of den ta l neglect. It has also been suggested 51 that sexu ally trans mitted diseases in children's mou t hs are a strong indi cation of sexual abuse. There are claims that child re n exposed to oral sexual activities over a co nside ra ble perio d of tim e may develop higher ca ries rates a nd erosion on the palatal surface of the maxill ary teeth. Den tal neglect in these patients may be pal1 of gen eral neglect. Although some reports in the literature have included the term 'neglect' in t he title, only a small number of these have discussed dental neglect as distinct from gen eral neg lect of the chi ld. An extraordinary case was reporied in the UK in 1990, whi ch escaped detection for several years. Three children in a family of six were found to ha ve miss ing permanent incisors. It tra nsp ired that their parents were extracting teeth as a punishment for bad beh aviour. 52 A 6-year-old Hi spa nic girl was reporied to have avulsed a tooth during physi cal abu se and the sw allowed tooth caused an oesophageal perforation resulting in a retropha ryn gea l abscess a nd a medi ast inal abscess co nta inin g the toot h. This ap pears to be a unique ma nifestation of dental neg lect. 5) The mou th is a frequent site of NAl in children, possi bly because of its psych ological signifi cance; it is the orga n of phonation , a nd v iolence may be used aga inst it in order to silence the child. 54 Contusi ons and bruises, which occurred in 43 per cent of a series of 14 cases reviewed , are the most common injuries to the mouth;42 lace ra tions and trauma to the t eeth eac h account for 28 per cent of injuries. Examination of the mouth should commence with the lips; the lips usually heal well follo wing min or injuries a nd scarring should alert the dentist to previous severe trauma. Bruised and swo llen lips or ab rasions at th e corners of the mouth are suggestive of bl ows from a fis t; NAls are espe cially likely if the bruises or injuri es are at varying stages of healing. The teeth themselves are vulnerable to blows to the mouth, from either a fist, a hand slap or a weapon. It must, of course, be remembered t hat children frequ ently avulse or fracture teeth in the course of normal play activities, but an unsa tisfacto ry history from the child or parent, or repeated atte ndance for treatment of injuries, should arous e the sus picion of the dent ist. The roo ts of the anterior teeth may be incompletely formed in children and, as a result, blows to th e face often result in complete avulsion and loss of the tooth rather than in fra ctu re of the roots. When teeth are missing, full radiographs of the jaw should be taken to rule out some abnormality of development or conditions of partial ano dontia; retained frac tured roots should be sought as these will require surgical removal. Mult iple healed micro frac tures of tooth roots may sometimes be v isible in intraora l radio graphs - these are al most always indicative of trauma by repeated abuse. Slight movements of the teeth may rupture delicate blood vessels pass ing through the apical fo ramina and, as a result, anterior teeth that survive trauma of this
-~~~
-
nature may become devitalized, especially if the root apices have completed their development. The crowns of the teeth are likely to become discoloured in the months fo llowing such trauma. A grey hu e usually develops ow ing to the pres ence of blood breakdown produ cts in the dentinal tubules, a nd the presence of discoloured anterior teeth for which a sa tisfactory ex pl anation is not forthcoming may suggest t hat NAI has occurred. The report of the au topsy findings in 29 fatal cases seen durin g a 2-year period at the London Hospital Medical Col lege included o ne of the first assessments of soft-tissue intraoral injury fo llowing child ab use. 5 Fifty per cent of vic tims had injuries to the head and neck, and 45 per cent of these had suffered damage to the upper lip region. This con sisted of laceratio ns to the mucosa of the inner aspect of the upper lip, close to the labi al frenulum ; these were so metim es so severe that th e reflected mucosa of the lip was completely tom away from that attached to the g ing iva. Tea ring of the labial frenulum is not uncommon in very small children, who may fall accidentally onto the mout h whilst learning to walk, but this lesion should arouse suspicion as to a possible non- accidental cause when it is found in children less tha n about 1 year old and more than ab out 2 years old. These injuri es can be produced by agg ressive bottle feeding or by forcing a spo on into the mouth, as is sometimes done by a frustrated parent whose child is a slow eater. The labial frenulum may also be damaged by blows to the mouth or by tearin g the upper lip away fro m the gingival attachment (Fig. 22. 1) (see a lso Fig. 8.6, p. 152). Following this report, the tom labial frenulum has been regarded as pa thognomonic of child abuse, but it must be noted t hat no later reports of intra oral da mage in cases of child abuse have indicated such a high incidence. Many paediatric dentists of considera ble experience worki ng in the UK have never seen a case of labial fren al damage that t hey could unequivocally relate to NAI. When it does occur, the condit ion is often associated with subluxation or fra cture of the upper centra l inciso rs, damage to the teeth being caused by the same blow.
Figure 22.1
Dam ag e to the labia l frenum may be caused by
forcible feeding or, as in this ca se, by a blow to the mouth. Teeth may be displa ced.
Bite marks I
Penetrating injuries to the palate, the vestibule and the floor of the mouth resulting from forced feeding or forcible introduction of spoons and other implements into the mouth have been described. Burns of the oral mucosa fol lowing enforced ingestion of hot or caustic liquids are seen quite commonly,27 as are intraoral cigarette burns and tear ing of the oral mucosa by a parent's finger placed in the mouth. Dentists who are unfamiliar with the range of oral pathologies of a non-accidental nature that may occur in the mouth should refer patients to a specialist if they are in doubt as to the cause of a particular lesion.
THE DENTIST ACCUSED OF CHILD ABUSE It is known that child abuse or NAIs to children are fre quently caused by an adult in close relationship to the child. It has been pointed out 55 that this may include a ch ild 's pro fessional adviser, such as a dentist. In the state of Tennessee in the USA malpractice insurance rates have increased by 1500 per cent between 1981 and 1987, and cases of child ab use/criminal assault have been reported against several dentists performing procedures on patients. Dental practi tioners have been accused of child abuse by using excessive restraint, for example by placing a hand over the mouth in order to deal with a difficult or noisy patient. This exercise was singled out by the Virginia Board of Dentistry as lead ing to reports of child abuse against dentists. Another risk is the non-consensual touching of a patient, even if this is aimed at improving a child's health. Questionnaires in the USA are currently being designed to discover how many dentists have been accused of battery or non-consensua l touching. The information is not easy to acquire, and until 1987 only one repol1 was available cov ering three cases. 55 These three incidents occurred during the course of normal t reatment and all three resulted in an arrest. As child abuse is a criminal offence, if it occurs in the dental surgery it is not covered by malpractice insur ance. 56 There is a danger of dentists becoming overdefen sive in these matters, and this may eventually be to the detriment of child patients.
BITE MARKS Bite marks have come to mean any injury or mark produced in flesh, foodstuffs or other material by the teeth and the surrounding soft tissues. 57 In non-living materials and in foodstuffs, the mark left is usually produced entirely by the teeth, but in bites left on human skin, the injury may be contributed to by pressure and suction and/or pressure from the tongue as well as from the teeth themselves. Recogni tion of bite marks as part of the multiple injuries sometimes seen in NAIs to children is a relatively new development. Attention appears to have been first drawn to the incidence of bite marks in NAls in a report of 13 forensic dental cases,
425
two of which involved teeth marks in battered babies. Trag ically, both children died. 58 Shortly afterwards, a case of a bite on a 17-month-old child was reported ;59 bite marks and a torn labial frenum, associated with battered child syndrome, were described . 5 The incidence of bites in victims was given as 12 per 100000 in the population of New York in 1977. 60 However, there are few studies on the frequency of bit ing associated with child abuse. In one study 61 nurses work ing in facilities that admitted child abuse patients were trained by forensic dentists to recognize bite marks on their patients. Bites were divided into incised bites without bruis ing, those incised with bruising or discolouration, marks with sucking bruises and marks in which the teeth had been d ragged over the tissue. Of 1100 children examined, 17 had evidence of bite mark abuse - an incidence of 1545 per 100000 sheltered chi ldren. Most of the bites had been inflicted on children between the ages of II and 15 years, but two were in ch ildren less than 3 years of age and four in chi ldren between 16 and 18. Most of the male victims were in the 4- to 10-year age group, whilst most females were in the 11- to 15-year-old age group. The most common site of bites was the head and neck region (43 per cent). The limbs and trunk were also common sites of attack. In an unpublished studyG2 of 99 cases of bitten children there were 169 bites. Thirty-three per cent were in children aged between 6 and J 8 months, 23 per cent in children aged between 18 months and 3 years but there were 17.2 per cent of bites in children under 6 months of age. There were no statistical differences in this study between the girls and the boys. Few studies have been carried out on the distribution of bite marks speCifically in chi ld abuse cases, but the anatomical distribution has been investigated in a series of 67 victims of varying ages. 63 Among this group, 13 were aged below 15 years. Female victims were most commonly bitten on the breasts, arms and legs, while the arms and shoulders were the most common sites in males. These results differ somewhat from those reported in the UK, where bite marks on the breasts of females were found to be more common than in the American sample. Bite marks on the arms were seen less frequently.64 Other studies on the anatomical distribution of bites include th ose on 122 injuries observed in a New York hos pital 65 and on 114 cases reported from Kansas.66 In both of these studies, bites were most often seen on the hands and fingers. It appears that predilection for the different parts of the body has changed in more recent times. In NAIs in children, the entire surface of the child's body may be attacked, resulting in bite marks on the cheeks, shoulders, chest, abdomen, arms, legs and buttocks. 67 Multiple bites at the same site have been described. In the Cardiff series,52 37 per cent were on the shoulder or arm, 31 per cent on the leg or buttock, 18 per cent on the head or neck region (of which 15 per cent were on the face) and 12 per cent were on the ab domen or back. Less than three per cent were on the chest region. There were no significant differences in
426 I
Dentist's role in child abuse and neglect
Injunes as isolated findings, with only 25 per cent of the children having associated injuries. Seventy per cent were caused by adults and 30 per cent by other children. The importance of a close working relationship between paedia tricians and forensic dentists is emphasized. Other studies have claimed that quite frequently there are other injuries, such as pinch marks, bruises or burns.72 Bite marks in the older child may be less punitive in nature and are more likely to be related to physical or sexual abuse. Distinguishing between adul t and child bite marks may not be straightfor wa rd . Using receiver operating characteristic analysis, it has been shown that the best decisi ons were made by senior or junior experts in the field. General dental practitioners and police officers were the least able to differentiate correctly between adul t and child bite marks. The effect of training was emphasized.73
Figure 22.2 Multiple bite marks are often seen. This child had 16 separa te injuries.
location between the males and females. Eighty per cent of th e children had more than one bi te, compared with only 20 per cent in a comparable adult group. Of these, the boys were twice as likely to have multiple bites as the girl s and one of the boys had 16 bites (Fig. 22.2). Bites on the hands and fingers may be ca used when th e child attempts to pro tect him- or herself. Some foren sic odon tologists believ e that bi te marks found on infants tend to be in locations di f feren t from those found in older children or adolescents and are meant as punishmen ts to the child in response to crying or soiling. These bites tend to be co ncentrated on the cheeks, arms, shoulders, buttocks or genital ia. Bites in tended as punishment for soilin g are most commonl y seen in the last two locations. There is considerable disagreement as to the location of bi te marks in NAl cases. Some observers take the view that bi tes on the breast are much more common in females, whilst bi tes on the arms are more common in males. G3 In the Cardiff survey, th is was true only in individuals over 16 years of age. The overall frequency of bite marks in children is not clear, but a three-month study of a sel ected juvenile popul ation,51 among sheltered children, demonstrated an inciden ce of 1545 bite marks per 100 000 population. This incidence is si milar to that of diseases such as go norrhoea. It seems unlikely that such an incidence wou ld be found in the popu lati on at large. Bite mark injuries constitute one per cent of all emergency department visits in the USA and are often associated with child ab use. 58 A recent report from th e USA69 looked at 101 bite mark cases that had been referred to Courts of Appeal. Seventeen per cent of the cases we re in children and all the male children had suffered bites to the genitalia, whereas in the femal e children bites were found at all locatio ns, including face, legs, arms and buttocks. Child bite marks from two centres in the UK were comp ared and it was found th at 1- to 5-year- olds were at greatest risk.70 The location of bite marks was simi lar in the children from both centres, the most common site of injury being the arm. In the UK,71 an incidence of child bites of 2.25 per 100 000 children per year was calculated. Many of these children had bite
-
-
-
Types of Bite Marks in Children Determinin g the aetiology of a human bite mark may be difficult. There is general agreement that there are two main patterns of bite marks. The first is an aggressive or anger bite in which the teeth, in the main, are used alone a nd the marks on the ski n most cl osely resemble the shape of the teeth mak ing them. The mark results fro m an attack or defence bite and is the ty pe most frequently seen in NAI to the child 72 This kind of bi te is a particularly aggressive one . No suction of lips and tongue is involved, th e tissues being bi tten directly between the teeth. On occasions, sub stantial abras ion marks may be associated with each tooth mark, suggestin g movement durin g the biting episode.5 The secon d type of bite mark is on e that is slowly, sadis ti cally an d deliberately inflicted . Suction or sometimes pres sure is applied to the soft tissues via the lips and to ngue. These types of mark are more often seen in abuse in older children with a sexual overlay. The injUly demonstrates bruising in the central part of the bite and sometimes in the peripheral areas, together wi th linear radiating abrasi ons caused by the incisal edges of the front teeth. The outlines of the teeth are usually quite clearly visibl e. The pressure exerted by teeth during biting may be considerable and may be as much as 11 kg (550 kPa) from the incisors an d pressure from the to ngue may rea ch at leas t Sib per square inch (55 kPa). Durin g suckling activity,14 suction as distinct from thrust from the tongue and lips, may reach a negative pres sure of 20 mmHg (2.75 kPa) in so me bi tes .75 The circum stances of the injury will vary a nd , although the present remit refers to NAl to th e child, it should be re membered that children may presen t wi th bite marks fo llowin g fights (61 per cent in one stud y) or pl ay (26 per cent).76,77 Most children with bite mark injuries are brought to the cas ualty or paediatric depal1ment by the parents or adul t guardians. The person who ab used the child and produced the bi tes is likely to be among th em.78 This, however, is not always the case, and the examiner should be aware of the fact that bites
-
---
- - - -
Bite marks I
427
may be produced on small children by siblings or play mates, either aggressively or as a means of punishment, or occasionally in sexual play. The methods of distinguishing bites produced by children from those produced by adults will be described later. The injury itself may consist of bmising, abrasions, lac erations, punctures or any mixture of these in a single bite mark. In a study carried out on 320 human bites in children in the USA,76 75 per cent of all bites were superficial abra sions, 13 per cent punctures and 11 per cent lacerations. This may be important since none of the abrasions became infected but 38 per cent of the punctures and 37 per cent of the lacerations did. In the Cardiff survey,62 82 per cent of bite marks in children were categorized as bruises, nine per cent as lacerations, eight per cent as abrasions or scratches and one per cent as amputation of tissue.
Self-inflicted Bites Self-inflicted bites may be emotional responses to extreme pain or a type of counterirritation to alleviate pain?9 It has been suggested that the arms may be pushed into a child's mouth by an assailant in order to stifle crying, so but self biting may also be a type of self-destructive behaviour, such as has been described in individuals who are mentally retarded or psychologically disturbed. Self-biting may occur in adults, and a case has been described in which a 51-year-old man inflicted a bite mark on the left wrist dur ing a fatal episode of myocardial ischaemia 79 It is obvious that self-inflicted bite marks can be made only on parts of the body that may be placed in the mouth, but one suicide victim is reported as having managed to self-inflict a bite upon the left breast. In alleged NAI cases it is clearly of great importance to determine whether self-infliction has occurred Bo or whether the bite might have been relatively innocently acquired during childhood play with a sibling or friend.
Animal Bites on Children NAI bites on the child may need to be distinguished from bites produced by domestic or wild animals, usually on unsupervised infants or older children. Most studies give the average age of children attacked by a dog as between 5.4 and 6.2 years. BJ.82 Animals may attack the face of a small child, producing very serious injury. It is usually reratively easy for a dentist to distinguish between a human and an animal bite because of the size of the dental arch, and more particularly the arrangement and size of the teeth - a carni vore such as a dog or a ferret,B3 for example, has large canines and diminutive incisors in both the upper and lower jaws. The classical bites from these animals present either as a four-point puncture wound from the canines or as tearing of the tissue in more aggressive attacks (Fig. 22.3). Small
Figure 22.3
Dog bites may result in four-point puncture
wounds or tearing of the tissues, as in this case.
animals, such as rats, have extremely sharp, razor-like inci sors. These can inflict deep and extensive lacerations which may be mistaken for injuries using sharp implements (Fig. 22.4).11 It may be important from a medicolegal point of view to determine which animal has made the bites since the parents or owners may be held responsible if it is a domestic animal. However, blame may be apportioned dif ferently if, for example, the injuries have been produced by rats that have not been exterminated by the local authority, as in the case shown in Fig. 22.4. Distinguishing between different animal bites requires a knowledge of comparative dental anatomy, but also a knowledge of the habits of com mon animals. Watch dogs tend to bite and hold their vic tims, whilst untrained wild dogs may move their heads as they bite, tearing the tissue and making the bite mark diffi cult to evaluate. B4 A case has been described in which the body of a 13-month-old baby was discovered with more than 80 puncture wounds, initially thought to have been caused by a mechanical instrument. It was eventually shown that the child had been injured and finally killed by two German Shepherds. 8s Although fatal animal bites are rare, two cases were described in Germany,BG and in the USA there were 238 deaths from dog attacks over a 20-year period. Most of the dogs were pit bull terriers and Rotweilers. 87 The author dealt with a case in south Wales where a pet German Shepherd savaged a three-month-old sleeping baby and dismembered the body. All this occurred in the presence of the sleeping father (Fig. 22.5).
Dentist's role in chi ld abuse and neglect
428 I
animals the child has been younger than 2 years old and usually less than 1 year of age.
Infections in Human and Animal Bites Children who survive NAls and who have suffered from either animal or human bites may develop serious infec tions. 9o.9 ! Aspirates from bite wounds in 39 children (21 with animal bites and 18 with human bites) were cultured for anaerobic and aerobic bacteria. 92 Aerobic bacteria alone were recovered in 7, or 18 per cent, of the wounds, anaerobic bac teria alone in 3, or 8 per cent, of the wounds and mixed bac teria in 29 cases, or 74 per cent. The most frequent isolates were Staphylococcus aureus, anaerobic cocci and Bacteroides spp., Pasturella multocida, Pseudomona s fluorescens were present in animal bites only. Human bites to the hand war rant special consideration 9J because of the special risk of severe infection. Recent data demonstrate that human bites occurring anywhere other than the hand present no more of a risk of infection than any other type of mammalian bite. Cultures of infected bite injuries in chi ldren may yield an average of five or more micro-organisms. 77 A case of primalY and recurrent herpes simplex infection was described in a paediatric nurse resulting from a human bite. 94 In 1988 the issue of transmission of HIV through human bites and scratches was raised . A total of 198 health care wo rkers were studied , 30 of whom were traumatized in this way while caring for an aggressive AIDS patient. 95 This single violent patient frequently bit his carers, his mouth contained both blood and saliva and he was HIV antibody and antigen positive. After 2.5 years of serial follow-up, all traumatized personnel were clinically normal. It appears, therefore, that the risk of transmission of HIV through this route should be low. Two cases were reported in 1996 ; in one case an HIV-positive female prostitute transmitted HIV to another individual by biting. In another report, two adult sisters had a violent fight in which several of the Hrv infected sister's teeth were knocked out. She then bit her HIV-negative sister, who contracted the virus. It appears that to transmit HIV the biter's mouth must conta in blood. 96 The current position was recently reviewed, and it was concluded that a bite from an mY-seropositive individual which breaks the skin or is associated with a previous injury, carries a risk of infection for the bitten individual. 97
Figure 22.4 Multiple bites from small anima ls may result in serious facial injury extending to loss of an eye. These injuries were caused by a rat.
Figure 22.5 A sleeping baby was savaged by a dog and then dismembered by the same animal.
Dogs are probably the most common animal to bite chi l dren,88 and most of the injuries involve the extremities. As a result of their small size, 33 per cent of bites in children under 5 years of age are to the head and neck, 88 and in some studies up to 67 per cent have been to the head and neck. 82 There have been instances of damage to the central nervous system caused by an English bull terrier, a Dober man pinscher and a Bengal t iger89 In nearly all the reported cases of injury to the head from dogs or similar
-
-
-
--
--
--
-
-
The Investigation of the Bite Marks Where bite marks are present in a case of NAl, they are usu ally to be found on the child victim. However, it should be rememb ered t hat occasionally they may be on an assai lant, if the victim has attempted to fight back or is attacking in self-defence. The same principles of investigation apply in both cases and, apart from the difficulties of examining a conscious sma ll child, the methods llsed to investigate bite
----
Bite marks I
marks on an adult can be applied routinely in most ch ild cases. Injuries caused by biting can be thought of as 'tool marks', the tools in this case being the biting edges of the teeth. In many ways they are a unique injUly in the sense that not only do they demonstrate injury which is difficult to explain away as an accident, but they are also one of the few injuries that can be related directly to the assailant. Bite marks in human flesh can be extremely difficult to interpret because of the dynamic nature of the injury. Distortion may be caused by the bite itself, by subsequent changes and by many other factors. It is important, therefore, that an expe rienced forensic dentist should be included in the team investigating cases of childhood abuse. This is still not the case in many child protection procedures produced by health authorities in the UK. As a general rule, bite mark injuries should always be carefully recorded and described before the teeth of a putative suspect are examined. This is because even experienced forensic odontologists may introduce an unconscious bias in to their interpretation if they already have a particular dentition in mind. The investigator should be attem pting to fit 'the injuries seen in the bite' to one of a number of suspects and should not be attempting to fit 'o ne particular dentition' to a specific injury.
The Bite Marks There are no detailed protocols for examinations of bite mark injuries that are acceptable throughout the world, and individual dentists have developed their own methods. Guidelines have been laid down by the American Board of Forensic Odontology.98 Once recognized as such, a bite mark on a child should always be investigated by a forensic odontologist if abuse is suspected. Children receive bite marks from adults, other children or animals, and they are a common problem in children presenting for medical attention. 99 When child abuse is suspected, the recognition of bite marks may pose significant problems for a medical clinician. It may be difficult to determine whether bruising or lacerations are caused by a bite because the marks are often incomplete, distorted and change over time. They may, on occasions, be confused with skin eruptions .6 ) Collaboration between pae diatricians and forensic dentists is a satisfactory solution to the assessment of bite marks where abuse is suspected. 71 After recording the usual details of the victim, the location of the bite mark on the body, be it living or dead, is carefully desClibed . At this stage it is important to note whether the bite mark is on a flat or a curved surface and whether it over lies soft tissue, fat or bone. Bite mark identification entails several steps: recognition of the wound, documentation, and interpretation . A human bite mark is usually identified because of the double arcade shape of the marks. If both upper and lower arches have left marks, the characteristics of the resulting bite include an elliptical or ovoid pattern containing tooth and arch marks,57 which are almost always
429
produced by the canine to canine of the upper arch opposed to canine to canine of the lower arch. Incisor teeth tend to leave rectangular bruises or incisions whe reas the canines leave puncture or triangular marks. A decision as to whether an injury is, in fact, a human bite mark depends upon its shape, colour and size and on the impressions made by indi vidual teeth. Bruising may be present not only rel ated to the biting edges of the teeth but also in the central area of the bite beca use of tongue thrust or suction by the assailant. This usu ally presents as petechial haemorrhage in the central area of the bite. If the marks of individual teeth in their correct align ment can be satisfactorily identified then it is possible to say at this stage, that the victim has suffered a human bite. Before the bite is touched, measured or photographed it is important to obtain evidence of saliva traces left on the skin surface by the assailant. These swabs may be used for subse quent blood grouping, usually limited to the ABO system only. A, Band/or H substances are secreted by some 75 per cent of the population. A negative result may therefore be due to the absence of saliva to a non-secretor status of the biter or to poor sampling. Nowadays swabs are taken for DNA analysis. Control swabs must be taken from swabs in other parts of the body where no bite marks are suspected 100 and it is often convenient to use the contralateral area of the body. A saline control swab should also be supplied. These specimens are needed to exclude contamination by secre tio ns from either the victim or the investigator, who should, of course, wear gloves throughout the procedure. The use of stelile cotton swabs moistened with steriJe water followed by dly swabs, the so-called double swab technique, is now rec ommended. 101 The importance of a standard and satisfactory tech niqu e cannot be over emphasized.
The Detail of the Bite Once an injury has been recognized as being caused by a human bite and saliva samples have been taken, the details of the mark should be recorded. A human arch mark may be identified when four or five individual marks of adja cent teeth are present. If the tips of the canines in both upper and lower arches ca n be recognized, then the inter canine width should be measu red. In a human bite, the intercanine width will lie between 2. 5 and 4.5 cm.70 If the intercanin e width is less than 3 cm, the bite will probably have been inflicted by a child. In this context, a child refers to any biter below the age of 12 whose anterior teeth are either deciduous teeth or mixed dentition. The two arcades of upper and lower teeth marks should be described and measured in terms of their curvature and overall size; in addition, marks left by individual teeth should be individ ually measured and described in an attempt to reconstruct the type of denti tion that could have made the mark. Miss ing teeth or areas where no bruises have occurred must be noted. The alignment of each bruise and the possibility of them being made by rotated teeth should also be examined.
430 I
Dentist's role in chi ld abuse and neglect
In general, the mark caused by the upper maxillary arch is often more diffuse and wider than that produced by the mandibular teeth. This results from the greate r surface area of the incisal edges of the upper teeth, but also because the maxilla is stable in relation to the skull. By contrast, the mandible is mobil e and acts as a cutting instnlment once the skin has been stabi lized against the upper teeth. Before the bite mark is recorded the forensic dentist should attempt to establish the circumstances surrou ndin g the bite. A slowly produced bite usua lly results in stretching of the skin over the cutting edges of the teeth and pressure or suction from the tongue. This will be reflected in the extent of bruising, both in the too th marks and in the surroundin g and central areas of the bite. It is said that a bite mark show ing these characteristics was probably produced under amorous circumstances. However, even so-called ' love bites' require considerable force to produce and can be painful to the recipient. A bite mark that consists only of tooth marks is of an aggressive nature and has probably been delivered rapidly. This is especially so if teeth have penetrated the sk in. The observer sho uld be wary of attributing a particular force or intent, such as aggression, in any particular bite mark case. It may sometimes be possibl e to offer a general opinion as to the timing of the bite. Too th marks that do not break the skin usually last from several minutes to 24 hours. 102 In cases where the skin is broken, the borders or edges wi ll last several days depending on the thickness of the tissue. Thin ner areas retain the marks 10ngeL IOJ It is clearly important that the forensic dentist be called in the early stages of the investigation of a bite mark case. Many investigators have recommended that bite marks be photographed at 24-hour intervals because their detail may improve in the first 2-3 days as swelling decreases and brui sing increases. 57 In add i tion, the outline and shape of a bite may change because of infection, oedema and discoloratio n of the skin. Accurate determination of the time of a laceration or brui se is diffi cult and complicated by individual valiation. The history of a bruise can be correlated to sequential colour changes, but is frau ght with difficulty, and only a gen eral comment should be made on these matters . The current position has been reviewed recently. 104.105 Quite different appearam:es will result if the bite was inflicted immediately prior to or immediately after death . If the bite mark was inflicted prior to death, providing that sufficient time has elapsed to allow extravasation of blood into the tissues, a clear bite mark may be seen that changes velY lit tle until the onset of putrefaction. Bites inflicted after death rarely result in bruising. Very considerable forces indeed are required to produce an appearance of bruising post mortem.
majority of cases, good-qu ality co lour and bl ack and white photographs will be used. Some fo rensic dentists argue that the photography should be carried out by a professional police photographer under the direction of the dentist, so that questions in court regarding scale and reproduction, colour balance and other photographic questions may be answered with authority. However, exp erienced forensic dentists prefer to take their own photographs in a standard ized manner as better results are often obtained. A 35-mm camera with a focal len gth lens of abo ut 100 mm and a ring flash and/or side flash is the most commonly used equip ment and the techniques involved have been described in detail. lOG Whatever equipmen t is used, photograp hs should include low-power distance shots of the bite mark and sur rounding tissues, which allow for orientation. A standard colour chart is often in clud ed along with the subject when colour film is used. In the case of black and white, ultra violet filters may be used, which may bring out details that are not obvious when viewed by conventional lighting. 107 Ultraviolet light requires a grea t deal of careful and expe rien ced interpretation as injuries made many years before the bite may be imaged using this technique a nd may con flict with the interpretation of a recent injUly. Whatever method of photography is used, photographs of the injury must be taken with a standard, rigid, L-shaped centimetre scale, arranged as close as possible to the injury, but not overlapping it. Because injuries are often on curved surfaces, it may be necessary to reposition the sca le a num ber of times in relation to different parts of the bite mark. In principle, the scale should be arranged in the same plan e as the portion of the injury being photographed and the plane of film of the camera should also be parallel to both scale and the portion of the injury being photographed. Inevitably, on a sharply curved surface, a number of photographs will be required around the curvature of the injury (Fig. 22.6). These procedures are essential to avoid introducing photo graphic distortion into a situation that is already distorted by
Recording the Bite Mark
Figure 22.6 Bite marks are often on cu rved su rfaces, such as the breast, and each portion of the injury should be photographed individually. The mark on the right is from the upper arch with only three incisors present.
It is important to record the injury for posterity, especially when legal proceedin gs may subsequen tly occur. In the
---
Bite marks I
curved sUifaces, oedema , positional distOliion and possibly putrefaction and post-mOliem changes. It is advisable to re-photograph marks on living victims every 24 hours for a number of days. One-to-on e en large ments should be made of both colour and black and white photographs for measurement and comparison purposes. In addition, larger prints at twice or four times the normal size may be prepared for demonstrations in court. lOB On co mple tion of photography, the marks should be examin ed care fully thro ugh the hand lens to determine whether there are foreign bodies in the depth of th e ma rks. These may include fragments of tooth, fractured from the assailant at the time of the attack, and they should be removed carefu lly and stored as evidence. 109 It is well recognized that de terioratio n may occur in a bite mark. This may be caused by dynamic activity during the bite, curvature and distortion of tis sues,JlO chan ge in the tissues after the bite or distortion introdu ced during photography. Care should be taken to minimi ze and record any distortion; there are methods avail a ble to correct the image to some extent. III Some odontologists recommend digital enhancement of bite mark images, 11 2 but this shou ld be done to a minimal ex tent, if at all; the method has been criticized in so me courts.
Impressions of the Bite Most bite marks in human tissue do not leave a three dimensional imprint, but are visible only by way of the bruising in the und erlying tissues. Impressio ns are valueless in such cases. If the bite is deeper, and particularly when the skin has been penetrated, an impression may be useful and shou ld be taken in a reco mmended si licon rubber-based dental impression material. It is advisable to back the impression with a rigid material, such as plaster of Paris, to minimi ze distortion on removal. Subsequent models should be cast in high-quality dental stone. They may be useful for measurement purposes and study models prepared from pos sible suspects may be compared with the cast of the injUiy.
Excision of Skin Samples In fatal NAI, further information concern in g a bite injUly may be acquired by excising th e wound and studying it histologic ally. Sections can be stain ed for iron and other blood break down produ cts an d examined for changes in the elastic and collagen fibres; they may also show impacted plaque from the biter's teeth (Fig. 22 .7). Classical features of a bite mark injUly are abrasion or penetration of the epithelium, compression and dis tOliion of underlying collagen fibres, oedematous spaces and extravasation of elythrocytes from the blood vessels. 11J Attempts have been made to detenlline the age of injuries by microscopic or biochemical examination. lo5 Enzyme histochemistry using serotonin and histamine, along with other histochemical methods, have been used. 114
431
Figure 22.7 Histological sections through a bite indi cate loss of epithelium, damage to connective tissue and, in this case, impaction of denta l plaque into the wound (arrow).
It has been suggested that bite marks in human skin in a victim who does not survive should be excised and preserved as evidence of the injury. There is a potential problem of shrinkage and distortion of such sa mples, and a recent study1 15 has shown that both contraction and expan sion of bite mark specimens ca n occur using the standard methods of fix ation and support. It seems, therefore, that standard techniques for stora ge and preservation of bite mark samples will not produce reliable dimensional accuracy.
The Suspect Once a suspect or suspects have been apprehended, it will be necessalY to examine their dentitions, to take impres sions a nd make dental casts . In England, the examination of suspects is governed by the Police and Criminal Evi dence Act. Written consent of the procedure should be acquired from the susp ect and witnessed, preferably in the presence of the suspect's solicitor. In Scotland, an impres sion made by virtue of a s heriff warrant would be admiss i ble evidence. It is advisab le to take a full dental history, especially if the suspect may have received dental trea t ment prior to and following the biting incident. Colour and black and white photograph s should be taken or ordered of the extra- and intraoral appearance of the suspec t. It may sometimes be useful to photograph the suspect with a scale in place show ing the maximum opening between the incisor teeth. Suitable blood and saliva samples should be taken from each suspect when a saliva swab has been taken from the victim. Comprehensive intraoral examinations shou ld be undertaken, charting the teeth present and any dam age, misalignment or fractures. Teeth tender to biting forces and any abnormal chewing activity should also be noted. Impressions in dental alginate should be taken of both upper and lower arches, and the occlusal relationship between maxilla and mandible must be record ed using a suitable technique. These impressions llluSt be attested, labelled and cast in high-quality dental ston e as quickly as
432 I
Dentist's role in child ab use and neglect
possible. The technician making the castings must record the details of the procedure. The base of the study models must be marked with the individual's name and also signed and dated by the technician. Some dentists invite a suspect to produce a samp le bite into a wax recording medium at the time of the examination.
Comparison of Evidence from the Bite and the Suspect The final comparison between the details of the bite mark injury and the dentition of a possible suspect is not a simple matter. Some a uthorities claim that transparent overlays of the biting edges of a suspect's teeth can be produced and directly compared with the bite mark. An accurate fit then warrants an opinion of guilt by a putative biter. However, transparencies produced by whatever meth od rarely, if ever, exactly fit the size and shape of a human bite mark. This is because of the distortion factors, which have been discussed earlier in this chapter. There is a danger in ass uming that bite marks are equivalent to fingerprints and that a reasona ble number of matching points wi ll prove the id entity of the biter. A more satisfactOlY ap proach is to describe the bite mark in detai l before looking at the details of a suspect's dentition a nd then to describe the suspect's dentition and to note the points of correspondence or deviation. The bite mark inju ry and the dentition of the suspects are then exam ined together, and an iterative process should enable the observer to determine whether or not the details of the injury can be explained satisfactorily by relating th em to the den tition of the suspect. If they can be so related, the forensic dentist shou ld base the strength of his or her final opinion on the si milarity of pattern between the injury and the dentition a nd should tl1en finally illustrate a nd confirn1 his or her opinion using transparent overl ays produced at the sa me scale. The non-scientific nature of this methodology has been emphasized, but it can work sat isfactorily in most cases. 98 Unusual features of bo th the mark and a suspect's dentition that would tend either to elimin ate or incriminate a suspect in the case should be noted. Thus, a single feature, such as a missing, rotated or fractured tooth leaving a com parable mark in the .bite, may be a much more useful pointer than any number of matching features in an arch of normal size and arrange ment (Fig. 22.6). Comparisons of dentitions and bite marks have used superimposition of colour slides, denta l articulators and the recording of bites in suitable media. Sophisticated scientific analysis such as micro co ntour mapping, scanning electron microscopy a nd reflex microscopy have been used. 11 6 In practice, the simpler and more obvious the method, the more likely the jury memb ers are to understand and appreciate the significance. The current vogue is to use computer analysis of the bite mark photograph and to produce transparent overlays digitally from sca nned images of the dentition.III, ll l These methods are sa id to be able to correct for common photographic
distOIiion and size discrepancies, eliminate examiner subjec tivity, provide better control of image visualization and pro v ide standardization of procedures. However, it should be remembered that these imaging methods are only as good as the original evidence and should not be used to enhance poor-qu ali ty evidence. 11 2 The final opinion is a very serious issue for possible suspects as well as for the victim and fam ily; in few areas of medical evidence is such a respons ibility placed upon the objectivity and honesty of the expert. No jury should be led to believe that bite mark analysis is sim ple and accurate and the expert must clearly indicate the pros and cons of each case.
REFERENCES
2
3
4 5 6 7
8
9
10 11
12
13
14
15
16 17
18
Knigh t B. The hi story of chil d abu se. Forensic Sci I11t 1986; 30: 13 5- 41. Kempe CH, Silverman FN, Steele BF, Droegemuell er W,
Silver HK. The battered child syndrome. lAMA 1962;
181:17- 24.
Welbu ry RR. Ch ild physical abuse. In Andreason JO, Andreason FM reds) Textbook alld C%ur Atlas of Traumatic Injuries to the Teeth, 3rd edn. Copenha gen: Munksgaard, 1994.
Mea do w RC. The ABC of Child Abuse. London: BMJ
Publis hing Group, 1997.
Cameron JM , J oh nso n HR , Camps FE. The battered ch ild
sy ndrome. Med Sci Law 1966 ; 6: 2-2 1.
Wagner GN. Bitemark identification in chil d ab us e cases.
Pediatr Dent 1906; 8 :96 -100.
Needl eman HL Orofacial trauma in child abuse: types,
prevalence, manageme nt, and the dental profession 's
in volvement. Pediatr Dent 1986; 8:71-80.
Wei bury RR, Murphy JM. The dental practitioner's role in protecting chi ldren from abuse. 3. Reporting a nd subsequent management of abuse. Br Dent 1 1998; 184:115 - 19. Kirschner RH, Stein RJ. The mistaken diagnosis of chi ld
abuse. A form of medical abuse 7 Am 1 Dis Child 1985;
139 :873-5.
John so n CF. Inflicted inju ry ve rsus accid e nta l injury. Pediatr Clill N Am 1990 ; 37:791-814. Whittaker OK. The principles of forens ic denti stry. 2. Non accidental injury, bite marks and arc haeo logy. Dent Upda te 1990; 17:386-90. Fonseca MA, Feigal RJ. Ten Bensel RW. Dental aspects of 1248 cases of child maltreatment on fi le at a major country hospital. Paediatr Dent 199 2; 14: 152-7. Rasm ussen P. Tannlegens forhoJd tilsykdo m hos barn, barnhemish andling og omsorgssvikt. In Odontoiogi. Copenhagen : Munksgaard Publishers, 1992, pp. 81-92. Speig ht N. Bruises and burns: accidental or non -accid enta l? In Accidents and Emergencies in Childhood. London: Royal CoJl ege of Phys ic ians of London , 1992, pp. 97 -103 . Welbu ry RR, Murphy JM. The denta l prac titi oner's ro le in protecting chil dre n from abuse. 2. The orofacia l sig ns of abuse. Br Dent 1 1998; 184:61 - 5. Simpson K. A Doctor's Guide to Court. London : Butterworth, 1967. Creighton SJ, Gallagh er B. Child Abuse Deaths: Informatio n Briefing, No.5 (revised). London: The National Society for the Prevention of Cru e lty to Children, 1988. Schm itt BD. Physical abuse. Specifics of clinical diagnosis. Paediatr Dent 1986; 8:83.
References I
19
20
Creighton SJ. The incidence of child abuse and neglect (Chapter 3). In Browne K, Davies C, Stratton P (eds) Early Prediction and Prevention of Child Abuse. Chichester: John Wiley Et Sons, 1988. Golder M. Non -accidental injury in children. Dellt Update
45 46 47
Fearn 1. Dental aspects of child abuse (Letter). Aust Dellt J 1987 ;
32:227. 22 Cappelleri JC, Eckenrode J, Powers .fL. The epidemiology of
child abuse: findings from the second National Incidence and Prevalence Study of Child Abuse and Neglect. Am J Publ Hlth 199 3 ; 83: 1622. 23
Hobson P. Child abuse: we can help (Letter). Br Dellt J 1987;
24
Child protection in ethics ill dentistry. BDA Advisory Semices, 1995, pp. 45-6. Department of Health. Working Together under the Children Act (1989), 48-49, London, HMSO, 1991. Depart ment of Health. Children alld Youllg People on Child Protection Registers, Year Ending Jist March 1995, England. London: Gov ernment Statistical Service, 1995, p. 19. Macintyre DR, Jones GM, Pinkney RCN. The rol e of the dental practitioner in the management of non-accidental injury to children. Br Dellt J 1986; 161:108-10. Kittle PE, Richardson DS. Parker JW. Examining for child abuse and child neglect. Pediatr Dellt 1986 ; 8:80-2. Casamassimo PS. Child sexual abuse and the paediatric dentist. Paediatr Dent 1986; 8: 102-6. Anasavice K1. Treatment regimes in preventive and resto rative dentistry. J Am Dent Assoc 1995 ; 126 :727-43. Ramos-Gomez F. Rothman D, Blain S. Knowledge and attitudes among California dental ca re providers regarding chi ld abuse and negl ect. J Am Dent Assoc 1998; 129:340-8. Beck JD , Weintraub JA, Disney JA. University of North Carolina. Caries Risk Assessment Study. Community Dellt Oral Epidemiol 1992; 20:313-21. Jessee SA. Physical manifestations of child abuse to the head. face and mouth: a hospital survey. ASDC J Dent Child 1995;
162:53.
25 26
27
28 29 30 31
32
33
62:245-9. 34 Rupp RP. Conditions to be consid ered in the differential
diagnosis of child abuse and neglect. Gen Dellt 1998; 46:96-100. 35 Jessee SA. Risk factors as determinants of dental neglect in children. ASDC J Den. Child 1998; 65:17-20. 36 Kellogg N and the Committee on Child Abuse and Neglect.
Oral and dental aspects of child abuse and neglect. Pediatrics 200 5; 116: 1565-8. 37 Sanders B, Brady FA, Johnson R. Injuries. In Saunders B (ed.)
Pediatric Oml and Ma.xilloJacia1 Surgery. St Louis: Mosby, 1979, pp. 330-400. 38 Fotana VJ, Dqnovan MD, Wong RJ. The maltreatment syndrome in children. N Eng I J iVIed 1963; 269:1389-94. 39 Skinner AE, Castle RL. A Retrospecti1Je Study. London: NSPCC, 1969. 40 Willging JP, Bower CM, Cotton RT. Physi cal abuse of children. A retrospective review and an otolaryngology perspective. Arch OtolarYl1gol - Head Neck Surg 1992; 118 :584-90.
Mehta MN. Physical abuse of abandoned ch ildren in India. Child Abuse Neg11982; 6:171-5. 42 Becker DB. Needleman HL , Kotelchuck M. Child abuse and dentistry: orofacial trauma and its recognition by dentists. J Am Dent Assoc 1978; 97 :24 -8. 43 Feldman KW. Patterned abusiv e bruises of the buttocks and pinn a. Pediatrics 1992; 90:633-6. 44 Wilson EF. Estimation of the age of cutaneous contusion in child abuse. Pediatrics 1977; 60:750-2. 41
Lenosky EF, Hunter KA. Specific patterns of inflicted burn injuries. J Tra uma 1977; 17:842-6. Kumar P. Child abuse by thermal injury - a retrospective survey. Bums 1984; 10:344-8. Prescott PRo Hair dlyers and non-accidental injury. Pediatrics 1990; 86:692-7.
1995; 22:75-80. 21
433
Hobbs CJ. Skull fracture and the diagnosi s of abuse. Arch Dis Child 1984; 59:246-52. 49 Worlock p. Stower M, Barbor P. Patterns of fractures in accidental and non -accidental injUly in cbildren: a comparative study. Br Med J 1986 ; 293: 100-2. 50 Kaban LB, Mulliken JB, Murray JE. Facial fractures in children. Plost Recollstr Surg 1977; 59: 15-20. 51 Kvaal 51. Oral injuries in cases of child abuse. Rev Beige [,vIed Dentaire 1993; 48:49-53. 52 Carrotte pv. An unusual case of child abuse. Br Dent J 1990; 48
168 :444-5.
Ablin DS. Reinhart MA. Esoph agea l perforation by a tooth in child abuse. Pedia tr Radiol 1992; 22:339-41. 54 Sperber N. The dual responsibility of dentistry in child abuse. Mt J Orthod 1981; 19:21-8. 55 Schuman NJ. Child abuse and the dental practitioner: discu ssio n and case rep0l1s. Quintessence 111t 1987; 53
18:619-22.
Troutman KC. Non-consensu al touching: your input needed. Am Acad Pediatr Dent 1987; 22:3. 57 MacDonald DG. Bite mark recognition an d interpretation. J Forens Sci Soc 1974; 14:299-233. 58 Hod son JJ. Forensic odontology in cases of battered baby syndrome. I Den. Res 1971; 50:6 56. 59 Sims BG, Grant JH, Cameron JM. Bite marks in the ' battered baby' sy ndrom e. Med Sci Law 1973; 13:207-10. 60 Marr JS, Beck AM. Lugo JA. An epidemiological study of the human bite. Public Health Rep 1979; 94:514-21. 61 Rawson RD, Koot A, Martin C et al. In cidence of bite marks in a selected juvenil e population: a preliminary report. J Foren s Sci 1984; 29:254-9. 62 Liddiard P. An analysis of bite mark cases in foren sic dental practice, with reference to dental anomalies affecting case outco mes. MS c, University of Wales, 2002. 63 Val e GL, Nogu ch i IT. Anatomical ciistribution of human bites in a series of 67 cases. J Forens Sci 1983; 28:61-9. 64 Harvey W. Denf'al Identification and Forensic Odontology. London: Henry Kinpton, 1976. p. 91. 65 Lowry TMcC. The surgical treatm ent of human bites. Al1nals Surg 1936 ; 104: 1103-6. 66 Speirs RF. Prevention of human bite infection. Surg Gynec Obstet 1941 ; 72:619. 67 Sims AP. Non-accidental injury in the child presenting as a suspected fracture of the zygo matic arch. Br Dent J 1985; 56
158 :292-3. 68 Liston PN, Tong DC, Firth NA, Ki ese r JA. Bite injuries:
pathophysiology, forensic analysi s and management. N Zealand Dent J 2001; 97:58-63. 69 Pretty IA. Sweet D. Anatomical lo catio n of bite marks and associated findings in 101 cases from the United States. J Forens Sci 2002; 45:812-14. 70 Simper RD, Whittaker OK. An analysis of bite marks presenting in two centres in the United Kingdo m. J Dent Res 1994; 7J :791.
Whittaker OK. Aitken M, Burfitt E, Sibert JR. Assessing bite marks in children: wo rking with a forensic dentist. Ambulatory Child Hlth 1997; 3:225. 72 Levine U. Bite mark ev idence. Dent Ciin North Am 1977; 7J
21: 145- 58. 7J Whittaker DK. BrickJey MR. Evans L. A comparison of the
ability of experts and non-experts to different iate between
434 I
74 75
76 77 78 79
80 81
82
83 84 85 86
87
88
89
90 91 92 93 94
Dentist's role in child abuse and neglect
adult and child human bite marks usin g receiver operating characteristic (ROC) ana lysis. Forflls Sc i Tilt 1998; 92 : 11-20. Ky dd WL. Quantitati ve analysis of fo rces o n the tongue. J Dellt Res 1956 ; 35:17 1- 4. Harvey W, Millington PF, Evans , JH et a l. Bite ma rk s - the clinical picture; physica l fea tures of skin a nd tong ue. Sta nd a rd and scan nin g d ec o'on m icrosco py. mt MicroJorlll J Leg Med 1973; 8:3. Baker MD, Moore SE. Human bites in chi ldren. A six -yea r expe rience. 1987; 141: 1285-90. Leung AK, Robson WL. Human bites in chi ldren. Pedilltr fJ-nerg Cllre 1992; 8:255-7. Levine LJ. The solution of a battered ch ild hom icide by dental ev id ence : report of case. J Am Dellt Assoc 1973; 87:1234-6. Walllick AJ, Biedrzycki L, Russanow C. Not all bite marks are associated with abuse, sexual activities or h om icid es : a case study of a self-inflicted bite mark. J Fo rflls Sci 1987; 32:788-92. And e rson W R, Hudson RP. Self- infli c ted bite marks in battered child syndrome. Forell s Sci 1976; 7 :7 1-4. Ca lkins CM, Bensard DD . Patrick DA. Ka rre r FIVI. Life threa tening dog attacks: a devas tating comb in at ion of penetrati ng a nd blunt injuri es. J Pedilltl· Su rg 200 1: 36 :1115-17. Gandhi RR. Liebman IVIA. Stafford BL, Stafford pw. Dog bite injuri es in c hildren: a preliminary survey. AIlI Surg 1999; 65:863-4. App lega te JA, Walhout MF. Childh oo d risks from the ferret. J Emerg Med 1998; 16:425-7. Triratana 1. Bite marks. J Dent Assoc Thllilllilil 1970; 20:259 -7 2. Class RT, Jordan FB, Andrews EE. Multipl e animal bite wounds: a case report. J forells Sci 19 75; 20: 305- 14. Rhe ul J , Bratz ke H, Fedderse n-Perterso n DU et al. Dea th ca used by 'attack' dog bites, Arch Kril1lillol 1998; 202: 140- 51. Sacks JJ , Sinclair L, Gilchrist Jet al. Breed s of dogs in vo lved in fa tal human attacks in the United States betwee n 1979 a nd 1998, J Am Vet Med Assoc 2000; 217:836 - 40. Belardi FG, Pascoe JM, Beegle ED, Pasteurell a l11ultocicla menin g itis in an infant following occipital dog bite, J fam P/'Gct 1982; 778:780-2. Ste inb ok P, Flodma rk 0, Scheifele Ow. Animal bites caus ing ce ntral nervous sys tem injury in children. A repolt of three cases. Per/ iat!· Neurosci 1985; 12:96-100. Callaham ML. Treatment of cOl11mon dog bites: infect ion risk factors. JACEP 1978; 7 :83-7. Marcy SM, Infectio ns d ue to dog and ca t bites. Pediatr IIIJecr Dis 1982; 1:35 1-6. Brook I. Mic robio logy of human and anima l bite wounds in childre n. Pediotr IIIJecr Dis J 1987; 6 :29-32. Griego RD , Rosen T, Orengo IF, Wol f JE. Dog, cat and human bites: a rev iew. JAm Acad Dermawl 1995: 33:1019- 29 . FUOItes L, Me lso n E. Prima lY and recurrent herpes sim plex infection in a paediatric nurse resul ting from a human bite (Comment). TnJecr COlli Hasp Epidemiol 1989; 120: 295.
95 Tsoukas CM, Hadjis T, Shuster J et al. Lack of transm iss io n of HIV throug h h uman bi tes and sc ratches. J ACl{uir 11I1JI1!llie Defic S.\'lId 1988; 1 :505-7 . 96 Anony mous. To tra nsmit HIV, biter mu st have blood in the
mouth. AIDS Policy Law 1996 ; 11:5.
97 Pre tty lAo And e rso n GS, Sweet OJ. Hum a n bites and the risk of human immunodeficiency virus transmiss ion . A m J Forells Med Porliol 1999; 20:232-9. 98 Arheart KL, Pretty IA. Results of th e 4th ABFO Bitema rk Workshop - 1999. Forells Sci 1111 200 I; 124: 104- 11. 99 Sc hweich P. Fleisher G. Human bite marks in children. Pediatr Emerg Cll re 1985; 1:51-3. 100 Pereira ilil. Possibilities and limitations of sa liva tests in forensic odonto logy. Br Dellt J 1971; 130:161. 101 Sweet 0, Lorente M, Lorente JA et al. An improved meth od to recover sa li va from hu man skin: th e double swab tec hnique. ) Forells Sci 1997: 42:320- 2. 102 Dinkel HI. th e use o f bite mark ev iden ce as a n in vesti gative a id. J Forells Sci 1974; 19:535 -47. 103 Sper be r ND . Bite ma rk s, o ral and racial injuries - h arb ingers of seve re c hild abuse 7 Pacdiatriciall 1989 ; 16: 207 - 11. 104 Dail ey JC. Bowers CM. Ageing of bite marks: a lite rature revi ew. J Forell5 Sci 1997; 42:792-5. 105 Van ez is P. Inte rpreting bruises at necropsy. J Oill Pllthol 2000; 54:348-55. 106 Summers R. Lewi n D. Forensic dental photography. In Cla rk DH (ed.) Practica l forellsic Odoiliology. Oxrord: Wright, J 992, p. 188. 107 Krauss TC, Warl e n Sc. The forensic science use of re fl ec ti ve ultravi o let photography. J forells Sci 1985; 30:262-8. 108 W h ittaker OK. MacDona ld DG. A Colour Atlas oj Forell sic DentiSTry. London: Wolfe Publications, 1989, p. 108. 109 Whit taker DK. Fo rensic dentistry in the ident ificatio n of victims a nd assaila nts. J Oill Forells IVIed 1995; 2: 145 -5/. 110 Sheasby DR, Mac Donald DG. A forensic classification of distortion in h uman bite marks. fore lls Sci II/rem 200 1; 122:75-8. 111 Bowers Cilil. J o ha nse n RJ. Dig ital analysis of bite marks and human id e ntifi cat ion. Dfllr Clill NAill 2001; 45:327-4 2. 112 Kara za lu s KP. Palm bach IT, Lee He. Digital enhancement of subquality bite mark photographs. J forells Sci 2001; 46:954 - 58. 113 Millington PF. Hi stological stucii es of skin carrying bite marks. J Forms Sc i Soc 1973: 14:239-40, 114 Raekallio J. Histological estimation of the age of injuries, In Perper JA, Wec ht CH (eds) Microscopic Dillgnosis ill Forensic Parliolog.\'. Springfield, IL: Cha rl es C Thomas, 1980, pp. 3-16. 115 Rothwell BR , Thie n AV Analys is of d isto rtion in preserveci bite mark sk in. J Forel1s Sci 200 1; 46:573. 116 Lightellll AJ, Coetzee WJC, Va ll Ni eke rk PJ, The identifica tion o f bite marks us ing th e re fl ex mi c ro sco pe. J Forms Odolltosrollliltol 1987; 5: 1- 8. 117 Sweet 0, Parhar 1Vl. Wood RE. Computer- based production of bite mark comparison overlays. J Forel1s Sc i 1998; 43: 1050-5.
-_.
- - - -- -
I
CHAPTER 23
I
PAEDIATRIC DENTAL IDENTIFICATION G Howard Moody
Introduction Comparison Facial reconstruction and dental profiling
435 435 440
II'JTRODUCTION
Accurate identification of the deceased by the relevant authorities is a human right enshrined in Article 6 of the United Nations Declaration of Human Rights (1948) - as this, for most people, will be crucial in coming to terms with the loss of a relative or friend, for the resolution of legal or insurance disputes and for attaining 'closure', enabling the mourning process and due ritual to proceed. The identification of individuals from their teeth is not a recent development. In AD49, Agrippina, mother of the future Roman Emperor Nero, ordered the murder of Lollia Paulina, whom she believed wanted to become the wife of Claudius, as this would have prevented Nero from becom ing Emperor. On being presented with Paulina 's head, Agrippina, to confirm its identity, 'op ened the mouth with her own hands and inspected the teeth which had certain peculiarities'.J In mediaeval times, illiterate apprentices sig nified thei r bondage to a master by biting into a piece of wax, leaving a unique mark and becoming 'indentured'. Worldwide, the task of identification is the responsibility of either the judiciary or the police; in Scandinavia, the police are responsible,2 while in the USA id entification is the role of the medical examiner. J In England, identification is the responsibility of the coron er and in Scotland of the procura tor fiscal. 4 The forensic dentist (odontologist) often has to provide these legal authorities with an opinion, assisting the identification using the techniques of comparison or profiling. • Comparison: the procedure involved is the comparison of post-mortem dental chartings, including
Age estimation References
441 444
radiographs, with ante-mortem dental reco rds to corroborate or deny an identity. This is the most common procedure. 5 • Profiling: the process whereby a forensic dentist, in co llaboration with colleagues, who may include forensic pathologists, anthropologists, archaeologists and biochemists, attempts to 'reconstruct' from human remains the age, sex, height and pOSSibly the ethnic origin of the deceased with a view to creating a possible liken ess that may cause so mebo dy to come forward and offer a possible id entification when the deceas ed is completely unknown to the authorities 6
COMPARISON While generally regarded as a common method of identi fication, the number of persons, which includes children, in the UK who are identified by their dental characteristics in anyone year is not known. There is no central data bank holding this information, and there is no evidence to suggest that all forensic pathologists refer evelY dental identification to a dental surgeon, let alo ne a forensic den tist. Corroboration involves the comparison of ante- and post-mortem records; the certainty of the identification is related to the degree of concordance between the two sets of records and whetl1er any discrepancies that occur are explicable. 7 To be va lid , comparative identification requires identifiable dental characteristics, the accurate recording of these characteristics and accurate ante- mortem data .
436 I
Paediatric dental identification
Dental Characteristics Dental characteristics are understood to be those features that contribute to a unique charting (Fig. 23.1). Such fea tures include decayed, missing, or restored teeth and fixed and removable prostheses (dentures). Restorations range from small white fillings through large amalgam and porce lain restorations, crowns, fillings, root canal fillings and implants. Given the complete adult dentition of 32 teeth, each with five surfaces, the wide range of dental materials, design and construction variations available with respect to prostheses and the variations in root patterns discernible on radiographs, the number of permutations is enormous. The statistical probability of concordant points on ante and post-mortem records has been calculated for extrac tions,a amalgam restorations,9 root morphologiO and digital radiographic scanning. ll ,1 2 A single unique dental feature may, on occasion, suffice,I3 but this is rare, and the number of concordant points that are considered sufficient to collec tively ascertain an identification remains controversial. 14 [n children, however, the situation is very different. In much of the industrialized world, dental decay is in sharp decline with a corresponding reduction in the number of
Figure 23.1 Maxilla retrieved from a shallow grave. Whilst not from a child, the teeth clearly demonstrate their value for the purposes of identifiCation. An ante-mortem chart of this dentition will demonstrate the appropriate missing and filled teeth. Whilst the upper left canine may have been extracted before death (there is loss of part of the bucca l plate consistent with a forceps extraction) the centra l incisors may either have been extracted or lost post mortem - possibly shaken out by urban foxes once the periodontal membrane had sufficiently decayed, rendering the teeth loose. The presence of the incisors in an ante-mortem charting would not necessarily invalidate the identification. Relying so/ely on computer matching is not advisable. Note also that remnants of the gingiva remain, suggesting that this is not an 'old' burial. Estimation of the time since burial is subjective, being influenced by soil temperature, bacteria in the soil and pH. These variables are to some extent interdependent.
decayed, missing and filled teeth (dmfl. The effect of this is that the dentition of individual children is becoming clini cally less characteristic (Fig. 23.2). Dental disease, as measured by the DMF, decreased in all social classes in the UK during the period 1978-98. 15 In 1996, Kas te et al 16 reported the assessment of dental caries in the USA as part of the Third National Health and Nutrition Examination Survey (1998-91). This survey yielded esti mates for over 58 million children aged 1-17 years. It was found that, among infants aged 12-23 months, only 0.8 per cent scored positively for early childhood caries while 62.1 per cent of children aged 2-9 years were caries-free in their deciduous dentition. One-quarter of the children and adoles cents aged 5-17 years with at least one permanent tooth accounted for about 80 per cent of the caries identified in permanent teeth. White et al 17 also describes how, in the USA, the number of children free from dental caries increased dramatically between 1963 and 1987. By 1987,
Figure 23.2 A caries- and restoration-free complete deciduous dentition: (a) maxilla, (b) mandible. There are no characteristic clinical features enabling identification. It is unlikely that there will be ante-mortem radiographs since the child is sign-free and thus not clinically justified.
---
Comparison I
Figure 23.3 Fragment of a maxilla. It is extremely improbable that two persons out of a (known) small populat ion of 270 persons wo uld both have a denta l charting correspo nding in every respect to these teeth and their restorations.
almost half of school children between the ages of 5 and 17 years were caries-free and , among those with caries, the number of teeth affected had also declined. Flinck et ai, 18 in Sweden, and Lawrence and Sheiham, 19 in Brazil, report sim ila r findi ngs. Following the Lockerbie air disaster in 1988, 208 of the 273 victims of were identitled solely, or in conjunction with other methods, by forensic dentists (Fig. 23.3). 20 The victims were mainly adult. A very different result may be envisaged if a similar major air disaster were to happen now, but involving numerous children of the same sex, of similar age, and who had been exposed throughout their lives to a non-cariogenic diet, fluoridated water supplies and flu orid e toothpaste. The observed differences between the individual dentitions might be minimal and dental records indicating only satisfactory 'check-ups' would be unhelpful. In this sit uation, identification would be achieved by comparison of DNA with DNA from tlrst-degree relatives. Dental pulp is an excellent source of DNA, being protected from heat and trauma by dentine, bone and the facia l muscles as well as offering a minimal risk of contamination. A useful adjunct to DNA comparison in this situation is the compa riso n of ante- and post-mortem dental radiographs; rad iographs taken at the time of a 'check-up ' may reveal characteristic root and crown patterns (Fig. 23.4) 2 1
Ante-mortem Procedures and Dental Records Ante-mortem procedures require the retrieval by the police of dental records, including radiographs, study models and clinical photographs, and rely on the clarity and accuracy of those records. The successfu l retrieva l of dental reco rds depends upon diligent police work. In a 'closed' situation such as an air disaster, the passenger list can be obta ined, relatives iden titled and dental records sought. However, an individual
--
~
-
437
Figure 23.4 Root canal fillings and root cu rvatures identified in an te-mortem radiographs can be matched against post-mortem findings. An experienced dental radiographer ca n adjust the position of th e skull or ja ws to the position in which the ante mortem radiographs are most likely to have been taken. This fa cilitates th e id en tifica tion of not on ly correspo nd ence of restorations but also root and ca nce ll ous bone morphology.
may visit several dentists over a lifetime. Multiple records may exist each recording only the work undertaken by one dentist. These will not necessarily include a record of pre vious treatment. Ret rieval of dental records of victims of an 'open' disas ter, such as a fire in a theatre or railw ay station, may take much longer since there will be no record of who was pres ent. Police may have to await contact from anxious rela tives enquiring about a person who has not returned home. The helpline telephone numbers issued by the police on such occasions help serve this purpose as their opera tors are trained to seek such information. In the case of the dis covery of a body of a long-deceased individu al, identifica tion will entail the retrieval of dental records of 'missing persons'; this may take many months. On receipt of the dental records, the forensic dentist cre ates a single dental chart on a form identical to that being used for the post-mortem charting (Fig. 23.5). This chart may be created manually or electronically. In theo ry, this should be straightforward, but in reality it is frequently very time-consuming and prone to errors related to the different methods of ante-mOliem chartin g; Ahlberg 22 records 40 dif ferent notations in use throughout the world . Incomplete, illegible and in accurate records compound the problem. While some states in the USA make legal provision for comprehensive charting of the teeth at dental consultation, there is no such requirement in the UK, where only work to be carried out need be charted.2J Records frequently do not include work carried out by a previous dentist and are thus incomplete. Ireland et al 24 have shown that even basic infor mation is often unrecorded, with only 7 per cent of carious cavities recorded and only 2 per ce nt of general dental prac titioners recording the occlusio n. Whatever system is used,
438 I
Paediatric dental identification
IDEN TAL Re f No
POSTMORTEM Ref No
0. 0 .8
FO RENAMES
SURNAM E AD D RESS
Tela,
() ~Hv ) D:.~) (Ill}) "'" ". 0 C~ 0) (v q~ ~ ~ ~D'O (0 (J)~) CD (~··D
~
100
DC(~
~
»
10:'
,I
~D
(0 0 CDdD G
,-
,'A
Figure 23.5 Asimple plastic lam inated dental char t fo r use at the locu s. An indeli ble pencil/pen should be used to prevent damp and dirty conditions fr om spoiling 'first sourc e' evidence. This record should be retained even after a 'cl ea n' copy for filin g, and later ante -mortem matching, has been created. Close examin ation , includin g rad iographs of the teeth and ja ws, should be undertaken in th e labora tory once the remains have been removed from the locus. Any laboratory findings, includ ing radiograp hs, may be added to the clean post-mortem record.
-_
~
.........
,
'
Figure 23.6
-
- ....
. -~ - ~,
..
~-
~.
An almost illeg ible ante-mortem dental chart.
illegib ility may m-ake interpreta tion almost imposs ible (Fi g. 23.6). Co mputerized records should red uce this problem. Among the common ante-mo rtem charting erro rs are transposit ion of left and right, misreco rdin g of a bu cca lly extended restoratio n as a ling ual extens ion, a tlrst bicuspid charted as a second bi cuspid tooth, and first an d sec ond permanent molar teeth cha rted as seco nd and th ird molars when on e molar is absent from the arch. Other errors includ e adding a surface to a resto ration or incorrect record ing of su rfaces fill ed, listing a tooth as restored when it is not and om itting to record a unique fe ature such as a sup ernumera ry too th or an apicectomy.7 A novel ante-mortem development designed to assist in the identification of child victim s of fires or mass di sasters
-- -
is known as 'toothprints'. This procedure, devised and developed by Tesini et aJ25.26 in conjunction with Kerr Dental, utilizes an arch-shaped thermopl astic warm wafer into which the child bites for 50 seconds. After 2-3 min utes' cooling, the wafer is sea led into a pl astic bag, thereby preserving both a very accu rate impression of the teeth and a suffi cient numb er of shed ep ithelial cells for fut ure DNA analysis. Good a nalytica l results after 3 yea rs appear to be achieved if the wafer is a ll owed to dry prior to sea ling in the zipl ock bag 27 Longitudinal trials will be necessary to determine how long the wafer remains a valuable source of DN A. This is important sin ce the de nta l record in the child's impression will beco me of decreasing value as teeth are shed, replaced by permanent teeth or restored.
Post-mortem Procedures The Illajority of denta l identification s are carried out on sin gle bodies. It is good practice for a forensic dentist to be called to the site (locus) where the body has been fou nd since the teeth necessary for identification may be loose or, follow ing a fire, brittle, and requi re expert handing to avoid dam age or loss th ereof. When remains are found in a shallow grave, dogs or foxes Illay have disturbed them, resul ting in loose teeth or dentures being displaced. A forensic dentist will be aware of these possibili ties and ensure that the site is meticulously inspected. If teeth or dentures are found in pl ace within the j aws, they should be photographed ;11 situ, ca re full y bagged, labelled and removed from the scene only with receipt of permi ssion from the investigating poli ce officer. In the mortuary, a detailed dental examinati on is usually best carried out afte r the completion of the post mo rtem examin ation, using a forens ic dental kit comp ris ing a full range of dental hand instruments, a portab le water pick to flu sh away debris and fas t-setting adhesives, su ch as cyan acrylate, to stab ilize tooth fragments. If den tures are present, they shou ld be photographed in position. If the deceased is beli eved to have had den tures but these are not immediately availab le then impressions of the jaws shou ld be taken in ord er to create plaster models upon which a later retrieved dentu re may be fitted. Making photographic records is importa nt. Thi s may best be achi eved by utilizing the servi ces of police photog raphers under the directio n and in the presence of the forensic dentists. However, where th is is not possible, a 35-mm single-lens reflex (SLR) digita l camera (10 mega pi xels or greater) with a dedicated flash facili ty will nor mally enable the dentist to obtain satisfactory illustrations. Accurate dental charting in the mOliualY is best achieved with the assista nce of a dental nurse or a seco nd forensic dentist. The author has a preference for Llsing pencil at this stage as charts may become soiled and ink may run . Whenever possible, the jaws should be radiographed. Small whi te (composite) fillings may be very hard to see in the mortuary but readily identified on radiographs, as are
----~---~-----------
---
Comparison
I
439
.:.~( ' /.......
1r ,-# Figure 23.8 Fire has destroyed the anterior teeth but the extent to which the soft tissues around the mouth protect the teeth is evidenced by the preservation of most of a plastic denture. There is no carbon trail across the palate. The victim died before the fire engu lfed him.
Figure 23.7 Victims offire are candidates for dental identificat ion, be ing vi sual ly unidentifiable but with teeth in tact.
root canal fillings and unerupted teeth. If suitabl e radio graphic facilities are not availabl e in the mortuary, permis sion may be granted to remove the jaws at autopsy and take them to an X-ray department. Permiss ion to do this may only be given if there is gross facial disfigurement preclud ing visual identificatio n. Disarticul ated jaws are simple to radiograph, multiple views may be taken to reproduce the orientation of any ante-mortem radiogra phs and thus facil itate matching. If permission is granted to remove the jaws then two criteria must be satisfied irrespective of technique: minimal further facial disfigurement and preservation of the apices of the tooth roots within the resection. An excel lent technique is described in detail by Whittaker and MacDonald. 28 Extreme care should be exercised when remov ing the jaws of fire victims (Fig. 23.7). Both the police and relatives of fire victims will want to know whether the deceased died before or after they were engulfed by fire. If the victim died before the fire, there often will be no soot deposit on the palate as evidence of smoke inhalation. It is imperative that those carrying out the autopsy do not inad vertently contaminate the palate with soot adheri ng to their gloves (Fig. 23.8). It is the author's practice, having gained access, to change gloves so as not to inadvertently carry soot onto the palate or tongue. Once the radiographs have been developed, post-morte m charting may be completed on a form identical to that upon wh ich the ante-mortem records have been transposed.
The procedure in a mass disaster is essentially the same, only the circumstances may necessitate working in a tem porary mortuary. There may also be a need to resist the pressure from relatives and the media for fast identifica tion, to keep the risk of errors to a minimum. After preparing the ante- and post-mortem records, the two records may then be compared. Where there is com plete concordance or an obvious gross discrepancy, there is little probl em. In other cases, an intelligent appreciation and interpretation of the data is necessary; this precludes simple 'computer matching'. In addition to the appreciation of the possible errors already described as possibly occur ring in ante-mortem charting, some other pitfalls should be considered. When teeth are extracted, a drifting of adjacent teeth may occur, and the resulting diastema and occasion ally associated partial rotation of a tooth may not be record ed. Where an ante-mol1em schematic drawing or computer generated icon appears on the record, the result may be a poor representation of the actual restoration. The number of restorations seen on post-mortem examination may be more or less than the number indicated in the ante-mortem chal1s. If greater, the causes may include additional work undertaken by a 'new' dentist or a simple clerical error. Reasons why the number of restorations may be less than indicated in the ante-mortem records include a post-charting emergency extraction, whi lst on vacation for examp le, a clerical error or, in the case of highly aesthetic restorations, a restoration that was not identified post mortem. Most forensic dentists have had the humblin g experience of returning to the disarticulated jaws and finding such restorations following a closer examination , possibly aided by ultraviolet light. However, if a filling is unam biguously indicated in the ante-mortem chal1s and radio graphs and is c1ea rl y absent post mortem, albeit with the
440 I
Paediatric dental identification
tooth present, then a mismatch in identification is the most likely exp lanation. Human error in recording and genuine disagreement about the identity of a tooth therefore prevent total reliance upon 'computer matching'. Aws 29 compared two computer programs, CAPMI (computer-assisted post-mortem identification) and DAV1D (disaster and victim identification), and found that both are a useful adjunct to comparison, especi ally in quickly iden tifying features indicative of a mismatch, but the final matching process requires a degree of human interpretation. In children without fillings , the matching of radiographs is of paramount importance. In a recent experimental eval uation, forensic dentists were provid ed with radiographs of teeth with no restora tions. For the purposes of identifica tion, we demonstrated that root morphology and alignment we re more usefu l than crown morphology for the purposes of identification. 21 Studies by Wood et a l ii demonstrate the value of digitized slices made from ante- and post-mortem radio graphs. The slices were manipulated using a computer graphics program and a quantitative comparison of the alignment of normal dental anatomical landm arks was undertaken. A perfect match was achieved in all 39 speci mens stud ied. In 1999, Wood et al 12 published a study of di gital den tal radiographic superimposition at various stages of devel opment of the dentition. The technique was app li ed to 25 cases in primary dentition phase, 25 in the mixed dentition phase and 25 in the permanent dentition phase. They found that their technique is reliable in the cas e of both primary and permanent dentitions but less so in the mixed dentition situation . Once the match ing process has been completed a repott must be made to the appropriate legal authority concluding with a statement of probability. The American Board of Forensic Odontology30 identifies four categories. 1. Positive. The ante- and post-moltem data match in sufficient detail a nd there are no irreconcilable differences. 2. Possible. The ante- and post-m ortem data have consistent features but, due to the quality of either the post-m ortem remains or the ante-mortem evidence, it is not possible t6 positively establi sh identifi cation. 3. In sufficient evidence. The avai lable evidence is inconclusive. 4. Exclusion. The ante- and post-mortem data are inconsistent and there are irreconci lable differences.
Facial Reconstruction The science of facial reconstruction is now very sophisti cated, using computer-guided morphometry. The interested reader is referred to the studies of Koelmeyer,31 Perper and Patterson,J2 Hill et al, 33 Philips and Smuts,34 Law and Bow ers,35 Sha h rom et al 36 and Tyrell et al. 37 While there is a strong scientific basis for morphologi cal accuracy, some features, such as colou r and length of hair or colour and texture of skin, can be based only upon probability, so that an exact likeness is not possible. How ever, in forensic work an 'exact' reconstlUction is not required. An approximate likeness may stimulate some one' s memory and enable him or her to go to the police to suggest a possible identity. The direct superimposition of photographs upon skulls in order to demonstrate a ' match' requires extreme care and attentio n to detail, especia lly as the quality of the photograph may not be high or recent. Where resources are meagre, how ever, it may be of some value. Brocklebank and Holmgren 38 assisted the Hong Kong authorities in body identification using this method when identification was required following gross disfigurement by fire (Figs 23.9 and 23. 10). Simpso n,39 noting the absence of negative controls in the Hong Kong cases, selected a skull from the collection of the Royal College of Surgeons of Edinburgh which, on the basis of size, morphology and dentition, was probably that of a middle-aged male Caucasian. Full fronta l photographs of 25 white members of staff of the Edinburgh Dental Institute aged between 23 and 60 years of age were superimposed upon the skull using the method described by Brocklebank and Holm gren. It was found that in two instances a 'convincing match' could be achieved. Despite much more soph isticated methods now being employed, further contro lled studies are required.
Dental Profiling Bodies fragmented by explosions or burnt beyond recog nition may be extremely difficult to identify, Dental
FACIAL RECONSTRUCTION AND DENTAL PROFILING In the absence of dental records of skeletonized remains, there have been many attempts to derive information from the skull and jaws with the aim of creating a 'profile ' of the deceased in li fe.
Figure 23 .9
A reassembled skull fo llow ing disintegration in a
fire. The body had been dumped in a skip and the n set alight. By kind permission of Dr L. Brocklebank.
--
~-
---
Age estimation I
441
Figure 23.10 Superimposition of a photograph of the suspected victim upon the skull. By kind permission of Dr L. Brocklebank. Figure 23.11 characteristics have proved to be of great value in their identification since the teeth are the most indestructible part of the body due to their mineralization, their partial encasement within bone and the low turnover rate of their constituent tissues. 4o The dental pulp may be an important source of uncontaminated DNA. Teeth may contribute information on age, sex, height and ethnicity.
AGE ESTIMATION Age estimation in children is conveniently divided into (1) prenatal and (2) from birth to 18 years of age.
PRENATAL In practice, it is unusual for a forensic dentist to estimate the age of a fetus, but an accurate estimation of age may be required for a prematurely born child when there are grounds to suspect infanticide. The examination of histological preparations of fetal and perinatal specimens provides very accurate estimations of age. 41 The data are based upon histological evidence of premineralization sequences coupled with mineralization and incremental patterns of enamel and dentine formation (Fig. 23.11). Radiographs can detect the earliest stages of calcifica tion in the developing primary dentition,42 provided the
Incremental striae.
fetal material is appropriately handled. Clement and Kossa 43 made this point succinctly in observing that calci fied tooth caps can be retrieved from the fetus throughout the second and third trimesters of pregnancy. Often these tooth caps are missing, since putrefaction liberates the tooth caps from the forming alveoli of the jaws. This is not surprising. The calcified part of the first permanent molar at birth, for example, is about the size of pinhead and weighs approximately 1.0 mg. For this reason, a complete radiographic survey of the fetus should be made before an autopsy in order to locate the tooth caps. However, radiography alone is insufficient to assess cal cification. There is a marked time lag in the identification of the initial stages of cusp calcification between those examined by histological and radiographic techniques. Histology (Fig. 23.12) is the more sensitive technique, early mineralization being detected up to 12 weeks before it becomes apparent on a radiograph.42 Since the primary and secondary (permanent) dentitions develop sequentially, it is possible to estimate the age of the child on any single defined tooth but it is much better practice to estimate the age from several teeth and average the result. With little routine experience in estimating age in this fashion, forensic dentists need to be aware of possi ble interobserver variation, the quality of the radiographs and the histological sections, and be cautious in their reporting.
442 I
Paediatric dental identification
Deciduous dentition
~ 2yearS
5 months
in utero
~ (+/-6mOS) o
7 months
in utero
Prenatal
~ O QwJ
~ -oooo dY1 Birth o
'0"1' e
~~ ~0 .' U .
=
~ moo", '" lc '0000= (+/-
Figure 23.12
2 mas)
,
' 4 years
~
[./- 9 m"i
Histology of a developing deciduous tooth.
~
In the author's view, dental age estimation in the peri and prenatal periods is not a routine procedure and should be carried out by two forensic dentists working in close collaboration with expert histologists and radiographers, initially examining the material independently but ulti mately presenting their findings in a joint report.
~ ~ o Infancy
Early childhood (pre-school age)
PFMP Figure 23.13
Figure 23.13 Schour and Massier charts for developing deciduous teeth.
BIRTH TO 18 YEARS OF AGE Possibly the most important question to be answered about a dead neonate is whether or not the child could have been born alive. An accurate estimate of age at the time of death is crucial to answering this question, and most authorities agree that data derived from the developing dentition provide the most accurate means of age estimation in a chiJd. 44 ,45 Ground sections of enamel reveal cross-striations in enamel prisms. The intervals between these striations, meas uring about 16 ~~m, represent the amount of enamel formed in each 24-hour period. 46 .47 Disturbances in cellular activity and calcium metabolism around the time of bilih produce a grossly accentuated striation, called the 'neonatal line' in both the deciduous teeth and the first permanent molars. The presence of the neonatal line there indicates that the child had been born alive and lived beyond the neonatal period. Using the neonatal line as a baseline, a count of the sub sequent striations should provide the child's age in days until such time as the enamel ceases to form. However, while the enamel is still forming, the leading edge is not calcified and therefore does not show an incremental pattern.
------------ -----------
18 months
(+/- 3 mas)
--
The number of days taken to produce the uncalcified form ing front is speculative and the potential error is taken into account by giving the age in days as between 20 days either side of the total count. From birth to 6 months the state of development and mineralization of the crowns and roots of the deciduous teeth and the first permanent molars can be determined radiographically, and the body need not be mutilated. 4s - 5o From 6 to 30 months, age may be estimated from the state of emergence of the deciduous teeth, provided that it is appreciated that there is a range around the mean time of emergence of about 2 months for the central incisors and 4 months for the lateral incisors, canines and molars. 51 Chronological and dental age do not, therefore, perfectly coincide and, as there may be interobserver variation in appreciating what constitutes 'emergence' into the mouth, the radiographic techniques developed by Demirjian and his colleagues,50,52,5] used in conjunction with Schour and MassIer charts (Fig. 23. J3),54 are the preferred methods.
-----------------
-----
-
Age estimation I
These charts are periodically updated by the American Den tal Association (available at: www.forensicdentistl),online. org/Forensic_pages_l/ eru ption2.htm). Recently, Maber et al 55 published an extensive and detailed review of the accuracy of age estimation of radi ographic methods using developing teeth. They concluded that the most accurate method is that of Willems, 55 using a modified Demirjian's technique. De Salvia et a[57 published a salutary paper in which they concluded, perhaps not surpris ingly, that the greatest limitation to the use of radiographic methods to determine the dental stage of development is operator experience. The case for more than one forensic dentist to make independent age estimation on the same material is implicit before conferring and submitting a joint report to the relevant authority. Age estimation of the living remains relatively uncom mon, but a recent paper by Garamendi et al 58 illustrates the value of dentistry in assisting immigration officials. By combining skeletal (carpus) radiographs and Demirjian's method, they were able to place 114 Moroccan males either above or below the age of 18 years. The results were con firmed by the Moroccan Embassy in Spain. The Schour and Massier charts are self-explanatol), and, for that reason, popular. Both deciduous and permanent dentitions are displayed i/1 situ, including root resorption sequences for the deciduous teeth. The drawings are life size, enabling direct comparison of either extracted teeth or radiographs. The chalis do not distinguish between males and females, but Ciapparelli 59 demonstrated that the corre spondence is very good for males, with females on average being aged 3-6 months younger than they actually are. Mornstad et al GO have suggested that these 'chart com parison' methods may be criticized as too subjective. In an attempt to address this concern, they made objective meas urements of developing teeth and correlated them with the subjects' chronological age. The structures measured were crown height, apex width and root length. With the aid of a multiple regression model, a linear relationship between some of these distances and age was shown. The method has a 95 per cent confidence interval (CI) of about ± 2 years around an estimated age. In a recent assessment of age estimation methods, Liver sidge et al 61 detei-mined that Monstrad's method is currently the most accurate. Since this paper was published, however, Foti et al G2 have described equations for age estimation in both living and dead children obtained with the help of step wise ascending multiple linear regression. The equations should be applied on the basis on the number of erupted teeth and tooth germs, which are detected on radiographs during clinical examination and in infant skeletal remains. Foti claims that this method permits age estimation until 20 years of age. Beyond the age of 20 years, the most accurate method of age estimation using teeth is not morphology but a bio chemical method, based upon aspartic acid racemization in dentine. 63 This method relies on the age-dependent non-enzymatic changes of L-form amino acids to D-form
443
Figure 23.14 Ground section of tooth. Tetracycline fluorescent bands are present, each indicative of an episode of tetracycline medication.
amino acids, a progression which shows a linear relationship with time. Analysis of aspartic acid in dentine is amongst the most reliable and accurate method to date, and work continues to refine the method yet fUliher. G4 Recently, Yekkal et al 65 published a pilot study, which evaluated the efficiency of high-performance liquid chromatography (HPLC) as dis tinct from gas chromatography. Their work suggests that HPLC may provide more stable derivatives of aspaliic acid for the estimation of racemization. Pioneering work based on the changes in human non collagenous proteins from dentine by Sajdok et al 66 demonstrated a linear correlation with increasing age, but the wide scatter of values in this initial work limits its cur rent usefulness. Should a tetracycline antibiotic be used to treat a febrile illness at a time when the crowns of teeth are still forming, it will bind to the hydroxyapatite crystals being laid down in the forming dentine and enamel. The bound tetracycline is fluorescent in ultraviolet light and visible in ground sec tions. From an examination of these F1 uorescent bands, the number of courses of antibiotic may be deduced and an estimate made of the age of the child at the time of admin istration. It may be possible to match these findings with the medical records of a missing child [Fig. 23.14).
Sex Determin ation Sex determination is normally achieved from the skull rather than the teeth. Prominent supraorbital ridges and mastoid processes may suggest masculinity, as do heavy rugged jaws with a tendency for the angle of the jaw to be a right angle G7 Other features in male skulls include promi nent muscle lines in the occipital area, square orbits with rounded margins, heavy, lateral arched zygomatic bones and large broad palates. G8 •59 These features are only sugges tive; children and bodies fragmented in fires and explosions
444 I
Paediatri c dental identification
may be very hard to sex from their skulls. 7o The cellular material in toot h pulp can be analysed for X and Y sex chromosomes/ I •n although vVhittaker et al n also showed that the determination becomes less reliable as putrefaction proceeds. Intact cell nuclei are needed in most methods, and this is clearly of limited use in burnt remains. 73 However, the protected dental pulp is an exploitable source of DNA because of its resistance towards physical a nd chemical exterior conditions. 74 Work by Urbani et al 75 has demon strated that dental pulp is an excellent source of DNA for gender determination, being retrievable and capable of an alysis from unprotected teeth heated up to 200°C for 15 minutes. With teeth embedded in bone and soft tissue a 100 per cent success rate was obtained from teeth heated up to 350°C for 15 minutes. Recently, Karaman 76 reported achieving 83 per cent accuracy in predicting sex usin g diagonal , as distinct from mesiobuccal, measurements of anterior teeth. However, this degree of accuracy could only suggest, not prove, the sex of skeletonized remains. Nevertheless, more sophisticated methods are not with out their problems. Kumar and Hegde,n using polymerase chain reaction techniques, were able to identify sex with 100 per cent accuracy from the pulps of deciduous teeth, but only up to 1 month after ex traction. After 6 months, the method was no longer effective.
In addition , neither of these studies makes a distinction between eruption and emergence. For a highly detailed and robust scientific study of odontometrics relating to compar ative and sexual dimorphism the reader is referred to Human Adult Odon tom etrics by JA Keisser. 8o Teeth have more recently caught the imagination of the Department of Anthropology of the University of Durham, where Buck and Vidarsdottir81 of the Evolution ary Anthro pology Research Group have investigated differences in the mandibular morphology of 'subadults' (i.e. not adults) using geometric analyses. One hundred and seventy-four mandibles from five morphologically distinct samples were digitized and subjected to Procruste's analysis. The results showed significa nt morphological differences between the samples and obtained cross-validation results of over 70 per cent accuracy in identification of unknown individuals using the complete mandibl e. They suggest that these tech niques could provide a method for the identification of race in a subadult individual. This appears to be, at first sight, a useful tool in demographic research, but application to pro filin g an individual skeleton can only result in a suggestion or, a t best, increased probability of gender determination.
REFERENCES
Height
2
Fazekas and Kosa 78 showed that all the bones of the fetus can be measured and that, by constructing regression curves for each bone, the size of the bone may be related to the height at death of the deceased. Clement and Kosa 43 claim that the basilar part of the occipital bone, the zygomatic and the temporal bones are excellent for this purpose, but the best and easiest bone to identify is the mandible. They state that the length in millimetres of the mandible is equal to the total crown-heel body length in centimetres, while the max illa is 1/20th the length of the body. At birth, with all the bones of the body much further developed, direct measure ments are possible_and these methods are superfluous.
3 4
5 6 7
8 9
Ethnicity
10
Prior to the advent of DNA analysis, the determination of ethnic origins from dental remains was unreliable because the dental a nd orofacial indicators were very imprecise and the overlap betvveen racial characteristics was so consider able that, realistically, they were of little value in determin ing the ethnic origin of anyone set of remains.7o Hurme 79 states that the eruptive process occurs 6 months to 1.5 years earlier in blacks than in whites, but this trend is of little value in profiling and it remains thus unless the circumstan tial evidence indicating a specific ethnic origin is strong.
--
-
II
12
13
14
-
Carey E. Dio's Romall History, vol. 8, book 61. London: William Heinem a nn, 192 5, p. 17. Solheim T, Lorentsen M, Sundnes P et a l. Th e 'S candinavian Star' feny disaster 1990 - a challenge to forensic odontology. lnf J Legal Med 199 2; 104:339-45. Delattre V. Forensic denta l id entification in the greater
Houston area . J Forensic Sci 200 1; 46:1379 - 84.
Knight B, Pearl H. The legal system in the British Isl es.
In Clark D (ed.) Pracrical Forens ic Odontology. Oxfo rd:
Butterwo n h-Hein emann, 199 2.
Hill IR. Id entification , the dentist and the coroner in Engl a nd and Wal es . J Foren sic Odontostomatol 1988; 6:77-82. Jones DG. Odontology often is the final piece in a grim puzzle. J Calif Dent Assoc 19 90; 26:650-1. Dailey J. Charting errors in ma ss disa ster denta l reco rds. In Bouers CM, Bel l GL (eds) Th e Manual oj Forensic Odontology, 3rd edn. Saratoga Springs, NY: Am erican Society of Forensic Odontology, 1995, pp. 250-8 . Kiesser- Neilson. Person Idelltijicatioll by Mea ns oJThe Teeth: A Practical Guide. Bristol: John Wright, 1977. Philips VM. Th e uniqu eness of amalgam restorations for identifi ca tion. J Foren sic Odol1tostol11atol 1983; 1 :33-8. Smith B. Recons truction of root morphology in ske letonised remains with postmortem dental lo ss. J Forensic Sc i 1992 ; 37: 176-84. Wood RE, Tai CE, Blenkins op B, Johns ton D. Digiti sed slice interpOSition in forens ic dental radiographic identification. A m J Forensic Med Pathol 1994; 15:70-0. Wood RE, Kirk NJ, Sweet DJ. Digita l dental radiographic identifica tion in the paediatric, mi xed and permanent dentitions. J Forensic Sci 1999; 44: 9 10-16. de Villi ers CJ, Philip s VM. Pe rson identification by me ans of a single unique dental feature. J Forens ic Odonrostomotol 1990 ; 16:17-19. Nonje CJ. Harris AJIiIP. Maxi llo-facia l radiology in forens ic dentistry: a review. J Forensic Odon tostol11a tol 1986; 4:2 9-38.
----
-
-~-
References I
15 Watt R, Sheiham A. In equalities in oral health: a review of the evidence and recommendation for action. Br Dellt J 1999; 187 :6- 12. 16 Kaste LM , Selwitz RH, Oldakowski RJ et al. Coronal caries in the primary and permanent dentition of children and adolescents 1-17 years of age: United States, 1988- 1991. J Den t Res 1996; 75:631-41. 17 White BA, Caplan OJ, Weintraub JA. A quarter century of changes in oral hea lth in the United States. J Dent Educ 1995 ; 59:19-57. 18 Flink A, Kallestal C, Holm AK et al. Distribution of caries in 12-yea r-old children in Sweden. Social and oral hea lth related behavioural patterns. Coml11l1ni~y De/If Hea ltft 1999; 16:160-5. 19 La wrence HP, Sheiham A. Caries progression in 12- to 16 year-old school children in tluoridated and deficient areas in Brazil. Communiry Dent Oral Epidemiol 1997; 25:402-11. 20 Moody GH, Busuttil A. Identification in the Lockerbie air disaster. Am J Forensic Med Pathol 1994; 15:63-9. 21 Sholl SA, Moody GH. Evaluation of dental radiographic identification: an experim ental study. Forensic Sci fnt 2001; 115:165-9. 22 Ahlberg JE. We must get our numbers right. FDI Newslerrer 1987; 158:8-9. 23 Ayton FE, Hill CM, Parfitt HN. The dental rol e in the identification of the victims of the Bradford City football ground fire. Br Dent J 1985; 159:262- 4. 24 Ireland RS, Harris RV, Pealing R. Clinical reco rd keeping by General Dental Practitioners piloting the Denplan 'Excel' accreditation programme. Br Dent J 2001; 191 :260-3. 25 Tesini DA. O'Malley KD, Schwartz S. Development of bite impression technique for use in identification of unknown and missing children. J Mass Dent Soc 1985; 34:61, 63 , 69, passim. 26 Tesini DA, Harte D, Crowley K. Dentistry's role in the identification of missing and unknown children: upd ate on the dental bite impression technique. J Mass Dent Soc 1999 ; 48:29-34. 27 Tesini DA. Comments on the Toothprints Bite Impression jo r Search and Identification oj Missing and Unknowl1 Children. 2003. http: //www.kerrdental.com/Toothprints/ Toothprints0f020Position20Paper 28 Whittaker DK, MacDonald D. A Colour Atlas oj Forell sic Dentistry. London: Wolfe Medical Publishers, 1989. 29 Aws G. An Overview and CompaTison ojCAPMI and DAVID Programmes. Dissertation for the Diploma in Forensic Odontology. Hatfield, UK: Uni ve rsity of Hertfordshire, 1999. 30 American Board of Forensic Odontology. Body identification guidelines. J Am Dent Assoc 1994 ; 125:1244-52. 31 Koe1myer TO. Video ca mera superimposition and facial reco nstru ction as an aid to identification. Am J Forensic Med Patho/1982; 3:45-8. 32 Perpe r JA, Patterson GT. Face imaging reconstructive morphography. A new method for physiognomic reconstruction. Am J Forensic Med Patho11988; 9:126-38. 33 Hill B, Macleod RM, Crothers A. Rebuilding the face of George Buchanan (1506 - 1582). J Audioli Media Med 1996; 19: 11-15. 34 Philips VM, Smuts NA. Facial reconstruction: utilization of computerized tomography to measure facial tissue thickness in a mixed race population. Forensic Sci Int 1996; 83:51. 35 Law CA, Bowers CM. Radiographic reconstruction of root morphology in skeletonised remains: a case study. J Forensic Sci 1996 ; 41: 514-17. 36 Shahrom AW, Vanezis P, Chapman RC et aJ. Techniques in facial reconstruction: computer-aided facial reconstruction using a laser scanner and video superimposition. Int J Legal Med 1996; 108:194-200.
445
37 Tyrell A, Eviso n, MP, Chamberlain AT, Green MA. Forensic three-dimensiona l facial reconst ruction: historical review and contemporary developments. J Forensic Sci 1997 ; 42 :653-61. 38 BrockJebank L, Holmgren CJ. Development of equipment for the standardization of skull photographs in perso nal identification. J Forensic Sci 1989; 31: 1373-9. 39 Simpson JDR. The investigation of an unknown skull with superimposition of known faces. Dissertation for a Master of Science Degree. Edinburgh: University of Edinburgh, 1998. 40 Malaver PC, Yunis YY. Different dental tissues as source of DNA for human identification in forensic cases. Croat Med J 2003; 44 :306-9. 41 Calonius PEB, Lun in M, Stout F. Histologic criteria for age estimation of the developing human dentition. Oral Surg 1970 ; 29:869-76. 42 Kraus BS, Jordan RE. The Human Dentition Bejore Birth. Philadelphia, PA: Lea and Febiger, 1965. 43 Clement JG, Kosa F. The fetal skeleton. In Clark DH (ed.) Practical Forensic Odontology. London: Wright, 1992, pp. 43-52. 44 Ashley KF. Identification of children in a mass disaster by estimation of dental age. Br Dent J J 970; 129: 167 -9. 45 Ciapparelli L. The chronology of dental development and age assessment. In Clark DH (ed.) Practical Forensic Odontology. London: Wright, 1992, pp. 22-42. 46 Schour L, Hoffman MM. Studies in tooth development, 2. The rate of ap position of enamel and dentine in man and other mammals. J Dent Res 1939; 18:161-75. 47 Noble HW. The estimation of age from the dentition. In Harvey W (ed.) Dental Identification and Forensic Odontology. London: Kimpton, 1976, pp. 28 - 35. 48 Moorrees CFA Fann ing EA, Hunt EE. Formation and resorption of three deciduous teeth in children. Am J Phys A l1thropol 1963; 21 :205-13. 49 Moorrees CFA. Fanning EA, Hunt EE. Age variation of formation stages for ten permanent teeth. J Dent Res 1963; 42: 1490-502. 50 Demerjian A. Postnatal grow th. In Falkner F, Tanner JM (eds) Dentition in HU1110n Growth. London: Bailliere Tindall , 1978, pp. 413 -44. 51 Koski M, Gam SM. Tooth eruption sequence in fo ssil and modern man. Am J Phys Anthropol 1957 ; 15:469-88. 52 Demiljian A. Goldstein H, Tanner JM. A new system of dental age assessment. Hum Bioi J973 ; 45:211-27. 53 Demerjian A, Goldstein H. New systems for dental maturity based upon seven and four teeth. Ann Hum Bioi 1976 ; 3:411-21. 54 Schour L, Massier Iv!. The development of the human dentition. JAm Dellt Assoc J941; 28: 11 53 - 60. 55 Maber M, Liversidge HM, Hector MP. Accuracy of age estimation of radiographic methods using developing teeth. Forensic Sci Int 2006; 159:S68-73. 56 Willems G, Van OImen A, Spiessens B, Carels C, Dental age in Belgian children: Demirjian's technique revisited. J Forensic Sci 2001; 46:893-5. 57 De Salvia A. Calzetta C, Orrico M, De Leo D. Third mandibular molar radiological development as an indi cator of chronological age in a European population. Foren.sic Sci Int 2004; 146:S9-1 2. 58 Garamendi PM, Landa M1, Ballesteros J, Solano MA . Reliability of the methods applied to assess age minority in living subjects around 18 years old. A survey on a Moroccan origin population. Forensic Sci Illt 2005; 154 :3 -12. 59 Ciapparelli L. The assessment of dental age in Essex school children using panoral radiographs with forensic app li cations. Dissertation for the Diploma in Forensic Odontology. Lond on : Hospital Medical College, 1985.
446 I
Paediatric dental identification
60 Monstrad H, Staaf Y, Welander U. Age es timation with the aid of tooth developm ent : a new method based on objecti ve measurements. SCll1l d J DCIII Res 1994; 102: 137 - 43. 61 Liversiclge Hrvl, Lyons F, Hector MP. The accuracy of three methods of age estimation usin g rad iog raphic meas urements of developing teeth. Forellsic Sci il1t 2003; 131 :22- 9. 62 Foti B, La lys L, Adalian P et al. New forensic approach to age es timation in ch ild re n based o n too th erup ti on. Forel1sic Sci ill I 2003; 132:49- 56. 63 Ritz -Timme S, Ca ttaneo C. Collins rvlJ et al. Age est im ation: the state of the art in relation to th e spe cific demand s of forensic pra cti ce.illl J Lego! M ed 2000 ; 113:1 29 - 36. 64 Arany S, Ohtani S, Yoshioka N. Gonmori K. Age esti mation from asparti c acid racemi za ri on of root dent in by internal standard meth od. Forellsic Sci 111 1 2004: 141: 127-30. 65 Yekkal R. Meers C. Van Schepdae l A et al. Rclc emi za rion of aspartic acid from human dentin in the estimation of chronologica l age. Fomlsi c Sc i i1l12006: 159:S89 - 94. 66 Sajdok J, Pilin A, Pudil F et al. A new method of age est im ation based on the changes in human non -col lageno us prote ins from dentin. ForeHsi(' Sci i ll l 2006; 156:245-9. 67 Aitchison J. Sex differences in teeth, jaws and skull s. Delli Prerel 1963 ; 14:52-7. 68 Biggerstaff RH. Craniofacial characteristics as dete rminants of age, sex and race in foren sic dentistry. Del/r Clil/ North Am J 997 ; 21 :85- 97. 69 Borha CT. Cra ni o facia l characteri stics as determ inants of age. race and sex in forensic dentistry - a hand s on guid e. J Forel/ sic Od OI110S(0))/lIto/ 1991; 9:47-61. 70 Harvey W. Effec ts of sex, race, hereditary and systematic disea se on oral tissues. In Harvey W [ed.) Del/ta! Jdel/rijicotiol/ III/d Forellsic Odol/to!ogy. Lond on: Henry Kimpron. 1976. pp.36-4 3.
71
SenD M. Ishi zu H. Sex iden tification of a human tooth. illl J Forel/sic Dml 1973 ; 1:8 - 1I. 72 Whittaker DK, Ll ewelyn DR , Jones RW. Sex determin ation from necrotic pu lp tissue. Br Dellt J 1975; 139:403- 5. 73 Murakami H, Yamam oto Y, Yosh ito me K et al. Forensic study of sex determination using PCR on tooth sampl es. Acta Med OkaYII/JI1I 2000 : 54:2 1-3 2. 74 Ga ill a rd H. Lucl es B, Kaess B, Ma ngin P. Th e use of the teeth in genetic fin gerprinting. Bull GrouJl i nt Reclr Sci SI011111tal OdO lltO / 1994; 37 :65-70. 75 Urbani C, lastrucci RD, Krame r B. The effec t of te mp era ture on sex determination usin g DNA-P CR analysis of dental pulp. J Forel1 sic OdOI/IOS I0111I1to!1999: 17:35-9.
76 Karaman F. Use of diagonal tee th measurem ents in predicting gender in a Turk ish population. J Forensic Sci 2006; 51 :630-5. 77 Kumar MG, Heg de AM. Sex ide ntification from exfoliated primary reeth - a PCR study. J Ciil1 Pedilltr Den12005 : 30:19 - 21. 78 Fazekas lG, Kossa F. Determ in ation of the body leng th and age of fe tuses on the basis of th e dimen sions of the facial bones. Chapte r X . In Fa zeka s IG, Kossa F (eds) Forensic Fetll! Os re% m'. Budapest: Akade miai Ki ado, 1978, pp. 16 5-96. 79 Hurm e va. Time and sequ ence ot- tooth el1lption. J Forel/s Sci 1957 ; 2: 371 -88. 80 Ki esser JA. HUIIIan Adulr OdOIIlollIclrics. Studies in Biol og ica l An thro pology. Camb rid ge: Cambridge Uni ve rsity Press, 1990. 81 Buck TJ, Vidarsdottir US. A proposed method for the id ent ifi ca tion of race in sub- adult skeletons: geo metric morpho metr ic analysis of ma ndibular morph ology. J Forel/ sic- Sci 2004; 49 : 11 59 -64.
I
CHAPTER 24
I
THE EXPERT WITNESS AND EXPERT TESTIMONY Anthony Busuttil
Introduction Mission statement of the expert Claim to expertise Professional witnesses Opinions Yes or no? Admissibility of expert evidence Communications from the expert witness
447 448 449 449 450 450 450 452
Physicians are not ordinary witnesses, but give judg ment rather than testimony. Digest - Emperor Justinian (AD483-565)
INTRODUCTION The courts, by the very diversity of cases that are brought to their attention, wi ll, from time to time, invariably require assistance from those who possess intimate and profound knowledge of specific and-specialized subjects, of which the court is not expected to have direct first-hand knowledge. This is particularly the case in the context of an adversarial system of law, as practised in Great Britain - and in most Commonwealth countries - and the USA, whether the case is being tried in the civil or criminal courts. For example, if the matter under consideration by the court is the safe construc tion of a bridge, it is unlikely that the court would know much about the specific matters related to this, unless mechanical engineers, architects, steel erectors and quantity surveyors ass ist them by providing them with basic details of constructional requirements, good practice and building techniques. Such persons are referred to as eJ.pert witnesses
Declaration by the expert in the report In the witness stand or box Pre-trial communication Conflict of interest Rules of evidence Conclusion Recent developments References
452 453 453 454 454 454 454 455
in that they are assisting the court through (Latin ex.) their acquired specialized skill, craft or profession (Latin periNa). WG Dickson stated that: 'Another exception to the general rule against examining witnesses on matters of opinion occurs wherever the issue involves scientific knowledge, or acquaintance with the rules of any trade, manufacture or business with which men of ordinary intelligence are not likely to be familiar. A foundation for such an examination must be laid by ascertaining whether the witness is a person of skill or an "expert" [the English term], under which is included those who have a theoretical acquain tan ce with the subject, as well as men who speak from practical know ledge:] Medical practitioners who are summon ed to the courts, criminal or civil, in order to assist with medical matters can therefore appear as expert witnesses. In criminal matters, their ex peliise is mainly required for the interpretation of traumatic changes and causes or modes of death; in civil cases, medical expertise is requ ired for cases of alleged medical or industrial negl igence and in li tigation relating to compensation for injury and disability. As with all other evidence given to the courts, the testi mony is taken under solemn oath or affirmation; thereby, it is empowered with a further, more robust, sea l of assurance as to its veracity, totality, professional integrity and impartiality.
*The mal e gender that is used throughout th is chapter also refers to the female gender in all instances.
448 I
Expert witness/expert testimony
As Cicero stated, 'the court is the temple of truth, and the search for truth is critical towards the conduct and conclu sion of a fair trial'; this should be the tenet that underlines the involvement of the expert witness. Similarly, in the preparation of cases for trials, be these criminal or civil cases, it is often useful for both the prose cuting auth orities and legal defence team (criminal) or, for the legal team putting the case together for one or other of the litigants involved (civil) , to acquire information on subjects with which they are not familiar. This would guide them with respect to the strength or weaknesses of the case and in coming to considered decisions as to whether to plead guilty to the charges, to settle out of court, to instruct further investigations to be carried out prior to the trial or to contest the trial. A very onerous duty and a heavy responsibility are placed on the expert. Initially he has to consider very carefully whether or not he possesses the qualities appropriate to fulfilling the expectations placed on him by the courts or by legal professionals in the partic ular matter under review.
MISSION STATEMENT OF THE EXPERT By its very nature, the criteria that expert evidence has to satisfy are the following: 1. It refers solely to specialized medical matters and themes, and not simply to matters that can be deduced from a common-sense or ordinary knowledge approach or by recourse to simple logic. It is on such specific issues, that the court - namely the ordinary 'ladies and gentlemen of the jUly' and/or the presiding judge(s) may require to be enlightened in order to reach the more appropriate verdict. 2. It has to be delivered only by those who are appropriately qualified to do so, both by virtue of an appropriate duration of first-hand specific experience in the matter under review and by having acquired their expel1ise by periods of study - these academic degrees and diplomas shou ld be related to the subject areas being addressed as, an expert. The direct expertise which is being sought should be kept in mind by the expert who is instructed, and, if it is felt that the subject is not within his direct professional knowledge and specific experience, then he should decline to act in the capacity of an exp e11 in the particular case. In the recent past, a trend developed in civil court rooms in which persons with a sound general expeliise spoke on more specific matters of which they had little first hand knowledge, and thereby brought the role of the expert witness into a certain level of disrepute. For example, a general surgeon cannot, under ordinary circumstances, be expected to speak from a vantage point of expertise on intricate stereotactic neurological procedures; if he does, he is quite likely to be overstretching his expertise. However, he is in a
position to give excellent expert advice on herniorrhaphy and its complications. The expert must appreciate the boundaries of his expertise and knowledge, and always ensure that he is not making pronouncements on subjects outside his experience. 3. The delivery of the evidence has to be such as to enable the court to appreciate and understand the specialized information that is being imparted to them; if this evidence is written down in the form of the report, it must be user-friendly, devoid of long, rambling sentences, and bereft of a surfeit of jargon and statistical calculations. It should be laid out logically, in sequentia l chapters, and it cannot be seen to be muddled, perhaps focally inconsistent, with conclus ions in it that are not fully justified and substantiated by wi?at appears earlier in the evidence. The information that is being imparted, albeit often quite complex and difficult, would somehow have to be titrated to a level that is readily comprehensible by the jury. The lawyers - and failing them the judge himself - will usually steer the expel1 in the correct direction while on the witness stand, and may ask him to elaborate on certain passages or chapters of his evidence, perhaps by giving analogies and examples that will assist the jury's comprehension. 4. This evidence has to be, as much as possible, f1awlessly dispassionate, totally unbi ased, non-partisan, carefully balanced and comprehensive. It must not contain views tailored to the advantage of the side instructing the expe11. In an adversarial system, whichever side calls on the services of the expert, the evide nce given to the court should be unsullied by leanings in favour of one or other side. Experts - including medical expe11 witnesses - who fail to achieve this level of probity have been variously referred to in a derogatory fashion as 'hired guns' and, worse, as 'gobs for hire'. The expert must always keep in mind that he is not there to win cases or to ensure that his 'side' triumphs against all odds; he is there to assist the court, to the best of his capabilities, in reaching the most appropriate decision or verdict, and no more. On many an occasion, the temptation may be very pronounced to veer away from this rule - but on the expert's head be it if he is found to have taken sides in this manner. 5. The data adduced in the expert's evidence must be accurate and up to date, both in terms of listing all the information given to him in his instructions by the legal personnel and in quoting the current peer reviewed literature. Using stale and superseded references in either books or journals that are later discovered to be so by the courts discredits the expert's testimony and his expertise. If the information given originally was somehow incomplete or in accurate, at least in the light of the evidence actually stated in court, and the conclusions reached are no longer
Professional witnesses I
tenable in the face of what was actually heard from the witness box. then the expert should honestly inform the instructing side that, as a direct consequence of this, his opinion has had to change. The courts, and eventually the instructing laYlryers , will accept sometimes quite reluctantly - that honesty is indeed the best policy. This implies that there may be occas io ns when it is expedient and indeed essential that, with the COUlt's permission. the expert actually sits in the court room and listens in to the factual evidence as it is being delivered. 6. All the data referred to by the expert must be properly acquired at clinical examination or necropsy or from available documents. To do otherwise would be fraudulent and. if the case proceeds to court, may be tanta mount to perjury. In this respect, the infamous instance of an expelt clinical neurologist who gave evidence on the findings of a clinical examination that he had never act ually personally conducted is perhaps a major sa lutary and admonitory lesson 2 The genera l judicial requirements of the expeli were summarized in the English civil case, National Justice Com pania Naviera SA v. Prudential Assurance Co. Ltd, tile Tk.arian Reefer (1993) 2 Lloyds Reports [2] 68, 81-82, and, after appeal, in (1995) 1 Lloyds Reports 455. These are par aphrased as fo llows: • Expert evidence must be an independent product of the expert uninfluenced by the exigencies of the litigation. • Expert evidence must be of independen t assistance to the court by way of objective unbiased opinion within the expert's expertise, wi th the expert never assuming the role of an advocate. • The expert witn ess must state the Facts or assumptions for the opinion expressed, never omitting material facts. • The expen witness must clearly indica te when a matter falls outside his ex pertise. • The expert witness must say if the data available are insufficient or if he cannot assert that the report contains the whole truth., • If the expert witness changes his opinion, he must co mmunicate this to the court and to the parties involved in the action. • If any photographs or other material are used to compile an expeli repoli, all of this must be made available to other parties.
CLAIM TO EXPERTISE Within the forensic arena, no matter how eminent the expert witness, aU aspects of the claimed expeliise and the testi mony given will be challenged by the legal parties involved in the trial - quite often themselves aided by 'experts' whom they have specifically hired to assist them and who may
449
indeed be present when the testimony is actually taken. The expert is never immune to criticism and to adverse com ment; he can, and most certainly will, be challenged in open court through searching quest ions and cross-examination. Evidence from other experts may be adduced to demonstrate that in the evidence given to the court, th ere were short comings, inadequacy of knowledge, flawed interpretations, a failure to consider all po ten tial options, and so on. All these matters should be considered prior to agreeing to prepare an expert report. There is no scope for arrogance or ex cathedra dogmatic pronouncements of the ipse dixit style; the court will be sure to query and to probe, and the expeli has to cultivate an unperturbed style that enables him to parry such ques tions and indicate why he disagrees with the propositions or scenarios that are being put to him , a nd to remain cool and collected, even under what may appear to be serious provo cation and verbal pressure from those who have been trained so to do. Thus, the opening gambit in couli is for th e expert to establish, to the satisfaction and probing of the court, the breadth and depth of his specific expertise, in order to pro vide reassurance. This is always the case, whether or not spe cific criteria are actually laid down by the courts delineating who can and who cannot be an ex pen. In continental juris dictions, the court has access to a list of persons who have been considered, vetted and accepted by the courts as pos sessing the appropriate standing to qu ali fy them as experts. In the USA, the Federal Rules of Evidence (Rules 104,40\-3, 702-6 - the Committee on the Judiciary, House of Represen tatives, US Government publication) establish the authority of the court to determine whether a person can be considered as an expert, whether his testimony can be heard, the type of evidence that can be admitted in court and the permitted scope of such testimony.] In the light of the Woolf Report, the position in Britain may already have changed in some respects towards the continental system.
PROFESSIONAL WITNESSES A fine line of distinction needs to be drawn between the so-called professional witnesses and expert witnesses.4 Any professional person, including a medical doctor, who is required to give testimony a bout professional matters is considered as a professional witness . Thus, for example, a young ARE department doctor w ho gives evidence about when he saw a patient wh o self-referred to his department - describing the nature of the injuries that he observed , with their measurements and locations, the number of sutures that he inserted, the presence and absence of frac tures - is giving professional evidence. Yet he should not be expected to be able to inform the court, from a position of expertise, whether the injuries he described were likely to have been caused in any particular manner or by any specific weapon, as this is unlikel y to be within his direct
450 I
Expert witness/ expert testi many
knowledge and ex perience. He is thus a professional and not an expert witness. This is a problem that has to be guarded against, both from the point of vi ew of the individual witness - who should have no hesitation at refusing to accept an expert's role if he feels unable to fulfil it - and from that of the courts - who should not expect every professional witness to transform himself into an expert witness on the spot when the court feels inclined to require this from him,s
OPINIONS The ability to assist the courts in matters not directly observed or personally done by the expert and the giving of op inion s highlights the only major co ncessions and privi leges that are given to the expert witness by the court; it is ass umed that, given his stated expertise, he can assess and weigh the information that he is provided with and present to the court balanced opil1io11s with regards to consistency, possibility, probability, ceI1ainty and likelihood of certain matters. He is not there to usurp the function of the court and take decisions, but sim ply to give evidence. Nor is he there to lecture and harangue tlle COU lt, no matter how emi nent his status and how sophisticated the subject matter that he is addressing. He cannot present ilie jury and judge with ready-made conclusions or take decisions on behalf of the court; he must at all costs remember that he is there only to assist them with infolmation , albeit of a specialized type and with opinions that will en able them to make up their own minds. Thus, the expert cannot proffer an opinion on the 'ultimate issue', i.e. the crux of the case, namely the central issues iliat will be deci ded by th e jury. In the light of R v. Stockwell (I 993} 97 CR App R 260, an expert in DNA profil ing cannot say to the court what he thinks the likelihood was that the defendant deposited blood or seme n at the crime scene as 'this requires considerati on of factors other than those within his area of expertise', and indeed this opinion wi ll have to be based on evidence that has been given in courts by other pa rties, which would have occurred when he was not present. This was brought to the fore when expert evidence was given relating to the degree of intoxication in a case of a woman who was alleging rape; there had been an allegation that she had been under the influence of alcohol. The expert referred to the complainant's level of intoxication as being commensurate witll that of 'someone who was not in control of her behaviour and judgement'. Although the expert was fully entitled to give this opinion on tlle matter of substance use in this case, the judge should have directed the jury that they were not bound by the expert's opinion and that they could make up their own minds (R v. Ugoh Et Others [2001] EWCA Clim 1381). In tlle case Re Hand R (Minors), the judge stated that, although the medical evidence was 'q uite indispensable', the 'final decision in the case is the judge's, and his alone'.
YES OR NO?
It is a fact with medical matters that often there are many possible ca uses or associations, although some of these are more probable than otllers, and some almost impossible; it is a matter of balanced judgement as to how a specific question about likelihood can be addressed and answered comprehen sively. It is therefore frequently the case that a cut-and-dried answer to a particular question asked of th e expert witness cannot be given, eith er in the affi rmative or in the negative. The answe r de pends on a numb er of facts and other factors, and can vary through a considerab le spectrum of possibili ties in the light of other prevailing and contributory circum stances. The expert should explain this to the court. A clear yes or 110 answer cannot be given with the best will in the world. Lavvyers may therefore ten d to consider doctors as being obtuse and sh ifty when they cannot provide the direct and curt answers that are being sought of them. Knock ing square pegs into round holes for the benefit of giv ing the lawyers ilie trite and curt responses requested by tllem does no on e any favours, particu larly the expert him self; one shou ld be as helpful as possible, bu t in many instances the telegrammatic answers required cannot honestly be given. Any bUllying by lawyers in this context should tllerefore be strenuously, but politely and resolutely, resisted, and if over persistent, the judge should be asked to intervene.
ADMISSIBILITY OF EXPERT EVIDENCE Paraphrasing C] King, the judge in the Australian case of R v. Bonythol1 1984 38 SASR 45, admissibility of expert evidence was held to involve the fo llowin g main issues: • Does the witness claiming expelt status possess sufficient knowl edge and experience to render his opinio n of substance and value in assisting the court to resolve the issues before it? • Does the subj ect matter of the evi den ce fall within the class of subject upon which expelt testimony sho uld be permitted, i.e. forms a part of a body of knowledge or human experience th at is sufficiently organized and recognized as to be accepted as reli able? • If new and unfamiliar techniques or technology have been used by the witness at arriving at his conclus ions, can the court be satisfied th at such techniques or technology have a sufficient scientific basis to render the result arrived at reliable and acceptabl e? • Will the judge, without experience in the subject matter on which the opinion is being sought, be ab le to for m a sound judgement without the assistance of the 'expert ' witnesses possessing spec ial knowledge and experi ence of that subject matter? In the US federal cou rts, adm issibility of expert evidence was, to a maj or degree, guided by the Frye test (Frye v. United States, 54 App. D.O. 46, 293 F. 1013, 1923), namely
Admissibility of expert evidence I
that the courts adm itted expert evidence deduced fro m weJl-recognized scientifi c principles or discoveries: ' the thing from whic h the ded uction is made must be suffi ciently establ ished to have ga ined general acceptance in the particul ar field in which it belongs'. The legal grounds of acceptab ili ty of expert evidence are even h igher a nd more demand ing than those of the expert's discipline 6 This 'general acceptance test' changed after the Daubert v. Merrell Dow Pharmaceuticals Inc (1993) 113 S Ct 2786 7 case and was replaced by a num ber of more flexible guidelines : • Can the theory or techni que on which t he evidence is based be tested independ ently? Has it been tested? • Can a determination be made of its fa lsifiab ili ty, i.e. can it be proved to be wrong if it is in fact wrong? • Has the theory or technique been subj ected to peer review and pub lic atio n? • What is currently known of its potential error rate? • Is the theo lY or technique generally accepted? The Kumbo Tire Co Ltd v. Carmichael (1999) 526 US 137 case perhaps enhanced further the flexibility of the test by stating that 'Daubert's list of specific factors neither neces sarily nor exclusively applies to all experts in all cases'. The US supreme co urt in Daubert's perhaps sum s up the 'proof of the pudding' in relation to expeli ev idence, namely 'vigoro us cross-exam ination, presentat ion of contrary evi dence and carefu l instruction on the burden of proof as being the t raditional means of attacking shaky but admis sib le evid ence'. In Brita in, detailed tests of admissibility of expert evi dence have not exercised the judiciary to any major extent; admissibility, to a large extent, has been based on a prag matic approach from the bench, erring on the si de of accept ance and acquiescence rather than otherwise, with evidence being ruled to be admissible if it is considered to be relevant and its probative va lue is considered to outweig h its prejudi cial effect. s Thus, in R v. Robb 93 Cr App R 161, the COLlli of Appea l upheld the trial judge's admiss ion of evidence given by a phonetician on voice recognition as expert, even though misgivings were expressed about the scientific basis for such ex pertise. , In the English family division, in the Re AB (child abuse; expert witnesses) (1995) 1 FLR 18 1 case the judge had to weigh the evidence before him in terms of fractures sus ta ined by a young child. An expeli for the defence claimed that a form of brittl e bone disease akin, but not identical , to the va rious types of genetica Lly documented brittl e bone disease existed that did no t have simila r hereditalY a nd familial characteristics of osteogenesis imperfecta but could account for the spontaneous occu rrence and accidenta l infliction of injuries, and thus exclude criminal damage and assau lt. 9 This 'temporary' brittle bone disease had not been substantiated by standard research studies or by pUbli ca tion s in peer-reviewed journals. In particular, the judge, in his criticism of this expert, indicated that he lacked obj ec tivity, omitted factors that fai led to support his theory and
451
continued to cite cases, altho ugh they had been t he subject of decisi ve judicial findings of non-accidental injUlY inflic tion. In summary, the judge considered this evidence to be unrel ia ble and inexpeli and he threw out his 'expert' evidence. In the context of child ab use, particularly sexual abuse, a considerabl e debate has arisen, particularly in the USA, concerning the weig ht tha t should be attribu ted to testi mony of such historical ab use because the subj ect could not recall spontaneously a ny such occurrences; expert psy cholo gists and psychotherap ists were required to gather information through repet itive interviewing and sessions under hyp nosis (sometimes combined with use of medica tion). These in te rviews we re carried out in alleged attempts to tap the so-called ' rep ressed-memory', ' recovered mem ory' or 'false-memory syndrome'. 10.11 The coulis a nd , per haps more teJlingly, the insurance companies who refused to pay compensatio n to those alleged ly stressed and trau matiz ed by suc h repressed occurren ces have not taken a very favourab le view of the entire mat te r. In civi l cases, in parts of Great Britain other than Scot land, there has been a move to ration alize the use of ex perts, particularly in personal injury compensation, and in a con sideration of the moves towards cond itiona l fees by law firms, namely the 'no win, no fee ' situation. In this field, the lawyers act ing for the claimant under the pre-action proto col set-up prepare the case, obtain a ll the relevant medical documentation abou t their client and submit this to the firm of so licitors acting for the person/company who will be defending the act ion. If possible, the matter is resolved at this point, before court proceedings are ini tiated. If there is no agreement reached at this stage, then the Civil Procedure Rules (SI 1998/3132 Part 35) and Practice Direction [1991] 1 WlR 1124, apply. Basically, these infer that the ex pert evi dence must be reasonably necessary (Rule 35.1), with the overriding duty o f the expert being to the co uli.) The court's permission is sought for the expert evidence4 a nd is then obtained as a w ritten report. 5 At this stage, the opposing party may put written questions to the expert. 6 The cOUli may direct that evidence will be required from only one expert,6 a single, joint expert who may be given instructions by both parties in th e litigation. s If the court believes that certain documentation should be further disclosed to the expert then it may direct SO.9 Such disclosed evidence may be adduced in court, I I but not undisclosed evidence. 12 Th e expeli may seek direction from the court. 13 The Daniels v. Walker [2000] 1 WlR 1382, CA, Lord WoolfMR case lai d out the proper approach for the cOUli . In the criminal cOUlis, the prosecution and the defence wi ll usually instruct their own experts, and it is here, perhaps, that the ogres of bi as and partisanship become manifest and have to be guarded against in the cut-and-thrust of the foren sic courtroom atmosphere and the persistent questioning by barristers and advocates. In some instances, th e judge wi/ I have seen, prior to the trial, the written repOlis from other experts, and he may decide that there is a sufficient agreement
452 I
Expert witness/expert testimony
and similarity between the two that perhaps a joint statemen t or minute of the points agreed to by both experts should be drafted with only the points of disagreement coming to the jury under their direction for assessment. In this instance, there should not be any disagreement about facts and the information collected as evidence. If this is shown to exist, there is a fundamental evidential problem that has to be resolved. Thus, for example, if there is a dis pute about the substantive clinical findin gs in a patient who has been examined separately, albeit on different occasions, by two do ctors or missed limb fractures in a utopsy findin gs then th ese matters require earl y and definitive resolution . Another fundamental problem relates to the breadth of the informatio n collected. A skeletal surveyor a blood clot ting scree n that has not been carried out by the prosecu tion expert when it is thou ght to be essentia l, omission of co l poscopic examination in a child abuse case, lack of histo logical examination on a utopsy blocks, failure to instruct toxicology or microbiol og ica l investi gations are all funda mental flaws that never come before a jUly but should be resolved prior to the court proceedings commencing.
COMMUNICATIONS FROM THE EXPERT WITNESS By and large the expert will communicate w ith th e instruct ing la wyers and with the courts in two ways: through a report and , eventually, in person (when evidence is given in court). In all instances, there should always be absolute clar ity in such communicatio ns, to the extent of using very ele mentary co mprehensibl e language. The ex pert 's reports should be dated, properly typed and form atted, with numbered pages and, when approp riate, numbered paragraphs. It should start by telling the readers of th e report'in whi ch matter this report is being produced, and who has instructed it and w hen. The reader would then want to be introduced to the report wliter in terms of how he claims expertise in the matter under revie w, and all the releva nt qualifications and experience should be listed; some prefer to attach t his a~ a mini-curriculum lJitae, which is entirely acceptable, and some suggest that it is preferable. If the a uthor has been instructed previously in simil a r cases and has also given evidence in court, this should be stated. If there are relev ant publications in the name of the expert these should be cited and, when relevant, copies of these publications should be attached to the report. Date of qualifi cation , General Medical Council (GMC) number and th e memberships a nd fello ws hips of va rious royal colleges, institutes and learned societies sho uld all be cited with the date of entry. The cunent post(s) held and previous relevant posts, with dates of appointments, should be given. (In Scotland , the date of bilth of the ex pert is a useful, but not essential, adjunct.) The report writer shou ld give some deta ils of his expertise in de al ing with th e specific case, for example the number of cases of child
---
sexual abuse in which he has examined the patient with a colposcope. The report will then have to cite what documentary and other evidence (photographs, histology slides, sketches) have been available for the compilation of the repOlt. If other inform ation has been used as background, then this should be included in the list. The various findings, if relevan t, in the clinical a nd any other reco rds ava ilable are then pn~c i sed or cited in fulJ as the case dictates. Obvious errors and inconsis tencies have to be referenced, for example if, in a particular, the nurse's notes refer to a left-sided pneumothorax but the doctor's record mentions a right-sided pneumothorax. It may be necessalY to indicate that the evidential bundle available is inco mplete, for example clinical photographs alluded to are not included, a sample sent for alcohol analysis has not been reported on or X-ray reports are not included. It is important to remember that those with legal training will have already gone through the notes page by page; it is not co nfidence inspiring if the la wyers point out the omission in yo ur repOlt of what turns out to be a vital bit of information that you have missed from you r perusal of the clinical notes. Having laid out all the facts , the co nclusion s should be derived logically and consistently. Of various co nclusions that are possibl e, there should be an indication of which one is the more likely. It should be kept in mind th at when referring to wounds one must be alert to the fact that lawyers may wish to know how muc h force was required, whether the wo und could be considered as life-threa tening or disfiguring, or whether the wo und could have been acci dentally inflicted or indeed se lf-inflicted. The options should be ranked and not mixed up in random order. One will be doing a major disservice to the instructing solicitors a nd to oneself - if the conclusio ns are jumbled up in any order and one has to admit in open court that the probability of one scenario is 1 in 1000 while tha t of another is 1 in 5. In coming to conclusions, one may have to cite the lit erature - joumals and books (Jatest edition) - to bolster up the arguments that are being made. These have to be cited in full. Ph otocopies of releva nt articles and chapters in books should be appended to yo ur report. The report has to be si gned, and it is a useful practice, to indicate that one has read the report carefully before sub miss ion and cOITected any typing and other errors in it, to sig n each p age individually. In spite of this - but obviou sly much more so without a check - a numb er of howlers do survive, even after the most meticulous trawling for errors in the rep ort, and one has to accept these candidly and apologetically.
DECLARATION BY THE EXPERT IN THE REPORT It is essential that the ex pelt witness shows that he has accepted his responsibilities and sbict duties to the COUlt full y and, as such, it is important that a decl aration is appended at the end of th e report to enunciate these principles.
-
~.
Pre-trial communication I . 453
This ' declaration' should include a statement to the effect that: • The report submitted, and any subsequent oral testimony which may have to be given later under oath or affirmation, is being given solely to assist the court and that such an overriding obligation takes full precedence over any obligations and loyalties the expert may have to those who have engaged him. • The report is a tru e and complete re port, and any opinions expressed in it are correct. • The report has been produced with all reasonable care to ensure that it meets the high standards expected by the courts. • In further reference to the completeness of the report, th e expert should indicate that he has cited all the sources of inform ation on which his report is based, whether it favours his arguments and opinions or othelwise. • The report is independent, unbiased and uninfluenced by the views of others. • It is ap preciated that th ere could be cross-examination on it. • Finally, it has to indicate that any professional fees due for expert services rendered by the witness, i.e. for compiling the report and subsequently presenting evidence, does not depend in any way on the final outcom e of the cas e. There have been regrettable instances in North America of experts producing favourable reports in return for large remuneration, particularly in high-profile cases, or in oth er instances where very large monetary settlements were in the sights of the lawyers. 12 - 14
IN THE WITNESS STAND OR BOX It is said that lawyers require experts who can articulate opinio ns clearly and pe rsuasively in court, can produce a polished but not patronizing presentation and who do not simply spout out raw technical detail. A speaker informs, and indeed persuades, by his, words, by his body language and through the modulations of his voice at the time in which he is speaking, with the words actually spoken per haps being the least effective communication. Body lan guage, in terms of facial expressions, posture, eye contact and hand movements, is important and conveys signa ls to the jurors and to the lawyers. To be able to do this well, you have to be very familiar with your script, and that includes all the other documen tary productions that may have to be referred to in your evidence. It impresses the jUly tha t you ac tually find imme diately the relevant paragraph in your report that refers to a particular question put to y ou , or th e page in the case records on which a particularly significant item has been recorded and, what is more, to be able to assist the court by helping court officials find the right place.
..
~
-
Think carefully before you a nswer a question and indeed allow an appropriate pause. Speak clearly and slowly enough that the judge's pen can keep up with you if he is recording the evidence in long hand. If you were un ab le to hear the question for one or other reason, or you wish the qu estion to be repeated or rephrased, do not hesitate to say so. Be careful to a nsw er the specific questions asked and not the qu estions that you would rather had been asked. The judge may intervene if you do not play by his rules. Whatever happens, keep your 'cool' and do not exhibit bad temper: do not scowl, sneer or sulk. You may be pro voked and tried ; thus, this advice is more easily' imparted than heeded. At the risk of being too nice, do not press your advantage by chea pening the opinions of others out of hand or by making cheap points against them, even if they deserve this. If, however, the question is objective and one is required to be critical of others, be fair in expressing your opinion, and give cogent an d competen t reasons for such criticisms. There is no place for arrogance and overconfidence ; be patient and courteous. Lord Avonside, in Assessor for Lothian Region v. Wilson, states that 'It can't be overemphasized that no committee or, for that matter, no court is in any way bound to accept expert evidence or to be dictated to by experts'. There is rarely scope for humour; the old adage that the judge is the only person in the court room who can make jokes, and that it is indeed mandatolY to laugh at them, is well made. The following two utterances by learned judges may illustrate this matter. In Rolland v. Lothian Health Board, Lord Ross stated: 'I am not disposed to place much weight on the evidence of Or H. He struck me as a witness who in this instance was unable to approach any question obj ectively; he was intent on defend ing his hospital and his staff and I found it difficult to accept some of the answers which he gave. Professor F gave his evi dence in what might not unfairly be described as a didactic manner; most judges, I imagine, dislike being lectured by w it nesses, and in tills respect I count myself among the majority'. In McVey v. Central Regional Council, RG McEwan sit ti ng as a tempo rary judge stated: '... He is a very experi enced surgeon and experienced witness. Whether this makes him unable or unwilling to accept that he could ever be wrong I know not. However, I find his at ti tud e less tha n helpful. He was patronizing almost to the point of discour tesy with the counsel for the pursuer and at times sarcastic ad personam over certain proper questions. Wisely counsel did not allo w herself to be provoked and responded only with courtesy. I might not have been so patient. For these and other reasons I did not have any co nfiden ce in his tes timony and where it differed from the other medical wit nesses, I had no hesitation in preferring them.'
PRE-TRIAL COMMUNICATION The expet1, to a much lesser extent but increasingly so - and long may the trend continue - will engage in pre-trial
454 I
Expert witness/expe rt testimony
conferences with the side which has instructed him. There, as in other communication, he should be absolutely dis passionate and ruthless but, without fear or favour, should give all the points in favour and those against. Holding back or camouflaging the problems at this stage will not do his side, or the case, any good; when the question is asked in court, he will have to concede (and blurt out what he may have politely, but reluctantly, tried to gloss over during a chummy cup of tea in chambers). The certainty, or otherwise, of the opinion given should be fully aired with counsel at this stage; this will go a long way in saving on expensive court time and, even better, preventing miscarriages ofjustice. Yet counsel should not assist the expert in any way in the preparation of his report. Lord Wilberforce at 256 in Whitehouse v. Jordan stated: 'While some degree of consul tation between experts and legal advisers is entirely proper, is it not necessary that expert evidence presented should be, and should be seen to be, the independ ent product of the expert, uninfluenced as to form or content by the exi gencies of li tigation. To the extent it is not, the evi dence is likely to be not only incorrect but also self-defeating'.
CONFLICT OF INTEREST The 'other' side in any forensic case may also wish to speak with you before the trial. Answer all the questions asked by them truthfully and honestly, and be willing to admit, if required , that certain aspects of the case had not been pre viously familiar to you and that on the basis of this fresh information there may be scope for revising the opinions already offered. The expert may find that he may thu s inadvertently be sucked into a conflict of interests l 5 . 16 between the parties involved. Perhaps the earthy statement of the Unites States Supreme court in Williams v. Florida should be kept in mind in this respect: 'The adversary sys tem of trial is hardly an end in itself; it is not a poker game in which the players enjoy an absolute right always to con ceal their cards until played '. In some other instances, there could be a more genuine conflict of interests in the m,ind of the expert, for example when giving evidence in a child abuse case knowing fuJJ well that the evidence may point to a member of the child's family, also a patient of the practice, and that there may be a disruption of that family unit or an unfavourable divorce settlement. The expert, thus, comes to lie on the horns of a clilemma; should his allegiance to the child override loy alty to other patients - the parents? The child's welfare may have to override obligations of medical confidentiality to the child (and his family) in the interests of public safety.
RULES OF EVIDENCE In the courtroom certain evidence can be excluded in certain well-defined circumstances. The policy reasons for such exclusions are basically ones in which confusion and
unfairness can be avoided; this will also prevent the judge or jury, especially the latter, from falling into error. Over the years, the rules of evidence have been relaxed; in Scotland, it was only in 1898 that the accused could give evidence before thejury trying him. In this respect, doctors often utilize infor mation obtained at second hand from others in coming to their decisions, i.e. hearsay evidence. As a generality the court will accept only the 'best evidence'. What constitutes hearsay is a legal matter to be determin ed by the judge.
CONCLUSION Perhaps the classic statement of Lord President Cooper as found in Davie v. Magistrates of Edinburgh sums up best the scope of expert evidence: Expert witnesses, however skilled or eminent, can give no more than evidence. They cannot usurp the func tions of the jury or judge sitting as a jury, any more than a technical assessor can substitute his advice for the judgement of the court ... Their duty is to furnish the judge or jury with the necessary scientific criteria for testing the accuracy of their conclusions, so as to enable the judge or jury to form their own independent judgement by application of these criteria to the facts proven in evidence. The scientific opinion evidence, if intelligibl e, convincing and tested, becomes a fa ctor (and often an important factor) for consideration along with the whole other evidence in the case, but the deci sion is for the judge or jury. In particular, the bare ipse dixit of a scientist, however eminent, upon a contro versial issue, will normally carry very little weight, for it cannot be tested by cross exa mination nor independ ently appraised, and the parties have invited the deci sion of a judicial tribunal and not an oracular pronouncement of an expert.
RECENT DEVELOPMENTS In Kariing v. Purdue (2004 SLT 1067), the Scottish courts were asked to consider a claim for damages against an expert wit ness. Mr Karling had been convicted of murder in 1995. Dr Basil Purdue was the defence expert asked to carry out a post-mortem examination of the victim. Dr Purdue's findings were consistent with those of the prosecution case. Mr Kar ling was convicted and sentenced to life imprisonment. In 200], fresh scientific evidence came to light indicating that there was no sound evidence on which to conclude that the victim had been suffocated. In an appeal, Mr KarJing's con viction was quashed. Subsequently, Mr Karling issued civil proceedings against the expert. He sought damages of £75000 on the grounds that, had the expert performed the post-mortem examination with sufficient care and compe tence, there would have been a reasonable chance that he wo uld have been acquitted of the charge. Purdue defended
References I
the actio n on the grounds of his abso lute immunity from suit. He argued that, in the absence of malice, no li ability could be attached to him, regardless of how incompetently th e post mOliem or the subsequent report was carried out. Mr Karling's respon se to this pl eading was that, by indi cating in his rep Oli that he was attaching his conclusions sepa rately, Dr Purdue was effectively adm itting that part of his report was intended for Mr Karling's le ga l team only. He argued further that this suggested that the expeli was act ing as an advisor, therefore owing Mr Karling a duty of care that was capable of standing separately from Dr Pur due's role as expert witness. In rejecting these arguments, the co urt said that the link in tim e and function with the criminal proceedings was more than suffici ently close to conclude that, in relation to his engagement to perform professional or expert services, Dr Purdue was immune from suit in Mr Karling's instance. The test imposed by the courts was whether the expert was preparing to give evi dence in proceedings. It was the proximity of th e criminal proceedings that, in effect, made Dr Purdu e's 'advisory' capacity indistinguishable from the work done in the pro vision of services as an expert witness. Another more recent impoltant development in the field of expert evidence was the decision of the Court of Appeal in ThE GEn eral Medical Council v. ProfEssor Sir Roy Meadow (Her Majesty's Attorney Generol iJ1ter[Jel1ing) (GMC v. Meadow) . In November 1999 , Sally Cl ark was tried for the murder of her two sons. Th e Crown relied. in palt, upon Professor Meadow's evidence to refute the proposi ti on that Mrs Cl ark 's children may have died fro m sudden infant dea th syndrome (SIDS), or co t death. Professor Sir Roy Meadow ventured into the field of statistics in givin g hi s opinion on the likelihood of occurrence of a second cot dea th in the same family; this invo lvement, sa id to have been erroneous and outs ide his field of expertise. resulted in proceedings before the fi tn ess to practise panel (FPP) of the GMC, with Professor Mead ow being found gUilty of serious professional misconduct and res ultin g in his era sure from the medical regis ter. Professor Meadow appealed to the High Court, whe re his appea l was a llo wed by Collins J. The finding of serious professional misconduct was qu ashed . Additionally, the judge commented that expel1 witnesses, by their status, enjoyed not merely immunity from civil suit, but also a wide, albei t not absolute, immu nity from disciplinary, regulatory or fitness to practi se pro ceed ings in relation to statements or ev id ence given by them in or for the purpose of legal proceedings. These com ments resulted in the Attorney General's interventi on to the COUlt of Appea l; it held unanimously that the immu nity identified by the judge did not exist.
REFERENCES Di ckson WG. A Treiltise all the Law of Eliidel1ce ill Scorlaml, 3rd edn. Edinburgh, UK: Scottish Law Commiss ion, No. 1887, 137 .4, paras 397 el sel].
2 3
4 5 6 7 8 9 10 1J 12 13
14
15
16
455
Lav in JH. Ev elyon e believed the pl ainti ff except his doctor. Med Ecol/ 1991; 68:34-4 1. Kaufmann HH. Th e expert witn ess. Neither Frye nor Daubert so lved th e prob lem: what ca n be don e? Sci Jmtice 200 1; 41 :7-20. Calvert-Smith D. Medica l evidence in criminal prosecutions. l'vI ed Leg J 2001 : 68 :11 7-2 9. Ca lvert- Smi th D. Medi cal evidence in criminal prosecutions. Med Leg J 2000: 68:117. Starrs JE. In the lan d of agog : an allegory for the exp ert witn ess. J forellsic Sci 1985; 30:289-307. Annas GJ. Scientific ev iden ce in the court room; the death of the FI")'e rule . N £lI gl J Med 1994: 330 : 101 8- 2 1. Mi tch ell J. Scientific evi den ce: pri nciple or pragmatism . Aied Leg J 200 1: 69: 117- 19. Ed itorial - Osteo genesis imperfecta, non-accidental injury and temporary brittle bo ne disease . Arch Dis Child 1995: 70:169. Brandon S. Recovered memories; some aspects of the controversy. Med Leg 199 9; 67: 25 -34. Brahams D. 'Repressed memories' and the law. La ncet 2000; 356:358. Katz J. The fallacy of the impart expert rev isi ted . Bull Am Aca d PS),ciliatr Law J9 92 ; 20:141-5 2. FacUo D. Buccia relli R. Peer review of the expe rt wi tness: an opportunity to improve our medical liability system. J Child Neu rol 1995: 10:403-4. Fisher CWo Dombrowski MP, Jaszczak SE et al. Th e expert witness: reo l issues and suggest ions. Am J Ob ster GYl1ecol 1995: J 72: 1792-800. Dona ldso n L. Kap lon C, Leung W-c, The medica l expert witness : time to regulate conflicts of interes t. Med Sci Law 1999 : 39 :1 1- 16. Ma narin B. Assess in g the expen:: a ca ll for reciprocal di sc lo sure in Canada. Med Sci Law 1999 ; 39: 17 -22.
Further Reading CASES • Nariollal Justice Campania Nau iera SA V. Prudelltial Assurance Co. Ltd, rlre Ikar iall Reefer (1993) 2. L1 0yds Reports 68, 8 1-82 and (19 95) 1. L1 0yds Reports 45 5. • R V. Srockwell (1993) 97 CR App R 260. • R v. Uljoh Et Orhers (2001) EWCA Crim 1381. • Re H an d E (Minors) CO Ult of Ap pea l (1996), 21 May. • R V. BO II,\'rllOli (1984) 38 SASR 45. • flye V. United Srates. 54 App. D.D. 46, 293 F. 1013 . 1923. • Dauberr v. iVlelTeli Do /U Phollllll ceu ricilis Inc (1993) 113 S Ct 2786. • KUlllbo Tire Co Lrd V. Carmichael (1999) 526 US 13 7. • R v. Robb 93 Cr App R 161. • Re AB (cliild abuse; expert lVili/esses) (1995) 1 FLR 181. • Daniels v. Walker (2 000) 1 WlR 1382, CA. • A ssess or fo r Lothian Region v. Wilso11 (1979) SC 34 1 lord Avonside at 349. • Rollalld v. Lothian Health Board, 27th August 1981 - unreported. • McVey v. Central Regiona l Co ull cil, 4th Ma rch 19 93 - unreported. • White/lOuse v. Jordan (198 1) 1 WlR 246. • Williams V. Florida (1970) 399, US 78. • Davie V. Magistrates of Edillburgh (1953) SC 34 at 40. • Karling v. Purdue (2004) SLT 1067 Coun of Session, Edinburgh. • Tire Gelleral iVIed ical Council V. Professor Sir Roy Meadow (Her Majes t."·s Artomey General ilitemel1iIJg) (GMC v. Meadow) Neutral Orar ion Nu mb er: (2006) EWCA Civ 13 90 Case No: CO/5763/2005.
I
I
APPENDIXA
CHILD PROTECTION
EXAMINATION FORMS
Form 1: Examination consent form CHILD PROTECTION EXAMINATIONS
CONSENT FORM
Child's name ..... .... ......... ... .. ............ ... .. .. ..... ..... ....... ........ .... .. ...... .. ...... .... ..... .. ...... .... .OOB ...... .. ... .... .. ...... .. ........... . Address ... ............ ... ... ......................... .. .......... ... ... ......... ... ........... .. ................ ............ ... ... ....... .. ... ............... ...... ..... . Unit no ..... ..... .... .... .. ..... ...... ... ...... ...... .. .... ......... ....... .. .. ... ....... ......... .. .... ..... .. .............. ... ....... .. ...... .. ........... .. .......... . Permission must be obtained from parent(s) or other(s) with responsibility for the child, and from the child where appropriate.
I give permission for: 1. Medical examination 2 . Collection of specimens for laboratory tests 3. Photography of clinical findings 4. CO of genital findings
Yes Yes Yes Yes
N/A N/A N/A N/A
No No No No
Photographs and CDs will be stored as part of the clinical records. They may be used to support clinical evidence of injury and may need to be shared with another Doctor involved in any court proceedings.
I give permission for photographs and CDs to be used to support clinical evidence in court proceedings. Yes No N/A Photographs and CD recordings can be used for teaching and training of other professionals working in Child Protection proceedings. Photographs and CDs used for this purpose are anonymized.
I give permission for anonymized photographs and CD recordings of my child to be used for teaching Yes No N/A and training of other professionals. I understand that the information from the medical examination will be shared with:
Social services Police GP Health visitor/school nurse
Verbal report Yes/ No Yes/No Yes/ No Yes/No
Letter Yes/No Yes/No Yes/No Yes/No
The procedure has been fully explained to me and I understand that I have the right to withdraw my con sent at any stage during the procedure. Signed .......... ........ ....... ..... ........ ....... ..... .. ... ...... ....... ..... .. .. ........ .. ...... .. ... ... ............ Date ........ ... ...... .. ... .. .. .. .. ......... ... ..
Name ............. ...... .... ................ ..... ..... ..... .... ....... ...... ... ....... .. .... .... ....... ..... .......... ..Parent/carer/professional
Examining doctor(s) ... .... ........ .. ..... .. ......... .. .............................................. .. .. ......... ... ............ .. ... .. ...... ... ............. .... ..
Signature .... ....... ..... ....... ... ............ ... ......... ...... .. ......... .... .... ...... ............... ...... ....... Oate .......... .... ....................... .... .. .
458 I
Appe ndix A: Child protection examin ation forms
Form 2: Examination form Confidential Medical Information
EDINBURGH AND LOTHIANS
CHILD PROTECTION OFFICE
Joint Paediatric/Forensic Examination/
Special ist Paediatric Examination of
Child who may have been Abused or Neglected
Name of examiner(s)
Place of examination
J
Child 's surname
Forename(s)
Known as Date of bi rth
Address
Sex CHI no. Date of examination
GP Address
Time of examination
Child accompanied to clinic by (please tick) Mother
D
Father
D
sw
D
Police
D
Other
D
School/nursery attended
BACKGROUND INFORMATION (e.g. previous concerns reo developmental delay, poor growth, possible episodes of NAI)
Family/social history N
o o
N
.c o U
o ~--------------------------------------------------------------------~
Page 1 of 12
1471
;?
Appe ndix A: Child protection exam ination forms I
459
ACCOUNT OF CIRCUMSTANCES LEADING TO REFERRAL
(a)
From referrer : Name:
(b)
Position:
From accompanying adult: Position :
Name:
(c)
From child:
Child 's terminology
In the presence of:
Breasts Vagina Anus Urethra Abdomen Semen
-
Penis Scrotum
CONSENT
(source, i.e . parent, young person, person holding parental righ ts) Date:
N
o o
N
Witnessed by:
Q)
.D
.8 u
Position
Name
o
L _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _~~ o
Page 2 of 12
1471
460 I
Append ix A: Child protect ion exa mination forms
PRESENTING SYMPTOMS
Has the child exhibited any behavioural symptoms:
Masturbation
D
Secondary wetting Soiling
D D
Aggression/anger Conduct disorder
D D
Sexualized play Other
D D
SYSTEMATIC ENQUIRY
N
o o
N
Q)
(Genitourinary enquiry) see over
.0
o
t5
o
~--------------------------------------------------------------------------------~~ Page 3 of 12
1471
Appendix A: Child protection examination forms I
461
GASTROINTESTINAL/GENITOURINARY ENQUIRY
Comment Frequency
Bowels
Consi stency of stool Pain on defaecation
Yes/No
Bleeding on defaecation/blood on wiping
Yes/No
Con stipation
Yes/No
Soiling
Yes/No
Mucus
Yes/No
Other
Yes/No
Previous UTI
Urinary tract
Genital
-
(suspected)
Yes/No
(con firmed )
Yes/No
Frequency
Yes/No
Dysuria
Yes/No
Urgency
Yes/No
Wetting - day/night
Yes/No
Other
Yes/No
Bleeding/blood on pants
Yes/No
Di scharge
Yes/No
Itch
Yes/No
Rash/inflammation
Yes/No
Circumcised
Yes/No
Menarche age Menstrual cycle Date of last period
Routine sanitary protection
Tampon/towel/both N
o o
N
Q; o
.0
ti
o o
~
Page 4 of 12
1471
462 I
Appendix A: Child protection examination forms
CLINICAL EXAMINATION People in attendance during examination
General physical appearance of child (note especially any evidence of infection, neglect or injury)
Demeanour/behaviour/impression of maturational status
Measurements Weight
kg
centile
Height
cm
centile
Head circumference
cm
centile
Findings on externa l physical examination (Tick appropriate boxes and where possible record findings. Omit if Preliminary Medical Assessment Form has already been completed.) Comment (a)
Skin and hair
(b)
Teeth
(c)
Eyes Ears, nose and throat
(d) (e)
-
Cardiovascular system
(f)
Blood pressure (if applicable)
(g)
Respiratory system
(h)
Alimentary system
(i)
Genitalia/testes
(j)
Nervous system
(k)
Locomotion/posture
Ig
(I)
Other
I~
Q;
.c 0 0 0 0 ~
Page 5 of 12
1471
--~.
Appendix A: Child protection examination forms I
463
.....-- ,,---- ®' I@ ~ l.\
.....c:::::'..: "4 ,---,
C\l
o
o
C\l
Q; .0
-----------------------------------p~~~~~--------------------------------~~
...
Page 6 of 12
1471 .
464 I
Appendix A: Child protection examination forms
(\J
o o(\J
Qj .0
o
o o ~--------------------------------------------------------------------~
Page 7 of 12
----- - --
-
-------------------------
1471
o
Appen dix A: Ch ild prot ecti on examinati on forms I
465
'"-'
N
o o
N
Q; o
.0
t5
o
~------------------------------------------------------------------------------~~ Page 8 of 12
1471
466 I
Appendix A: Child protection examination forms
Mons pubis Prepuce of clitoris Clitoris Frenulum of clitoris Orifice of urethra
\
Labium majus (labia majora)
\
I
Vestibule (area between labia minora)
I
)
Labium minus (labia minora)
/
Hymen (covering introitus of vagina) Vest ibular fossa Fourchette Posterior commissure
Genital examination Position Method Magnification __________
Video
Yes/No
Identification:
Video no.
I
Serial no.
.'.~:- ' J C\J
o o
C\J
OJ .D
o tl
o o
Page 9 of 12
1471
--
-
---
Appendix A: Chi ld protection examination forms I
467
(
u\ Genital examination
Position Method Magnification Video
Yes/No
Identification: Video no. Serial no.
Examination of anus
Position Method Magnification Video
u
Yes/No
Identification: Video no. Serial no.
(\J
o o
(\J
Q; .D
.2 u
o
o ~--------------------------------------------------------------------------------~~
Page 10 of 12
1471
468 I
Appendi x A: Child protection examin ation forms
INVESTIGATIONS
D
Photography
Taken by: Police: Processed at: Medical staff: Identifier:
D Radiographs D Coagulation profile
Routine specialist:
D D
Serum HIV/hepatitis Swabs
Site:
D Urine D Pregnancy test
D Other
CONCLUSION/OPINION
TREATMENT/ACTION REQUIRED
C\J
o o
C\J
Q; o
.D
t5
o
o ~------------------------------------------------------------------------------~~
Page 11 of 12
1471
Appendix A: Child protection examination forms I
469
INFORMATION SHARING Time Parent
PresenUcontacUno contact
Social worker
PresenUcontacUno contact
Police
Present/contacUno contact
General practitioner
ContacUno contact
HV/school nurse
ContacUno contact
Name
Present tel. no.
AGREED PLAN 1. Further investigation of possible abuse. 2.
Need for further assessment of medical/developmental problems. (NB: If child is looked after/accommodated, Part 2 of the 'Initial Medical Assessment' will be required.)
3. Other action required: Circle
as appropriate
1. 2. 3.
4. 5. 6.
7.
4. Report sent to: Circle as
appropriate
1. 2. 3.
4. 5.
Admit to hospital Refer for Child Protection Order Refer for Child Assessment Order Refer to Reporter
Discuss with Procurator Fiscal Contact GP Other (specify)
General practitioner Social worker Reporter
Procurator fiscal Hospital records
Signed Name in block letters
Date Designation
Signed Name in block letters
Time completed
Date Designation
Time completed
AUDIT
(\J
o o
(\J
Q)
.a o tl
o
o ~--------------------------------------------------------------------------------~~
Page 12 of 12
1471
Table 1
Fetal autopsy standards at
Age (GW) Maceration"
12
0-3
0-3
0-3
0-3
0-1
(em)
(em)
(g)
(g)
(g)
2 (g)
3 (g)
0-1
2-3
(g)
(g)
Spleen
0-3 (g)
0-1
2
3
(g)
(g)
(g)
0-1 (g)
Kidneys
Adrenals
2-3
0-1
2-3
0-1
2-3
(g)
(g)
(g)
(g)
(g)
0.11
29.6
4.8
1.5
1.4
1.3
0.6
0.9
0. 10
0.D3
0.01
0 .25
0.19
0.04
74.9
7.4
7.2
7.2
7.2
0.9
0.9
0. 14
0.06
0.02
0.15
0.75
0.78
0.18
12
8.7
11.8
8.5
37.4
6.5
2.0
1.7
1.7
1.2
1.2
0.20
0.04
0.02 003
0.08 0.03
0.3
0.2
0.17
0.17
01
0.1
0.18
0.18
0.Q4
0.14
0.4
0.3
0.3
0.2
~
to r m
3
7.2
7.8
7.2
74.9
7.4
7.2
7.2
7.2
0.9
0.9
014
006
15
9.9
13.7
9.8
53.0
9.1
2.9
2.4
2.3
2.0
1.5
0.3
0.05
3
7.2
7.8
7.2
74.9
2.5
7.2
7.2
7.2
0.9
0.9
0.1
0.06
0.06
0.06
0.04
0.04
0.7
0.7
02
0.2
Mean
18
11 .1
15.6
11 .1
76.5
12.7
4.2
3.3
3.2
2.9
2.1
0.5
0.07
0.08
0.06
0.06
0.17
0.6
0.5
0.5
0.3
SO
3 21 3 24
7.2
7.8
7.2
78.5
3.9
7.2
7.2
7.2
0.9
0.9
07
0.06
0.06
0.06
0.06
0.06
0.3
0.3
02
0.2
Vl
12.4
17.5
12.4
17.3
5.9
4.5
4.2
3.9
2.7
0.8
0.11
0. 12
0.09
0.09
0.17
0.9
0.8
0.6
0.4
~~
Mean Mean Mean Mean Mean
SO Mean
SO Mean
SO 23
Mean
24
Mean
SO SO Mean
SO Mean
SO 27
0-3
Heart Thymus
7.7
SO
26
0-3
(mm) (em)
Lungs
7. 1
SO
25
Liver
7.7
SO
22
Brain
9.8
SO
21
Body
7.7
SO
20
HOC
7.4
SO
19
CH
3
Mean
18
CR
9
14
17
FL
Mean Mean
16
weeks of gestation'
SO 13
15
72-27
Mean
SO
3 27 3 30 3 33 3 36 3 39 3 41
7.3
108
007
0.05
7.8
7.3
47
5.4
7.5
7.5
7.5
7.2
1.2
0.2
0.06
0.06
0.06
0.08
0.08
0.4
04
0.3
0.3
19.3
13.6
147
22.9
8. 1
6.1
5.4
5.1
3.5
1.0
0.18
0.18
0.12
0.13
0.16
1.3
1.1
0.8
0.5
7.3
7.9
7.3
53
6.9
3.0
3.0
3.0
7.7
1.7
0.4
0.06
0.06
0.06
0.12
0.72
06
04
04
14.7
21.1
14.8
194
29.4
10.7
7.9
6.8
6.4
4.4
1.4
0.3
0.3
0.2
0. 19
0.15
06 1.8
1.5
1.0
0.7
13.5
7.3
7.9
7.3
65
8.4
4.5
4.5
4.5
2.3
2.3
0.5
0.2
0.2
02
0.17
0.17
0.8
0.8
0.4
0.4
15.9
22.9
16.0
249
37.0
13.8
10. 1
8.4
7.9
5.4
1.7
0.4
0.4
0.3
0.3
0.15
2.4
2.0
1.2
0.8
7.3
7.9
7.3
78
9.8
6.0
6.0
6.0
2.8
2.8
0.7
0.3
0.3
0.3
0.2
0.22
1.0
1.0
0.5
05
17.0
24.6
17.2
312
45.5
17.2
12.5
10.2
9.5
6.5
2.1
0.6
0.5
0.3
0.4
0.17
3.0
2.5
1.4
1.0
7.4
7.9
7.4
92
77.3
7.5
7.5
7.5
3.4
3.4
08
0.4
0.4
04
0.3
0.29
1.2
1.2
0.6
0.6
18.2
26.3
18.3
382
55.0
21.1
15.2
12.3
11 .2
7.8
2.6
0.8
0.7
0.4
0.5
0.22
3.8
3.1
1.7
1.2
7.4
7.4
707
728
9.0
9.0
9.0
4.0
4.0
1.0
0.5
0.5
0.5
0.4
0.36
7.4
7.4
0.7
0.7
19.3
20 28.0
19.4
461
65.4
25.5
18.2
14.5
13.1
9.2
3.1
1.0
0.9
0.6
0.7
0.3
4.6
3.8
1.9
1.4
7.4
2.0
7.4
722
74.3
70.4
70.4
70.4
4.6
4.6
1.7
0.6
0.6
0.6
0.4
0.4
1.6
1.6
20.4
29.6
20.5
547
76.9
30.2
21.6
16.9
15. 1
10.7
3.6
1.3
1.1
0.7
0.9
0.4
5.5
4.6
08 2.2
08 1.6
0.8
0.8
08
0.5
0.5
7.9
1.9
08
0.8
4
7.5
2.0
7.4
722
75.8
77.9
77.9
77.9
5.3
5.3
1.3
44
21.5
31.2
21.6
641
89.3
35.4
25.2
19.5
17.3
12.4
4.2
1.6
1.3
0.8
1.1
0.6
6.5
5.5
2.5
1.8
4
7.5
2.0
7.5
737
77. 2
73.4
73.4
73.4
5.9
5.9
7.4
09
0.9
0.9
0.6
0.6
2.1
2.1
09
09
47
22.6
32.8
22.6
743
103
41.1
29.1
22.3
19.6
14.1
4.9
1.9
1.6
1.0
1.4
0.8
7.6
6.4
2.8
2.0
4
7.5
2.7
7.5
754
79
74.9
74.9
74.9
6.6
6.6
1.6
1.1
1.1
1.1
0.7
0.7
2.4
2.4
7.0
1.0
50
23.6
34.3
23.6
853
11 7
47.1
33.4
25.3
22.0
16.0
1.9
1.2
1.7
1.1
8.8
7.4
3.1
2.3
7.5
2.7
7.5
77l
20
76.4
76.4
76.4
7. 3
7.3
5.6 1.7
2.3
4
7.2
7.2
7.2
0.9
0.9
2.7
2.7
1.1
1.1
52
24.7
35.8
24.5
971
133
53.6
37.9
28.6
24.6
18.0
6.3
2.6
2.2
1.4
2. 1
1.4
10.1
8.4
3.4
2.5
4
7.6
2.7
7.5
788
22
77.9
77.9
77.9
8.0
8.0
1.8
1.4
1.4
1.4
1.0
1.0
3.0
3.0
1.2
7.2
CH, crown-heel lengt h; CR, crow n-rump length ; FL, foot length ; GW, weeks of gestation; HOC, head circumference; SO, standard devi ation. "0, none; 1, mild; 2, moderate; 3, marked.
Vl
o '"T"j
tJ~
»m
b6 I ---'
~><
mto
» Vl
c
~
~
m
Z
--l Vl
"" -.j I\)
Table 2
Fetal autapsy standards at 28-43 weeks af gestatian 1
Age (GW)
»
FL
CR
CH
HDC
Body
Brain
Liver
0-3 (mm)
0-3 (em)
0-3 (em)
0-3 (em)
0-3 (g)
0-3 (g)
0-1 (g)
Lungs
Heart
Thymus
Kidneys
Spleen
"0 "0 C1l
Adrenals
:::J Q
Maceration'
2 (g)
3
0-1
2-3
0-3
0-1
2
3
0-1
2-3
0-1
2-3
0-1
2-3
(g)
(g)
(g)
(g)
(g)
(g)
(g)
(g)
(g)
(g)
(g)
(g)
(g)
x'
rn n;i IT
C1l (fl
28
Mean
SO 29
Mean
SO 30
Mean
SO 31
Mean
SO 32
Mean
SO 33
Mean
SO 34
Mean
SO 35
Mean
SO 36
Mean
SO 37
Mean
SO 38
Mean
39
Mean
40
Mean
SO SO SO 41
Mean
SO 42
Mean
SO 43
Mean
SO
55
25.7
37.3
25.5
1096
149
60.6
42.7
32.0
27.4
20.2
7.1
3.1
2.5
1.6
2.5
1.8
11.4
9.6
3.7
2.8
S
4
1.6
2.2
1.6
206
23
19.3
19.3
19.3
8.7
8.7
2.0
1.6
1.6
1.6
1.1
1.1
3.3
3.3
1.3
1.3
ill
57
26.7
38.7
26.4
1230
166
67.9
47.8
35.6
30.2
22.5
7.9
3.5
2.9
1.8
3.0
2.2
12.9
10.8
4.1
3.1
en
:::J Q
Cll
4
1.6
2.2
1.6
225
25
20.8
20.8
20.8
9.5
9.5
2.1
1.8
7.8
1.8
1.3
1.3
3.6
3.6
1.4
1.4
0.
60
27.7
40.1
27.2
1371
185
75.7
53.3
39.4
33.2
24.9
8.7
4.0
3.3
2.1
3.6
2.7
14.4
12.1
4.5
3.4
:3
4
1.6
2.2
7.6
244
26
1.4
(fl
28.1
22.3
22.3
22.3
10.2
10.2
2.3
2.1
2.1
2.1
1.4
1.4
3.9
3.9
1.4
C
62
28.7
41.4
1520
204
83.9
59.0
43.4
36.3
27.4
9.6
4.5
3.7
2.3
4.2
3.3
16.0
13.4
4.8
3.8
4
1.7
2.2
1.7
264
28
23.8
23.8
23.8
lJ.0
lJ.0
2.4
2.3
2.3
2.3
1.6
1.6
4.3
4.3
1.5
1.5
64
29.7
42.8
28.9
1677
224
92.6
65.0
47.6
39.6
30.0
10.6
5.0
4.2
2.6
4.8
3.9
17.7
14.9
5.2
4.1
4
1.7
2.3
1.7
285
29
67
30.6
44.0
29.7
1842
245
4
1.7
2.3
1.7
306
31
69
31.6
45.3
30.5
2015
268
4 71
25.3 102 27 111
25.3
25.3
lJ.8
ll.8
2.6
2.5
2.5
2.5
1.8
1.8
4.6
4.6
1.6
1.6
71.3
52.1
43.0
32.8
11.6
5.6
4.6
2.9
5.5
4.5
19.5
16.4
5.6
4.5
26.7
26.7
12.6
12.6
2.7
2.8
2.8
2.8
1.9
7.9
5.0
5.0
1.7
1.7
77.9
56.7
46.6
35.7
12.6
6.2
5.1
3.2
6.3
5.2
21.4
18.0
6.0
4.8
1.8
2.3
1.7
328
32
28
28.2
28.2
13.5
13.5
2.9
3.1
3.1
3.1
2.1
2.1
5.4
5.4
1.8
1.8
46.5
31.2
2195
291
121
84.8
61.5
50.3
38.7
13.7
6.9
5.7
3.5
7.2
6.0
23.3
19.6
6.5
5.2
5
1.8
2.3
1.8
350
33
30
29.7
29.7
14.3
14.3
3.0
3.3
3.3
3.3
2.3
2.3
5.8
5.8
1.9
1.9
73
33.4
47.7
31.9
2383
315
132
92.1
66.5
54.1
41.9
14.8
7.5
6.2
3.8
8.1
6.7
25.4
21.4
6.9
5.6
5
1.8
2.4
1.8
373
35
31
34.3
48.9
32.6
2580
340
142
31.2 100
37.2
15.2
15.2
3.2
3.6
3.6
3.6
2.5
2.5
6.2
6.2
2.0
2.0
71.7
58.1
45.1
16.0
8.2
6.8
4.2
9.1
7.5
27.5
23.2
7.4
6.0
5
1.8
2.4
7.8
397
36
33
33
32.7
16.1
16.1
3.3
3.9
3.9
3.9
2.7
2.7
6.6
6.6
2.1
2.1
78
35.2
50.0
33.2
2784
366
154
107
77.2
62.2
48.5
17.2
8.9
7.4
3.9
10.1
8.3
29.8
25.0
7.8
6.5
5
1.9
80
36.1
2.4 51.1
7.8
421
38
34
34
34.2
17.0
17.0
3.4
4.2
4.2
4.2
3.0
3.0
7.1
7.J
2.2
2.2
33.8
2996
394
165
116
82.8
66.5
52.1
18.5
9.7
8.0
5.0
11.2
9.1
32.1
27.0
8.3
6.9
5
1.9
2.4
7.9
446
39
36
36
35.6
18.0
18.0
3.6
4.6
4.6
4.6
3.2
3.2
7.5
7.5
2.3
2.3
82
37.0
52.1
34.4
3215
422
177
124
88.6
70.9
55.7
19.8
10.5
8.6
5.4
12.4
9.9
34.5
29.0
8.8
7.4
5
1.9
2.5
7.9
471
41
37
37
37.1
18.9
18.9
3.7
4.9
4.9
4.9
3.4
3.4
8.0
8.0
2.4
2.4
84
37.8
53.1
35.0
3443
451
190
133
94.6
75.4
59.5
21.2
11.3
9.3
5.8
13.7
10.7
37.0
31.1
9.3
7.9
5
7.9
2.5
1.9
497
42
39
39
86
38.6
54.1
35.5
3678
481
203
142
38.6 101
19.9
19.9
3.9
5.3
5.3
5.3
3.7
3.7
8.4
8.4
2.5
2.5
80.1
63.4
22.5
12.2
10.0
6.2
15.0
11.5
39.6
33.3
9.9
8.4
5
2.0
2.5
2.0
524
44
40
40
40
20.9
20.9
4.0
5.6
5.6
5.6
4.0
4.0
8.9
8.9
2.6
2.6
88
39.4
55.0
36.0
3922
512
216
151
107
84.9
67.4
24.0
13.1
10.7
6.6
16.4
12.2
42.2
35.5
10.4
8.9
5
2.0
2.5
2.0
551
45
42
42
42
27.9
21.9
4.2
6.0
6.0
6.0
4.2
4.2
9.4
9.4
2.7
2.7
CH, crown-heel length; CR, crown-rump length; FL, foot length; GW, weeks of gestation; HDC, head circumference; SD, standard deviation. '0, none; 1, mild; 2, moderate; 3, marked.
CD
:3 C1l
:::J
32.5
76
C1l
Cll
u;
Appendix B: Tables of sta ndard measu rements I
Table 3
Mean weights and percentiles for singleton placentas 2 No. of cases
Gestational
90th
75th
Mean singleton
25th
10th
age (weeks)
percentile
percentile
placental wt.
percentile
percentile
172 191 211 233 256 280 305 331 357 384 411 438 464 491 516 542 566 589 611 632 651
158 175 193 212 233 255 278 302 327 352 377 403 428 453 477 501 524 547 567 587 605
143 157 172 189 208 227 248 270 293 316 340 364 387 411 434 457 478 499 519 537 553
128 138 151 166 182 200 219 238 259 281 303 325 347 369 391 412 432 452 470 487 502
114 122 133 145 159 175 192 210 229 249 269 290 311 331 352 372 391 409 426 442 456
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
473
700,-----r----,-----,----,-----r----.----~----r_--
3 6 7 9 19 14 9 16 11 12 14 24 30 32 44 36 32 62 103 193 87
__----~----~ 90th percentile
600~----r----+----~----+---~~---+----~----~--_4~~~~--~
75th percentile Mean singleton
500 ~----r----+----~----+---~-----+----~~_=~--~~~~~--~
25th percentile 10th percentile
200~--~~~~~~~~~q-----~---+----~----+_--~----_+-----
100~----+----1----~----+-----~---+----~----+_--~----~----~
22
24
26
28
30
32
Ges tat ion [weeks) Singleton placental weights and ranges.
34
36
38
40
474 I
Appendix B: Tables of standard measurements
Table 4
Umbilical cord length from 20 to 47 weeks' gestation 3
Gestational age (weeks)
n
16 27
20-21 22-23 24-25 26- 27 28- 29 30-31 32-33 34-35
38 59 80 11 3 337 857 3153 10083 13 841 4797
36-37 38-39 40-41 42-43 44-45 46-47
1450 492
Umbilical cord length (cm)
32.4 :!: 36.4 :!: 40.1 :!: 42.5:!: 45.0 :!: 47.6:!: 50.2 :!: 52.5 :!:
8.6 9.0 10.1 11 .3 9.7 11.3 12.1 11.2
55.6 :!: 57.4 :!: 59.6 :!: 60.3 :!: 60.4 :!: 60.5:!:
12.6 12.6 12.6 12 .7 12.7 13.0
Total no. of cases 35779. Data represent mean ::': 1 SD. n = no. of cases.
----
--------
=-~-:~
~
-
-+
Appendix B: Tables of standard measurements I
Body weight, height and head circumference, boys, birth to 1 year. (Courtesy of the Child Growth Foundation. Printed by Harlow Printing Ltd, Maxwell St, South Shields, NE33 4PU.j
Table 5
O-lyr
I
II
NAME ..... ..................... ................ ..
D.O .B. ..... .I ...... ! ...... Correct by......... ..... for prematurity
u
. 3'2 34 3,6 3'8 46
.
45
2
1
6
J
1~ . ·
'/l
I
38
::
I I
I
I
I
1.-- -1
I
i.
'
8
I rfj}
I
---l1:=.:
f i'
I
I .•
I
I ;
(
/
" f/1)1-~'
/!.y
I
t IYI/
/
t~ /, 1 /{1 ;1/
I
II
I 51
99.•• , "
50 49
98 th .1SI
-1
:~: -, :~
'~_ 2:: -I'
~.
.' I
46 45
2nd -
.
, ; . -......1"
m '
,
II __ - :: :: - I·
.
8 5 0 52
I j -1 -llJ-1 ____
•
'1 j 1 ,~-1 ___ . ~/ I ;--<~1 ::r:l> --. 11 j / .
8 10 12
l'
,
II
40
,
'7
<
pre-term
39
I
)-
:; 111 41
32 31 3'6 3'814'0 4 2 -~4 46
28 3
" 84
I
82
99 .• ' b
II, : I .-1---1
"I
. '
---- .
_1~1~1--' 9;1h
80
:i ::1-:1 ~;:~ ~ ;~:~ :: :: IVt)jfl ./r~,~ rl~: ' ' Iii J J.3 1:1 ~ -i:~~f __ ~~;: ~ L1- -~ ___~-
j/! / I :,
I' "
32 31 30 29 28 27 26
.
I' t
/1 /'
11 1 1
I I
;.-- ~
""......,
, , '
'I
5lt
--1-
0.4,"
68 66 64 62
I
y
5~;S1' ;,--- , , ):" ./ /~ -.--l l4~ ' /1/ '1/ y j , I
--1-
1/1; t.-V: .: : ; . _> >-_:'--/--1-= - .1 t /-:--1-::'//1: ,'''',::_ --,=1--'-' I t;1.< . !.-' --I -J -] _' J -1/ Cf/' j l,. '
/t/
'
24
I
!I
-'
I .(1 f/. «( ~ I, / / I , / f ()--, 1· / / -. 1/ '/ (" --
/
25
/V'1/ '
I
I I
60
.
58 14
I 13
9 9.6 1b.
..(
98 tb
12
~ 1;-(
Slit·
:~ I
11 7Slh
10
50 t b
-
10
~5th
9
9 9th -~-
81'
-'
71
61
2 nd
.- .
8
O.4tb
7
_i ...-' pre-term
6
5
5
3~ 32 3 4 36 38 40 42~4 46 + 8 50 52 8
28
3 2
D a Le
9
Ag t.:
*
~ tcasllrem("n1
~ ;J. m t·
: :
~eeks /? months
-'-~8·~' ~, I O
,
11 1'4
1~~8
1
20 1 2 2'4
!
: : :
.-
475
476 I
Appendix B: Tables of standard measurements
Table 6 Body weight and height, boys, 7-5 years. (Courtesy of the Child Growth Foundation. Printed by Harlow Printing Ltd, Maxwell St, South Shields, Nf334PUj
,
I
2'/,
2
1
1-5yrs
3'/,
3
4'/,
4
5
130
years --l-
125
996th
120
98th
115
115
gist
75tb
-
11 0
50th
105
110
105
25tb Sib
100
2.d
--
95
-- 100
a.4th
95
90
85
80
---~
I
/'
-
26
9 9,6th
./
.. /
75
25
'"
./
70
T
98th
L
65 I
"'"
~--
I
1... _ _ _._---+-_ _ _ _ _ _ _
23
/~
I
60
'" 24
./
/"
r
22
9 l st
--t
21
20
75t h
19
20
19
~
18
-
---- 18
5 0th
.-+
17
25 th
16
17
16
9t h
15
15
2nd
, 14
14 OAth
13
12
11
11
10
10
9
8
7
-~---+-
i
6 f---. 1
-- -+
years
1'/,
2
2'/,
-......J
3
3'/,
4
4 ' /,
5
Appendix B: Tables of standard measurements I
Body weight ond height, boys, 5-78 yeors. (Courtesy of the Child Growth Foundation. Printed by Harlow Printing Ltd, Maxwell St, South Shields, NE33 4PU.;
Table 7
20 5 6 5 200 - - - + - -
8
11
10
9
12
13
14
16
2 05
17
200 ..- _
/""-1l-./ I
With provision for school reception class
/
-
99.6t h - :
/
-
--- -
.' ....
h / /'
1 75
/- : 1/ 170
-
19 0
91&t -
185
98th
/'
/
r--t>'
180
. 75th
'
180
. ..
17 5
5 0th
/
25 . h
170
77
4/
16 5
/
160 /
+7 v'
/
150
,( /
145
L--/'
/ //
/
/
/
.4-
/ 1-/
.-/,
/.-
././~
r-
./
155
7'
- y
/'
105
~
17'
--'
/ /
/ 4
.
.-.
.., 99.6' h -
_.- - + - -
100 95
/'
./
/
./
./
160
/,V
.L __/ /'
/ / L/
125
;-
/-
/
/ ,/ ,Lo/
130
/-I~
/
/'
/' / /t-/
135
....--::i - ./ r-. __ 0.4'h --'
/
~
140
165 2n d
/~
/
9th
~-
.
/
.L-/
155
195
1
~~./
NAME .... .... . ...... ..... ... ........ .... .
D.O.B. ... ... 1 .. . . .1 .. .. ..
185
98th
-- 90
/ 85
/
L. / / '7
120 115
15
5-18yrs
195 ' 190
7
/
/
75
/ 75th
110
. 70
105
50t h
100
25th
95
90
---
65 60
9t h ----
55
2nd
50
O.4th
45 40
40
35
35
30
30
---
.....--.
25
25
20
20
15
15
years
10 5
6
7
8
9
10
11
10 12
13
14
15
16
17
477
4 78 I Appendix B: Tables of standard measurements
Table 8 Body weight, height and head circumference, girls, birth to 7 year. (Courtesy of the Child Growth Foundation. Printed by Harlow Printing Ltd, Maxwell St, South Shields, NE334PU.j
O-lyr
14 16
~8
Q)
h 0
241' 28 3b 32
0
C!;
3~
36 38140 42 h4 46 ,f8 50 52
®
@
@
6
50
NAME O.O.B. . .. f .... .I ... .
Correct by ..
20
weeks /6 :months
9 9.Sl h 9 8th
48
91sl
. .. .. .. for prematurity
7 3th
32 3436 38
49
2
50th 3
-
47 46
25th
45 ·
9tl1
45
44
20d
44
43 ,
O.-1lh
43
42
pre-t e rm
40
42
.//
41
39
/ / /
38
37
36
.'.
84
82
99 .6 1h BSIII
9Ist 75th
35
34
,
33
74
2Gth
72
Olll
O.4t h
f
f
70
68
6
I·
30 ,
78
'76
50th
2 tl d
/
31
28 /
1
/
32
29
/
80
64
/
62
;
60
27
58
/
26
25
52
24
56
54
52
50
48
13
46/
9fU;:h
.;.4
12
42
9~l h
4()
11
!J ls.t
10
75a;
10
50tb
9
9
25 th ~){ h
8
8
2nl!
7
6
O.·l lh
pre-term
5
/
4
,. /
3
r r
2
6
/
5
85052
,/
3
,
;'
/
2
weeks / j> m onths
1
32 34 36 38
EO D
2
7
~ 8~ 12~14 ~8 6
10
16
20 l2 2 4
i
t t-,- ---+---I--+-----+
o ___ T o
o'
____
__
Appendix B: Tables of s ta ndard measurements I Body weight and height, girls 7-5 years. (Courtesy of the Child Growth Foundation. Printed by Harlo w Printing Ltd, Maxwell St, South Shields, NE33 4PU.j
Tabl e 9
,
,
1
2
1-
2 '/,
3
1-5yrs
3'/,
4
4 '/,
5 25
years
120
99.6 tb
NAME ............... .. ......... . .
98 th
D.O.B. .. .... / .. .... / .... ..
115
91 . t 75 th
110
110
5 Qth
105
28 th
105
9t h
100
2nd
OAth
95 90
27 9 O.6t h
26 25 24
98tb
60
23
55
22 9Is t
21 20
20
75t h
19
19
18
18
50t h
.-'/
17 16
~
15
17
----- ..... ---
2:,l h
16 9 th
15
14
2nd
14
13
OA t h
13
12
12
11
11
10
10
9
--
8
,.. ,..
9
;;-
8
7
7
6
years 1
1 '/,
*Measurement: H i e
2
2 '/,
=H c~ d Circtlmfcrt'llc('. L = LCIl!!t1 l. W = W cl~ hl. H = 3
3 '/,
4
4 '/,
6
liCl,~il!
5
4 79
480 I
Appendix B: Tables of standard measurements
Body weight ond height, girls, 5-78 yeors. (Courtesy of the Child Growth Foundotion. Printed by Harlow Printing Ltd, MaxwellSt, South Shields, NE334PU.j
Tabl e 10
195
5
6
7
9
8
190
10
12
11
14
13
15
16
17
---, 195
. 190
I ,
185
5-1Syrs
180 f
With provision for school reception class
175
NAME .. ...... ... ... ... ...... ... ... .. ... .. .
-; 185
99 .6th .=, _
O.O.B . ...... 1...... 1 ..... .
--
170
--' 98th
__
~
175
g lst _
170
76th
165
_ - - - --
160
50th _
9t h
150
/
145
155
/
/
165
160
25th
155
180
2nd
150
/
-
O.4th
145
/' /'
140
.7
/ /
/
135
130
95
L
125
90
/' _99.6 t l>
120
. 85
115
80
98th
110
75
105
..-:._ 91st -
/'
100
70 65
,; 75 th
95
60
/ 50t h
90
55
50
/ 9th
,;
2nd
"
40
_
35
/' /'
30
25
/'
/
15
/'
Y /'
/'
//
.
~
/
,.'
~-
__ - - - - - -
-
..'
.
-~--::-
:.:.:--:--:::- -::: .5. ,... -~-
--
_lo-
~
---
6
~
25
-
20
.....
.
.
15
years
10
5
7
40
30
.-----.-
/"
//'
~ .'
/
45
35
,,/,,/
~ -- ------...- -~
•
20
0 .4th
./
_
8
9
10
11
10
12
13
14
15
16
17
Appendix B: Tables of standard measurements I
Tabl e 11
Organ weights in infants, newborn to 72 months 4
Organ
No. in group
Mean weight (g)
Newborn Brain Heart Right lung Left lung Liver Spleen Right kidney Left kidney Thymus
13 16 12 12 16 16 14 14 16
325 18 29 24 109 8 11 11 11
69 65 63 63 74 74 74 74 68
489 27 55 46 176 16 17 17 21
102 100 94 94 102 103 104 104 90
569 31 61 52 193 17 19 19 28
81 80 83 83 85 85 84 83 76
651 33 67 57 223 19 21 22 31
44 40 37 37 44 44 43 43 38
691 33 73 60 232
1 month Brain Heart Right lung Left lung Liver Spleen Ri ght kidney Left kidney Thymus 2
months
Brain Heart Right lung Left lung Liver Spleen Right kidney Left kidney Thymus 3 months Brain Heart Right lung Left lu ng Liver Spleen Right kidney Left kid ney Thym us 4 months Brain Heart Right lung Left lung Li ver Spleen Right kidney Left kid ney Thym us
22
21 21 29
(Continued)
481
482 I
Appendix B: Tables of standard measurements
Tabl e 11
(Continued)
Organ
No. in group
Mean weight (g)
26 24 23 23 25 26 26 26 22
682 34 69 56 246 22 23 26 25
17 15 16 16 15 17 18 18 16
766 37
12 11 10 10 12 12 12 12 9
763 37 70 57 285 26 24 27 32
8 8 7 7 8 8 8 8 6
852 45 74 63 340 31 30 29 32
12 12
925 45 87 79 342 45 28 28 26
5 months
Bra in Heart Right lung Left lung Liver Spleen Right kidney Left kidney Thymus 6 months
Bra in Heart Right lung Left lung Liver Spleen Right kidney Left kidney Thymus
72
61 280 24 22 22 32
7 months Brain Heart Right lung Left lung Liver Spleen Rig ht kidn ey Left kidney Thymus
8 months Brain Heart Rig ht lung Left lung Liver Spleen Right kidney Left kidney Thymus
9 month s Bra in Heart Right lung Left lung Liver spleen Right kidney Left kidney Thymus
11 11
12 12 12 12 10
(Continued)
Appendix B: Tables of standard measurements I
Table 11
(Continued)
No. in group
Mean weight (g)
10 10 9 9 10 9 10 10 8
986 47 89 76 369 36 30 35 28
Left kidney Th ymus
8 9 8 8 9 9 9 9 9
935 52 104 104 388 39 35 35 33
12 months Brain Heart Right lung Left lung Liver Spleen Right kidney Left kidney Thymus
3 3 3 3 3 3 2 2 2
920 51 95 86 405 42 30 33 16
Organ
10 months Brain Heart Right lung Left lung Liver Spleen Right kidney Left kidney
Thymus 11 months Brain Heart Right lung Left lung Liver
Spleen Right kidney
Oa ta from ref. 4.
Sudden infant death syndrome (SIOS) and non-SIOS aggregated; no significant difference found between
groups in original analysis.
483
484 .
Appendix B: Tables of standard measurements
Table 12
Major organ weights (g) by age (7 -7 8 years) and sex:5
Age (years)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Brain
Lungs (right
Heart
Liver
Spleen
Kidneys
(right + left)
+ left)
M
F
M
F
971 1076 1179 1290 1275 1313 1338 1294 1360 1378 1348 1383 1382 1356 1407 1419 1409 1426
894 1012 1076 1156 1206 1225 1265 1208 1226 1247 1259 1256 1243 1318 1271 1300 1254 1312
107.3 245.9 304.7 314.2 260.6 399.5 365.4 405.0 376.4 474.5 465.6 458.8 504.5 692.8 691.7 747.3 776.9 874.7
175.3 244.3 265.5 311.7 319.9 357.5 404.4 382.1 358.4 571.2 535.0 681.7 602.3 517.0 708.3 626.5 694.5 654.9
M
54 63 73
83 95 103 110 122 132 144 157 180 202 238 258 282 300 310
F
M
F
M
48 62 71 80 90 100 113 126 140 154 168 188 207 226 238 243 247 250
400 460 510 555 595 630 665 715 770 850 950 1050 1150 1240 1315 1380 1450 1510
390 450 500 550 590 635 685 745 810 880 960 1080 1180 1270 1330 1360 1380 1395
35 42 48 53 58 62 64 68 73 82 91 101 111
127 135 145 152 157
F
34 41 47 52 57 62 67 71 77
85 93 103 112 120 127 134 140 146
M
72
85 93 100 106 112 120 128 138 150 164 178 196 212 229 244 260 270
F
65 75 84 93 102 112 123 135 148 163 180 195 210 222 230 236 240 244
Data from ref. 5.
F, female; M, male.
References
2
Maroun LL, Graem N. Autopsy standards of body parameters and organ weights in non-macerated and macerated human fetuses. Pediatr Del! Pathol 2005; 8:204-17. Pinar H, Sung CJ, Oyer CE, Singer DB. Reference values for singleton and twin placental weights. Pediatr Pathol Lab Med 1996; 16:901-7.
3 4
5
Naeye RD. Umbilical cord length: clinical significance. J Pediatr 1985; 107: 278-81. Thompson WS, Cohle SO. Fifteen-year retrospective study of infant organ weights and revision of standard weight tables. J Forel1sic Sci 2004; 49: 1- 11. Altman PL, Dittmer OS. Growth: Including Reproduction and Morphological De1!elopment. Washington DC: Federation of Am erica n Societies for Experimental Biology, 1962.
--
I
INDEX
I
Note: For purposes of conciseness, the terms 'abuse' and 'non-accidental' imply suspected and/or actual abuse. Figures and tables are comprehensively referred to from th e text. Therefore, significant material in figures and tables have only been given a page reference in the absence of their concomitant mention in the text referring to that figure. Abbreviations: SIDS, sudden infant death syndrome; SUD], sudden unexpected death in infancy. abando ned body (baby or fetus) 140-1
identifying 413
abdome n (and contents)
maternal trauma as cause of fetal
death
blunt 190-2, 192-3
penetrating 192
neonatal post-mortem examination
160
abortion
definition 181
induced/deliberate 181
offence of procurement 181
abortuses 141
abscess, retro ph ary ngeal 240- 1
abuse
child see child ab use
drug see drugs, recreational
acdde ntal injury 20, 70, 336-44
bone, vs non-accidental 20, 70, 71
causes and mechanisms 338
consequen ces 338
death due to
bath tub drowning 3 53-4
epid emiology 336, 337
epidemiology 336-7
falls see falls
fundus ha emorrhage due to 131
head see head
in hosp ital 372-3
prevention 341
asphyxial deaths 334
role of carers 341
proneness 33 7-8
road traffic-related see road traffic
accidents use of term 'accident' 33 7
acetaminophen poisoning 149
acho ndroplasia 236
acromion fracture 59
activated partial thromboplastin time
(aPTI/PTIK) 80,81
with other haemato logica l
abnormalities 89
prolonged , isolated 86
ac ute metabolic autopsy 119-20
acute phase response, ante-mortem
109
acyl-CoA deficiency )16
long cha in 244
medium chain see medium chain
acyl-CoA deficiency short chain 244
Addison's disease 113, 128
adnexa, ocular, non-accidental injury
125-8
adolescen ts/teenagers/ older chi ldren
drug ab use 256-7 , 341-2
anabolic steroids 258
testing 272 -3
volatile substances see volatile
substan ces
epilepsy and bath-tub drowning
355
see also school-age children
adrenal glands
fetal weight 451
injury 68
insufficiency/hypoplasia 113, 238
lipid dep letion and SIDS 212
adulteration, biological specim en 269-70
adverse drug reactions 367
age
dental estim ation 441-3
gestational see gestational age
agricultural injuries 341,393-4
air bags 386
air em bolism 364
airways
lower, narrowing in asthma 243
obstruction
foreign body 330-]
imposed 332-3
upper 207
peripheral, resistance, increase with
immersion 350
upper 240-1
abnormalities 240
infections 240, 241
obstruction 207
post-mortem examination 155
post-mortem exa mi na tio n, live
birth determination 185
see also inhalation; respirato ry tract
albumin, post-mortem evaluation 108
alcohol (ethanol)
fetal exposure to 267,274
fire deaths relating to 318, 321
immersion victims, testing 347
post-mortem diffus ion 270
road traffic accident
testing child 389
testing driver 388
alleles 39 7-8
allergy see an aphylaxis ;
hypersensitivity
Allitt, Beverley Gait 27 4, 374
alloimmune thrombocytopenia,
neonatal 98
Alport's syndrome 95
alternating current injury 32 7
alveoli
in immersion 350
SUD! 210,215,216
amel ogenin 400-1
amino acid disorders 115
sudden death 117
am niotic fluid toxicology 267
amphetamine tests, false-positives 268
amyloid -beta precursor protein see
beta-amyloid precursor protein
ana bolic steroids 258, 271-2
anaemia, aplastic 99
anaesthesia 370-1
dental 371
manslaughter 367
anaphylaxis 109-10, 239-40
drug-induced 367-8
ana to mical variants (normal)
genitalia 29-36
skeletal 70
486 I
Index
ane urysms, intracranial 236
an hidrotic ectodermal dysplasia,
X-linked 235
animal bites 42 7- 8
anterior segment (eye) injury 126
anthro pometry see body
anticoagulants 98
anticonvu lsant (antiepileptic) dru gs
aspirin 99
assessme nt, clini ca l and medical, in
abuse see child abuse
asthma 342
acute 243
'at-risk' register, death of child on
146
atherosclerosis, premature coronary
243
249
antidi uretic hormone, syndrome of
inappropriate sec retion 366
a ntiepiieptic drugs 243
antiplasmin deficiency 93
a ntiplatel et drugs 99
a nus and perianal area, examination
11-12,36-7,39-40
acute/healing/healed trauma 37- 8
boys 27, 467
girls 26
sig ns of abuse 39-40
in sum 2 17
aortic stenosis, congenital 227
aplastic anaemia 99
apolipoprotein A genotype and
traumatic brain injury 312
apoptosis, brain cells (in early life)
167
apparent life- threatening events
(ALTE; near miss events)
haemosid erin as marker 216
respiratory cha in complex deficiency
246
SUDI/STDS preceded by 202, 202 - 3
arousal shakin g, retin al haemorrhage
132
arrhythmias, sudden death associated
wi th 232, 365
epilepsy and 243
in hospital 365
due to adve rse dru g reactions 367
immersion 350
infants (SUOT) 207, 232
arrhythmoge nic right ventri cu lar cardiomyopathy 229- 30
arsenic exposure 233
arte ritis, coronary 231
in stantaneous infant death 205-6
ascorbic acid (vita min C) deficiency
and scurvy 71,94
asp artic ac id, dentine 443
asphyxial deaths 329-35
immersion in non-water media 349
preve ntion 333 -4
sleeping infants see sleeping
enviro nment aspiratio n see inhalation
atrioventricular block 232
atrioventricul ar node, endod erma l
heterotopia 233
attention deficit hyperactivity disorder
(ADHD), methylphenidate in
257- 8, 272-3
autopsy see post-mortem examinatio n axon al damage/injury 168-9
diffuse, road traffic victims 392
nerve root 176
babies see infan ts and babies
babygram (whole-body radiograph) 48
sudden death (infancy/early
childhood) 146
bacteria l in fect ion
ante-mortem 110
bites 428
trachea 24 1
bacteriological samples, sudden death 148-9
'b angungu t' 249
basal gangl ia, hypoxic-ischaemic
damage 172
bath sa lt flu orescein 353
bath tub and spa pool drownings
353 -6
accidental 353-4
non -acc idental 354- 6
battered child syndrome see physical ab use
Bayley's Motor Scale 17
bean - bags, polystyrene-filled 204
bedding and SUDl/STD S 140, 203
bed-sharing see co-sleeping
behaviour in abuse
assessing 460
negative 14-15
Bernard-Soulier syndrome 78, 81, 82,
94
0- ox id ation see fat ty acid 0- oxidation
defects
beta -amyloid precursor protein 169
tra umatic brain inju ry 3 12
bicycles see cyclists
biochem ical tests on post-mortem
specimens 106-26, 149-50
factors affectin g 107
ge nera l eva lu ation 107-8
immersion victims 347
infants/young children 149-50
technical considerations 117- 20
birth
concealment of delivery or 181, 182,
352
live, determination 183-7
see also de livery; labour
birth trauma 173-6
neurological 173-6, 289, 300-2
mechanisms of brain injury 304
sp inal cord 176,302,312
post-moltem examination 161- 2,
188-90
retinal haemorrhages 133
skeleta l see skeleton
SUD Tand 218
bite marks 425-32
detail of 429-30
impressions with see impressions
pattern/distribution 18, 425-6
recording 430-1
self-inflicted 19, 427
suspect 431-2
comparison evidence from bite and
suspect 432
types 426-7
bladder, STDS 210
Bla nd-White-Garl and syndrome
231
bleedi ng disorders (inc!. diatheses) 80,
82-101
bruising 76-7
drug-induced 98 - 9
evaluation/m easurements 82-94
of primary haemostasis 82
fundus haemorrhages 132
bleeding patient, evaluation 82-6
see also blood, loss; haemo rrh age
bleeding time 82
blood
clotting see coagulation
flow, cerebral (CBF), mon itorin g 291,
292, 292-3
loss, estimation in sudden death 152
sa mples
acute metabolic autopsy ll9
collection 11 8
to xicology 263
tests see haematological tests blood vessel see vasculature blunt abdomin al t rauma, matern al, causing fetal death 190-2,
192-3
~ - -
-
--
~r:.
Index I
body in fet al /perin ata l death,
measurements (inc!.
anthropometlY) 187
in fire fatalities 319-20
iden tifying 320, 412
hei ght and weight see heigh t; weigh t
identifying see ident ification
in sudden unexpected death 139
cavities s ee ab domen; chest
clothing see clothing
family wish to view 142, 389
measureme nts (inc!. anthropometry)
150
road traffi c accid en t 389-90
to tal surface area, burns and 323
bone brittle see brittle bone disease facial, fetal hei ght estimation from 444
injury
accidental vs non-accidental 20,
70,71
fractures see fractures
post- moliem ex amination 163
metabolic disease 7l, 94
bone marrow failure sy ndromes
99-100
bone scan, ab use 51
bowel see intestine
b.ox jellyfish 358
boys
growth charts 475-7
sex ual abuse, examination for 467
technique 27
brain
damage see brain injury;
neurological lesions
herniations 311
hypoxia see hypox ia
imaging see neuroim aging
infarction 310-11
in early life 176-7
malformations 169-71,237
oedema see oedema
post-mortem exam ination and
findings 156-7
neonatal 160
SlDS 211
weight see subheading be/ow
temperat ure, head injury 292
weight 484
fetal 471
neonatal and infa nt 481-3
see also intracranial path ology and entries Linder cerebral
brain injury, trau matic 63-7,
282-3 17
genetic factors influencing
recovery 312
non-accidenta l head injury and
63-7
neu rological presentations 295-6
ocular injury associated with
127-8 , 130-1
primary mechan isms 302-7
secondary mechanis ms 307-11
see also head injury
brainstem
damage 172
bi rth -rel ated 302
SIDS, histology 211
breastfeeding
asphyxial death 204. 372
infant poiso nin g 257
breech presentation 188
brittle bone disease
inherited see osteogenesis
imperfecta
temporary 73
bronchiolitis, acute 241
bronchopulmonary dysplasia 210, 248
hospital de ath 372
bruises
diseases causing 76-7
in fetal/perinatal death 189
non-accidental 294
assessment for 17 - 18, 19
importance of sha pe 4
in sudden death
infants 214-15
observed at scene 140
bucket, drowning in 353
bucket-handle fracture 53 , 54, 57
bullous impetigo 20
burns 5, 323-5
eyelid 125
pathological changes 324-5
patterns 20
skin conditions simulating 20
C-reac tive protein, post-mortem 109
calcification, infantile coronary 232
callosal agenesis 170
callus formation 52
calorie intake, post-mortem evaluation
108
Camino fibreoptic pressure monitoring
291
CAMPI 440
cancer/malignancy, bone marrow
infiltration 100
487
capillalY electrophoresis and capillalY zone electrophoresis 261
capilialY reactions, bra in 168
caput succedan eum 302
car(s) an d other vehicles
accidents and collisions see road traffic accidents
entrapment in 330
inspection following collision
388
car seats, child 386-7
carbohydrate meta bolism, sudden
death 117
carbon monoxide
haemoglobin and 320, 322 -3
poisoning 139
carboxyhaemoglobin 320, 322-3
cardiomegaly. myocardial infarc tio n
w ith 233
cardiomyop athy, su dden death 155,
156, 228, 228-9, 229 -30
infant 206-7
cardiopulmonary resuscitation see
resuscitation
cardiovascu lar disease/disorders see
heart and entries above
carers and hea lth professionals
perpetrating abuse or death 274,
373-5
preventing accidents 337-8
carnitine deficiency 244
carnitine palmitoyltransferase type 11
deficiency 244
catheter-related deaths 368
cell death, brain (in early life) 167
central venous catheter-related deaths
368
cephalic presentation 188
cepha1(o)haematoma 218, 301
cerebral artery
infarction 176-7
middle (MCA), in head injury
flo w velocity assess me nt 293
occlusion 290
spasm 293
cerebral blood flow (CBF) monitorin g
29 1, 29 2, 29 2-3
cerebral function an alys ing monitor,
head injury 292
cerebral hae morrhage
birth-related 161, 190
sudden death 236-7
cerebral oedema see oedema
cerebral palsy 167
cerebral perfusion, infarction due to
impairment of 310-11
488 I
Index
cerebral perfusion pressure (CPP) in
neurotrauma 309
monitoring 291
cerebral sinovenous thrombosis 177
cerebrospina l fluid , post-mortem 120
cerebrovascular autoregulatory
mechanisms 291
certification of death, SillS/SUm 218-19
cervica l spine and spina l cord trauma
311-12
perinatal 193, 311
road traffic victims (=whiplash
injury) 393
cervicomedullary syndrome (whiplash
shaking injury) 283, 306, 312
hyperacute 295
Chediak-Higashi syndrome 81, 92 , 95
chemical asp hyxi a 333
chemical burns 324
non-accidental 326
chemical ionization 262
chest (thorax) viscera
post-mortem examination and
dissection 154
neonatal 160
traum a in abus e 68
child abuse 1-105, 125-36,457-69
alerting signs 2- 3
clinical and medical assessment
J-23
~ow chart 5-6
forms 11 , 17,457-69
formu lating an opinion 2 1
interpreting findings 17- 20
need for comprehensiveness 4
rol e 7
definition 2
dentist and see dentist
documentation see docum entation
dru g-facilita ted 273
expert witnesses see expert
witnesses
investigations 12
legal issues 1-2, 4, 7 -8, 22
medical conditions simulating see
medical conditions
ongoing health care 21
prevalence 2
types 3
neglect see neglect
physical see physical abuse
sex ual see sex ua l abuse
Child Assessment Orders 7
ch ild destru ctio n
definition 181
law 181
child-killing immersion sy ndromes 351,352- 3 see aIso filicid e; homicide; infanticide; neonaticide
chi ld protection forms II , 17, 458-69
Child Protection Orders 7
Child Protection Services 2
Children Acts (inc!. Scotland) 1, 7
sig nificant harm in 4
chime rism 410
chinon 161
chlamydial infection 42
chloride leve ls, post-mortem 110, III
chromatography
gas see gas chromatography
liquid see liquid chromato graphy
micellar electrokinetic capi llary
(MECC) 261
chromosom es 393-4
see also X chromosome;
Y chromoso me
chro nic disease, sudden death 226
cigarettes
burns 326
smoking see smoking
Clauss fibrinogen assay 81, 82
clavicular fracture 59
at birth 71 , 161,218 clinical assessme nt, abuse see child abuse clinical presentation see presenting history clothing of decedent bagging for police in road traffic accident 389
in hospital , road traffic accident 388
at scene of sudden death
remo va l from body ]]9 road traffi c accide nt 388
clotting see coag ula tion
clozapine 272
coagulation (clotting) 79-80
activation 100
classical path way 80
defects 96 -7
pe tech ial haemo rrhages and 329
diss eminated intravascular (DIC) 90,
101
inhibitors, tests showing 86, 87, 88,
89
initia tion and amplification 79-80
normal, and normal platelet count,
bleeding with 90-4
platelet enhancement of 79
tests 77
coagulation fa ctors 79-80
complexes 79
deficien cies 80, 83, 84,90, 92 - 3,
96-7
tests revealing 86, 87, 88, 89
inhibitors, tests show ing 86, 87 , 88,
89
coca ine, environmental exposure 268,
272
prenatal 274
codeine, breastfeeding mother 257
coll oid cyst, 3rd ventricle 237
colposcopy 11-12,27,39
video 7, 8, 11 , 27
combined paternity index 409
compression head injury 302
computer-assis ted dental identification
systems 440
computer tomography (CT) of head 12,
50, 61-2
accidental injury 293
mild 289-90
non-accidenta.l injury 12, 50, 61-2,
299-300
conception, prod ucts of 141
confi dentiali ty 8
confl ict of in terest, expert witnesses
454
congenital infections
brain 177
syphilis 71
conge nital malformations/anomalies/ abnormal ities
brain 169- 71 , 237
genitalia 41
heart 226-7
cyanotic 83 - 4, 101
drowning and 3 56
instantaneous infant death due to
205
perinatal death due to 193
rapid infant death due to 205
in-hospital neonatal death 372
upper airway 193, 240
congenital platelet disorders 94-5
con ing 310, 369
conj unctival petechiae, SUD! 213 -14
consciousness, assessing level of 282
conse nt (inci. informed .consent) in
abuse to examination 8, 457, 459
sexual abuse 27
contaminatio n of sampl es,
environmental
DNA samples 403
hair samples 268, 268 -9
contusions
brain 294
Index I
birth-related 302
imaging 63
spinal cord 311
copper deficiency 71
copy number
low see low co py number ana lysis variation (CNV) 398
core temperature, head injury 292
co rneal damage 126
coronary artery pa thology in SUDI
227,230-1,232-3,364
anomalies 227, 230-1
embolism 364
instantaneous infant death 205-6
occlusive disease 232-3
petechial haemorrhages 210
premature atherosclerosis 249
coroner 143, 190, 225, 345
hospital death 375, 377
corpus callosum, agenesis 170
cortex, cerebral , hypox ic-isch ae mic
damage 172
co-sleeping and bed-sharing 203,
204-5
overlaying risk 203, 215, 331
wedging risk 204, 331
see also 'dead in bed ' syndrome
cot 203
baby found dead in (and 'cot death')
140, 198
abandoning of term 'cot death' 199
pathology 206, 208
bedside 204
characteristics 202
see also sudden infant death
syndrome
cotinine, neona ta l hair testing 273-4
court, expert witness see expert
wi tness Cozart® drug detection system (DDS) 260
crack (cocaine), hair testing 272
cranium see skull
creatinine 114
creeks, drowning 357-8
crime scene manager 138-9
Criminal Hearing System (Scotland) 7
Criminal Procedure (Scotland) Act
(1995) 7
crimping bruising 19
Crohn's disease, anogenital 41
crown-heel length (CHL), fetal autopsy
471-2
crown-rump length (CRL), fetal
autopsy 471-2
Cushing phenomenon 291
cyanide (fires) 320, 322, 323
cyanotic congenital heart disease
83-4, 101
cyclists/bicyclists 388, 392
helmets 338-9, 392-3
cyst
brain 168, 171-2, 237
laryngeal 207
cystathione B syn thase deficiency, thromboemb olism 235
cystic fibrosis 397-8
cystic medial necrosis 235
cytochrome P450 (CYP) enzymes in
toxicology 258, 263
cytokines and burns 325
cytomegalovi rus
parotitis 250
SIDS/SUDI 211
489
degradation, drug, post-mortem 270-1
dehydration, ante-mortem 110-11
deliberate acts see non-accidental acts
delivery
concea lment of birth or 18], 182,
352
instrumental see instrumental
delivery
intracrani al haemorrhage and mode
of 174,17 5
unattended 182-3
dental identification 435-46
comparison technique 435, 435-49
profiling technique 435,440-1
dentine, aspartic acid 443
dentist and child abuse 420-34
accused of abuse 425
examini ng for/reporting of abuse
420-2
Dandy-Walker 237
dating of fractures , radiological 59-60
skull 60-1
DAVID 440
'dead in bed' syndrome and insulin
dependent diabetes mellitus 245
death(s)
accidental see accidental injury
adequa te cause of, in SU D! 200-1
asphyxial see asphyxial deaths
certifi cation in SIDS/SUDI 218-19
dyadic see dyadic deaths
feta I see fetus
head injury 283, 284-8
heat-induced see fire fatalities;
thermal injury or death
immersion see drowning
infant see infant
mUltiple 141- 2
perinatal see perinatal death
scene of see scene
time of see time
traumatic, relative ranking order of
346
see also cell death; ch ild-killing; eutha nasia; filicide; homicide; infanticide; manslaughter; neonaticide; serial killing; suicide decedent body see body clothing see clothing deciduous teeth, age estim ation 442,
443
decomposition, electrolyte
abnormaliti es 111
decongestants, nasal, affecting
amphetamine tests 268
dentistry, deaths in 371
Denver Developmental Screening Test
14
depressed skull fractures 294
depression, post-natal 352
dermati tis, napkin 217
developmental assessment (in abuse)
12-15
diabetes mellitus 112-13,243-4
sudden death 243 -4
diaphragmatic hernia, co ngenital
239
diaphyseal fractures 52, 53
at birth 71
dia to ms, immersion victims 347
diazepam and breastfeeding 257
diffuse axonal injury, road tra ffi c
victims 392
diffusion-weighted MRI of head 50,
299-300
digital dental radiographic
identification 440
digi tal injuries 59
dilated cardiomyopathy 229
dilute Russe ll's viper venom time 87
direct current injury 327
disaster and victim identification 440
diseases see medical conditi<;>ns
dissection, infants/older children
152-6
disseminated intravascu lar coagulation
(DIC) 90, 101
distribution, drug, post-mortem 270
diving reflex 349-50
bath epilepsy and 355
DNA 396,397
databases 414
490 I
Index
DNA (continued) mitochondrial see mitochondrial DNA DNA tests 395-419
dental identification and 437
dental pulp, sex determination 444
fatty acid oxidation defects 117,
246 history and techniques 398-402
identifying body (inci. missing
persons) 411-13
immersion victims 347
samples for 119, 402-3
mixed 406-7
problems and solutions 406-7
sexual abuse 403-7
twins and 396
see also molecular methods
documentatio nl records
in abuse 16-17
babies/infants/young children
145-6
bite marks 430- J
fundus haemorrhages 133-4
scene of sudden death 138,
145-6
dental
post-mortem 438-40
retrieval of patient records 437-8
expert witness 452
declaration by expert in report
452-3
road traffic collision 387
see also photographic evidence;
reporting; video recordings
dodecanoic acid 116
dog bites 427, 428
doping agents 258,271-2
screening for 27 J-2
Doppler, transcranial, head injury
292-3
dosage errors 366-7
drains, drownings 358
drowning and near-drowning
(immersion) 332, 345-61
aftermath 358-9
causes 345-6
classification 352-9
by site 352-9
unlawful child-killing 352-3
epidemiology 337, 345
immersion syndromes 346,
351-9
investigations 349
medium 346-9
pathophysiology 349-50
drugs recreational (illicit drugs)
accidental ingestion 342
coercive administration 272
environmental exposure to 268,
272
fetal exposure to 257, 267, 273-4
fire fatalities relating to 318,321
older children see adolescents
road traffic accident 388-9, 389
therapeu tic use
death due to errors 266-7
paediatric pharmacology 258-9
see also toxic substances
'dry' drowning 348
dura
brain
haemorrhage see intradural
haemorrhage
in sudden death 157
spinal cord, in sudden death 158
dyadic deaths (homicide-suicide)
141-2
drowning of child 352,355 dysfibrinogenemia see fibrinogen dysrhythmias see arrhythmias ecchymoses 88, 89
echo viral infection 209
ectodermal dysplasia, X-linked
hypohidroticlanhid rotic 235
eczema (dermatitis), napkin 217
EEG, head injury 292
electrical inj ury/burns
child 326
pregnant woman 192
electroencephalography, head injury
292
electrokinetic capillary methods
269
electrolytes, post-mortem 110-11
immersion victims 347
electron transport chain (oxidative
phosphorylation; respiratory
chain) defects 116, 244, 246-7
electrophoresis, ca pillary 1ca pillary
zone 261
electrophysiological monitoring, head
injury 292
electroretinography 134
ELISA, drugs 259
embolism
hospital fatalities 364
pulmonary 235, 364
embryo transfer mix-ups 410
emotional abuse, assessment 15-16
emotional trauma see psychological stress
encephalitis 238
encephalopathy
cardiac arrest 172
haemorrhagic shock (syndrome of)
238
pancreatitis (acute)-related 248-9
of Reye's syndrome 238-9
traumatic/non-accidental 283,
295-6
endocardial fibroelastosis (EFE) 229
sub-EFE 207
endocrine disorders 112-14
endodermal heterotopia, AV node 233
endotoxin, bacterial 110
endotracheal tubes 368
energy metabolism, disorders 114, 116
England, road traffic casual ties and
fatalities 390
enterocolitis, necrotizing 212
entrapment asphyxia 330
envenoming and drowning 358
environment
drug exposure from 268-9, 272
sample contamination see
contamination
sleeping
good 203
suboptimal 203-4
enzyme assays
in fatty acid oxidation defects, liver
116-17
fibroblast cultures see fibroblast
cultures
enzyme immunoassays (ElAs) for
drugs in oral fluids 259
enzyme-linked immunosorbent assays
for drugs 259
enzyme-multiplied immunoassay
technique for drugs 259
epidural haematoma see extradural
haematoma/haemorrhage
epiglottitis, acute 240
epileptic seizures see seizures
epiphyseal plate injury, radiology 56
epistaxis (nasal haemorrhage)
sum and history of 203
SUD! and presence of 214
erythrocyte sedimentation rate,
post-mortem 109
escharotomy 325
ethanol see alcohol
ethnicity, dental remains 444
euthanasia by bath tub drowning 354,
355-6
-
_.
~-~
~ -:, "
.- I
,
-
_.
Index I evidence (forensic) from bi te and suspect, co mparison of 432
expert witn ess see expert witness scene of sud de n death 138 road traffic collision 387 sex ual abuse 27- 8 see also sa mpl es Exclusion Orders 7 expert witnesses (a nd testimon y) 447-55
abuse physical 45 1 sexual 43, 45 1 admiss ibili ty of evid ence 450 - 2 claim to expert ise 449 communica tion s from 452 pre-trial 453 - 4 conflict of interes t 454 mission statement of expert 448-9 opinion 450 rece nt devel op ments 4 54-5 rul e of evidence 454 in witness sta nd /box 4 53 yes or no answers? 450 external exa min ation in fetal/perinata l death 189 non-acciden ta l head inj ury 294- 5 in sudden dea th 150 - 2 fi ndings of questionable significance 21 7 neonatal 158- 9 at scene 140 SlDS vs no n- accide nta l injury 212-15
see also physical exam ina tion extradural (epidural) haema tom a/haemo rrhage bra in 6 1 neonatal 160 neo natal bhih-related 176, 302 post-mortem exami na tion 156, 160 traumatic (non-b irth-rela ted) 294 sp ina l cord , birth-related 176 eye 125- 36 non- accidental inju ry 125- 36 scope 125-8 post- mo rtem examination 157-8 eye witness, roa d traffi c co llision 387
eyelids, non-acciden ta l inj ury 125 fa bri cated/in d uced/sim ula ted ill ness (FlI ; Munchausen's sy ndrome by proxy) 3 assessment 16
by carers 274 in hospi ta l 373
face feta l height estima tion from bones of 444
injury in abuse, dentists and 42 2-3 reco nstructio n 440 fa ctor V (a nd Va) 79 defi Ciency 89, 97 fac tor VI I (an d VIla) 79, 80 defi ciency 80, 89, 97 factor VIII (and VIlla) 80 deficiency see ha em ophilia, type A inhibi to rs 88 factor IX (an d IXa) 79, 79-80, 80 defi ciency see haemophili a, type B factor X (and Xa) 79, 80 detlci ency 89, 97 factor XI (and Xla) 80 deficiency (hae mophilia C) 80 , 84, 88, 97
factor XII 80 defici ency 80, 88, 92-3, 97 factor XIll deficiency 83, 84, 88, 90, 93
faeces (stool s), live birth determination 185
failure to th ri ve (n on-organi c), assessment of 13-14 falls, accidenta l 33 8, 339-40 from height 338, 33 9-40 mortaliti es 33 7 short fall s 298, 338 , 339 false-negative drug screens 269 false-posi tive dru g screens 268-9 fa lx, neona ta l exa mination 160 fa mili al searchi ng 4 13-1 4 fa mily see pa rents/immediate family fa mily history ab used child 10 bl eeding child 84 dea th (perinata l/in fan t/young chil d) 146
fa rmin g injuri es 34 1, 393 -4 fasc iitis, necro tizing 248 fat embolism 364 see aIso lipid fat her, alleged see paterni ty testing fa tty acid f)-o xidatio n defec ts 11 4-17 , 244-6
sudd en death 114-1 7, 244- 6
infa nt 114, 208 fatty acid ethyl esters 26 7 fatty change, liver, SID S 211-12 Fechtner's synd ro me 95 fe male children see girls
491
femoral fracture 49, 54, 55 SIDS 215 fertili ty treatment mix -ups 410 fetus (prenatal child) abandoned remain s see abandoned body
age see gestati ona l age
damage (= intrauterine damage)
169- 73
brain 300 roa d traffic accid ents 192-3, 193, 393
deat h (=intrauteri ne death) 180-97 defini tion 180 giving cause of 193 - 4 maternal death as ca use of 190-3 natural causes 19 3 post- mortem exam ination see post- mortem exa mination see also abortion ; abOIi uses drug exposure 257,267,273 -4 viability assessmen t 187 see also entries under co nge nital fever (py rex ia) in brain tra uma 292
flbreoptic pressure moni torin g of intracra nial pressure 291 fibri noge n abnormal levels (dysfibrinoge nemia) 81, 88, 89-90, 97
co ngenital causes 97 measurements 81-2 fi brob last cultu res (for enzyme analysis) 119 fa tty acid oxidation defects 117, 208
fibroelastosis, endocardial see endocardial fi broelastosis fi bro ma , cardiac 23 4 fibromus cul ar dysplasia 233 fibular fra ctures 57 filicide 353 in hospital 373- 5 fi re fata lities (house fire deaths) 31 8- 27
dental id entifica tion 439
epidemiology 3 18 - 19
pathologiSt's role 3 19-27
scene 141 , 3 19- 20
firearm deaths 337 fi reworks 32 4 fi rst-degree burns 32 3 fixation, organs 163 FLAIR MRI, head 64 fl as h injuri es 326 flotation test 185
492 I
Index
fluid administration
burns 325
deaths associated with 368, 371
fluid-attenuated inversion recovery (FLAIR) MRI, head 64
fluorescein, bath salt 353
fluorescent polarization immunoassay
(FPlA), drugs 259
follow-up skeletal survey 48
foot length (FL), fetal autopsy 150,
types 128-9
see also retinal haemorrhages
gamma-hydroxybutyric acid 273
gas chromatography (GC) 261
and mass spectrophotometry (GC
MS) 261-2
gas embolism 364
gastrointestinal tract
injury 68
sudden death relating to 238
Gaucher's disease 100
General Medical Council and abuse 8
genes 397-8
mutated 410-11
genetiC (inherited) disease
bleeding due to 84, 90-3, 94, 94-5,
187,471-2
forcers), shaking injury to head 303
forceps delivery, intracranial
haemorrhage 174
forearm fractures, bilateral 53
foreign body inhalation, asphyxial
death 330-1
forensic , meaning of term 7
forensic specialists in abuse 22
joint working with paediatricians
96-7, 99
metabolic see metabolic disease
thrombophilias 364
genetic (inherited) factors, recovery from traumatic brain injury 312
genetic material, inheritance 395-8
genetic variation 396, 397-8
genitalia (and their examination)
in abuse 11-12
acute/healing/healed trauma 37-8
boys see boys
child protection examination form
6-7,8-9, 11,458-60
fossa, posterior, damage at birth 176
fossa navicularis 29, 31
signs of abuse/injury 38,39
fourchette, posterior 39
normal findings 29-30
fractures 51-60
birth 71 , 161-2, 174, 189,218
head injury and non-skull fractures
461,466-7
298
non-accidental
accidental vs 20, 70, 71, 72
facial bones, dentist's role 423
post- mortem examination for 154,
girls see girls
mimics of abuse 40-1
normal anatomy and variants 29-36
in sudden death 152
Germany, child car restraints 386-7
germinal matrix haemorrhage 172-3
gestational (fetal) age
autopsy standard measurements by
155, 163
radiology 50,51-60
dating see dating
long-bone 52-6
sudden death (infant/early
childhood) 146
SIDS 215,218
see also specific bones
France, child car restraints 387
frenulum, labial, injury 151,213,423,
471-2
424
17
growth
assessment in abuse 12-15
charts 475-80
growth plate (epiphyseal plate), in
abuse, injury, radiology 56
haemangioma, cardiac 234-5
haematological disease 20, 76-105,
132
haematological tests 81-2, 86-90
bruised child 77
first-line 85-6
patterns of abnormal results 86-90
haematoma (and associated haemorrhage) see cephalohaematoma; intracranial haematoma/haemorrhage haemoglobin
carbon monoxide and 320, 322-3
HbAlc 112-13
haemophagocytic Iymphohistiocytosis
84-5
haemophilia 84, 85, 96-7
family histolY 84
type A (factor VIII deficiency) 80,
88,96-7
sex and 83
type B (factor IX deficiency) 80,
96-7
dental estimation 441-2
facial bone estimation 444
girls
growth charts 478-80
sexual abuse, examination 29-36,
466
fresh water, immersion in 347-8,357-8
frontal collisions 393
head-on 393
froth, post-immersion formation 350
fult-thickness burns 324
fundus [manifestations of abuse) 126-7
haemorrhage 128 - 34
differential diagnosis 131-3
examination with 133-4
forces required 130-1
mechanisms 129-30
glutaric aciduria 70
glycoproteins, platelet 78
gpIb deficiency 93
gpIIb-llla complex deficiency 93
glycosylated haemoglobin 112-13
gonococcal infection 42
grasping marks/injuries 295
grey platelet syndrome 95
Griffiths Development Scales 14,
findings 38-9
technique 26
Glanzmann 's thrombasthenia 78, 84,
93-4
Glasgow Coma Scale 282-3
gliosis 168
glucose
brain supply, inadequate 310
levels, assessment 112-13
vitreous humour 1.12, 113, 150, 244
see also hypoglycaemia
sex and 83
type C (factor XI deficiency) 80, 84,
88, 97
Haemophilus influenzae 248
type b and acute epiglottitis 240
haemorrhage
fundus see fundus
in hospital, fatal 363
intracranial see intracranial
haematoma/haemorrhage
nasal see epistaxis
petechial see petechiae
pulmonary, sudden death and 240
SIDS 215-16
retroplacental 192
subaponeurotic/subgaleal 161,
173-4, 301
~ -
-
-----
-
-
-.=
Index I
subconjunctival 126
subperiosteal 161
sudden death du e to 240
see also bleeding
haemorrhagic disease of the newborn
98
haemorrhagic shoc k encephalopathy
syndrom e 238
haemosiderin 168
haemosideros is, idiopathic pulmonary
215
haemostasis 77 - 8,78-81
primalY 78-80
measurements 82
secondary 80-1
hair samples (toxicology) 263-6, 272-3
applications 272-3
fetal exposure determination
273-4
environmental contamination 268,
268-9
hanging by ligatu re 332
head
circumference (He)
charts 475-80
fetal autopsy 471-2
occipitofrontal 150
imaging see neuroim ag ing
moulding, excessive 188
head injury 60-7, 282 - 317
accidental 283, 293
cycling, and its prevention 338-9,
392-3
mechanisms of brain injury 304
coagulation abnorm alities and
101
death, scene 141
definition 282
differential diagnosiS 69-70
epidemiology 283-94
imaging see radiology
non-accidental 63-7, 283-4,
294-300
dentists and 422-3
diagnosis 296-8
investigations 299-300
mechanisms of brain injury 305
severity, classification 283
see also brain injury; neurolo gical
lesions
head-on collisions 393
headspace technique (in toxicology)
260, 262
headwear/helmets
cycling 338-9,392-3
winter sports 341
health professionals see carers and health profess ion a ls heart arrest
with dysrhythmias 365
encephalopathy 172
perioperative 369,370,371
arrhythmias see arrhythmi as
cond uction abnormalities 232
histological examination 162
examination in sudden death
external 154
his tological, for conduction
abnormalities 162
internal 155-6
muscle see myocardium
surgery, death 370
weight 484
fetal 471
neonatal and infant 481-3
heart block 232
heart disea se 226-35
congenital see congenital malformations
drowning an d 354, 356
neurogeni c 249
non-congenital/in general, sudden
death due to 226-35
instantaneous infant death 205-7
heat
injUly or death due to see thermal
injUly or dea th
loss with burns 325
height
child's (stature/length), assessment
abuse 12-13
charts 475-80
post-mortem 150, 444
falls from 338 , 339- 40
see also crown-heel length ; crown-rump length ; foot length; umbilical cord exa mination helmets see headwear
Henoch-Schon lein purpura 20, 84, 85,
94
heparin 86, 86-7, 98-9
hepatic injury/problems see liver
heredity see entries under genetic
Hermansky-Pudlak syndrome 81, 82,
94-5
herniations, brain 3 11
heroin, coercive administration
272
high-molecular weight kininogens
(HMWKs) 80
deficiency 97
493
high-performance liquid
chromatography 260-1
coupled with mass
spectrophotometry 262
high-tension electrical injuries 326
histiocytoid cardiomyopathy
206, 230
histology (microscopic ap pearance)
of brain injury in early life, timing
by 167
immersion victims 347
lungs in fetal / perina ta l death 185-6
prenatal dental , age estimation via
441
subdural haem ato m a 153
in sudden death 162-3
SUm/SIDS 210-11,218
history-taking
abuse 9-10, 25-6
bleeding child 83-4
death (perinatal /infant/you ng child)
146
SUDr see subheading below sum 200,201-2
adverse historical factors 202-3
HIV (human immunodeficiency virus)
infection 42
human bites/scratches and risk of
428
homicide (inc!. murder)
bath tub drowning 355
by fire 321
in hospital 373-5
scene of death 142
see also filicide; infanticide;
n eonaticide; serial killing homicide-s uicide see dyadic deaths homocysti n uria, thromboembolism 235
horse-riding 341
hospita l, sudden death 362-84
causes 363-75
defin ition and frequency 362-3
investigation 375-7
hot fluid burns see scalds house fires see fire fatalities HPV 41 ,42,404 human immunodeficiency virus see HIV human papillomavirus (HPV) 41,42, 404
humeral fracture 55, 57
hyaloid, posterior 129
hydroc ephalus 237
shun ts 369
hydrogen cyanide (fires) 322, 323
494 IIr Index
3-hydroxyacyl CoA dehydrogen ase deficiency, long chain 244 hymen exa mination 26 measurements 35-6 normal appearance/variants 29 -31, 32-5 configuration 32 ga pi ng/ narro w i ng/ a tten ua tion 34-5 notches/ c1efts/ tra nsectio ns 33-4 ridges/bumps/tags/bands 32-3 signs of abuse 39 hyp ernatraemic dehydration III hypersensitivity/allergy 342 see also anaphyl ax is hypeliension, pulmonary arterial 242 hype li hermia, malignant 371 hypertonic vitreou s humour III hyp ertrophic cardiomyopathy 228-9 familial 228
s udd en death 156, 228-9
infant 206-7,228 hypofibIinogenemiaseefibrinogen hypo g lycaemia 112, 113 neonatal/infant 177 hy pohidrotic ectodermal dys plasia , X- linked 235 hyp otensio n, matern a l, due to injUly 192 hypo thermia immers ion 349 induced, head injUly 292 hypotonic vitreous humour 111 hypovol ae mic shock with burns 325 hypoxaemia, head injury 290 hypoxa nthine, time of death estim ation 111-1 2 hypoxi a (inc!. brain) \09 drowning 349, 350 hypoxic-ischaemic injury 167-8, 172 non- accidenta l causes 297 in shaken baby syndrome 307 ICD see International Classificat ion of Diseases ice, spOliS injuIies 340-1 identification of body/body remains 411-13
dental see dental identifica tion
DNA techniques 411-13
fire fatality 320,4 12
missing persons 411-13
road traffic fatality 389
imaging see rad iology
immers ion into hot water 326 see also drowning and near drowning immune thrombocytopenia 96 neonatal 98 immunoassays for drugs 259-60 false-positives 268 oral fluids 266 imp act injuries to head 302 non-accidental, signs of 294 road traffic accident 388, 390, 391, 392, 393 speed of impact see speed see also shaken imp act syndrome impac tion fractures 53 impetigo, bullous 20 impress ions (w ith bite marks) bite 431 suspect 431 in vitro fertilization mix-up 4\0 inborn errors of met a bo lism see metabo lic disease indu ced illness see fabricated illness infant(s) and babies (up to I year) death
adequacy of cause of, in SUDI
200-1 bath tub drownings 354 bucket/pa il drownings 353 insta ntaneo us, path ology 205-7 post-moliem examination, babies/infants/young c hildren 145-65 rapid (in recognised illn ess), path o logy 206, 207-8 sudden unexpected see sudde n infant death syndrome; sudden natural death; sudd en unexpected death in infancy swimmin g pool drownin gs 356 see also in fa nticide genetic metabol ic disea se 114, 115 growth chalis boys 475
girls 478
neurological abno rmaliti es see
neurological les ions
newborn see neonates
organ weigh ts 481-3
infanticide 352-3 definition
England and Wales 181
USA/elsewhere 181
drowning 351,352-3,355
scene 141
infections 247-8 ante-moliem \09, 110 bite injuries 428 brain, young children 177 fetal see congenital infec tions fundus haemorrhages seconda ry to 132 sexually-transmitted, tests for 41-2, 404 sudden death 247-8
airw ay infections 241-2
in hospital 372
neonates 209, 372
see also specific pathogens/diseases inflamm ation ante-m ortem \09 SUD! 2 10-11 inflammatory mediators, burns 325 informatio n from abused child 10-11 death (peI"inatal/infant/young child) 146 post-mortem exc hange of (between professionals) 163-4 recording see documentation from road traffic accident involvees 388-9 see also histOly-taking; in terv iew ing informed consent see consent inhalation (aspiration) foreign body, asphyxial death 330-1
of gastIic contents see sto mac h
sm oke 318,320-3
solvents see vo latile substances
water 332
inhelitance see entries under genetic injury/ traum a ch ild accidental see ac cidental injury asphyxia due to 330 birth see birth trauma non-accidental see physical abuse relative rankin g order of deaths due to 346 sce ne of death from 141 maternal , fet a l death due to 190-3
see also specifIc sites and types of injury inqu ests and inquiri es 143 instrumental delivery intracranial ha emo rrhage 174 sk ull fractures 30 1 insulin, levels 244 insulin-dependent diabetes melli tus 244, 245
Index I
intensive care unit
head injury 290-4
road traffic victim dying in 390
intentional acts see non-accidental acts interagency dimensions see multi-agency and multidisciplinary dimensions International Classification of Diseases (ICD)
head injury 283
sudden death - cause unknown
198
interviewing at scene of road traffic
accident 388
intestine (bowel)
enquiry on protection examination
form 461
injury 68
sudden death relating to 239
intracranial haematoma/haemorrhage in abuse 61-3,296,297,306-7
chronic 296, 306
fundus haemorrhage and 130
birth-related 71, 174-5, 189-90, 302
imaging 71,289
differential diagnosis of cause 70
maternal abdominal trauma causing
191-2
perinatal 172-3
birth-related see subheading above
death due to 189-90, 193
in sudden death 236-7, 240
estimation of blood loss 152
in sudden death, examination for
156
infant 218
neonatal 160, 161
traumatic 293-4
birth-related see subheading above
non-accidental see abuse
(subheading above) intracranial pathology, sudden death
237-8
intracranial pressure, raised 130
head injury 291-2, 307
intradural haemorrhage 174, 307
birth-related 174
intrauterine damage and death see
fetus
ischaemic stroke in mild head injury
290
see also hypoxic-ischaemic injury
islets, pancreatic, sum 218
isotope bone scan, abuse 51
Italy, child car restraints 387
jaws, radiography see radiography
jellyfish venom and drowning 358
joint bleeding 88, 89
joint working in abuse cases,
paediatricians- forensic specialists
6-7, 8-9, 11 , 458-69
see also mUlti-agency and
mlJ1tidisciplinary dimensions
karyotype 396
Kawasaki's disease 231-2
instantaneous infant death 205, 206
keratin matrix, drug isolation from
265
ketamine 272
ketones 113
kidney
disease and failure
ante-mortem 113-14
bleeding in 83, 100-1
injury 68
SIDS 211
weight 484
fetal 471
neonatal and infant 481-3
kininogens see high-molecular weight
kininogens
labia minora and majora, examination 26
normal anatomy 29,31
signs of abuse 39
labial frenulum, injury 151,213,423,
424
laboratory tests
biochemical tests on post-mortem
specimens see biochemical tests
in haematological disorders see
haematological tests
microbiological, sudden death
(infants/early childhood) 148-9
samples/specimens for see samples
in sexual abuse, interpreting findings
42-3
labour, prolonged or difficult 187-8
lakes, drowning 357-8
larynx
cyst 207
malformations 193
spasm with immersion 350
laser, portable, road traffic fatality
389
legal issues (inc], legislation)
abuse 1-2, 4, 7-8, 22
dentists'role 421
cycle helmets 393
495
DNA samples 403
fetal/perinatal death 181-2
Leigh's syndrome 177
length see height
lens damage 126
leucodystrophies 238
leucoencephalopathy
multicystic 171-2
telencephalic 171
leucomalacia, periventricular 169, 171,
211
leukaemia 84, 99
fundus haemorrhages 132
lichen sclerosus et atrophicus 20, 40
life support systems, infant deaths
following over 12 hours on 199-200
ligature, hanging by 332
lighting, scene of sudden death 139
lightning 327
likelihood ratios, DNA evidence 402
linea vestibularis 41
lipid (fat)
accumulation in liver, SIDS 211-12
depletion in adrenals, SIDS 212
overload in intravenous feeding
364
see also fat embolism
lipid storage disorders 100, 177
liquid chromatography
high-performance see high
performance liquid
chromatography
ultraperformance 261
liquid-liquid extraction (in toxicology)
260
liver
failure 100-1
injury
abuse 68
at birth 162
siderophages 217
in sudden death, assessment and
findings 154
SIDS 211-12,217
tests, fatty acid oxidation defects
116-17
weight 484
fetal 471-2
neonatal and infant 481-3
local anaesthestics in dentistry 371
long-bone fractures, radiology 52-6
long chain acyl-CoA deficiency 244
long chain 3-hydroxyacyl CoA
dehydrogenase deficiency 244
long QT syndrome 207, 232, 365
496 II: Index
low copy number (LCN) analysis 400
sex ual assault 404
lumba r puncture 368-9
lun gs
in drowning, pathology 350,351
in Feta l/perinata l death,
examination/assessment 185-6
maturity 187
haemorrhage see haemorrhage
oedema, postoperative 371
in sudden death 215-16,240
examination 154, 155, 156
histology 162
inFant, pathology 210
weight 484
Fetal 471 - 2
neonatal and inFant 481-3
lupus anticoagulant 87, 89
lymphadenopathy, SUDI/SlDS 210
lymphoblastiC leukaemia, fundus
haemorrhages 132
lymphohistiocytosis, haemophagocytic
84-5
maceration
Fetal, measuring 471-2
neonatal 158-9, 184
macrocyto pathies 95
macrophage response,
hypox ic-ischaemic injury 167-8
magnetic resonance imag in g (MRi) of
head 12,50-1 , 61-7
accidental injury 293
birth injury 289
non-accidental injury 12, 61-7,
299-300
malabsorption syndromes 101
male children see boys
malignancy, bone marrow infiltration
100
malignant hyperthermia 371
malnutrition, post-mortem assessment
108
manslaughter, drug/anaesthetic errors
366, 367
Marfan's syndrome 235
Marshall CT score 293, 294
mass disaster
dental identification 439
fire 321
mass spectrophotometry
gas chromatography and (GC-MS)
261-2
liquid chromatography and (LC-MS)
262
time-oF-flight 262
mast cell tryptase 109, 110, 239
maternal issues see mother
mattresses 203, 204
May-Hegglin anomaly 81,95
Meadow, Professor Roy 16, 199, 202,
274, 455
meconium
staining 159, 187-8, 193
in stomach 185
toxicology 267
median perineal raphe, congenital
abnormalities 41
medical assessment, abuse see child
abuse
medical conditions/illnesses/ diseases
death due to 225-55
bath tub drownings 356
infant see subheading below
infant death due to 225-55
rapid 207-8
simulating/mim icking abuse 20,
69-70, 71-3,76-105, 13 2
haematological 20, 76-105, 132
sexual abuse 20,40- 1
see also sudden natural death
medical devices/procedures 368-71
medical history
family 10
perinatal/inFant/young child death
146
SUDI 201-2
medium chai n acyl-CoA deficiency
116, 117, 239, 245-6
sudden death 239, 245-6
infants 208
meningitis
coning 369
menin gococcal see meningococcal
infection
meningococcal (N. meningitidis)
infection 247
meningitis 247
fundus ha emorrhages 132
metabolic disease 177
of bone 71,94
genetic (inborn errors of metabolism)
fundus haemorrhages 132,177, 204-7
Reye-like 239
sudden unexpected death see
sudden unexpected death post-mortem investigations 150
acute metabolic autopsy 119-20
inFants/yo un g children 150
metaphyses Fractures 53-6
at birth 71
in scurvy 71
normal variants 70
methadone 273
methylamphetamine, prenatal
exposure 274
methylphenidate 257-8,272-3
micellar electrokinetic capillalY
chromatography (lVlECC) 261
microbiological samples, sudden death
(infants/ea rly childhood) 148-9
micro dialysis catheter 293
microscopic appearance see histo logy
midbrain shearing injury 307
mineralization, brain 168
missing persons, DNA tests on body
411-13
mission statement of expert 448-9
mitochondrial DNA 397, 407-8
analysis 407-8
abandoned body (baby/fetus ) 413
paternity tests 411
diseases of 244-7
babies 177
cardiomyopathy 230
sudden death 244-7
molecular methods, post-mortem 150
see also DNA tests
molecularly imprinted polymers (in
toxicology) 260
monitoring (physiological)
death to failure to monitor 365-6
in head injury 290-4
monozygotic twins 396
Monstrad's dental age estimation
method 443
mosaicism 410
mother
DNA tests with abandoned baby/fetal remains 413
filicide 353
infanticide 352
pregnant
complete separation from, definition 181
drug abuse 257, 267, 273-4
injury causing fetal death 190-3
neonaticide by drowning 352
motor control in abuse
assessmen t 14, 17
ocular 128
motor vehicle accidents see road traffic accidents; vehicles
mouth see oral cavity
mucocutaneo us bleeding 88
multi-agency and multidisci plinary
dimen s ions
abuse 5-6
sudden death
Index I
infan t (SUD1) 201
and post-mortem information
163-4
see also joint working
multichannel EEG, head injulY 292
multicystic leucoencephalopathy
171-2
mummified remains 183
Mun chause n's syndrome by proxy see
fabricated illn ess
murder see homicide
muscle, cardiac see myocardium
mutations 410-11
myelodysplasia 96, 99 - 100
myocarditis 110, 227-8
myocardium (ca rdi ac muscl e)
examination in sudden death 155,
156 infarction with cardiomegaly 233
neurogenic disease 249
tumours 233-5
myristic acid oxidation ass ay
117, 208
myxoma, cardiac 234
napkin dermatitis 217
nasal decongestants a ffe cting
amphetamine tests 268
nasal haemorrhage see epistaxis
nasogastric tubes 368
natural death see sudden natural dea th
nea r miss events see app arent life-
threatening events neck
dissection 154
injury
muscles at birth 161
see also cervical spi ne
necropsy see post-mortem
ex amination
necrosis, brai n cells (in early life) 167
necrotizing enterocolitis 212
necrotizing fasciitis 248
needles, multipl e, penetration by 69
neglect
assessment 16, 458-69
dental 422, 424
Neisseria meningitidis see
meningococcal infection
neonates/newborns
bleeding disorders 97-8
dea th/sudd en de ath 249, 371-2
abandoned body see ab andoned bod y
causes 209
definition 180
in hospital 371-2
post-m0l1em examination 158-62
dental age estimation 442
drug tests on hair 273-4
organ weight 481
neonaticide 181 , 352
drowning 351,352
in hospital 372
neoplasms see tumours nerve fibre layer of retina, h aemorrha ge into 128
nerve injury, bi!1h-related 302
nerve root injury, birth-related 176
neu roblastoma 84
neurogenic heart disease 249
neuroimaging (imaging of head/brain)
60-7, 3 10
in abuse 12,60-7,299-300
CT 12,50, 61-2, 299-300
MRI 12,50-1,61-7,299-300
ultrasound (infant) 50
accidental head injLllY 289-90, 293
birth injury 289
neuroin tensive care, head injulY
290-4
neurological lesions/damage 166-79
electricity-induced 327
infants/early life 166-79
at bil1h see birth trauma clinical manifestations 167
traumatic see brain inju ry
see also brain; periph eral nerve
injury; spinal cord neurones
death 167
migration disorders 170, 217
newborns see neonates
nicotine, neonatal hair testing 273-4
9/11 disaster (2001), identitying
remains 412
non - a cci denta 1/ in te n ti 0 n a 1/ deli bera te
acts
biological speci men alteration 269-70
drownin g 354-6
injury see physical abuse
pOisoning by health professionals
274
non-steroidal anti-inflammatory drugs
99
noxious gases in fires 322
nucleic acid amplification tech niques
(incl. PCR) 400, 403, 407
sexually-transmitted infections 404
nutritional status of deceased 107-8
observing the scene of sudden death
139-40
obstetric events, death related to 248
497
see also pregnancy ocular problems see eye oedema (swelling) bra in/cerebral 66-7,309-10,364-5
ea rly life 167
in hospital 364- 5
imaging 66-7,310
immersion v ictims 351
traumatic causes 66, 309-10
pulmonary, postoperative 371
Offences Against the Person Act (1861)
181
ongoing health care, abuse 21
ophthalmology see eye
opiate tests, false-positives 268
opinion in suspected abuse
expert wit nesses 450
formulating 21
optic nerve lesions 127,1 28
oral cavity (mouth)
examination in sudden death 150-1
fluids in toxicology 266
injuries in abuse, dentists' role
423-4
surgery, deaths 371
organ(s) see viscera
organic acid disorders 115
sudden death 114, 117
orogastric tu bes 368
osteochondrodysplasia 236
osteogenesis imperfecta (brittle bone
disease) 20, 71- 3
expert witnesses and 451
types I-IV 72
osteomyelitis 71
ostial ste nosis 230
overheating, sleeping infant 203, 204,
205
overlaying 203, 2 15, 33 1
~ - ox id at ion see fatty acid ~ -o x idation
defects oxidative phosphorylation (electron
transport ch ai n) defects 116, 244,
246-7
oxygenation
brain, inadequate 310
monitoring status in head injury 29 0
pacemaker
failure 369
post-mortem assessment 377
paediatricians/ pa ediatric doctors in
abuse 22
joint working with forensic
specialists 6-7,8-9,11,458-69
paedophilia-related de aths 142
pail, drowning in 353
498 I
Index
pancreas
histology in sudden death 244
injury 68
islets, SUDl 218
pan creatitis, acute, en ce phalopathy associated with 248-9
paracetamol poison in g 149
paren ts/i m med ia te family
in sudden death, considera tio ns 142-3
ro ad tra ffic acciden t 389
in suspected a buse, presenting
history from 10
see also carers; mother
parotitis, cytomegalovirus 250
pa rtial -thic kness burns 324
partial thromboplastin time with
kaolin see activated partial
th rombo pl astin time
particulates (sm oke) in a irway 320,
32 1
patch testing, sweat 266-7
paternity testing 408- 10, 412
mtDNA 411
mutations affecting 410
Y chromosome 411
pedestrians (in car acc idents) 386,
38 7,388,390,391.392
run over by reversi ng car 330
veh icle inspectio n 388
pelvic fra ctu re 59
penetrating injuries/ tra um a
head 294. 302
ma te rn al abd omen . causing feta l
death 19 2
oral cavity. in abuse 425
v iscera 68-9
penile injury 15
perfusion. cerebral see cerebral perfus ion ; cerebral perfusion pressure perianal area see anus and perianal
a re a
perinatal death 180-97
p ost-mortem exa mination 145-6 5
perin eal raphe. median. co ngen ital
abnormalities 41
perios tea l reaction 51-2
in frac ture repair 50
physiological 51-2. 70
peripheral nerve inju ry. birth-related
302
periventricular leu comalacia 169, 171 .
171 ,2 11
perso nnel
health see carers a nd health
professionals
J t scene of sudden dea t h 138
in tervi ewin g 388
petechiae 84
in fetal/perin ata l dea th
cutan eo us 188
pulmonary 185
sh owers 18
in sudden de at h
in asphyx ia 329-30
neonates 159
observed at scen e 140
in sudd en death in infan ts
(SUDI/SIDS)
co njuncti va l, possibility of non
accidenta l injury 213-14
thymic/ pulmon aJy/coronary artelY
210
ph alangeal injuries 59
pha rm acok in etics 258, 269
Pha rmChek'" 266-7
ph aryn gea l perforation 68- 9
phospholipid an tibody (lupus
anticoagu lan t) 87,89 pho togra phic evidence bite injury
marks 430- 1
susp ec t 431
on child protection examination forms 468
dental iden tificat ion 438
fun du s hae morrh age \ 33 -4
ro ad tra ffi c accident 38 7
sex ual abuse 28
sudd en death 147-8
infants/youn g children 147-8
scene of 138
physical abuse/violence (non acc ide nta l/in te ntional inj ury ; battered child syndrome) child 47-76, 125- 36, 420- 34
airway obs truction 332 -3
assessing pattern s of injury 17-l[l
deaths per 100 000 337
dentist a nd see dentist
differen tial diagnOSis 69 -73
drownin g 354, 354-5
expert witnesses 451
head injury see head injury
medica l co nditions simulating see
medical conditi ons
ocul ar invo lvement see eye
radiolo gy see radiology
SID S and p ath ol ogy raiSing
co ncerns 212-17
spinal injury 312
ther ma l injuries see th erm al injury
mother, ca usin g fetal death 191
see also specific parts of body and types of injury e.g. bruises; frac ture physical exa min ati on in ab us e 11- 12
on child protection exa mination
form 462-7
sex ual a buse 25
te chn iq ue 26-7
see a/so externa l exam inat ion
physiological monitoring see
monitoring
pigmentation, hair, dntg concentration
an d 264, 265
pillows 203. 204
place nta
abruption 192
DNA tests with aband oned
ba by/fetus 4 13
exa min at ion 160-1
we ight 473
plas mi no ge n ac ti vator inhibito r-l deficiency 93
plastic bag asphyxia 331
platelets 78-9
acti vation 78
adhesio n 78
aggregation 78
measurement 82
diso rders 78. 93-4
drug-induced 99
dysfunctional 82 , 99
morphological 8 1, 94-6
neonatal 97-8
numerical 90. 94-6, 97-8
sto ra ge po ol 92
enhancem en t of coagul ation 79
secretion/ rel ease 78-9
measureme nt 82
pl ayground/recreation al area injuries
339,340
fa lls 340
poisoning see toxic substances;
veno ms
polari ty of drugs a nd hair sam ples
264
polycystic kidney disease, autosomal
dominant, subarachnoid
hae morrha ge 236
p olymerase chain reaction (PCR) see nucleic ac id a mplificatio n
p olymicro gy ria 170-1 , 177
polymorphisms 399
restri ction fragment length 398-9
short tandem repeat 399
Sin g le nucl eotide (SNP) 405, 41 3
Index I
polystyrene-filled bean-bags 204
polythene bag asphyxia 331
porencephaly 170
posterior seg ment (eye) injury 126-7
post-mortem
CT scan 50
drug changes 270-1
skeletal survey 48-9, 146
specimens, biochemica l tests
106-26
post-moliem examination (autopsy;
necropsy)
babies/infants/young children 145-65
SUDI 200, 205-18
birth trauma 161-2, 188-90
dental identification and 438-9
fetal/perinatal death 183-90,194-5
injuries 188-90
live birth determination 183-6
standard measurements 471-2
hospital death 375-9
intraoral. in abuse 424
road traffic accident 389-90
technical considerations at time of
117-20
post-mortem interva l 111-12
post-nata l depression 352
potassium levels, post-mortem 110,
111
time of death estimation 111-12
prealbumin, post-mortem evaluation
108
pre-excitation (Wolff-Parkinson
White) syndrome 232, 236
pregnancy
concealed 181. 182,352
drug abuse in 257, 267, 273-4
termination see a bortion
see also fetus; mother; obstetric
events
prekaJikrein 80, 81
deficiency 88, 97
premature babies see pre-term babies
prenatal child see fetus
pre-retinal hae morrhages 128-9
pre-school children see toddlers and
pre-school children
presentation (fetal) 188
presenting history/clin ical presentation
9-10,460
bleeding child 83-4
from parent 10
pre- term/premature babies 248
body weight/height and head
circumference charts
boys 475
girls 478
bronchopulmonary dysplasia 210
sudden death 210, 248
procalcitonin 109
professional witness 449-50
proforma 11, 17,458-69
prone sleeping position 203
protein , post-mortem evaluation 108
protein C deficiency 132
prothrombin deficiency 89, 97
prothrombin time (PT) 81
prolonged 89-90
psychological features of sexual abuse
25
psychological maltre atment (emotional
abuse), assessment 15-16
psychological stress and emotional
trauma
immediate family in sudden death
142- 3
road traffic collision witnesses 388
psychometric tests in abuse 14
public swimming pool drownings 357
pugilistic attitude of body (fires) 319
pulmonary arterial hypertension 242
pulmonary embolism 235, 364
pulmonary non-vascular tissue see
lung pulmonary vascular vasoco nstriction, reflex 350
pulmonalY veno -occlusi ve disease 242
pulp, dental, DNA from 444
pul se oximetry, head injury 290
purpura 83
Henoch-Schonlein 20 , 84, 85, 94
idi opathic thrombocytopenic 94, 96
showers of 18
Purtscher retinopathy 129
pyrexia in brain trauma 292
pyruvate dehydrogenase deficiency
177
QT interval, prolonged (long QT
syndrome) 207, 232, 365
499
abuse 12,47-76
differential diagnosis 69-73
head see neuroimaging
radiologist's role 47
head injury see neuroim aging
immersion victims 347
mod a lities 48-51
see also specific modalities road traffic fatal ity 389
sudden death 146-7
neonatal 158
radiolluclid e bone scan, abuse 51
random match probability 401-2
raphe, median perineal , congenital
abnormalities 41
rear-end collisions 393
records see documentation
recovery room, death in 371
recreational a rea see playground
recreational drugs see drugs
rectal perforation 68
rectal tempera ture 146
referral in abus e, initial 5
reflex anal dilatation 36-7,40
reflex pulmonary vasc ular
vas oconstriction 350
renal injuly/c1 isord ers see kidney
reperfusion injury 212
reporting of abuse
dentists 420-1
written report 16-17
see also docum e ntation
respiratory chain (electron transport
chain; oxidative phosphorylation)
defects 116, 244, 246-7
respiratory depression as adverse sedation event 367
respiratory syncytial virus 242
res piratory trac t
inh a lation into see inh alat ion
in sudden death 240-2 , 243
SUDI and respiratory symp toms
202
see also airways; lung
restraints, child (in cars) 386-7,388,
radicul ar (nerve roots) injury, birth
related 176
radiography
dental Ua ws and teeth) 438-9
prenatal age estimation 441
skeletal survey see skeleton, survey
see also digital dental radiographic
identificat ion
radioimmunoassay for drugs (RIA)
259
radiology 47-76
~ ~ ------~---------~
391
restricted-access material s (RAMs) 260
restriction fragment len gth
pol ymorph isms 398-9
resuscitation (cardiopulmonary; CPR)
changes induced by 212
retinal haemorrhages 129, 132
rib fractures 58,212
immersion victims 351
road traffic acc ident, unsuccessful
388
-----
""'"
500 I
In dex
retinal detachme nt 126
retinal haemorrhages (and over/under
retina)
cardiopulmonary res uscitation
ca using 129, 132
differentia l diagnosis 131
exami nation with 133-4
mech an isms 129
non- accidenta l 296
shaking injury see shaking
retinal haemorrhages (and over/ under
retina) 128-9 13 2, 133
retroph aryngea l abscess 240-1
retro placental haemorrhage 192
reverse suspens ion 333
Rey e's and Reye- like syndrome
238-9
Reyn eJ Developmental La nguage
Scale 14
rhabd omyoma, card iac 233 - 4
rheumatic fever, acute 228
rib fractures 56-8
at birth 71, 161-2,215
CPR-related 58 , 212
post-mortem examination for 154,
155, 161-2
radio logy 56-8
sudden death (infants/early
childhood) 146
SlDS and 212,215
rickets 71 , 215
ri ding injuries 341
ri vers, drowning 357-8
road traffic (motor vehic le) accidents
338, 385-94
asphyxia l death 330
co llision (MYC) 338, 38 7-9 1
fetal injuries and dea th 192- 3,
193,393
conseq uences 338
fatal 385-90
asp hyx ial death 330
epidemiology and causes 385 -7
investigation 38 7-90
pedestrians in see pedestrians
root (tooth) pattern 437
Roth spots 128
route of drug adm inistratio n, wrong
366
RSV 242
Russell's viper venom time, dilute 87
safety (seat) belts 386, 387
sa liva traces with bites, evidence of
429
sa lt/sea water immersion 348, 358
samples/sp ecim ens
collection and processing (for
laboratory tests) 12
for biochemical tests in sudden
death 149
for DNA evidence see DNA tests
for drug detection 260, 262-7
for haematological tests, pitfalls 86
hospi tal deaths 376, 377
for microbiological tests in sudden
death 148 - 9
sa li va traces w ith bites 429
sexual abuse 86
for toxicological tests in sudden
death 149
po st-mortem, biochemical tests see
biochem ica l tests
skin, with bites 431
scalds (hot fluids/ liquids) 324, 355
non-accidental 235-7, 355
scalp
injury
birth-related 173-4
non-accidental 294
neonatal post-mortem examination 159
scapu lar fract ure 59
scene
fire fatalities 141,319-20 sudden unexpected death 13 7-44
asphyxia l deaths 330
babies/young children 145-6
definition 140
good look around 139-40
management 137-8
road traffic collision 38 7
sens itivity and stress of
investigation 142-3
sequence of events 139
schizencephaly 170-1
school-age children
growth charts
boys 477
girls 480
see also adol escents
organ weights 484
Schour and Massier charts 442-3
scintigraphy, abuse 51
Scotland
abuse
bite injury 431
legal dimensions 2, 3, 7
concealment of birth 181
expert witnesses 454- 5
road traffi c casualties and fatalities
390
screening tests
coagulation 8 1
drugs 259-61
doping agents 271-2
pitfalls and limi tations 267-71
ScUlVY 71 , 94
sea/salt-water immersion 348, 358
seats and seat belts (carl, child
386 - 7
Sebastian platelet syndrome 95
second-degree burns 323
security, scene of sudden death 13 9
sedation, adverse o utcomes 367
seizures (epileptic)
retin al haemorrhage 132
sudden death 242-3
by drowning 354, 35 5, 35 7
SUDr and history of epilepsy 203
self-inflicted bite marks 19, 427
self-strangul ation 331-2
semen
evidence of 404, 405
samples 28, 403
septic shock 363-4
septicaemia
H. injluenzae 248
meningococcal 247
serial killing, carers/health
profeSSionals 274, 373 - 5
sewers, drow nings 358
sex
denta l determination 443-4
haemophilia and 83
sex chromosomes see X chromosome;
Y chromosome
sexual abuse 24- 46
definiti on 24
diagnosis/investigation 7, 24-46
conditions simulating 20, 40-1
consisten t vocabulary 29
DNA tests 403 - 7
interpreting findings 42-3
'jigsaw'in 4
skills and experience for 28-9
drug-facilitated 273
epidemiology 2, 24- 5
expe rt witnesses 43, 451
forensic ev idence 27 - 8
sex ually -transm itted in fectio ns , tests
41-2, 404
's haken baby ' syndrome 283, 297
death scen e 141
diagnosis 297
encephalop athy 295, 295-6
hypoxic ischaemi a in 307
shaken impact syndrome 283, 296
. ..
_...l.~_
--
.
~,:'- _I
Index I
shaking 303-6
cervical spine injury see
cervicomedullary syndrome
head injury 303-6
biomechanical determinants 303
mechanism 303
predisposing factors 303
pathophysiology 306
retinal haemorrhage 296, 29 7
arousal shaking 132
unilateral 134
shearing injury, brain 65-6, 307
imaging 65-6
shock
haemorrhagic shock encephalopathy
syndrome 238
hypovolaemic, with burns 325
septi c 363-4
spinal 311-12
toxic shock syndrome 110, 248, 364
short chain acyl-CoA defici ency 244
short tandem repeats (STRs) 399-402
identification of body remains 411,
413
interpreting data 401-2
mutations in 410
paternity testing 408
sexual assault 404
technique 399-401
shunts in hydrocephalus 369
sickle cell disease 240
side collisions 390-1, 393
side s leeping position 203
siderophages
hepatic, SlDS 217
pulmonalY, SlDS 2 15-16
significant harm 4
signs of life, definition 181
simulated illness see fabricated illness
single nucleotide polymorphisms
(SNPs) 405, 413
sinovenous thrombosis, cerebral 177
skeletal dyspl asia 236
skeleton
birth trauma 71
post-mol1em examination 161-2
injuries 51-60
at birth see subheading above
head injury and 298
non-accidental 51 -60, 294-5
see also fractures
normal variants 70
survey (in abuse) 12, 48-51
follow-up 48
post-mortem 48-9, 146
skiing sports injuries 340- 1
#
-
skin
burns see burns
disorders 20
excision of samples with bites 431
see also external exami nation
skull (cranium)
birth-related injury 17 3-4
fractures 60-1, 294, 298
biomechanics 298
birth-related 174, 189,301
non-accidental 60-1 , 294-5
short falls as cause 298
in sudden death (infants/early
childhood) 146
normal variants 70
post-mo rtem exa mination, neon atal
159
birth-related 161
post-mortem opening 156
neonatal 159-60
slap, pa tte rn of injury 18
sleeping environment (SUDI/SIDS and)
203-5
asphyxial deaths 204, 205, 331
in hospital 372
small bowel, sudden dea th relating to
239
smoke (from fire)
detectors/alarms 318-19
inhalation 318,320-3
poisoning 322
smoking
house fires related to 318
passive
neonatal hair testing 2 73-4
SIDS and 203-4
smothering
accid ental (=overlaying) 203 , 215,
33 1
non-accidental 203
snow, spol1S injuries 340-1
social histOlY 10
sodium levels, post-mortem 110-11,
150
sodium val pro ate 99
sofa, s leeping on 205
soft-tissue injury 69
solid-phase ext raction and
microextraction (in toxicology) 260
solvents see vo latile substances
soot inhalation 320,321,322
spa pool drownings see bath tub
drown ings
Spain, child car restraints 387
specimens see samples
--------------------
~-
501
speed (vehicle) of impact
deaths related to 386
determining 387
limit, casualties by 392
sperm/spermatozoa 395
evidence of 404, 405
samples 28
spinal cord injury 311-12
birth-related 176, 302, 312
cervical see cervical spine and
spinal cord trauma
wit hout radiographic abnormalities
(SCIWORA) 311
post-mortem examination 157-8
neonatal 160
spin al trauma 58-9
cervical see cervical spine
perinatal 193
spiral fracture of tibia 53
spleen
injUlY 68
weight 484
fet a l 471-2
neonatal and infant 471-3
sports
dopin g agents see doping agents
win ter, injuries 340-1
sta ture see height
steroids, anabolic 258, 2 71-2
still birth, defin itio n 180-1
stomach
aspiration (of foreign material) fro m
anaesthesia-related 370
SIDS an d 2 16,2 17-18
live birth determination, examination 185
rupture 68
in sudden death
extern al examination 154
infant 2 10
internal examination 15 6
stonefish 358
stools, live birth determina tion
185
strangulation
acc idental 331-2
bruising patterns 18
Streptococcus group A
0-hemolytic 247-8
invasive 248
stress see psychological stress stroke
developing brain 176-7
ischaemic, in mild head injury 290
--------------
502 I
Index
subaponeurotic (subgaJeaJ)
haemo rrhage J 61, 173-4, 30 I
subarachnoid haemorrhage 63, 307
birth-rela te d J 76, 190, 302
fundus haemorrhage and 130,
130-1
post-mortem examination of neonate
for 160
sudden death 236, 237
subconjunctival haemorrhage 126
subdural haematoma/haemorrhage
in abuse 61-3,71,296,297,306
chronic 296, 306
fundus haemorrhage and J30-1
at birth 71,175-6,189,218,289, 302
location 175
natural history 175-6
differential diagnosis of cause 71
perinatal death due to 189, 19 3
in maternal abdominal trauma
191-2
predisposing conditions 306
in sudden death 156
histology 153
infant 218
neonatal 160, 161
traumatic 294
non-accidental see abuse
(subheading above)
subendocardial fibroelastosis 207
subgaleal haemorrhage 161 ,
173, 301
subperiosteal haemorrhage 161
subretinal haemorrhages 128-9
substance misuse see drugs
sudden death - cause unknow n,
leo use of term 198
sudden infant death syndrome (SIDS)
198-224
epidemiology 201-3
expert witnesses 455
genetic metabo lic defects 114,
208
hypoxia 109
non-SIDS or 140
post-mo rtem findings compatible
with conclusion of 209-12
predicting 201
terminology/definitions 140,
198-203
tox icology and 274
see aIso infants ; sudden unexpected
death in pinfancy
sudden natural death 225-55
fetus 193
in hospital 363, 363-5
infant 225-55
pathology 205-9
'intermediate' pathology 249-50
unexplained
distinction from explained sudden
death 226
older children 249
see also medical conditions
sudden une xpected death 137-65
diabetes mellitus 113
genetic metabolic disorders 114-17,
149, 244-7
infant 114, 208
neonate in hospital 372
in hospital see hospital
nervou s system examination 166
older childre n 142,249
post-mortem examination
145-65
scene of see scene
sudden unexpected death in infancy
(SUD! - inc!. babies) 198-224
clothing 139
epidemiology 201-3
genetic metabolic disease 114, 208
histolo gica I examin at ion 162
pathology 205-18
findings compatible with SIDS
209-12
findings of questionable
significance 217-18
findings raising possibility of
non-accidental injury 212-17
terminology/definitions 198-203
see also sudden infant death
syndrome suicid e by perpetrator of homicide see dyadic death in hospital 373
supin e sleeping position 203
surfactant in water inhalation 350
surgery, deaths assoc iated wi th
369-71
investigation 375
sweat, toxicology 266-7
swelling see oedema
swimming pool drownings 356- 7
syphilis, congenital 71
systematic enquiry in abuse 460
systemic diseases, bleeding tendency
100-1
systemic lupus erythematosus, AV
block 232
systemic respon se to burns 325
technetium-99m scan, abuse 51
teenagers see adolescents
teeth see dental identification; dentist;
dentistry tem pera ture abnormal see hypelthermia; hypothermia; pyrexia body
in head injury, monitoring 291-2
rectal, in sudden death 146
environmental, sleeping infant 203
water, drowning and 348-9
see also heat; thermal injury
tentorium, neonatal examination 160
Terso n's syndrome 130
tet racyclin e and dental age estimatio n
443
THO I (short tandem repeat) 388
thalamus, hypoxic-ischaemic dam age
172
therapeutic misadventures, death 363,
366-71
thermal (heat-induced) injury or death
318-28
non-accidenta l 325-7
dentist's role 423
pathologist's role 319-27
patterns 20
see also burns
th ird-d egree burns 323-4
thorax see chest
thrombasthenia, Glanz mann 's 78, 84,
93-4
thrombin time (TT) 81
long 89-90
thrombocytopenia
immune see immune
thrombocytop enia
neonatal 97-8
thrombocytopenic purpura, idiopathic
94, 96
thromboembolism 235,364
thromboplastin see activated paltial
thromboplastin time
thrombosis, cerebral sinovenous 177
thromophilias, inherited 364
thymus
in sudden death 154
infant 210,211
weight
fetal 471-2
neonatal and infant 481-3
tibial fractures 53, 54, 55, 56, 57
time (determination)
of brain damage in early life 167
of death 111-12
- - :,
._ .
. Index I
of sudd en death 139
infant 199
ti me-o f-fli g ht mass spec trophotometry
262
toddlers and pre-sc hool children (1-4
year olds)
bath tub drownings 354
bucket/pail drow nings 3 53
dental age estimation 442- 3
growth cha rts
boys 476
g irls 479
organ weights 484
swi mming pool d rowni ngs 356-7
toilets, deli very into UlJ, 352
touch in g, non-consensua l 42 5
tox ic shock syndro me 110, 248, 364
toxic sub sta nces (inc l. drugs and
poison s) 256-81, 342
bleeding due to 98-9
epidemiology 256-8, 337
fi res 322
infants/early childh oo d deaths 149
sta tistics on death s from 337
testing 259 -74
ca rer-associated seria l killin g 374
immersion victi ms 347
pitfalls and limitations 267 -71
sa mpl es 260, 262-7
specific app lication s 271-4
s pecific cases 274
techniqu es 259-62
see also drugs; venoms
toxicity due to do sage errors 366-7
tox ins, bac terial, assays 110
t race evidence, road traffic collision
387
tracheal ab normal iti es 240
tracheal tub es 368
tra cheitis, bacterial 241
tracheostomy 368
tra ctor-related injuries 34 1
transc ranial Doppler, head inju ry
292 -3
tran sp lantation, organ donation for
390
tran spo rt-related deaths 337
traum a see injury
tren ches, drownings 358
tryptase, mast cell 109 , 110, 239
tub ero us scl eros is, cardi ovascular
pat holo gy 233, 236
tumours
cardiac 233-5
embol ism 364
intracranial 23 7
twins
monozygoti c 396
simultaneous SUD! 201
Uhl's anomaly 23 0
ultrape rform ance liquid
chromatography 261
ultrasound 49-50
see also transcranial Dopple r
um bilical cord examination 159
insertion/s tump 184- 5, J8 8
length 474
'unasce rtained' or 'und eterm ined'
ca use of dea th 219, 377
uncloth ing body
road traffic fatal ity in hospital,
retrieval 388
scene of sudd en death 139
'undetermined' cause of death 219,
377
United States, repo rting of ab use 421
uraemi a, vitreous humour a nalysis III
urea (and urea nitrogen), post-mortem
111 , 114, 11 8, 150
urea cyc le disorders 11 5
sudd en death 117
urethral prolapse 41
urin ary bladder, SJDS 2 10
urinary tract enquiry on protection
exa mination form 461
urine sa mples
collection 118- 19
tests
acute metabo lic a utop sy J] 9
dru gs 262, 262 -3, 272
in fatty acid oxidation defects 116
USA, rep0l1ing of ab use 42 1
vacuum delivery, intracranial
haemorrha ge 174
vaginal examination 27
norm a l an atomy and vari ants (incl.
vest ibule/posterior fo urchette)
29 - 30
signs of abuse 39
see also vulvovag initis
variant anatomy see a natomica l
vari ants
vas cular endothelial growt h factor 109
vasc ulature
ca theter-related deat hs 368
electrical injury 327
oc ular, damage in non-accide nta l
injury 129
vasocon strictio n, refl ex pulmonary
vascular 350
503
VEGF 109
vehi cles
agricul tural 341
road see car; ro ad traffic acc idents
ve noms, sea creatures, a nd drowning
358
veno-o ccl usive di sease, pulmona ry
242
venous ca theter-related deaths 368
venous embolism
air 364
thrombotic 364
venous infa rction, brain parenchyma
173
ve nous pressure, intraocul ar, raised
129
ven ous throm bosis, cerebral 177
vent ilation, ass isted, cha nge due to
21 2
ventouse delivery, intrac rani al
haemorrhage 17 4
ventricles (brain), 3rd, colloid
cyst 23 7
ventricles (hea rt)
arrhythmogenic righ t ventri cu lar
cardiomyopathy 229-30
septal defect 227
v entriculoatrial and
ventriculoperiton eal shunts 369
vernix caseosa, dru g test in g 274
vertebral fra cture 58 - 9
vestibule, vaginal
congenital ab normalities 4J
normal findings 29-30
video recordings
col poscopy 7, 8, 11 , 27
sce ne of sudde n death 138
surveillan ce in suspected
ab use 332
v iolence, scene o f death from 142
see also physical abuse
vira l infection
ante-mortem 110
bites 428
bra in 177
neo natal, causi ng death 209
virological samples, sudden death
148-9
v isceralorga ns
donation for transplantation 390
injuries 68-9
post-mortem rete ntio n 163
road traffic acc ident 389- 90
SJDS 215
surgical traction on, causi ng ca rdiac
arres t 369
504 I
Index
viscera/organs (continued) weigh t (post-mortem) 150
fetal 471-2
neonatal/infant 481-3
visual pathways, non-accidental injury
127-8
visual sequelae of ocular non
accidental injury. long-term 125,
134
vitamin C deficiency and scurvy 71,
94
vitamin K
deficiency 98
malabsorption 101
vitreous humour 117-18,271
acute metabolic autopsy 120
electrolytes 110-11
glucose 112, 113, 150, 244
haemorrhage into 129
vocal cords, SIDS 210
volatile substances (incl. solvents)
analysis 260, 262
inhalation/abuse 262,333,341-2
road traffic fatalities 389
von Willebrand disease 82, 90-2
family history 84
neonatal 97
types 1-3 90-2
von Willebrand factor 79
qualitative abnormalities 90, 91
quantitative abnormalities (incl.
deficiency) 88, 90, 90-2
vulval signs of abuse 39
vulvovaginitis 40
Wales, road traffic casualties and fatalities 390
warfarin 89, 90
water
hot see scalds
loss with burns 324
submersion in see drowning and
near-drowning; immersion
waterbeds 331
watersiled damage 172, 311
wedging 204, 331
weight assessment
child
abuse 13
chalts 475-80
post-mortem 150
organs see viscera placenta 473
whiplash injury, road accidents 393
whiplash syndrome (whiplash shaking
injury) see cervicomedullary syndrome white matter
non-traumatic lesions 171-2
shearing injuries 307
see also leucodystrophies;
periventricular leucomalacia whole-body radiograph of baby see babygram
Williams ' syndrome 227
winter sports injuries 340-1
Wiskott-Aldrich syndrome 81,92,94
witness
expert see expert witness
eye, road traffic collision 387
professional 449-50
Wolff-Parkinson-White syndrome
232, 236
Working together 421
World Trade Center disaster (2001),
identifying remains 412
X chromosome 395
Y chromosome regions homologous
to 400-\
X-ray see radiography; skeleton, survey
Y chromosome 395
abandoned baby/fetal remains 413
paternity testing 411
STR analysis 405-6
sexual abuse 404, 405-6
X chromosome regions homologous
to 400-1
young children see infants; neonates;
toddlers and pre-school children
-~
-
.