Textbook of Rapid Response Systems: Concept and Implementation

Textbook of Rapid Response Systems

Michael A. DeVita    Ken Hillman Rinaldo Bellomo ●

Editors

Textbook of Rapid Response Systems Concept and Implementation

Editors Michael A. DeVita, MD Professor, Critical Care Medicine University of Pittsburgh School of Medicine Department of Critical Medicine Pittsburgh, PA USA Rinaldo Bellomo, MD Chair, Australian and New Zealand Intensive Care Research Centre Professor, Faculty of Medicine University of Melbourne Honorary Professor, Faculty of Medicine Monash University Honorary Professor, Faculty of Medicine The University of Sydney Honorary Principal Research Fellow Howard Florey Institute University of Melbourne Director of Intensive Care Research Staff Specialist in Intensie Care Department of Intensive Care Austin Hospital Melbourne, Australia

Ken Hillman, MBBS, MD, FRCA (Eng), FCICM Professor of Intensive Care Director of the Simpson Centre for Health System Research The University of New South Wales The Australian Institute of Health Innovation The University of New South Wales Liverpool, Australia

ISBN 978-0-387-92852-4 e-ISBN 978-0-387-92853-1 DOI 10.1007/978-0-387-92853-1 Springer New York Dordrecht Heidelberg London © Springer Science+Business Media, LLC 2011 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)

To my wife Sharon, and my children Lizzie, Chris, Tim, and Annie who are my life, and to my parents for challenging me and preparing me to achieve. I am indebted to my teachers, Dennis Greenbaum, Jim Snyder, Ake Grenvik, and Richard Simmons, who have taught me what critical care excellence is. MAD To Sue Williams, a loyal and committed colleague of 20 years. KH To the many patients who have suffered and suffer because of an outdated approach to acute hospital care. May this book help make their time in hospital safer. RB

Preface

When we created the Manual of Medical Emergency Teams (MET), Implementation and Outcome Measurement, the concept was relatively new. The term MET was not well recognized and had different meanings to different people. The Institute for Healthcare Improvement (IHI) was just starting to promote the same process using different terminology: the Rapid Response Team (RRT). We created our manual to provide readers with the tools to create teams of their own, and with the resources needed to measurably succeed. Since then, even though it was a mere four years ago, much has happened. Two consensus conferences have been held. The first defined the terminology and the process better. At that meeting, it was recognized for the first time that the intervention is much more than a team. It is a system, the Rapid Response System (RRS), and it has four components, without which the team is unlikely to succeed. A unifying set of terms was created, and a lexicon for describing interventions and reporting outcomes was proposed and is now being utilized more frequently. The aim was to promote the ability to compare interventions between organizations and more accurately analyze results. We assumed that it is possible that an intervention that has nursing responders might have different results than interventions that have physician responders. Differing support structures in the administrative or quality improvement limbs might also impact outcomes. The ability to compare, we believed, was essential to understanding the influence of various components of the system and would enable improvement and growth internationally. At the second consensus conference, which was focused on defining and improving the ability to detect crises outside of the critical care setting (that is, the “afferent limb”), the need for monitoring was explored and a classification system was proposed. Participants recognized the differences between continuous and intermittent monitoring in terms of costs, equipment, and capabilities. Two major purposes of monitoring (continuous or intermittent) were recognized. First, it enables prognostic risk stratification. There is now ample data to show that those with abnormal vital signs are more likely to suffer a serious or lethal event. This risk stratification can facilitate the ability to move patients to a level of staffing and equipment to promote safety. Of course, these prognostic systems are not ­perfect. Patients are identified who in fact do not have an event, while others who will have an event are not recognized. As a consequence, many at the conference promoted vii

viii

Preface

the second function of a monitoring system, the ability to detect critical deteriorations as they occur. This function, of course, requires a continuous monitoring system. Such a system costs more, but it may reduce unexpected mortality and morbidity in hospitalized patients. Future studies will clarify the cost and the benefit of the two approaches. We hope our textbook clearly recounts the issues involved and promotes needed investigations. A second major event that has transpired is the addition by The Joint Commission for the Accreditation of Healthcare Organizations (JCAHO) in North America of a patient safety goal that includes the principles of the Rapid Response System. Healthcare providers must have the capability to detect patients with sudden clinical deterioration and a system in place to rapidly respond to the patient’s needs. These days, most hospitals in the United States have some sort of RRS, with some teams being nurse-led (RRT), and others physician-led (MET). Still others did not create a new team but “hot-wired” existing systems to promote the goal. The use of the RRS is spreading in Europe, Asia, and Central America. In short, in just a few years, it is becoming a global standard. Recognizing the importance of the system, reports are published in major journals with increasing frequency, and it is becoming difficult to keep abreast of the field as it is growing so rapidly. A lively debate exists regarding the benefit of the RRS. In this textbook, we have tried to capture that debate, including potential emerging applications of the system. A year ago, we decided to update our manual to make it current with recent publications. However, as the project progressed, it became obvious that the field had expanded so significantly that more than an update was required. Instead, we chose to create a new work, the first “Textbook of Rapid Response Systems.” We added over a dozen chapters, recruited many new researchers in the field, and have tried to create a comprehensive resource for the clinician, administrator, and researcher. As in our first book, we tried to make each chapter capable of standing alone, yet tightly integrated with the others into a cohesive whole. The RRS concept is now being applied to a number of problems and processes in hospitals that require a “short circuit” to timely and expert help. When we each created our RRS at our home institutions, none of us recognized the many problems that might be amenable to the approach or the potential applications of the system. In this textbook we describe an RRS to rapidly and safely find lost, eloped, or wandering patients, another to immediately support staff who have suffered mental or psychological trauma in the course of their work, and yet another to provide support for suddenly dangerous patients, staff, or visitors. These problems share the common need for early recognition that an RRS provides using additional expertise and hands (afferent limb), and a planned, systematic approach to provide support (efferent limb). The ability to recognize the need, to capture and analyze data to guide change (quality improvement limb), and to provide resources (administrative limb) is also necessary to each application’s success. In a sense, the RRS is becoming, in some hospitals, a “system of systems.” In this textbook, we hope to provide readers with the tools they need to create systems that emulate some of the work reported in the book, and to discuss the expansion of the RRS to address other critical needs.

Contents

Part I  RRSs and Patient Safety   1  Rapid Response Systems History and Terminology............................. Bradford D. Winters and Michael DeVita

3

  2  RRS’s General Principles........................................................................ Ajay D. Rao and Michael DeVita

13

  3  Measuring and Improving Safety........................................................... Bradford D. Winters, Peter J. Pronovost, Marlene Miller, and Elizabeth A. Hunt

19

  4  Integrating a Rapid Response System into a Patient Safety Program........................................................................ John Gosbee

37

  5  Acute Hospitalist Medicine and the Rapid Response System.............. David J. McAdams

47

  6  Medical Trainees and Patient Safety...................................................... Stephen W. Lam and Arthas Flabouris

55

  7  Rapid Response Systems: A Review of the Evidence........................... Bradford D. Winters and Julius C. Pham

65

  8  Healthcare Systems and Their (Lack of ) Integration.......................... Ken Hillman, Jeffrey Braithwaite, and Jack Chen

79

  9  Creating Process and Policy Change in Healthcare.............................. Stuart F. Reynolds and Bernard Lawless

87

ix

x

Contents

10  The Challenge of Predicting In-Hospital Cardiac Arrests and Deaths................................................................................... Michael Buist

93

11  The Meaning of Vital Signs..................................................................... 109 John Kellett, Breda Deane, and Margaret Gleeson 12  Matching Illness Severity with Level of Care........................................ 125 Gary B. Smith and Juliane Kause 13  Causes of Failure to Rescue..................................................................... 141 Marilyn Hravnak, Andrea Schmid, Lora Ott, and Michael R. Pinsky Part II  Creating an RRS 14  Impact of Hospital Size and Location on Feasibility of RRS............................................................................... 153 Daryl Jones and Rinaldo Bellomo 15  Barriers to the Implementation of RRS................................................. 163 Michael A. DeVita and Ken Hillman 16  An Overview of the Afferent Limb......................................................... 177 Gary B. Smith and David R. Prytherch 17  The Impact of Delayed RRS Activation................................................. 189 Daryl Jones, Michael Haase, and Rinaldo Bellomo 18  The Case for Family Activation of the RRS.......................................... 197 Helen Haskell 19  RRT: Nurse-Led RRSs............................................................................ 207 Kathy D. Duncan and Christiane Levine 20  MET: Physician-Led RRSs..................................................................... 221 Daryl Jones and Rinaldo Bellomo 21  Pediatric RRSs......................................................................................... 231 James Tibballs and Richard J. Brilli 22  Sepsis Response Team.............................................................................. 245 Emanuel P. Rivers, David Amponsah, and Victor Coba

Contents

xi

23  Other Efferent Limb Teams: (BAT, DAT, M, H, and Trauma)............ 253 Daniel Shearn and Michael DeVita 24  Other Efferent Limb Teams: Crisis Response for Obstetric Patients............................................................................... 263 Gabriella G. Gosman, Hyagriv N. Simhan, Karen Stein, Patricia Dalby, and Marie Baldisseri 25  Personnel Resources for Responding Teams......................................... 275 Andrew W. Murray, Michael A. DeVita, and John J. Shaefer III 26  Equipment, Medications, and Supplies for an RRS............................. 291 Edgar Delgado, Wendeline J. Grbach, Joanne Kowiatek, and Michael DeVita 27  The Administrative Limb........................................................................ 313 Daryl Jones and Rinaldo Bellomo 28  The Second Victim................................................................................... 321 Susan D. Scott, Laura E. Hirschinger, Myra McCoig, Karen Cox, Kristin Hahn-Cover, and Leslie W. Hall Part III  Monitoring of Efficacy and New Challenges 29  RRSs in Teaching Hospitals.................................................................... 333 Max Bell and David Konrad 30  The Nurse’s View of RRS........................................................................ 341 Donna Goldsmith and Nicolette C. Mininni 31  Resident Training and RRSs................................................................... 347 Geoffrey K. Lighthall 32  Optimizing RRSs Through Simulation.................................................. 357 Melinda Fiedor Hamilton, Elizabeth A. Hunt, and Michael A. DeVita 33  Evaluating Effectiveness of Complex System Interventions................ 371 Jack Chen 34  RRS Education for Ward Staff............................................................... 381 John R. Welch and Gary B. Smith 35  Standardized Process and Outcome Assessment Tool.......................... 397 Gabriella Jäderling and David Konrad

xii

Contents

36  The Impact of RRSs on Choosing “Not-for-Resuscitation” Status............................................................... 405 Arthas Flabouris and Jack Chen 37  The Costs and the Savings....................................................................... 415 Dana Edelson and Rinaldo Bellomo Index.................................................................................................................. 429

Contributors

David Amponsah, MD Ultrasound Director, Senior Staff Physician, Department of Emergency Medicine, Henry Ford Hospital, Detroit MI, USA Marie Baldisseri, MD Associate Professor, Department of Critical Care Medicine, University of Pittsburgh School of Medicine, Pittsburgh PA, USA Max Bell, MD, PhD, DESA Department of Anesthesia and Intensive Care, Karolinska University Hospital, Solna, Stockholm, Sweden Rinaldo Bellomo, MD Department of Intensive Care, Austin Hospital, Heidelberg, 3084, VIC, Australia Jeffrey Braithwaite, PhD, MBA, FCHSM, FAIM Professor and Director, The Australian Institute of Health Innovation, Centre for Clinical Governance Research, The University of New South Wales, Sydney NSW, Australia Richard J. Brilli, MD, FCCM, FAAP Chief Medical Officer, Nationwide Children’s Hospital, Columbus OH, USA Michael Buist, MB, ChB, FRACP, FJFICM, MD Rural Clinical School, University of Tasmania, Private Bag 3513, Burnie, TAS 7320, Australia Jack Chen, MBBS, PhD, MBA The Simpson Centre for Health Systems Research, The University of New South Wales, Sydney NSW, Australia and The Australian Institute of Health Innovation, The University of New South Wales, Sydney NSW, Australia Victor Coba, MD Senior Staff Physician, Emergency Medicine, Critical Care Medicine, Henry Ford Hospital, Detroit MI, USA xiii

xiv

Contributors

Karen Cox, PhD, RN Manager, Quality Improvement and Patient Safety Office of Clinical Effectiveness, University of Missouri Health Center, Columbia MO, USA Patricia Dalby, MD Assistant Professor, Department of Anesthesiology, Magee-Womens Hospital, University of Pittsburgh School of Medicine, Pittsburgh PA, USA Breda Deane Data Collection Officer, Nenagh Hospital, Nenagh Co. Tipperary, Ireland Edgar Delgado, BS, RRT Director, Respiratory Care, UPMC Presbyterian Shadyside, Pittsburgh PA, USA Kathy D. Duncan, RN Faculty, Institute for Healthcare Improvement, Cambridge MA, USA Dana Edelson, MD, MS Assistant Professor, Department of Medicine, University of Chicago Medical Center, Chicago IL, USA Arthas Flabouris, MBBS, MD, FCICM, FANZCA Intensive Care Unit, Royal Adelaide Hospital, North Terrace, Adelaide, SA 5001, Australia Margaret Gleeson Clinical Nurse Manager, Nenagh Hospital, Nenagh Co. Tipperary, Ireland Donna Goldsmith, RN, MN, CCRN Austin Health, Heidelberg, VIC, Australia John Gosbee, MD, MS Department of Obstetrics, Gynecology and Reproductive Sciences, University of Pittsburgh School of Medicine, Magee-Womens Hospital, Room 2314, 300 Halket Street, Pittsburgh, PA 15213, USA Gabriella G. Gosman, MD Associate Professor, Department of Obstetrics, Gynecology and Reproductive Sciences, Magee-Womens Hospital, University of Pittsburgh School of Medicine, Pittsburgh PA, USA Wendeline J. Grbach, MSN, RN, CCRN, CLNC Curriculum Developer for Simulation Education, UPMC Presbyterian Shadyside School of Nursing, Pittsburgh PA, USA Michael Haase, MD Assistant Professor, Department of Nephrology and Intensive Care, Charité – University Medicine Berlin, Berlin, Germany Kristin Hahn-Cover, MD Medical Director, Office of Clinical Effectiveness, University of Missouri Health Center, Columbia MO, USA

Contributors

Leslie W. Hall, MD Chief Medical Officer, Senior Associate Director of Clinical Affairs, Dean’s Office, University of Missouri Health Center, Columbia MO, USA Melinda F. Hamilton, MD, MSc Department of Critical Care Medicine, University of Pittsburgh Medical Center Director, Pediatric Simulation, Peter M. Winter Institute for Simulation, Education, and Research (WISER) and Children’s Hospital of Pittsburgh Simulation Center, Pittsburgh, PA, USA Helen Haskell, MA Mothers Against Medical Error, 155 S. Bull St., Columbia, SC 29205, USA Laura E. Hirschinger, RN, MSN, AHN-BC Clinical Improvement Specialist, Office of Clinical Effectiveness, University of Missouri Health Center, Columbia MO, USA Marilyn Hravnak, RN, PhD, ACNP-BC, FCCM, FAAN School of Nursing, University of Pittsburgh, 3500 Victoria Street, 336 Victoria Building, Pittsburgh, PA 15261, USA Elizabeth A. Hunt, MD, MPH, PhD Pediatric Intensivist, Assistant Professor, Department of Anesthesiology and Critical Care Medicine, The Johns Hopkins University School of Medicine, Baltimore MD, USA Gabriella Jäderling, MD, DESA Department of Anesthesia and Intensive Care, Karolinska University Hospital, Solna Stockholm, Sweden Daryl Jones, MD, BSc (Hons), MBBS, FRACP, FCICM Department of Intensive Care, Austin Hospital, Studley Road, Heidelberg, VIC 3084, Australia Juliane Kause, BSc, MBBS, MRCP, AKC Consultant in Critical Care Medicine and Acute Medicine, Medical Assessment Unit and Department of Critical Care, Queen Alexandra Hospital, Portsmouth Hospitals NHS Trust, Portsmouth Hampshire, UK John Kellett, MD Department of Medicine, Nenagh Hospital, Nenagh, County Tipperary, Ireland David Konrad, MD, PhD, DEAA Director of Thoracic ICU, Department of Cardiothoracic Surgery and Anesthesiology, Karolinska University Hospital, Solna Stockholm, Sweden Joanne G. Kowiatek, RPH, MPM Pharmacy Manager, Medication Patient Safety, Department of Pharmacy and Therapeutics, University of Pittsburgh Medical Center Presbyterian Shadyside, Pittsburgh PA, USA

xv

xvi

Contributors

Stephen W. Lam, MBBS (Hons), FRACP, FJFICM University of Adelaide, Intensive Care Unit, Royal Adelaide Hospital, North Tce, Adelaide, SA 5001, Australia Bernard Lawless, MD, MHSc, FRCS Assistant Professor, Adjunct Scientist, Provincial Lead, Department of Surgery, St. Michael’s Hospital, University of Toronto, Toronto ON, Canada and Li Ka Shing Knowledge Institute, St. Michael’s Hospital, University of Toronto, Toronto ON, Canada and Critical Care and Trauma, Ontario Ministry of Health and Long-Term Care, St. Michael’s Hospital, University of Toronto, Toronto ON, Canada Christiane Levine, RN, BSN, BS Program Manager of Quality and Patient Safety, Children’s Healthcare of Atlanta at Hughes Spalding, Atlanta GA, USA Geoffrey K. Lighthall, PhD, MD Associate Professor of Anesthesia and Critical Care, Department of Anesthesia, Stanford University School of Medicine, Palo Alto CA 94305, USA David J. McAdams, MD, MS, SFHM, FACP Department of Internal Medicine, University of Pittsburgh Medical Center, 200 Lothrop Street, MUH 826E, Pittsburgh, PA, USA Myra McCoig Coordinator, Risk Management, University of Missouri Health Center, Columbia MO, USA Marlene Miller, MD, MSc Vice Chair of Quality and Safety Initiatives, Associate Professor, Johns Hopkins Children’s Center, Johns Hopkins University, Baltimore MD, USA and Department of Pediatrics, Johns Hopkins University, Baltimore MD, USA and Department of Health Policy and Management, Bloomberg School of Public Health, Johns Hopkins University, Baltimore MD, USA Nicolette C. Mininni, RN, MEd, CCRN Advanced Practice Nurse, Critical Care Department of Nursing Education and Research, UPMC Shadyside, Pittsburgh PA, USA Andrew W. Murray, MD, ChB Department of Anesthesiology, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213, USA

Contributors

xvii

Lora Ott, RN, MSN, BSN School of Nursing, University of Pittsburgh, Pittsburgh PA, USA Julius C. Pham, MD Assistant Professor, Departments of Emergency Medicine, and Anesthesia and Critical Care Medicine, The Johns Hopkins University School of Medicine, Baltimore MD, USA Michael R. Pinsky, MD, FCCP, FCCM University of Pittsburgh, Pittsburgh PA, USA Peter J. Provonost, MD, PhD, FCCM Professor, Director, Departments of Anesthesiology and Critical Care, Surgery, and Health Policy and Management, The Johns Hopkins University School of Medicine, Baltimore MD, USA and Quality and Safety Research Group, Medical Center for Innovations in Quality Patient Care, Baltimore MD, USA David R. Prytherch, MD, PhD, MIPEM, CSci Senior Research Fellow, Queen Alexandra Hospital, Portsmouth Hospitals NHS Trust, Portsmouth Hampshire, UK Ajay D. Rao, MD Division of Endocrinology, Diabetes, and Hypertension, Brigham and Women’s Hospital, Boston MA, USA Stuart F. Reynolds, MD, FCCP Associate Clinical Professor of Critical Care Medicine, University of Alberta Hospital, 3C1.12 WMC, 8440 112 Street, Edmonton, Alberta, Canada Emanuel P. Rivers, MD Clinical Professor, Vice Chairman, and Research Director, Emergency Medicine and Surgery, Henry Ford Hospital, Wayne State University, Detroit MI, USA Andrea Schmid, RN, PhD School of Nursing, University of Pittsburgh, Pittsburgh PA, USA Susan D. Scott, RN, MSN, CLNC University of Missouri Health System, Office of Clinical Effectiveness, MU Sinclair School of Nursing, Doctoral Student, Columbia, MO, USA John J. Shaefer, III, MD Director, Professor, Lewis W. Haskell Blackman Endowed Chair, Medical University of South Carolina, Charleston SC, USA and Healthcare Simulation, South Carolina, Medical University of South Carolina, Charleston SC, USA and

xviii

Contributors

Department of Anesthesia and Perioperative Medicine, Medical University of South Carolina, Charleston SC, USA Daniel Shearn, RN, MSN UPMC Presbyterian Hospital, Pittsburgh PA 15213, USA Hyagriv N. Simhan, MD, MSCR Associate Professor, Department of Obstetrics, Gynecology, and Reproductive Systems, Magee-Womens Hospital, University of Pittsburgh School of Medicine, Pittsburgh PA, USA Gary B. Smith, MD, FRCAP Consultant in Critical Care Medicine, Department of Intensive Care Medicine, Queen Alexandra Hospital, Portsmouth, UK Karen Stein RN, BSN, MSED, CCRN Clinical Education Specialist, Department of Patient Care Services, Magee-Womens Hospital, University of Pittsburgh School of Medicine, Pittsburgh PA, USA James Tibballs, BMedSc, MBBS, MEd, MBA, MD, MHlth & MedLaw, GDipArts, FANZCA, FJFICM, FACLM Associate Professor, Intensive Care Physician, and Resuscitation Officer, Department of Pediatrics, Pediatric Intensive Care Unit, Royal Children’s Hospital, Melbourne VIC, Australia and Department of Pharmacology, University of Melbourne, Melbourne VIC, Australia John Welch, RN, BSc, MSc Consultant Nurse, Department of Critical Care, University College London Hospitals, NHS Foundation Trust, London, UK Bradford D. Winters, MD, PhD Assistant Professor, Departments of Anesthesiology, Critical Care Medicine and Surgery, The Johns Hopkins University School of Medicine, Meyer 297 600 N Wolfe Street, Baltimore, MD 21298, USA e-mail: [email protected]

Part I

RRSs and Patient Safety

Chapter 1

Rapid Response Systems History and Terminology Bradford D. Winters and Michael DeVita

Keywords  Rapid • Response • Systems • History • Terminology

Principles The Rapid Response System (RRS) concept has matured substantially since its inception in the early 1990s when critical care physicians, primarily in Australia, Pittsburgh, PA, and the UK started asking some crucial questions regarding patients who deteriorated and often arrested on general hospital wards prior to their admission to the ICU. Specifically, they asked exactly what is happening to general hospital ward patients in the minutes and hours prior to their cardio-respiratory arrests and whether we can do something to intervene and halt these deteriorations before the patient arrests or nearly arrests. This was a sea-change in thought and perspective since, at that time, resources focused on resuscitation were primarily concerned with how to improve performance of CPR and ACLS rather than preventing the event to start with. Critical Care physicians were well aware, in a general sense, that patients admitted or readmitted to the ICU from the general ward uncommonly went from “just fine” to critically ill. This sense was confirmed by early studies that clearly showed that arrests and deteriorations were not sudden but rather commonly heralded by long periods of obvious hemodynamic and respiratory instability that was often unappreciated by general ward providers.1–17 The development of critical illness on the general ward was rarely “sudden,” only suddenly recognized. Given this result, critical care physicians reasoned that if we could create usable criteria for general ward staff to use in the early recognition of impending deterioration and empower these staff members to bring a team of critical care physicians

B.D. Winters (*) Departments of Anesthesiology, Critical Care Medicine and Surgery, The Johns Hopkins University School of Medicine, Meyer 297 600 N Wolfe Street, Baltimore, MD 21298, USA e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_1, © Springer Science+Business Media, LLC 2011

3

4

B.D. Winters and M. DeVita

and/or nurses outside of the ICU to the bedside, we could improve the outcomes of these patients. Data from these studies of the antecedents to arrest provided the basis for developing the physiological criteria that were put into place for general ward staff to use as a guide for making the decision to call for help.13–17 Intensive Care Unit staff formed the team that would come to the patient’s bedside to evaluate, stabilize and help create a new care and triage plan. Through sheer will, and often working in isolation without the support of the health community, these critical care clinicians created a new patient safety and quality initiative, long before patient safety and quality became a national and international concern and had the attention of the public and policymakers.18–49 These early programs were often referred to as Medical Emergency Teams (METs), although other terms, such as Condition C Teams and Critical Care Outreach Teams were also used. This linking defined activation criteria to a response team and empowering the ward staff to summon that team, has become a powerful patient safety and quality initiative that has enjoyed wide adoption in the US, Australia, New Zealand, Canada and the UK and ever-increasing acceptance around the world, such that the first International Conference on Medical Emergency Teams was held in Pittsburgh, PA, in May, 2005. Subsequent conferences have been held each year again in Pittsburgh and also in Toronto, Canada, and Copenhagen, Denmark. The RRS concept has reached such significance that the Institutes for Healthcare Improvement (IHI) included Rapid Response Systems as one of its six “planks” for improving patient in its “100,000 Lives” Campaign.50 Additionally, the essential principles of the RRS have recently been embraced by the Joint Commission (the accreditation organization for US hospitals) as a mandate for American hospitals in the form of National Patient Safety Goal 16 and 16A.51 While this goal does not specifically ask hospitals to create response teams or dictate activation criteria, it requires hospitals to develop systems that “improve recognition and response to changes in a patient’s condition with the organization selecting suitable methods that enable healthcare staff members to directly request additional assistance from a specially trained individual or individuals when a patient’s condition appears to be deteriorating.” Clearly, while the Joint Commission mandate does not demand RRSs by name, RRSs are the logical solution for ­meeting this requirement and providing patients with the safety net they need to help prevent medical deteriorations from progressing and degenerating into an arrest and death. While other solutions have been proposed, such as increased nurse-to-patient ratios, hospitalist services and others, none has the practicality and body of evidence in their favor like RRSs. One of the primary goals of RRSs is to prevent cardio-respiratory arrest and therefore the very high mortality known to be associated with such in-hospital events. Since the physiological instability that precedes the arrest is usually evident for substantial periods of time prior to the arrest, there is significant face validity to the notion that RRSs should result in a reduction in the incidence of cardiac arrest and mortality. Additionally, RRSs should be able to, through early recognition and intervention, reduce unanticipated ICU admission. By catching problems early in

1  Rapid Response Systems History and Terminology

5

their course, it is envisioned that patients not only won’t arrest but also may not even require ICU care and be able to be managed on the general ward. This helps to keep ICU beds open for other patients and improve through-put. Likewise, with reductions in serious deteriorations and complications, length of stay should also decrease. Even when patients still require transfer to an ICU as a result of their deterioration, in principle, having the early intervention afforded by an RRS should have the patient arriving in the ICU in better condition than without such a system. The expected benefit in this circumstance is reduced ICU and hospital mortality and reduced ICU and length of hospital stay. The foundation for achieving these goals is an underlying principle and strength of RRSs, namely that RRSs address the mismatch between the patient’s needs and the available resources on the general wards.52 This imbalance between what the patient needs (human resources, monitoring, specialized equipment and medications) and what the general ward can provide (staffing, monitoring and policy limitations) are at the center of these deteriorations. RRSs are commonly activated by nurses who determine that the patient is seriously ill and cannot be managed under the current circumstances. These circumstances may include staff/acuity limitations, inadequate care plans, and/or new events such as sepsis. Through rapid assessment and intervention, new plans can be developed and communicated to the ward staff and primary service and resource/needs imbalances accounted for resulting in an effective triage and care plan for the patient. Often the patient requires triage to a higher level of care to achieve a re-balancing of resources and needs, but if needs are reduced through RRS intervention, that re-balancing can be achieved with the patient remaining on the ward. Many of these goals and benefits have been realized through the implementation of RRSs, while some have been less successful and yet others not well evaluated.18–49 While outcome measures such as mortality are important to clinicians, regulators and patients, other measures of RRS success and positive impact need to also be considered. Some of these include process of care measures (such as meeting sepsis management guidelines and appropriate institution of Not for Resuscitation status),53–55 patient and nursing satisfaction56–58 and especially the value RRSs bring to staff education in the recognition and management of the critically ill patient who presents as such outside the walls of the ICU.59,60 In fact, this last goal and benefit of RRSs may be the most under-appreciated, although in many ways the most important. RRSs change culture and culture is a crucial element of the health systems in which we work. RRSs are not just teams, they are systems in themselves that include a component that emphasizes and educates in the early detection of problems and a component that responds to the call for help. The RRS functions within a greater system that spans from the patient, through the providers, their environment and up to the departmental, hospital and institutional and even governmental level. This realization requires that RRSs have two additional components besides the activation process and a responding team. The first is an evaluative element that continuously assesses the performance of the RRS and helps to inform the hospital Quality Improvement (QI) process.52 Institutions such as the University of Pittsburgh have used their RRS to scrutinize all of their arrest and MET calls in an ongoing QI process that has had great impact for their

6

B.D. Winters and M. DeVita

hospital.61 The second component is a governance and administration structure.52 This helps to develop, implement and, most crucially, maintain and improve the RRS program. The work of the evaluation/QI element and the governance/­ administrative structure has been greatly improved in the last 2 years with the addition of RRS data fields to the American Heart Association’s (AHA) National Registration for CPR database.62 From this database useful reports and comparisons can be ­generated to support the RRS. While these two additional components are not absolutely essential to having an RRS, they enhance its effectiveness, role and status within the hospital system and are well worth consideration. RRSs have become a great agent for change, encouraging and empowering ward staff to ask for help for their patients. The archaic concept, often held and promulgated by physicians and occasionally others, that calling for help is “a sign of weakness” is washed away by the RRS as it changes the culture to an understanding that the patient and his well-being is the primary concern of all providers and that calling for help is the sign of the wise and caring clinician.

Terminology It is important to have a clear understanding of the terminology for Rapid Response Systems so as to get the most from this book and any review of RRS literature. Historically, the early RRSs were most commonly called “Medical Emergency Teams” or METs, although other terms were also used, including Medical Emergency Response Teams (MERT), Patient-at-Risk Teams (PART), Critical Care Outreach Teams (CCOT), and eventually Rapid Response Teams (RRTs). Some of these terms are used interchangeably in places such as Australia where RRT and MET often mean the same thing. While hospitals and institutions often create specific names for their programs based on local preferences and the desire to use something memorable to encourage utilization, consensus has been developed on specific terminology that should be used when reporting and sharing information and data in the public forum (publications, research articles etc).52,63 The term Rapid Response System refers to the entire system for responding to all patients with a critical medical problem. Most broadly, this can include the Cardiac Arrest Team (Code Team) and the MET as well as other specialized teams that may exist within the hospital, such as a Difficult Airway Response Team, although most commonly and preferably the term RRS is used to refer to systems that seek to prevent deterioration and arrests rather than respond to arrests. This term encompasses both the recognition process (the activation criteria and the activation process) and the responding team. These two sub-components are referred to as the Afferent and Efferent Limbs of the RRS, respectively. Additional confusion may arise when the Code Team and the MET are one and the same in terms of personnel but take on different roles depending on the patient’s situation (arrest versus deterioration).

1  Rapid Response Systems History and Terminology

7

The historical MET and similar systems are now defined through consensus based on the team structure and functionality.52 Teams that include physicians along with nurses, but may also include respiratory therapists and others, are properly called Medical Emergency Teams, which have full capability for assessment, treatment and triage planning, while teams that do not include physicians as responders and rely on nurses and others only are referred to as Rapid Response Teams (RRTs). These nurse-led RRTs often have physician consultation available but the physician does not respond to the bedside as a member of the initial response. RRTs often provide an intermediate range of capability since nurses cannot write orders for therapy. An exception to this is the Nurse Practitioner in the US, who can write a range of orders. As such, true RRTs are able to assess and provide some level of stabilization but if needs and resources are severely out of balance, the patient is likely to require triage to a higher level of care. Teams that provide follow-up service and surveillance on patients discharged from the ICU on a regular basis, as well as response to any general ward patient that may or may not have been in an ICU previously, are described as Critical Care Outreach Teams. These teams are often staffed by nurses and therefore their response to deteriorating patients would be an RRT-type response team. Other terms, such as Patient-at-Risk Team, may be used as the local name for the program and hospitals may choose to call their system an RRT even though it has physicians as responders; such names should only be for local use and should be avoided in the literature. So that proper comparisons may be made, the preferred nomenclature should describe the response component of the RRS as an MET, RRT or CCOT-based on these consensus definitions. As mentioned previously, another important terminology distinction to consider is the difference between the process of recognition that the patient is deteriorating and that of the teams that respond. The process and criteria used for triggering the call for help is called the Afferent Limb, while the response to that call, the team, is the Efferent Limb. While both work in concert as a system, their separate nomenclature and consideration is important. Many think of the Efferent Limb as the RRS but the Afferent Limb is equally important, if not more so since this is where the recognition is made that the patient needs help. Providing a responding team is of less benefit if the patient has already progressed to the point where arrest is imminent; the earlier the recognition is made, the better. Some have argued that it may not even matter who comes to help the patient (a critical care team, a hospitalist team or the primary service) as long as it is recognized that help is needed early enough to make a difference. The importance of early recognition was addressed by the first Consensus Conference on Medical Emergency Teams and then further emphasized by the special Afferent Limb Consensus Conference convened ahead of the third International Conference on Medical Emergency Teams. In these forums, the question of how RRSs might improve their identification of seriously ill general ward patients was considered and debated. The report of the first Consensus Conference52 indicates that RRS should use clear methods of detection for identifying “emergent unmet patient needs” and deteriorations. Objective criteria are preferred and several identification systems exist, including direct vital sign parameters and various scoring systems.52, 64–81 The upcoming report of the Afferent

8

B.D. Winters and M. DeVita

Limb Consensus Conference suggests that technological solutions are going to be essential to better monitoring and detection in the mostly ambulatory general ward patient population but that we still are struggling to determine what best needs to be monitored and how. This is an area of very active research with many exciting possibilities likely to result in the next few years.82–84 The Efferent Limb also continues as an area of active research. New education modalities and strategies such as simulation are being used to improve team performance and function and prepare teams for unusual or rare scenarios. The kind of team that makes up the Efferent Limb probably does matter ‑ not so much by what their title is, but rather by how well they are prepared and how well they work as team members.85–87

Summary RRSs have grown substantially since their inception almost two decades ago to become a robust strategy for improving patient care and healthcare culture. Clinicians who support these initiatives have striven to be evidence-based and thorough, resulting in an ever-expanding body of literature and experience that points to RRSs as a successful systems-based solution to the problem of deteriorating general ward patients and the imbalance of resources necessary to care for them. Clear definitions and nomenclature have aided this process. By working to improve the Afferent and Efferent Limbs of RRSs through methods best suited to their uniqueness and melding them into an effective system, RRSs can continue to be a developing and dynamic patient safety and quality of care improvement paradigm.

References 1. Sax FL, Charlson ME. Medical patients at high risk for catastrophic deterioration. Crit Care Med. 1987;15(5):510–515. 2. Schein RM, Hazday N, Pena M, Robin BH, Sprung CL. Clinical antecedents to in-hospital cardiopulmonary arrest. Chest. 1990;98(6):1388–1392. 3. Bedell SE, Deitz DC, Leeman D, Delbanco TL. Incidence and characteristics of preventable iatrogenic cardiac arrests. J Am Med Assoc. 1991;265(21):2815–2820. 4. Daffurn K, Lee A, Hillman KM, Bishop GF, Bauman A. Do nurses know when to summon emergency assistance? Intensive Crit Care Nurs. 1994;10(2):115–120. 5. Smith AF, Wood J. Can some in-hospital cardio-respiratory arrests be prevented? A prospective survey. Resuscitation. 1998;37(3):133–137. 6. Buist MD, Jarmolowski E, Burton PR, Bernard SA, Waxman BP, Anderson J. Recognising clinical instability in hospital patients before cardiac arrest or unplanned admission to intensive care. A pilot study in a tertiary-care hospital. Med J Aust. 1999;171(1):22–25. 7. Hillman KM, Bristow PJ, Chey T, Daffurn K, et al. Antecedents to hospital deaths. Intern Med J. 2001;31(6):343–348.

1  Rapid Response Systems History and Terminology

9

8. Hodgetts TJ, Kenward G, Vlachonikolis IG, et  al. Incidence, location and reasons for avoidable in-hospital cardiac arrest in a district general hospital. Resuscitation. 2002;54(2):115–123. 9. Kause J, Smith G, Prytherch D, et al. A comparison of antecedents to cardiac arrests, deaths and emergency intensive care admissions in Australia, New Zealand, and the United Kingdom ‑ the ACADEMIA study. Resuscitation. 2004;62(3):275–282. 10. Hillman K, Bristow PJ, Chey T, et  al. Duration of life-threatening antecedents prior to ­intensive care admission. Intensive Care Med. 2002;28:1629–1634. 11. Franklin C, Mathew J. Developing strategies to prevent in-hospital cardiac arrest: analyzing responses of physicians and nurses in the hours before the event. Crit Care Med. 1994;22(2):244–247. 12. McGloin H, Adam SK, Singer M. Unexpected deaths and referrals to intensive care of patients on general wards. Are some cases potentially avoidable? J R Coll Physicians Lond. 1999;33(37):255–259. 13. Goldhill DR, White SA, Sumner A. Physiological values and procedures in the 24 h before ICU admissions from the ward. Anaesthesia. 1999;54(6):529–534. 14. Morgan RJM, Williams F, Wright MM. An early warning scoring system for detecting ­developing critical illness. Clin Intensive Care. 1997;8:100. 15. Stenhouse C, Coates S, Tivey M, Allsop P, Parker T. Prospective evaluation of a Modified Early Warning Score to aid earlier detection of patients developing critical illness on a general surgical ward. Br J Anaesth. 2000;84:663P. 16. Subbe CP, Kruger M, Rutherford P, Gemmel L. Validation of a modified Early Warning Score in medical admissions. Q J Med. 2001;94(10):521–526. 17. Hodgetts TJ, Kenward G, Vlachonikolis IG, Payne S, Castle N. The identification of risk ­factors for cardiac arrest and formulation of activation criteria to alert a medical emergency team. Resuscitation. 2002;54(2):125–131. 18. Lee A, Bishop G, Hillman K, Daffurn K. The medical emergency team. Anaesth Intensive Care. 1995;23:183–186. 19. Goldhill DR, Worthington L, Mulcahy A, Tarling M, Sumner A. The patient-at-risk team: identifying and managing seriously ill ward patients. Anaesthesia. 1999;54(2):853–860. 20. Bristow PJ, Hillman KM, Chey T, et al. Rates of in-hospital arrests, deaths and intensive care admissions: the effect of a medical emergency team. Med J Aust. 2000;173(5):236–240. 21. Buist MD, Moore GE, Bernard SA, Waxman BP, Anderson JN, Nguyen TV. Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. Br Med J. 2002;324(7334):387–390. 22. Ball C, Kirkby M, Williams S. Effect of the critical care outreach team on patient survival to discharge from hospital and readmission to critical care: non-randomised population based study. Br Med J. 2003;327(7422):1014–1016. 23. Leary T, Ridley S. Impact of an outreach team on re-admissions to a critical care unit. Anaesthesia. 2003;58(4):328–332. 24. Bellomo R, Goldsmith D, Uchino S, et al. A prospective before-and-after trial of a medical emergency team. Med J Aust. 2003;179(6):283–287. 25. Kenwood G, Castle N, Hodgetts T, Shaikh L. Evaluation of a medical emergency team one year after implementation. Resuscitation. 2004;61(3):257–263. 26. Priestley G, Watson W, Rashidian A, et  al. Introducing critical care outreach: a ward-­ randomised trial of phased introduction in a general hospital. Intensive Care Med. 2004;30(7): 1398–1404. 27. Bellomo R, Goldsmith D, Uchino S, et al. Prospective controlled trial of effect of medical emergency team on postoperative morbidity and mortality rates. Crit Care Med. 2004;32(4):916–921. 28. DeVita MA, Braithwaite RS, Mahidhara R, Stuart S, Foraida M, Simmons RL. Use of medical emergency team responses to reduce hospital cardiopulmonary arrests. Qual Saf Health Care. 2004;13(4):251–254.

10

B.D. Winters and M. DeVita

29. Garcea G, Thomasset S, McClelland L, Leslie A, Berry DP. Impact of a critical care outreach team on critical care readmissions and mortality. Acta Anaesthesiol Scand. 2004;48(9):1096–1100. 30. MERIT Study Investigators. Introduction of the medical emergency team (MET) system: a cluster-randomised controlled trial. Lancet. 2005;365(9477):2091–2097. 31. Jones D, Bellomo R, Bates S, et al. Long term effect of a medical emergency team on cardiac arrests in a teaching hospital. Crit Care. 2005;9(6):R808–R815. 32. Jones D, Opdam H, Egi M, et al. Long-term effect of a medical emergency team on mortality in a teaching hospital. Resuscitation. 2007;74(2):235–241. 33. Jolley J, Bendyk H, Holaday B, Lombardozzi KAK, Harmon C. Rapid response teams: do they make a difference? Dimens Crit Care Nurs. 2007;26(6):253–260. 34. Dacey MJ, Mirza ER, Wilcox V, et al. The effect of a rapid response team on major clinical outcome measures in a community hospital. Crit Care Med. 2007;35(9):2076–2082. 35. Chan PS, Khalid A, Longmore LS, Berg RA, Kosiborod M, Spertus JA. Hospital-wide code rates and mortality before and after implementation of a rapid response team. J Am Med Assoc. 2008;300(21):2506–2513. 36. Tibballs J, Kinney S, Duke T, Oakely E, Hennessy M. Reduction of pediatric in-patient­ ­cardiac arrest and death with a medical emergency team: preliminary results. Arch Dis Child. 2005;90(11):1148–1152. 37. Tibballs J, Kinney S. Reduction of hospital mortality and of preventable cardiac arrest and death on introduction of a pediatric medical emergency team. Pediatr Crit Care Med. 2009;10(3):306–312. 38. Brilli RJ, Gibson R, Luria JW, et al. Implementation of a medical emergency team in a large pediatric teaching hospital prevents respiratory and cardiopulmonary arrests outside the ­intensive care unit. Pediatr Crit Care Med. 2007;8(3):236–246. 39. Sharek PJ, Parast M, Leong K, et  al. Effect of a rapid response team on hospital-wide ­mortality and code rates outside the ICU in a children’s hospital. J Am Med Assoc. 2007;298(19):2267–2274. 40. Zenker P, Schlesinger A, Hauck M, et al. Implementation and impact of a rapid response team in a children’s hospital. Joint Comm J Qual Patient Saf. 2007;33(7):418–425. 41. Buist M, Harrison J, Abaloz E, Van Dyke S. Six-year audit of cardiac arrests and medical emergency team calls in an Australian teaching hospital. Br Med J. 2007;335(7631): 1210–1212. 42. Jones D, Egi M, Bellomo R, Goldsmith D. Effect of the medical emergency team on long-term mortality following major surgery. Crit Care. 2007;11(1):R12. 43. Mailey J, Digiovine B, Baillod D, Gnam G, Jordan J, Rubinfeld I. Reducing hospital standardized mortality rate with early interventions. J Trauma Nurs. 2006;13(4):178–182. 44. Tolchin S, Brush R, Lange P, Bates P, Garbo JJ. Eliminating preventable death at Ascension Health. Joint Comm J Qual Patient Saf. 2007;33(3):145–154. 45. Offner P, Heit J, Roberts R. Implementation of a Rapid Response Team decreases cardiac arrest outside of the Intensive Care Unit. J Trauma. 2007;62(5):1223–1228. 46. Dacey M, Mizra R, Wilcox V, et al. The effect of a Rapid Response Team on major clinical outcome measures in a community hospital. Crit Care Med. 2007;35(9):2076–2082. 47. Story D, Shelton A, Poustie S, Colin-Thome N, McIntrye R, McNicol P. Effect of an anesthesia department-led critical care outreach and acute pain service on postoperative serious adverse events. Anesthesia. 2006;61:24–28. 48. King E, Horvath R, Shulkin D. Establishing a Rapid Response Team (RRT) in an academic hospital: one year’s experience. J Hosp Med. 2006;1(5):296–305. 49. Hunt EA, Zimmer KP, Rinke ML, et al. Transition from a traditional code team to a medical emergency team and categorization of cardiopulmonary arrests in a children’s center. Arch Pediatr Adolesc Med. 2008;162(2):117–122. 50. 100K lives campaign. www.ihi.org/IHI/Programs/Campaign/Campaign. Accessed 10.07.09. 51. Joint Commission National Patient Safety Goals. www.jointcommission.org/patientsafety/ nationalpatientsafetygoals/. Accessed 10.01.10.

1  Rapid Response Systems History and Terminology

11

52. Devita M, Bellomo R, Hillmam K, et al. Findings of the first consensus conference on medical emergency teams. Crit Care Med. 2006;34(9):2463–2478. 53. Sebat F, Musthafa AA, Johnson D, et al. Effect of a rapid response system for patients in shock on time to treatment and mortality during 5 years. Crit Care Med. 2007;35(11): 2568–2575. 54. Jones DA, McIntyre T, Baldwin I, Mercer I, Kattula A, Bellomo R. The medical emergency team and end-of-life care: a pilot study. Crit Care Resusc. 2007;9(2):151–156. 55. Chen J, Flabouris A, Bellomo R, Hillman K, Finfer S. MERIT study investigators for the Simpson Center and the ANZICS Clinical Trials Group, the Medical Emergency Team System and not-forresuscitation orders: results from the MERIT study. Resuscitation. 2008;79(3):391–397. 56. Jones D, Baldwin I, McIntyre T, et al. Nurses’ attitudes to a medical emergency team service in a teaching hospital. Qual Saf Health Care. 2006;15(6):427–432. 57. Galhotra S, Scholle CC, Dew MA, Mininni NC, Clermont G, DeVita MA. Medical emergency teams: a strategy for improving patient care and nursing work environments. J Adv Nurs. 2006;55(2):180–187. 58. Salamonson Y, van Heere B, Everett B, Davison P. Voices from the floor: nurses’ perceptions of the medical emergency team. Intensive Crit Care Nurs. 2006;22(3):138–143. 59. Buist M, Bellomo R. MET. The emergency medical team or the medical education team? Crit Care Resusc. 2004;6:88–91. 60. Jones D, Bates S, Warrillow S, Goldsmith D, et al. Effect of an education programme on the utilization of a medical emergency team in a teaching hospital. Intern Med J. 2006;36(4):231–236. 61. Braithewaite RS, Devita MA, Mahidhara R, et al. Use of medical emergency teams (MET) responses to detect medical errors. Qual Saf Health Care. 2004;13:255–259. 62. American Heart Association National Registry for CPR. www.nrcpr.org. Accessed 23.01.10. 63. Cretikos M, Parr M, Hillman K, et al. Guidelines for the uniform reporting of data for Medical Emergency Teams. Resuscitation. 2006;68:11–25. 64. Subbe CP, Davies RG, Williams E, Rutherford P, Gemmell L. Effect of introducing the Modified Early Warning score on clinical outcomes, cardiopulmonary arrests and intensive care utilisation in acute medical admissions. Anesthesia. 2003;58(8):797–802. 65. Goldhill DR, McNarry AF. Physiological abnormalities in early warning scores are related to mortality in adult inpatients. Br J Anaesth. 2004;92(6):882–884. 66. Goldhill DR, McNarry AF, Mandersloot G, McGinley A. A physiologically-based early warning score for ward patients: the association between score and outcome. Anaesthesia. 2005;60(6):547–553. 67. Sharpley JT, Holden JC. Introducing an early warning scoring system in a district general hospital. Nurs Crit Care. 2004;9(3):98–103. 68. Gardner-Thorpe J, Love N, Wrightson J, Walsh S, Keeling N. The value of Modified Early Warning Score (MEWS) in surgical in-patients: a prospective observational study. Ann R Coll Surg Engl. 2006;88(6):571–575. 69. Jacques T, Harrison G, McLaws M, Kilborn G. Signs of critical conditions and emergency response (SOCCER): a model for predicting adverse events in the inpatient setting. Resuscitation. 2006;69:175–183. 70. Harrison GA, Jacques T, McLaws ML, Kilborn G. Combinations of early signs of critical illness predict in-hospital death – the SOCCER study (signs of critical conditions and emergency responses). Resuscitation. 2006;71(3):327–334. 71. Jacques T, Harrison GA, McLaws ML, Kilborn G. Signs of critical conditions and emergency responses (SOCCER): a model for predicting adverse events in the in-patient setting. Resuscitation. 2006;69(2):175–183. 72. Subbe CP, Hibbs R, Williams E, Rutherford P, Gemmel L. ASSIST: a screening tool for ­critically ill patients on general medical wards. Intensive Care Med. 2002;28(suppl):S21. 73. Haines C, Perrott M, Weir P. Promoting care for acutely ill children. Development and evaluation of a paediatric early warning tool. Intensive Crit Care Nurs. 2006;22(2):73–81. 74. Duncan H, Hutchison J, Parshuram CS. The pediatric early warning system score: a severity of illness score to predict urgent medical need in hospitalised children. J Crit Care. 2006;21(13):271–279.

12

B.D. Winters and M. DeVita

75. Subbe CP, Gao H, Harrison DA. Reproducibility of physiological track-and-trigger warning systems for identifying at-risk patients on the ward. Intensive Care Med. 2007;33(4):619–624. 76. Bell MB, Konrad D, Granath F, Ekbom A, Martling CR. Prevalence and sensitivity of MET-criteria in a Scandinavian University Hospital. Resuscitation. 2006;70(1):66–73. 77. Green A, Williams A. An evaluation of an early warning clinical marker referral tool. Intensive Crit Care Nurs. 2006;22:274–282. 78. Cretikos M, Chen J, Hillman K, Bellomo R, Finfer S, Flabouris A. MERIT study investigators. The objective medical emergency team activation criteria: a case-control study. Resuscitation. 2007;73(1):62–72. 79. Smith GB, Prytherch DR, Schmidt PE, Featherstone PI, Higgins B. A review, and performance evaluation, of single-parameter “track-and-trigger” systems. Resuscitation. 2008;79(1): 11–21. 80. Smith GB, Prytherch DR, Schmidt PE, Featherstone PI. Review and performance evaluation of aggregate weighted “track-and-trigger” systems. Resuscitation. 2008;77(2):170–179. 81. Santiano N, Young L, Hillman K, et al. Analysis of medical emergency team calls comparing subjective to “objective” call criteria. Resuscitation. 2009;80(1):44–49. 82. Smith GB, Prytherch DR, Schmidt P, Featherstone PI, et  al. Hospital-wide physiological surveillance. A new approach to the early identification and management of the sick patient. Resuscitation. 2006;71(1):19–28. 83. Watkinson PJ, Barber VS, Price JD, Hann A, et al. A randomised controlled trial of the effect of continuous electronic physiological monitoring on the adverse event rate in high risk ­medical and surgical patients. Anaesthesia. 2006;61(11):1031–1039. 84. Tarassenko L, Hann A, Young D. Integrated monitoring and analysis for early warning of patient deterioration. Br J Anaesth. 2006;97(1):64–68. 85. DeVita MA, Schaefer J, Lutz J, Wang H, Dongilli T. Improving medical emergency team (MET) performance using a novel curriculum and a computerized human patient simulator. Qual Saf Health Care. 2005;14(5):326–331. 86. Wallin CJ, Meurling L, Hedman L, Hedegård J, Felländer-Tsai L. Target-focused medical emergency team training using a human patient simulator: effects on behaviour and attitude. Med Educ. 2007;41(2):173–180. 87. Jones D, Duke G, Green J, et al. Medical Emergency Team syndromes and an approach to their management. Crit Care. 2006;10(1):R30.

Chapter 2

RRS’s General Principles Ajay D. Rao and Michael DeVita

Keywords  Overview • Rapid • Response • System

Introduction Studies focused on hospital outcomes have proven to be conflicting with regards to the Rapid Response System irrespective of whether the responders are Medical Emergency Teams (METs) or Rapid Response Teams (RRTs). On one hand, several studies have shown improved hospital outcomes,1–4 yet on the other, there have been some studies that have also shown negative outcomes.5–7 Despite the literature, many hospitals have incorporated METs and RRTs to promote patient safety. In order to understand the benefits of any form of Rapid Response System requires not only information about the constituents of the response team, but also knowledge of the triggers that set off the cascade of responses. Specifically, understanding the afferent limb (i.e., the calling criteria and the triggering mechanism) can shed light on why (or why not) the efferent limb (the MET or RRT) is effective.8 Indeed, some have argued that the most important part of the RRS is a robust triggering mechanism to reap most of the benefits of an RRS.9 In addition, one might expect that a hospital’s quality improvement efforts may be entwined with events discovered in performing a response, or in a retrospective review of the events. Finally, different hospitals may bring different material and personnel resources to bear in either the response or the review of events. All these factors contribute to the practice of responding to and preventing future patient crisis events. Because of these considerations, experts have concluded that the response ‑ RRT or MET ‑ is a system event and not exclusively a team event. In this chapter, we will describe and provide an overview of the Rapid Response System.

M. DeVita (*) Clinical Professor, Department of Critical Care Medicine, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213, USA e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_2, © Springer Science+Business Media, LLC 2011

13

14

A.D. Rao and M. DeVita

Overview In 2004, experts convened to describe and create a common terminology for teams that respond to patient deterioration outside the ICU. At that time, the participants of the consensus conference recognized a critical observation: All organizations that had successfully created a planned team response had also made a series of interventions in their organizations beyond the creation of a team. They recognized and defined the Rapid Response System.8 The RRS has four components: the afferent limb, the efferent limb, the quality improvement limb, and the administrative limb (Fig. 2.1). The afferent (or event detection and response trigger) limb is one of the most important components of the rapid response system (RRS). This is most likely related to the fact that many failures within the health system’s ability to manage deteriorating patients originate in the afferent limb. In other words, a patient may deteriorate without rescue if he or she is not assessed, or if assessed, the person doing the assessment does not recognize a critical state, or if recognized, a call for help not made, or if made, the responders not arrive (Fig. 2.2). Each step in the chain can be easily broken, all leading to the same outcome ‑ failure to respond to a patient in crisis. In the sense that the afferent limb is responsible for identifying a crisis and triggering a response, it may be both the most important component because without it there is no response. It may also be the

Afferent Limb Trigger

Efferent Limb MET/RRT/CCO specialized resources

Event detection Urgent Un-met Patient Need

Administration oversees all functions Data collection and analysis for Process Improvement

Crisis Resolved

cardiac Arrest Team Trauma Team Stroke Team

Data acquisition point

Fig.  2.1  Rapid Response System structure. When patients have critical unmet needs and as a result are at risk for imminent danger, the afferent limb detects the event and triggers a systemic response. The response provides resources to stabilize and triage the patient to a location where services meet the patient’s needs. Data are collected to determine event rate, resources needed and outcomes, and to enable an analysis of events to prevent or prepare for future events. An administrative mechanism is needed to oversee all components and to provide resources to facilitate the system. MET medical emergency team, RRT rapid response team, CCO critical care outreach

2  RRS’s General Principles

15

Clinical deterioration

Observation by staff

Recognition of abnormality

Knowledge how to call

Knowledge who to call

Decision to call for help

Call for help

Reliable calling mechanism

Team receives alert

Fig. 2.2  Steps in the afferent limb

most error-prone component of the rapid response system because of the many steps in the process, and the fact that patients are often not continuously observed. Logistically, the afferent limb can be divided into three main components: (1)  the selection/diagnostic/triggering criteria; (2) human and/or technologic ­monitoring (with alarm limits); and (3) a mechanism for triggering response. Each one of these components poses unique challenges to any hospital’s attempt to create an RRS as they all break with the traditional hospital culture. These are described more fully in Chaps. 14, 15 and 18, which also deal with the afferent limb. We will note here, however, that a subjective assessment can be a valuable tool for RRS triggering. This relies on the staff’s feelings of being “worried” about a patient. This criterion is usually used in conjunction with the first two systems, but this system is effective in allowing nurses to call for help based on their clinical instincts. Santiano10 has shown that while “nurse worried” is the most common calling criterion, most patients who have a call based on this criterion also met at least one of the criteria in the single parameter system. Others have shown that the reliability of crisis detection and team triggering is increased if objective signs are used.11,12 In actuality, although most published criteria consist of a variety of ­objective measures, many organizations have also included an option for a provider who is “worried.” This essentially represents a “fallback” position to support a staff member who may be unsure of making a call. In a study from Australia, over one-half of “worried” calls ended up falling into one of the other objective calling ­criteria categories.10 Furthermore, it is important that triggering an RRS response should be the inherent responsibility of all hospital personnel. It has been up to each organization to determine which criteria will be used, and, more importantly, to educate the staff accordingly. In general, the most common hospital personnel responsible for triggering these criteria have been nurses. The RRS has enabled nurses to exert their own independent judgment in calling for help through a preset system that goes “around” usual hospital hierarchy.13 The efferent limb consists of the resources that are brought to bear: both personnel and equipment. Thus the efferent limb is more than the response team. The components of the efferent limb, including the types of teams that may be created to deal

16

A.D. Rao and M. DeVita

with a host of different critical situations, are described in Chaps. 19–26. The intention is to restore balance to a situation where the patient’s needs exceed the resources available at a time when the patient’s status is critical and time is of the essence. For most crises, it intends to enable critical care resources to get to the patient’s bedside without bureaucratic, sociological, or logistical barriers. Nevertheless, nurses often continue to choose to call the patient’s primary physician, regardless of how long an RRS has been in place.14 Examples of critically unbalanced resource-to-needs situations include an acute trauma or fall, acute hemorrhage, a lost patient, an obstetric emergency, an unruly patient or family member, or even a staff member who is despondent over committing an error that caused patient harm. To make patients as safe as possible, all these needs should have a planned triggering mechanism and response. It is unlikely that facilities will create all the teams we have identified in this textbook at the outset of creating their RRS. It is only through continuous quality efforts that development of the resources becomes recognized as a need. We and others have found that events that trigger the RRS are examples of situations in which patients were in danger.15 To prevent future dangerous situations, one must have a quality apparatus attached to the RRS. This quality limb should both attempt to prevent critical events as well as improve the response to those events. Both require careful data collection, analysis, and provide feedback to responsible administrative and clinical personnel. These individuals or teams can create needed new processes to attend to both preventive and responsive measures. This limb is addressed in Chaps. 2, 3, 5, 7, 11, 22, and 23. The final component of the Rapid Response System is the Administrative Limb. This limb is responsible for marshalling both personnel and equipment resources. When we at the University of Pittsburgh first started working to improve our ­cardiac arrest response program, we focused on both organizing the responders and overcoming equipment barriers. We found many different types of defibrillators, which led to inefficiencies in equipment use, and costly repairs because of the need to maintain spare parts inventories as well as education needs for our engineers. We worked with our hospital administrators to create a business plan to standardize our defibrillators and crash cart contents, resulting in lower costs and better use of the equipment. Without a strong quality improvement limb to collect and report data, and administrative limb responsive to patient needs, the accomplishment would not have occurred. Since that time, we recognized the need to prevent cardiac arrests by responding sooner to patient deteriorations. Our clinical decision that a response team of critical care clinicians was needed would not have been an implementable conclusion without the resources (including internal ­marketing campaigns) that only administrators and clinical leaders could marshal. It is now being recognized that capturing vital sign data (as frequently as possible) is needed if every patient deterioration is to be assessed and recognized. Some are calling for continuous vital sign monitoring for traditionally unmonitored patients.16,17 Intermittent monitoring provides the requirement of a human-technology interaction, and is usually lower cost. Of course, abnormalities may occur in between assessments, allowing the potential for missing an evolving event for minutes to

2  RRS’s General Principles

17

hours, depending upon the monitoring interval and the specific time of the occurrence relative to the monitoring timetable. Continuous monitoring would likely prevent (or reduce) the occurrence of undetected deterioration; however, it does have major drawbacks worth considering. But moving to an “all patients ­monitored” hospital will require a strong administrative limb to deal with cost and cultural obstacles.

Summary In summary, the Rapid Response System is indeed an integrated system of care. At its best, it requires development and integration of the four limbs of the system. Each limb has a specific goal and place in the chair of care. Each is important to all other pathways of the rapid response system. Most importantly, the degree of development and integration directly affects the effectiveness of the other limbs.

References 1. Bellomo R, Goldsmith D, Uchino S, et al. A prospective before-and-after trial of a medical emergency team. Med J Aust. 2003;179:283–287. 2. Bellomo R, Goldsmith D, Uchino S, et al. Prospective controlled trial of effect of medical emergency team on postoperative morbidity and mortality rates. Crit Care Med. 2004;32:916–921. 3. Buist MD, Moore GE, Bernard SA, Waxman BP, Anderson JN, Nguyen TV. Effects of a ­medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. Br Med J. 2002;324:387–390. 4. DeVita MA, Braithwaite RS, Mahidhara R, Stuart S, Foraida M, Simmons RL. Use of medical emergency team responses to reduce hospital cardiopulmonary arrests. Qual Saf Health Care. 2004;13:251–254. 5. Hillman K, Chen J, Cretikos M, et al. MERIT Study Investigators. Introduction of the medical emergency team (MET) system: a cluster-randomized controlled trial. Lancet. 2005;365: 2091–2097. 6. Kenward G, Castle N, Hodgetts T, Shaikh L. Evaluation of a medical emergency team one year after implementation. Resuscitation. 2004;61:257–263. 7. Chan PS, Khalid A, Longmore LS, Berg RA, Kosiborod M, Spertus JA. Hospital-wide code rates and mortality before and after implementation of a rapid response team. J Am Med Assoc. 2008;300:2506–2513. 8. DeVita MA, Bellomo R, Hillman K, et  al. Findings of the first consensus conference on ­medical emergency teams. Crit Care Med. 2006;34(9):2463–2478. 9. Howell, M. D., Zullo, N., Folcarelli, P., Aronson, M., Moorman, D., Mottley, L., Yang, J., & Sands, K. Monitoring and adjusting a rapid response team implementation using a novel informatics tool. 24th International Conference on Quality in Health Care (ISQuA). 30 September‑3 October 2007 (Boston, MA). 10. Santiano N, Young L, Hillman K, et al. Analysis of Medical Emergency Team calls comparing subjective to “objective” call criteria. Resuscitation. 2009;80(1):44–49. 11. Sakai T, DeVita M. Rapid response system. J Anesth. 2009;23(3):403–408. 12. Foraida MI, DeVita MA, Braithwaite RS, Stuart SA, Mori-Brooks M, Simmons RI. Improving the utilization of medical crisis teams (Condition C) at an urban tertiary care hospital. J Crit Care. 2003;18:87–94.

18

A.D. Rao and M. DeVita

13. Tee A, Calzavacca P, Licari E, Goldsmith D, Bellomo R. Bench-to-bedside review: the MET syndrome ‑ the challenges of researching and adopting medical emergency teams. Crit Care. 2008;12(1):205. 14. Jones D, Baldwin I, McIntyre T, et al. Nurses’ attitudes to a medical emergency team service in a teaching hospital. Qual Saf Health Care. 2006;15:427–432. 15. Braithwaite RS, DeVita MA, Mahidhara R, Stuart S, Simmons RL. Use of Medical Emergency Team (MET) responses to detect medical errors. Qual Saf Health Care. 2004;13:255–259. 16. Taenzer AH, Pyke JB, Mcgrath SP, Bilke GT. Impact of pulse oximetry surveillance on rescue events and intensive care unit transfers. Anesthesiology. 2010;112:282–287. 17. DeVita MA, Smith GB, Adam S, et al. Consensus conference on the afferent limb of rapid response systems: identifying patients at risk. Resuscitation. 2010;81:375–382.

Chapter 3

Measuring and Improving Safety Bradford D. Winters, Peter J. Pronovost, Marlene Miller, and Elizabeth A. Hunt

Keywords  Measuring • Improving • Safety

Introduction It has been a decade since the Institute of Medicine raised the call to action for patient safety in their landmark report, To Err Is Human, which brought to the public’s attention the significant problems with patient safety in our healthcare system.1 This call to action has been heeded by many healthcare leaders, who are actively and vigorously addressing patient safety issues. The healthcare community has worked to educate themselves on methods to improve safety, and strived to execute interventions toward the goal of improving patient safety.2,3 We still have much to do and the science of safety needs to mature rapidly to meet the needs of patients. We especially need to develop effective methods for evaluating the impact of our interventions so that we know what works and where to best invest our resources and answer the question, “Are patients safer?” Ten years later, patients and families want to know if we have made progress. They want to know if the care they are receiving is safer. They want to know if they or their loved ones are less likely to be injured or die as a result of medical care. They do not want our perceptions or predictions. They want evidence to support the notion that we are succeeding. How do we know that our patients are safer and our efforts to improve patient safety are working? Measuring and improving safety is difficult. Not all safety measures lend themselves to examination by randomized, blind, placebo-controlled trials and specific rate reductions. Moreover, the context in which an intervention is implemented can

B.D. Winters (*) Departments of Anesthesiology, Critical Care Medicine and Surgery, The Johns Hopkins University School of Medicine, Meyer 297 600 N Wolfe Street, Baltimore, MD 21298, USA e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_3, © Springer Science+Business Media, LLC 2011

19

20

B.D. Winters et al.

influence its effect. While some are, we have come to understand that the foundation for success in improvement in-patient safety is the active change from a culture of tolerance for unsafe care to a culture that seeks to eliminate preventable harm. Changing culture is a formidable challenge. Measuring that change and the impact it exerts is almost equally difficult but essential to answering our patients and their families when they ask, “Am I going to be safe?”.2,3 This chapter provides an overview of the issues in measuring patient safety and presents a framework for measuring and improving safety. It is important to recognize that safety is a component of the broader concept of “quality,” which also includes care that is effective, efficient, patient-centered, timely, and equitable.4 The boundaries between these concepts often overlap, and measures can often bridge several categories. For example, is the failure to use an evidence-based therapy a safety measure - a mistake of omission - or an effectiveness measure? Is a complication, such as a catheter-related bloodstream infection that also increases length of stay, a safety or effectiveness measure? For these examples, the answer may be either or both. In terms of understanding whether we have made a positive impact, the distinction is less important than having a valid measure. Thus, in this chapter, we will use the term “safety” to refer to both safety and effectiveness.

Approach for the Organizational Evaluation of Patient Safety Donabedian’s approach to measuring quality of care ­ evaluating how we organize care (the structures), what we do (the processes), and the results we obtain (the outcomes) - also provides a framework for institutions to measure safety.5 Many institutional efforts to improve safety focus on structural measures, such as policies and procedures.6 Institutions may also measure processes and outcomes, although these are generally more difficult to develop and collect than structural measures. For example, organizations may measure how often certain aspects of safe and effective care were performed (a process), or how often certain complications occurred (an outcome).7,8 Unfortunately, not all institutions measure all three. While process and outcome measures are generally preferable to structural measures, they are not sufficient. Generally, process and outcome measures are rates that include a numerator and denominator, but not all measures of safety can, or should, be presented as rates. For example, a single episode of potential harm or actual harm (such as the failure to rescue a single deteriorating patient) may be statistically insignificant in terms of a measure such as in-hospital mortality, but due to the circumstances that led to that failure may be sufficient to trigger an organizational change of great impact. If organizations do not recognize and learn from such single episodes, they fail to maximize opportunities to improve safety. Analysis of multiple single episodes may lead to recognition of patterns that ­further provide insights into opportunities for safety interventions. Additionally, measurement of rates is resource-intensive and not feasible for every type of ­medical error.

3  Measuring and Improving Safety

21

Along with the ability to learn, many other aspects of an organization’s culture have a significant impact on safety.9,10 In aviation, changes in culture, as opposed to technology, have been responsible for most of the advancements in safety over the last two decades.9,11 Within healthcare, communication failures are a common cause of sentinel events.12 (www.jointcommission.org). Indeed, communication patterns within an organization are a reflection of the existing culture. Thus, the measure of both organizational learning and culture may provide insight into an organization’s measure of safety. W. Edwards Deming once said, “There is no true value of anything that is measured; change the method of measurement and you change the result.” The same concept applies to measuring safety. In the absence of universally agreed-upon standard definitions and methods to measure patient safety, including methods for risk adjustment (e.g., healthcare-acquired infections),13 it is unlikely that national measures of patient safety will be effectively achieved. There are multiple ways to measure each area of patient safety. For medication safety, we can have a structural measure, such as the presence of computerized physician order entry; a process measure, such as prescribing errors; or an outcome measure, such as adverse drug events. Moreover, each category (structure, process, or outcome) can be measured in multiple ways. For example, the methods of surveillance for evaluating adverse drug events - many of which use self-reported events, with the numerator being how the adverse event is defined and the denomi­ nator being either patient, number of patient days, or dose - vary widely (see Table 3.1).14–18 Which method is most “correct” is unclear. The result may vary

Table 3.1  Sample of methods to measure medication Number Study studied Numerator Denominator Record Disabling Leape, et al. 30,195 reviewed/ records adverse NEJM admission events (1991) Prescribing Number 289,411 Lesar, errors of orders medication Briceland written orders/ JAMA 1 year (1990) Prescribing Medication 1 year of Lesar, errors orders prescribing Briceland, written errors Stein, detected JAMA and (1997) averted by pharmacist Number Adverse Cullen et al. 4,031 adult of patient drug admissions Crit Care days events over Med 6 months (1997)

Assessed by Physician Reviewer

Rate of events 3.7 per 100 admissions

Physicians

3.13 errors for each 1,000 orders 3.99 errors Pharmacists, per retrospectively 1,000 evaluated orders by a physician and 2 pharmacists Self-report by nurse and pharmacists, daily review of all charts by nurse investigators

19 events per 1,000 ICU patient days

22

B.D. Winters et al.

but they may all be “correct.” In the absence of standardized definitions, comparisons within and between institutions are problematic.19,20 Even with standard definitions, there is concern that comparing outcomes among hospitals is neither scientifically sound nor fair, with differences influenced by insufficient risk adjustment and random error rather than variations in-patient safety.8,19–21 In the face of the movement towards “pay-for-performance” driven measures, this is especially worrisome. In addition, the context in which an intervention is implemented (its staff, resources, leadership, etc.) can influence the impact of an intervention. It is naive to believe that interventions will be equally effective in all healthcare organizations in which they are implemented. Context matters. Yet context should not be viewed as static but rather as dynamic leading to better understanding of mechanism. In basic science, how genes turn on and off, how proteins fold, and how receptors bind was once context. These are now mechanisms and the target for therapies. Likewise, how culture influences outcomes, how leadership drives improvement, and how teamwork impact patients is context and with robust science. These will increasingly become targets for interventions. Based on this background, our approach to evaluating patient safety at the organizational level has four components and prompts the institution to answer the following four questions: 1. How often do we harm patients? 2. How often do patients receive the interventions they should? 3. How often do we learn from our mistakes; and 4. How well have we created a culture of patient safety? This framework is presented in Table 3.2. Table 3.2  Framework for an institutional scorecard for patient safety and effectiveness Example from department Domain Definition of anesthesiology How often do we harm Measures of health careBloodstream infections patients? acquired infections Surgical site infections Using either nationally validated How often do patients process measures, or a receive the interventions validated process to develop a they should? measure, what percentage of patients receive evidence-based interventions

Use of peri-operative beta blockers Elevation of head of bed in mechanically ventilated patients Rates of postoperative hypothermia

How often do we learn from our mistakes?

What percentage of months does each area learn from mistakes

Monitor percentage of months in which the department creates a shared story

How well have we created a culture of patient safety?

Annual assessment of safety culture at the unit level

Percentage change in culture scores for each care area

3  Measuring and Improving Safety

23

Measuring Defects To measure safety, we often estimate reliability in defects per unit, or Sigma, with 1 Sigma defined as defects per units of 10, 2 Sigma as defects per unit of hundreds, 3 Sigma defects per thousand, 4 Sigma defects per ten thousand, 5 Sigma defects per hundred thousand, and 6 Sigma defects per million. Measuring safety is difficult, and the methods are evolving.8 Often we are not clear regarding the unit of analysis for the denominator. In Rapid Response Systems (RRSs), for example, is the appropriate denominator the total number of hospital admissions, or the number of admissions who have no “Do Not Resuscitate” orders? The defect rate can be influenced significantly by the chosen denominator. Measures are often chosen based on their ease of collection (commonly from institutional databases that are kept to satisfy regulatory requirements). They may not reflect the true impact of the safety intervention. There may be measures that better define the impact of the intervention. For example, in RRSs, mortality data is usually analyzed based on total in-hospital mortality because it is easy to glean from commonly kept databases. Moreover, the staff may not be aware of the quality and safety measures collected by the institution and these measures may lack meaning for the frontline staff expected to use them to improve safety. System-level measures need to be meaningful to the workers in their local areas. In our zeal to improve patient safety we run the risk of creating measures of safety as though the goals were based on increasing the number of identified defects rather than actually learning from those defects. In this haste, we have often compromised validity. Many organizations use rates of self-reported adverse drug events as a measure of safety without recognizing that, as for all outcome measures, variations in the method of data collection/definition/data quality, case-mix, and quality, as well as chance, influence outcomes.19 Changes in these rates may reflect an improved culture of safety but they do not indicate that we have learned from them. Moreover, variations in data quality and case-mix are likely to be far greater than the variation in safety, which limits our ability to make inferences about quality of care from these measures. Measures of safety and quality must be important, usable, scientifically sound, and feasible. Importance and usability are value judgments that are typically made by the group, institution, or organization that decides to measure a particular area. Importance is relative and usability refers to its ease of use by those who seek to directly improve safety. Scientifically sound refers to validity and reliability. An indicator is deemed valid if the following criteria are met (www.rand. org)22: • Adequate scientific evidence or professional consensus exists supporting the indicator. • There are identifiable health benefits to patients who receive care specified by the indicator. • Based on experience, health professionals with significantly higher rates of adherence to an indicator would be considered higher-quality providers.

24

B.D. Winters et al.

• Most factors that determine adherence to an indicator are under the control of the health professional (or are subject to influence by the health professional, such as smoking cessation). An indicator is considered to be feasible if 22: • The information necessary to determine adherence is likely to be found in a ­typical medical record. • Estimates of adherence to the indicator based on medical record data are likely to be reliable and unbiased. • A reliable measure produces similar results when measurement is repeated. In many efforts to measure quality of care and safety, the measures are collected without the support of additional staff. With such human resource limitations, the feasibility of a measure figures prominently in its success. To measure quality, we need valid numerators (defects) and denominators (risk pool). To be scientifically sound, both the numerator and denominator must be valid and reliable. Yet there are challenges in measuring both. What constitutes a defect is not well defined in most healthcare areas limiting the ability to measure a numerator. For example, substantial evidence suggests that controlling blood sugar in patients in an intensive care unit (ICU) reduces mortality, yet we do it infrequently. But what might constitute a defect in glucose control? Is it a single high blood sugar, two high sugars, or the average sugar over some period above a defined threshold? In addition, it is unclear what the unit of analysis should be for the denominator. The choice of denominator can change performance by several Sigmas. For ­example, aviation and anesthesiology changed its denominators from minutes flown to takeoffs and landings, and minutes of care to a case, respectively. Thus, if an average flight was over 100 miles, or an average anesthesia case 100 min, the defect rate could change 2 Sigmas without any change in safety. Consider also ways to measure rates of failed extubation in the ICU: Should the denominator be the patient, the ventilator day, or the number of attempted extubations? Unfortunately there are often trade-offs between validity and feasibility of data collection. Often the better the validity of the measure, the greater the workload required to collect it. This may increase both human and other resource needs to the point where feasibility suffers. The converse is also often true where easily collected measures may not be the most valid. It is also important to distinguish whether we are measuring the reliability of a process (what we do) or an outcome (the results we get). While commercial aviation is believed to perform at 6 Sigmas for crashes (outcome), it performs at 1 or 2 Sigmas for on-time departures. Intuitively, outcome measures are more appealing than process measures, yet measuring outcomes pose added risk for bias that often leads to little or no useful information.19,23 Process measures may, like on-time departures, have a more direct impact on the common experience in healthcare. Reliability of an outcome measure can be influenced by variations in the methods of surveillance, in methods of data collection and definitions, in case-mix, in true variation in safety, and random error.23 Among institutions, variation in quality is

3  Measuring and Improving Safety

25

often significantly smaller than variation of other variables. In healthcare, we need to work toward standardized measures of reliability. One excellent example is the National Nosocomial Infections Surveillance (NNIS) program (now called the NHSN) that provides standardized methods to monitor rates of health care-acquired infections.13 Evidence-based processes of care (defects of omission) lend themselves to monitoring rates. However, we currently only have a limited number of validated process measures. A more diverse group of quality measures is needed and these should be appropriately monitored as defect rates. In addition, healthcare organizations need to recognize that the value of some defects lies solely in learning from the numerator; the costs of obtaining an appropriate denominator, even if methodologically feasible, may be prohibitive. Measures of safety need to be valid, yet we can learn from defects that lack denominators. How do we select measures? W. E. Deming again provides some guidance. Measures should be selected to optimize learning: that is, ensure the measure has face validity - is it important to the person expected to use the data? To develop measures that are clinically meaningful, we need the combined input of diverse and independent sources and apply methodological rigor. For example, the exposure risk for a failed extubation is an attempted extubation. Yet rates of failed extubation are often presented using patients or ventilator days as the denominator. 24 To ­estimate feasibility, first test-run the data collection tools. Moreover, the measurement of safety should be approached with the same rigor as that applied in clinical research. Whether we are measuring bloodstream infections as part of a federally funded trial or for hospital safety efforts, we need a valid measure of infections. Much research is needed to advance the science of measuring defects. Given this, what are some possible measures of safety for RRSs? Unanticipated ICU admission, in-hospital mortality and the incidence of cardio-respiratory arrest are commonly used measures, and while they all seem to meet many of the criteria, each has potential problems. For example, unanticipated ICU admission lacks validity. While the intuitive goal would be to reduce this rate, we do not know whether an increase or decrease is high-quality care. Patients who are prevented from deteriorating to arrest may still be sick enough to require ICU admission and since those who arrest and die go to the morgue, an increase in this measure may, in fact, be an indicator of better quality care. Additionally, local culture and general ward resources may have a strong influence on this outcome, leading to wideranging results across institutions. On the other hand, use of chest compressions or intubations may be an appropriate numerator for defects. In addition to the numerator, we need to consider an appropriate denominator or risk group. Although hospital admissions are used commonly as the denominator, patient days may be a more valid denominator. A patient’s risk for arrest is influenced by, among other things, the length of time he is in the hospital. The longer a patient is in the hospital, the greater the risk. Hospital mortality and length of stay may be measures of safety for RRSs but, as with all outcome measures, case-mix will significantly influence these outcomes making comparisons among hospitals difficult to interpret.23 As long as a hospital does not change a product line, case-mix within a hospital is relatively constant, making changes in mortality rate within a hospital

26

B.D. Winters et al.

potentially important and measurable. However, even an apparently unbiased outcome such as in-hospital mortality may not be using the most valid denominator. A certain portion of hospital mortality is not preventable and cannot be expected to be reduced by RRSs. Defining what is preventable and non-preventable is much more difficult and this choice of denominator may lead to different results. Much more effort is needed to produce scientifically sound and feasible measures of safety for RRS. Process measures are an alternative to outcome measures that have only recently been explored as an opportunity for RRS to have a positive impact. Recent evidence suggests that RRSs can have a positive impact on process measures such as appropriate institution of “Not for Resuscitation” status25 and implementing early goal-­ directed fluid resuscitation for severe sepsis and septic shock.26 These types of patient safety and quality measures deserve more research, especially in the area of RRSs. We can also measure context. There are multiple tools to measure culture and scores on culture that can be linked to process and outcomes to better understand context. In addition, we have developed standardized tools (called the team checkup tool) to measure leadership and team behaviors that can also be linked to process and outcome measures. Finally, qualitative analysis of context (asking staff to reflect on what worked and what did not) can help develop conceptual models and hypothesis that can later be tested.

How Might We Improve Safety? There are many examples of nearly flawless, highly reliable systems from nuclear power to commercial airlines to auto racing pit crews. These systems have in common standardized processes, specified roles, high levels of training and education, and a culture dedicated to high performance. Processes are repeatedly rehearsed until a high degree of fidelity is achieved. Thus, despite the high-risk nature of many of these systems, their safety and performance rates are excellent. How might these systems inform patient safety? Virtually all organizations are aware of the need to improve patient safety, and most have committed to doing so. Yet fewer have a clear plan of attack to accomplish this goal. The drive to improve patient safety is still new in healthcare despite the decade since the IOM’s call to action, and culture change takes effort and time. We must view healthcare delivery as a science as well as an art if we are to improve safety. Standardization of processes and applying scientific methods can help achieve higher levels of fidelity and reduce the defects that imperil our patients. Here we present an overview of measuring and reducing defects in healthcare and suggest some potential system-level measures of safety.

A Framework to Improve Reliability In healthcare, most of our processes are between 1 and 2 Sigmas. For a wide variety of processes, patients can rely on receiving the interventions they should half the time,

3  Measuring and Improving Safety

27

or 1 Sigma.27 For some outcomes, defects are 2 to 3 Sigmas - for example, catheter -related bloodstream infection rates and rates of ventilator-associated pneumonia are typically between 1 and 20 per 1,000 catheter or ventilator days, respectively.13,28 Nevertheless, there are some notable exceptions in anesthesiology and in transfusion medicine that are estimated to be 4 or 5 Sigmas (defects per 10,000 or 100,000).29,30 Improving reliability first depends on creating a culture of safety where the entire care team makes the patient their “North Star,” according to which they ­create and implement common goals. A culture of safety allows all members of the care team to speak up when they have concerns and listen when others voice ­concerns. It also creates a “zero tolerance” attitude towards defects. Next comes standardization, specifying what is done and when it should be done.31–33 This contrasts with traditional practice in which the “art of medicine” and “eminence-based medicine” trumps the science and “evidence-based medicine” - individual caregiver practice is unstructured and at times appears chaotic (i.e., care givers do what they want, when they want). In the ICU, the therapies that a patient receives depend more on who is making the rounds, rather than what the evidence suggests resulting in defects in the 1–2 Sigma range. Blood banking approaches the reliability of non-healthcare settings such as commercial aviation because they are standardized. Without standardization, reliability will remain at 10−1 imparting a significant toll on patients. An important aspect of standardization is to simplify or reduce complexity. Every step is a process that has an independent probability of failure. As such, processes that have five steps are more likely to fail than those that have 4, 3, or 2 steps. An analogy is the telephone game, in which a story is told through a series of people: the risk factors for getting a garbled story (a defect) at the end are defined by how complex the story is and how many people it passes through. While this may be an oversimplification, as there are feedback loops that may catch mistakes if we reduce the number of steps in a process, we have a higher probability of improving reliability. Undoubtedly this is an oversimplification, since there are feedback loops that may catch mistakes. Nevertheless, it is helpful to consider ­simplification when we examine our work processes. For example, in our effort to eliminate catheter-related bloodstream infections (CRBSI), our team noted that the operator often had to obtain the equipment for insertion from many different sources. This often led to missing or erroneous items requiring the operator to break sterility in order to get them, or, worse, do without. The “line cart” was instituted as a mobile central location for all supplies necessary for central catheter insertion. This reduces complexity and contributes to the operator’s adherence to our central line insertion guidelines. A second aspect of standardization is ensuring that evidence is translated into practice. Evidence-based therapies that can reduce harm are often not translated into practice because, while significant research efforts are made in developing the therapies, little effort is made to determine how to best deliver these therapies. One model34 designed to address this problem is based on five key components: 1. Focus on systems rather than individual patients; 2. Engage local multi-disciplinary teams to take ownership of the initiative;

28

B.D. Winters et al.

3. Create centralized support for the technical aspects; 4. Encourage local adaptation; and 5. Create a collaborative culture within the local unit and larger system. The specific steps of the model include summarization of the evidence, identifying local barriers to implementation, performance measurement and ensuring all patients receive the intervention. Successful use of this model to support standardization depends also on creating and sustaining the culture of safety. This can be achieved by using the Comprehensive Unit Safety Program (CUSP).35 The CUSP includes five steps: 1. Educate staff on the science of safety; 2. Identify hazards; 3. Partner a senior executive the CUSP team; 4. Learn from one defect per month; and 5. Implement teamwork tools that address communication struggles. The learning objectives include the following: 1. Understand that safety is a property of the system; 2. Understand basic principles of safe design (these include standardizing work, creating independent checks [checklists] for key processes, and learning from mistakes); 3. Recognize that the principles of safe design apply to technical as well as team work; and 4. Understand that teams make wise decisions when there is diverse and independent input. Using strategies such as CUSP can help cement these approaches and concepts into the institutional culture. When this occurs, safety becomes of paramount importance to the whole system. Third, we need to identify and learn from defects as described in the CUSP model. This involves creating independent checks to identify defects. A significant challenge we face in healthcare is developing a shared definition or concept of a defect. To illustrate, we developed a glucose protocol for our ICU, but were only capturing about 80% of patients. To improve reliability, we needed an independent check to identify defects but we had not defined a defect. Although we could have defined it in multiple ways, we decided that in the morning during the shift change, the nurses would review a patient’s glucose and if two blood sugars were out of range, they would talk to the physician and implement the protocol. We defined the defect first and then created an independent check to identify it. To learn from defects, we need to investigate what went wrong and make recommendations for improvement. We have developed a tool kit to learn from a defect. This tool kit (see Table 3.3) helps uncover what happened, why it happened, and what must be done to fix the defect. It differs from Root Cause Analysis (RCA), commonly used by institutions for evaluating sentinel and other critical events, in that while it seeks to answer these questions, it also addresses mitigating factors that may have ameliorated the harm. This has value for application to other defects and processes besides the one at hand.

3  Measuring and Improving Safety

29

Table 3.3  How to investigate a defect Problem statement:  Health care organizations could increase the extent to which they learn from defects What is a defect?  A defect is any clinical or operational event or situation that you would not want to happen again. These could include incidents that you believe caused a patient harm or put patients at risk for significant harm Purpose of tool:  The purpose of this tool is to provide a structured approach to help caregivers and administrators identify the types of systems that contributed to the defect and follow up to ensure safety improvements are achieved Who should use this tool? Clinical departmental designee at morbidity and mortality rounds Patient care areas as part of the Comprehensive Unit-based Safety Program (CUSP) All staff involved in the delivery of care related to this defect should be present when this defect is evaluated. At a minimum, this should include the physician, nurse, and administrator, and others as appropriate (e.g., medication defect includes pharmacy, equipment defect includes clinical engineering) How to use this tool:  Complete this tool on at least 1 defect per month. In addition, departments should investigate all of the following defects: liability claims, sentinel events, events for which risk management is notified, case presented to morbidity and mortality rounds and health care–acquired infections Investigation process I. Provide a clear, thorough, and objective explanation of what happened. II. Review the list of factors that contributed to the incident and check off those that negatively contributed and positively contributed to the impact of the incident. Negative contributing factors are those that harmed or increased risk of harm for the patient; positive contributing factors limited the impact of harm III. Describe how you will reduce the likelihood of this defect happening again by completing the table. List what you will do, who will lead the intervention, when you will follow up on the intervention’s progress, and how you will know risk reduction has been achieved Investigation process I. What happened?  (Reconstruct the timeline and explain what happened. For this investigation, put yourself in the place of those involved in the event as it were unfolding, to understand what they were thinking and the reasoning behind their actions/decisions when the event occurred.) An African American male >65 years old was admitted to a cardiac surgical ICU in the early morning hours. The patient was status-post cardiac surgery and on dialysis at the time of the incident. Within two hours of admission to the ICU it was clear that the patient needed a transvenous pacing wire. The wire was threaded using an IJ Cordis sheath, which is a stocked item in the ICU and standard for pulmonary artery catheters, but not the right size for a transvenous pacing wire. The sheath that matched the pacing wire was not stocked in this ICU, because transvenous pacing wires are used infrequently. The wire was threaded and placed in the ventricle but staff soon realized that the sheath did not properly seal over the wire, thus introducing risk of an air embolus. Since the wire was pacing the patient at 100%, there was no possibility for removal at that time. To reduce the patient’s risk of embolus, the bedside nurse and resident sealed the sheath using gauze and tape II. Why did it happen? Below is a framework to help you review and evaluate your case Please read each contributing factor and evaluate whether it was involved, and if so, whether it contributed negatively (increased harm) or positively (reduced impact of harm) to the incident (continued)

30

B.D. Winters et al.

Table 3.3  (continued) Contributing factors (example)

Negatively contributed

Positively contributed

Patient factors Patient was acutely ill or agitated (Elderly patient in renal failure, secondary to congestive heart failure.) There was a language barrier (Patient did not speak English.) There were personal or social issues (Patient declined therapy.) Task factors Was there a protocol available to guide therapy? (Protocol for mixing medication concentrations is posted above the medication bin.)

XX

Were test results available to help make care decision? (Stat blood glucose results were sent in 20 min.) Were tests results accurate? (Four diagnostic tests done; only magnetic resonance imaging [MRI] results needed quickly — results faxed.) Caregiver factors Was the caregiver fatigued? (Tired at the end of a double shift, nurse forgot to take a blood pressure reading.) Did the caregiver’s outlook/perception of own professional role impact on this event? (Doctor followed up to make sure cardiac consultation was done expeditiously.) Was the physical or mental health of the caregiver a factor? (Caregiver was having personal issues and missed hearing a verbal order.) Team factors Was verbal or written communication during handoff clear, accurate, clinically relevant, and goal-directed? (Oncoming care team was debriefed by outgoing staff regarding patient’s condition.) Was verbal or written communication during care clear, accurate, clinically relevant, and goal-directed? (Staff was comfortable expressing concern regarding high medication dose.) Was verbal or written communication during crisis clear, accurate, clinically relevant and goal-directed? (Team leader quickly explained and directed the team regarding the plan of action.) Was there a cohesive team structure with an identified and communicative leader? (Attending physician gave clear instructions to the team.) Training and education factors Was the caregiver knowledgeable, skilled, and competent? (Nurse knew dose ordered was not standard for that medication.) Did the caregiver follow the established protocol? (Provider pulled protocol to ensure steps were followed.) Did the caregiver seek supervision or help? (New nurse asked preceptor to help mix medication concentration.)

XX

(continued)

3  Measuring and Improving Safety

31

Table 3.3  (continued) Information technology/computerized physician order entry factors Did the computer/software program generate an error? (Heparin was chosen, but Digoxin printed on the order sheet.) Did the computer/software malfunction? (Computer shut down in the middle of provider’s order entry.) Did the user check what was entered to make sure it was correct? (Caregiver initially chose 0.25 mg, but caught error and changed it to 0.025 mg.) Local environment factors Was adequate equipment available and was it working properly? (There were two extra ventilators stocked and recently serviced by clinical engineering.) Was operational (administrative and managerial) support adequate? (Unit clerk out sick, but extra clerk sent to cover from another unit.) Was the physical environment conducive to enhancing patient care? (All beds were visible from the nurse’s station.) Was enough staff on the unit to care for patient volume? (Nurse ratio was 1:1.) Was there a good mix of skilled and new staff? (A nurse orientee was shadowing a senior nurse and an extra nurse was on to cover the senior nurse’s responsibilities.) Did workload impact the provision of good care? (Nurse caring for 3 patients because nurse went home sick.)

XX

Institutional environment factors Were adequate financial resources available? (Unit requested experienced patient transport team for critically ill patients, and one was made available the next day.) Were laboratory technicians adequately in-serviced/educated? (Lab technician was fully aware of complications related to thallium injection.) Was there adequate staffing in the laboratory to run results? (There were 3 dedicated laboratory technicians to run stat results.) Were pharmacists adequately in-service/educated? (Pharmacists knew and followed the protocol for stat medication orders.) Did pharmacy have a good infrastructure (policy, procedures)? (It was standard policy to have a second pharmacist do an independent check before dispensing medications.) Was there adequate pharmacy staffing? (There was a pharmacist dedicated to the ICU.) Does hospital administration work with the units regarding what and how to support their needs? (Guidelines established to hold new ICU admissions in the emergency department when beds are not available in the ICU.) III. How will you reduce the likelihood of this defect happening again?

Specific things to be done to reduce the risk of the defect

Who will lead this effort?

Follow-up date

Bedside nurse called Central Supply and requested pacing wires and matching sheaths be packaged together

Bedside nurse

1 week

How will you determine the risk is reduced? (action items) Supplies are packaged together

32

B.D. Winters et al.

These steps: 1. Create a culture of safety; 2. Standardize what and when actions are done; and 3. Identify and learn from defects - provide a framework to improve reliability. Transfusion medicine offers an example of how the application of these principles created a high reliability process: using discharge data, the estimated incidence of a transfusion reaction in healthcare is 4 per 100,000. How did they achieve such success? They standardized, created independent checks for key processes, and learned from defects (see Table 3.3). Physicians often resist standardization and outside regulators, such as the Joint Commission. All too often they are relied upon to force the change in culture. This is in fact the case for RRSs. It took the Joint Commission requiring, as one of its patient safety goals for 2009, that hospitals implement better systems to respond to deteriorating patients on general hospital wards (www.jointcommission.org). While they did not explicitly require Medical Emergency or Rapid Response Teams (RRTs), these are a logical choice to meet this requirement. The concept of physician autonomy is deeply ingrained in the practice of medicine and is often at odds with the need to standardize practice. When physicians are asked to relinquish their autonomy in order to standardize practice, we need to be sure that those standards are just, wise and supported by the best evidence possible. The process by which these standards are developed needs to be transparent with the full account of risks, benefits and cost estimates considered.36 Although regulations may be an important driver for standardization and culture change, there are far too many processes to rely on this strategy alone. Indeed, we need the courage of leaders within our healthcare systems to support standardization and culture change. To date, most efforts to improve reliability of evidence-based therapies in healthcare have focused on practice guidelines: a series of conditional probability, or “if yes then ‘x’” statements.34 The Centers for Disease Control and Prevention’s (CDC’s) guidelines for preventing catheter-related bloodstream infections, a nearly 100-page document (www.cdc.gov), is one example. It is not surprising that the use of guidelines alone has met with little success.37,38 Under time pressure, it is difficult for care givers to think in terms of conditional probabilities.39 An additional problem is that most guidelines have been developed for physicians, ignoring other members of the care team who could provide an independent check. A checklist is one tool to help standardize work processes and increase reliability. Checklists have led to significant improvements in aviation, nuclear power, and rail safety. Useful checklists transform a complex diagnostic/­ therapeutic decision into a series of simple yes/no tasks. It is first necessary to identify which parts of a task are “mission critical,” especially those supported by strong evidence, and develop measures for those tasks. In addition to the “line cart” for central line insertion, we developed a checklist based on five key ­processes distilled from the CDC guidelines for preventing catheter-related infection and virtually eliminated this defect in our ICUs.28 This kind of checklist can

3  Measuring and Improving Safety

33

also be used to monitor performance, with each item serving as a process measure of quality of care.7,40 Measurement then becomes a tool to improve performance.

Why RRSs Can Improve Safety Rapid Response Systems are well grounded in the science of safety outlined above. In many, and perhaps even most, sentinel events, someone either did not speak up, or spoke up but was not heeded because of a hierarchical or punitive culture (people had previously been reprimanded when they spoke up). With RRSs, frontline staff are empowered - indeed encouraged - to call the MET or RRT when they are concerned. This requires a strong culture of safety. Frontline staff is also educated to call for help based on a standardized set of parameters, in the absence of which the trigger for calling for help happens too late once an arrest has occurred. The RRS identifies problems early, when there is still time to recover from them. As such, RRSs are based on sound safety theory and would be expected to improve safety.

Summary The science of measuring safety is gradually maturing. Some measures of safety lend themselves to rates, while others do not. We have described an approach for organizations to answer the question, “Are patients safer?” We also have summarized the issues regarding measuring and improving reliability, and provided a framework for improving safety. With these measures, we defer to the wisdom of care givers and administrators to identify and mitigate safety concerns, but also attempt to provide a framework to assist the caregiver with safety efforts. The need to improve quality and safety is significant, and hospitals are learning how to accomplish this goal. Rapid Response Systems are grounded in safety theory and offer the promise to reduce patient harm. While imperfect, the current data would suggest that they do. We hope that practical strategies such as those proposed here will help move safety and quality efforts forward.

References 1. Kohn LT, Corrigan JM, Donaldson MS, eds. To Err Human: Building a Safer Health System. Washington, DC: National Academies Press; 2000. 2. Altman DE, Clancy C, Blendon RJ. Improving patient safety – five years after the IOM report. N Engl J Med. 2004;351(20):2041–2043. 3. Wachter RM. The end of the beginning: patient safety five years after “To Err Is Human.” Health Aff. 2004:1–12.

34

B.D. Winters et al.

4. Crossing the Quality Chasm: a New Health System for the 21st Century. Committee on Quality of Health Care in America, Institute of Medicine. Washington, DC: National Academies Press; 2001. 5. Donabedian A. Evaluating the quality of medical care. Millbank Mem Fund Q. 1966;44(3): 166–206. 6. Paine LA, Baker DR, Rosenstein B, Pronovost P. The Johns Hopkins Hospital: identifying and addressing risks and safety issues. Jt Comm J Qual Saf. 2004;30(10):543–550. 7. Haya R, Rubin H, Pronovost PJ, Diette G. The advantages and disadvantages of process-based measures of health care quality. Int J Qual Health Care. 2001;13(6):469–474. 8. Pronovost PJ, Nolan T, Zeger S, Miller M, Rubin H. How can clinicians measure safety and quality in acute care? Lancet. 2004;363(9414):1061–1067. 9. Sexton JB, Thomas EJ, Helmreich RL. Error, stress, and teamwork in medicine and aviation: cross-sectional surveys. BMJ. 2000;320(7237):745–749. 10. Shortell SM, Marsteller JA, Lin M, et al. The role of perceived team effectiveness in improving chronic illness care. Med Care. 2004;42(11):1040–1048. 11. Sexton JB, Klinect JR. The link between safety attitudes and observed performance in flight operations. In: Proceedings of the 11th International Symposium on Aviation Psychology. Columbus, OH: The Ohio State University; 2001:7–13. 12. Pronovost PJ, Weast B, Bishop K, et al. Senior executive adopt-a-work unit: a model for safety improvement. Jt Comm J Qual Saf. 2004;30(27):59–68. 13. NNIS System. National Nosocomial Infections Surveillance (NNIS) system report, data ­summary from January 1992 through June 2003, issued August 2003. Am J Infect Control. 2003;31(8):481–498. 14. Bates DW, Leape LL, Cullen DJ, et  al. Effect of computerized physician order entry and a  team intervention on prevention of serious medication errors. JAMA. 1998;280(15): 1311–1316. 15. Flynn EA, Barker KN, Pepper GA, Bates DW, Mikeal RL. Comparison of methods for detecting medication errors in 36 hospitals and skilled-nursing facilities. Am J Health Syst Pharm. 2002;59(4):436–446. 16. Lesar TS, Briceland LL, Delcoure K, Parmalee JC, Masta-Gornic V, Pohl H. Medication prescribing errors in a teaching hospital. JAMA. 1990;263(17):2329–2334. 17. Lesar TS, Briceland L, Stein DS. Factors related to errors in medication prescribing. JAMA. 1997;277(4):312–317. 18. Leape LL, Cullen DJ, Clapp MD, et  al. Pharmacist participation on physician rounds and adverse drug events in the intensive care unit. JAMA. 1999;282(3):267–270. 19. Lilford RJ, Mohammed MA, Braunholtz D, Hofer TP. The measurement of active errors: methodological issues. Qual Saf Health Care. 2003;12(suppl 2):ii8–ii12. 20. Hayward RA, Hofer TP. Estimating hospital deaths due to medical errors: preventability is in the eye of the reviewer. JAMA. 2004;286(4):415–420. 21. Cook DJ, Montori VM, McMullin JP, Finfer SR, Rocker GM. Improving patients’ safety locally: changing clinician behaviour. Lancet. 2004;363(9416):1224–1230. 22. Brook RA. The RAND/UCLA Appropriateness Method. Methodology Perspectives. AHCPR 95–0009. Rockville, MD: Public Health Service; 1994:59–70. 23. Lilford R, Mohammed MA, Spiegelhalter D, Thomson R. Use and misuse of process and outcome data in managing performance of acute medical care: avoiding institutional stigma. Lancet. 2004;363(9415):1147–1154. 24. Pronovost PJ, Jenckes M, To M, et al. Reducing failed extubations in the intensive care unit. Jt Comm J Qual Improv. 2002;28(11):595–604. 25. Jones DA, McIntyre T, Baldwin I, Mercer I, Kattula A, Bellomo R. The medical emergency team and end-of-life care: a pilot study. Crit Care Resusc. 2007;9(2):151–156. 26. Sebat F, Musthafa AA, Johnson D. Effect of a rapid response system for patients in shock on time to treatment and mortality during 5 years. Crit Care Med. 2007;35(11):2568–2575. 27. McGlynn EA, Asch SM, Adams J, et al. The quality of health care delivered to adults in the United States. N Engl J Med. 2003;348(26):2635–2645.

3  Measuring and Improving Safety

35

28. Berenholtz SM, Pronovost PJ, Lipsett PA, et  al. Eliminating catheter-related bloodstream infections in the intensive care unit. Crit Care Med. 2004;32(10):2014–2020. 29. Romano PS, Geppert JJ, Davies S, Miller MR, Elixhauser A, McDonald KM. A national profile of patient safety in U.S. hospitals. Health Aff. 2003;22(2):154–166. 30. Zhan C, Miller MR. Excess length of stay, charges, and mortality attributable to medical injuries during hospitalization. JAMA. 2004;290(14):1868–1874. 31. Reason J. Combating omission errors through task analysis and good reminders. Qual Saf Health Care. 2002;11(1):40–44. 32. Reason J. Managing the risks of organizational accidents. Aldershot: Ashgate Publishing Ltd; 1997. 33. Weick KE, Sutcliffe KM. Managing the Unexpected: Assuring High Performance in an Age of Complexity. San Francisco: Jossey-Bass; 2001. 34. Pronovost PJ, Berenholtz SM, Needham D. Translating evidence into practice: a model for large scale knowledge translation. BMJ. 2008;377:a1714. 35. Pronovost PJ, Berenholtz SM, Goeschel C. Improving patient safety in intensive care units in Michigan. J Crit Care. 2008;23(2):207–221. 36. Matthews SC, Pronovosot PJ. Physician autonomy and informed decision making: Finding the balance for patient safety and quality. JAMA. 2008;300(24):2913–2915. 37. Grol R. Improving the quality of medical care: Building bridges among professional pride, payer profit, and patient satisfaction. JAMA. 2001;286(20):2578–2585. 38. Gross PA, Greenfield S, Cretin S, et  al. Optimal methods for guideline implementation: ­conclusions from Leeds Castle meeting. Med Care. 2001;39(8 suppl 2):II85–II92. 39. Klein GA. Sources of Power: How People Make Decisions. Cambridge Massachusetts Institute of Technology: MIT Press; 1999. 40. Haya R, Rubin H, Pronovost P, Diette G. Methodology matters. From a process of care to a  measure: The development and testing of a quality indicator. Int J Qual Health Care. 2001;13(6):489–496.

Chapter 4

Integrating a Rapid Response System into a Patient Safety Program John Gosbee

Keywords  Integrating • MET • Patient • Safety • Program • Rapid • Response

Overview Since at least the publication of the Institute of Medicine’s report To Err Is Human,1 most healthcare organizations have been struggling to find and eliminate hazards. Their struggle arises from the complex mixture of issues that plague any organization dealing with the seemingly easy problems to be solved by a new safety program. Many healthcare organizations soon realize they are dealing with organizational psychology issues that require tools from change management. Somewhat fewer facilities are aware of the problems ingrained in human factors engineering of systems, devices, and tools. We will define these terms and how they apply to Rapid Response Systems (RRS) and their efferent limb, the Medical Emergency Team (MET) and Rapid Response Team (RRT) throughout this chapter. An MET or RRT response is not just a wonderful tool to improve morbidity and mortality associated with hospital medical crises and cardiopulmonary resuscitation, it is also an indirect tool to address the struggles to improve quality and safety throughout a healthcare organization. Conceptually and empirically, most hospitals will likely need RRS programs due to findings from human factors engineering and healthcare.

J. Gosbee (*) Human Factors Engineering & Healthcare Specialist, Red Forest Consulting, LLC and University of Michigan Health System, 3812 Lake Pointe Lane, Ann Arbor, Michigan, USA e-mail: [email protected]

M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_4, © Springer Science+Business Media, LLC 2011

37

38

J. Gosbee 

Creating and Sustaining Safety The difficulty of creating and sustaining a patient safety program cannot be underestimated. Logistic and strategic questions quickly overwhelm the best and brightest: What are the most frequent or remediable adverse events that hurt and kill patients? Why did these adverse events occur? What can we do about the root causes of these events? What sources can provide effective remedies? Why are people so resistant to using safety remedies? To answer these questions, hospitals have used safety methods required by regulatory organizations (e.g., Joint Commission on Accreditation of Healthcare Organizations (JCAHO)) or governments (e.g., state departments of health). These methods include root-cause analysis (RCA); failure mode and effect analysis (FMEA); and traditional quality improvement tools.2,3 These safety and quality approaches work best in organizations that are developing a so-called “high-­ reliability organization.”4 An RRS can complement RCA and FMEA activities that will be described more fully. An RRS can also provide the tangible proof that the organization is serious about the “safety culture” described for high-reliability organizations. Specifically, certain aspects of RRSs are especially suited to meet many of the criteria in a specific model of organizational change described by Rodgers,5 a model that has been accepted for many decades. This theory looks at factors such as perceived relative advantage, compatibility with existing values, and norms and trial-ability.

Definition and Relevance of Human Factors Engineering The human factors engineering field is several decades old and has been applied in various organizations and domains when they face design, personnel, and policy issues such as those surrounding an RRS.6 Briefly, human factors engineering is the discipline that studies human capabilities and limitations and applies that information to safe, effective, and comfortable system design.7,8 It includes the design of tools, machines, and systems that take into account human capabilities, limitations, and characteristics. Ergonomics, usability engineering, and user-centered design are considered synonymous or closely related to human factors engineering, which is based on design-related aspects of several biomedical disciplines. From a systems perspective, a person is receiving input from a “clinical assessment machine,” processing that input, and creating an output that goes to the “healthcare machine.” Anthropometrics and biomechanics cover most of the physical aspects of input and output. The science of sensation and perception is related to input to the person. Cognitive psychology, which covers models and theories of human performance, memory, and attention, relates to the processing of the input and ­initiating the output. Observations and studies regularly conclude that many design issues thwart even the best attempts at resuscitation and the application of critical care expertise.9

4  Integrating a Rapid Response System into a Patient Safety Program

39

Some researchers have seen problems with using defibrillators ‑even those made for novices.10 Others have identified design problems with defibrillators, even when testing individuals like paramedics who use them often.11 The layout and human factors aspects of the medication drawers in many crash carts can add minutes of delay to well-intentioned and -motivated clinicians and their ability to retrieve key medications.12 Lack of proper transitions of care and teamwork during and following resuscitation exists even in the best clinical care.13 The breakdowns and missed opportunities are accentuated by time pressure, design of devices, and even the layout and furnishings of the resuscitation area.14 All of this evidence points to the need to use METs or RRTs to avoid or abort crises ‑even if the most highly skilled personnel and fully staffed settings are available.

The MET as a Driving Force for a Patient Safety Program For many reasons, an RRS and events that the METs or RRTs respond to can be a key driving force for a hospital safety and quality program. First, there are the difficulties and limitations of commonly used safety methods. Second, an RRS is a broad-sweeping safety initiative that impacts and is visible to many sites in the healthcare organization, and to many types of professionals and personnel. Third, as a safety or quality activity, RRSs have the most successful change attributes that are cited in change management theory and practice.

Root Cause Analysis Most healthcare organizations perform many RCAs per year due to JCAHO requirements and the general standard of practice in the patient safety movement. The basics of the RCA include: 1 . Deciding when to do an RCA 2. Figuring out what happened (e.g., people and devices involved, sequence of events) 3. Making decisions about root causes and contributing factors 4. Developing remedies or action plans, and approaches to measure effectiveness 5. Convincing and selling to management and staff, and then (hopefully) implementing the action plans There may be more than one team or individual doing these general steps. Depending on the event to be studied, each general step might take hours or weeks to complete. An MET or RRT can aid all five general steps in the RCA process, but has the most effect on two troublesome steps: deciding when to do an RCA, and convincing, selling, and implementing action plans. When initiating an RCA on an adverse event, the hospital has to know several things besides just the severity of that specific

40

J. Gosbee 

event: What is the severity of events like the one in question, what is the frequency of this kind of adverse event, and how often is the event severe? In a robust RRS program, there are several adverse events that lead to the response team being called. Braithwaite et al identified 31% of MET events that were associated with medical errors (adverse events and close calls).15 In their hospital system experience, they found 18.4 events per 1,000 hospital admissions. This is similar to the experience of Australian hospital systems,16 which reported five MET responses per 1,000 admissions. Thus, most hospital systems could have a rather large body of knowledge about adverse events and close calls. This data would be several times larger than that obtained from auditing just cardiac arrest events, because crises seem to occur five to ten times more frequently. If the operators log all MET/RRT calls they make, the MET/RRT call itself provides two functions: it provides additional resources to prevent that patient from dying, and it enables creation of a database of crisis events to fuel patient safety reviews. Thus MET/RRT calls may help overcome perhaps the most difficult aspect of performing quality indicator activities: finding errors worth fixing, since people are notoriously poor at recognizing errors as they occur, or reporting them when they do recognize them. As mentioned previously, the events detected by MET/RRT are quite diverse. This can aid in determining which of the many types of reported events upon which a hospital should do an RCA. Braithwaite et al15 identified 67.5% of the 114 adverse event-associated MET/RRTs as diagnostic errors, such as incorrect or delayed diagnosis or delay or incorrect action following monitoring or test data. They also saw 59.6% related to treatment errors, including problems during or following surgery and medication administration. Finally, 26.3% of the MET events included problems arising from “prevention”; examples of prevention problems encompassed prophylactic treatment (e.g., anticoagulation for deep vein thrombosis) and telemetry monitoring issues for patients with hyperkalemia. The richness and diversity of data from events or vulnerabilities that lead to an MET activation can help the RCA team throughout their process. In short, people do poor root-cause analyses; some studies have demonstrated that developing accurate and specific root causes and contributing factors is problematic. Carroll et al17 looked at problem-investigation teams in nuclear power plants and chemical plants. In a quantitative analysis of 27 RCA teams at three different nuclear power plants, they found “a disappointing level of depth and completeness, insight, and clarity,” The researchers were able to make some correlation between some attributes of the team members and the deeper and clearer RCAs. Their analysis found that “more training in teamwork” and “more varied plant experience” were the strongest positive predicting attributes. These attributes may be more widespread or likely to be increased in organizations that accept the central concepts behind RRS processes, such as calling for help early.18 Another troublesome step in RCA is the final one: convincing, selling, and implementing the action plans. Action plan implementation is the step where healthcare organizations and RCA teams discover how hard it is to change a system, and how resistant personnel can be if the organization has not embraced a culture of safety. It is difficult to convince managers and frontline personnel that

4  Integrating a Rapid Response System into a Patient Safety Program

41

just one event or close call is serious enough for them to change. In essence, one must convince caregivers that their current workflow is wrong and a new, untested work plan is better. With data from adverse events leading to the creation of an MET or RRT, and workflow changes directed at preventing repeats of actual neardeath events, it is much more likely to change mindsets and practices. For instance, an RCA about an empty oxygen cylinder almost being used during patient transport might result in an action to purchase cylinders with a more direct indication of contents (indicator valves). However the same problem that caused a  near-death event from hypoxemia would more highly motivate not only the ­bedside caregivers, but also the administrators who oversee purchasing choices and materials management. Furthermore, procurement committees will likely accept the additional cost if there were more than one event. Because MET/RRT responders tend to find similar types of errors, creating lists of similar events is not difficult. For example, if the organization had five MET/RRT events that involved confusion over oxygen cylinder levels, the organization would be less likely to blame an individual and more likely to blame the system. Also, frontline respiratory therapists, purchasers, transport personnel, nursing staff, administrators, and physicians are more likely to become allies instead of naysayers if they know about all five events.

Failure Mode and Effect Analysis Most healthcare organizations perform at least one FMEA per year due to JCAHO requirements and the general standard of practice in the patient safety movement. The general steps of a FMEA include19: 1 . Identifying and prioritizing a high-risk process 2. Flowcharting the process and subprocesses 3. For each subprocess, developing potential failure modes and prioritizing based on risk (risk priority is usually the product of severity, frequency, and detectability) 4. For each failure mode cause, identifying actions to remedy them 5. Convincing, selling, and then (hopefully) implementing the action plans There may be more than one team or individual performing these general steps. Depending on the process to be studied, each general step might take hours or weeks to complete. Much less has been written about the problems or shortcomings in applying FMEA in healthcare.20 However, since FMEA shares many attributes with RCA, many of the ideas and findings about the complementary role of the RRS are likely true. In addition, some safety professionals think that there are many ways to inadvertently misuse FMEA as a safety tool.21 One could infer that data from the events proceeding RRS events and close calls seem invaluable for the FMEA team in all five general steps listed above.

42

J. Gosbee 

For example, the RRS data can be used to develop realistic failure modes and failure mode causes. An FMEA team might only consider two failure modes in the monitoring and treatment of conscious cardiac care unit patients: 1 . Malfunction of physiologic monitors 2. The patient falling while getting out of bed. RRS data would have provided them with a third failure mode such as that described by Braithwaite: bradycardia/ asystole from delay in pacemaker placement. Or the team might not conceptualize the failure mode cause of “permanent pacemaker placement considered a low priority consult.” Having a body of RRS data-associated adverse events and their causes will make the FMEA process more efficient The process of convincing and selling the action plan from FMEA is more difficult than RCA, since many healthcare personnel may not be motivated to change when no “real” event occurred. But just as the RRS data helps sell and convince personnel to try and accept action plans from RCAs (as described above), it will help promote FMEA action plans.

Safety Culture and High-Reliability Organizations MET/RRTs can also provide the tangible proof that the organization is serious about the safety culture described for high-reliability organizations. MET/RRT characteristics meet many of the criteria in one specific model of organizational change described by Rodgers.5 This model has been accepted and applied by many organizations for many years. The theory looks at five crucial factors for organizational change to occur: 1 . Perceived relative advantage 2. Perceived as compatible with existing values and norms 3. Perceived low (or lack of ) complexity 4. Trial-ability (ease of doing it on a trial basis) 5. Observe-ability (visibility of the change to non-experts) In contrast, one can see that the required safety method of RCA fares only average in each of the five organizational change factors. For those convinced that safety is an issue, they would positively perceive the relative advantage RCA (factor 1). However, many novices would see the several-hour process of RCA as burdensome when they believe the remedies, such as enforcing policies, are apparent in minutes of analysis. RCA is also perceived as somewhat compatible with existing values/norms (factor 2) and high complexity (factor 3). Since RCA team meetings often take place weekly and last 2 h, busy clinicians would say that trial-ability is low (factor 4). Finally, as evidenced in Carroll’s work,17 many RCAs result in training or policy changes that will not be viewed as really much of a change (factor 5).

4  Integrating a Rapid Response System into a Patient Safety Program

43

AN RRS comes out much better than RCA when judged against each of the five crucial factors for organizational change. The following analysis assumes that management of the organization is serious about rewarding, not punishing, providers for calling for help early. Providers see a large relative advantage to have an outside team deal with the clinical issues of their patient efficiently and effectively (factor 1). Certainly if an organization using an RRS (either RRT or MET) has nearly the success with decreasing mortality and crisis code frequency as some U.S. and Australian hospitals, the relative advantage will be clear to all stakeholders, from the boardroom to the classroom. Calling for help early, accurate and reliable team communication, and teamwork are the hallmarks of high-reliability organizations as well as RRS. Thus, Rodgers’s second factor, compatibility of values and norms, will increase as the organization moves forward. In contrast, organizations that do not value teamwork and constructive critique will have dissonance with this crucial factor for change. METs and RRTs will have to be perceived as being of low complexity. This book contains many examples of the importance of making METs and RRTs seamless with code teams, paging systems, and other hospital ward activities (see Chaps. 12, 17, and 18). Also, short lists of understandable, objective criteria for activating the RRS increase organizational acceptance. There are mixed aspects of RRSs when judging against the last two factors, trialability and observe-ability of success in organizational change (factors 4 and 5). Some providers are unaffected by changes in the organization due to MET/RRT implementation; for them, MET/RRTs seem easy to try out. However, an MET or RRT trial may not seem so easy or non-threatening to personnel in critical care units, existing crisis code teams, and others who play a role in solving urgent or emergent problems. Non-experts can understand fewer codes for acutely ill ward patients more easily, if they are made aware of this. Changes in morbidity in critical care areas may not be as appreciated by some management. However, in many facilities, quality and outcome measures of various types are now tracked by many non-direct care providers, so visibility of better outcomes would be higher and organizational change more likely in those settings of transparency.

Patient Safety Overall General concepts discussed in patient safety communities include normalization of deviance and normalization of complexity ‑overly complex phrases that get at the observations in healthcare of something being out of place or hard to use. The general observations, if correct, also conspire to permit certain complication rates for procedures and patient interventions. Careful analysis of events antecedent to an RRS activation provides a stark set of data about how healthcare devices have grown overly complex, or standard policies needlessly convoluted. Once again, the attention of healthcare personnel is focused on these issues by the criticality of the event (respiratory or cardiovascular distress) and the fact that many of the events recur.

44

J. Gosbee 

Implementation of METs or RRTs will tend to move the norm; instead of trying to prevent deaths, organizations have a tool to prevent crises, and to help prevent those crises that do occur from turning into a death. Calling for help early can be promoted to become the norm instead of the exception.22

Summary Using the prescribed patient safety tools of RCA or FMEA has only partially helped healthcare organizations with the struggle to find and eliminate system vulnerabilities. Organizations soon find that easy problems with easy answers are not common. Fortunately, the Rapid Response System, and in particular the response teams MET and RRTs provide a service that is complementary and supportive of other safety methods, since up to 30% of team activations are associated with patient safety issues. AN RRS is also needed since there are many ingrained problems with human factors engineering of systems, devices, and tools involved in resuscitation. That is, there are intransigent constraints to the maximum effectiveness in dealing with cardiopulmonary emergencies. Most important are the RRS features that organizational theory and observation support as crucial success factors to turning an organization into a highly reliable one, with a solid safety culture.

References 1. Kohn LT, Corrigan JM, Donaldson MS, eds. To Err Is Human: Building a Safer Health System. Washington, DC: National Academies Press; 2000. 2. Bagian JP, Gosbee JW, Lee CZ, Williams L, McKnight SD, Mannos DM. VA’s root cause analysis system in action. Jt Comm J Qual Improv. 2002;28:531–545. 3. Stalhandske E, DeRosier J, Patail B, Gosbee JW. How to make the most of failure mode and effect analysis. Biomed Instrum Technol. 2003;37:96–102. 4. Weick KE, Sutcliffe KM. Managing the Unexpected: Assuring High Performance in the Age of Complexity. San Francisco, CA: Jossey Bass; 2002. 5. Rodgers E. Diffusions of Innovations. 5th ed. New York: Free Press; 2003. 6. Sanders MS, McCormick EJ. Human Factors in Engineering and Design. 7th ed. New York: McGraw-Hill; 1993. 7. Gosbee JW, Lin L. The role of human factors engineering in medical device and medical system errors. In: Vincent C, ed. Clinical Risk Management: Enhancing Patient Safety. 2nd ed. London: BMJ; 2001. 8. Gosbee JW. Introduction to the human factors engineering series. Jt Comm J Qual Saf. 2004;30:215–219. 9. Donchin Y, Gopher D, Olin M, et al. A look into the nature and causes of human errors in the intensive care unit. Crit Care Med. 1995;23:294–300. 10. Mattei LC, McKay U, Lepper MW, Soar J. Do nurses and physiotherapists require training to use an automated external defibrillator? Resuscitation. 2002;53:277–280. 11. Fairbanks RJ, Shah MN, Caplan S, Marks A, Bishop P. Defibrillator usability study among paramedics. In: Proceedings of the Human Factors and Ergonomics Society 47th Annual Meeting. Santa Monica, CA: Human Factors and Ergonomics Society; 2004.

4  Integrating a Rapid Response System into a Patient Safety Program

45

12. McLaughlin RC. Redesigning the crash cart: usability testing improves one facility’s medication drawers. Am J Nurs. 2003;103:64A, 64D, 64G–64H. 13. Xiao Y, Hunter A, Mackenzie CF, Jeffries NJ, Horst R. The LOTAS Group. Task complexity in emergency medical care and its implications for team coordination. Hum Factors. 1996;38:636–645. 14. Xiao Y. Human and technology factors in coordination in emergencies. In: Medicine, Technology and Human Factors in Trauma Care: A Civilian and Military Perspective. Baltimore, MD: National Study Center; 2001:16–21. 15. Braithwaite RS, DeVita MA, Mahidhara R, Simmons RL, Stuart S, Foraida M. Medical Emergency Response Improvement Team (MERIT). Use of medical emergency team (MET) responses to detect medical errors. Qual Saf Health Care. 2004;13:255–259. 16. Bellomo R, Goldsmith D, Uchino S, et al. A prospective before-and-after trial of a medical emergency team. Med J Aust. 2003;179:283–287. 17. Carroll JS, Rudolph JW, Hatakenaka S. Learning from high hazard organizations. In: Staw B, Kramer R, eds. Research in Organizational Behavior. Greenwich, CT: JAI; 2003. 18. Carroll JS, Rudolph JW, Hatakenaka S. Lessons learned from non-medical industries: root cause analysis as culture change at a chemical plant. Qual Saf Health Care. 2002;11:266–269. 19. ECRI. An introduction to FMEA. Using failure mode and effects analysis to meet JCAHO’s proactive risk assessment requirement. Failure modes and effect analysis. Health Devices. 2002;31:223–226. 20. DeRosier J, Stalhandske E, Bagian JP, Nudell T. Using health care failure mode and effect analysis: the VA National Center for Patient Safety’s prospective risk analysis system. Jt Comm J Qual Improv. 2002;28:248–267. 21. Spath P. Worst practices used in conducting FMEA projects. Hosp Peer Rev. 2004;29:114–116. 22. Foraida M, DeVita MA, Braithwaite RS, Stuart SA, Brooks MM, Simmons RL. Improving the utilization of medical crisis teams (Condition C) at an urban tertiary care hospital. J Crit Care. 2003;18:87–94.

Chapter 5

Acute Hospitalist Medicine and the Rapid Response System David J. McAdams

Keywords  Acute • Medicine • Hospitalist • Specialty • Rapid • Response • System

History of the Hospitalist Movement Traditionally, the Primary Care Physician (PCP), usually an internist or family practice physician, has been responsible for outpatient and inpatient care. There have been many forces in healthcare that have pushed toward a separation of care provided to patients in both of these locations. Changes in hospital management systems, hospital size, increasing severity of patient illness, and out-of-control healthcare costs have all been integral in the push towards an inpatient physiciandriven care model.1 Within the context of these changes, there has been a growing sense of PCP dissatisfaction in the ability to provide timely and efficient care to both their outpatient and inpatient populations. This has ultimately given birth to the hospital medicine “specialist.” This emerging specialty is defined, much like Critical Care and Emergency Medicine, by the site of care rather than a disease, patient population, or organ-system.1 While physicians with inpatient care duties have existed both in North America and Europe for some time, the appearance of the hospital physician is a newer phenomenon. Certainly, the ever-present “house-officer” has had a place in history as the physician who essentially lives in the hospital; however, this role has been mainly restricted to the medical trainee with little experience and much responsibility, and has served as a rite of passage to become the more senior and less-present attending physician. In distinction, however, the hospitalist physician is a more experienced physician, not under the same training and hierarchical constraints as

D.J. McAdams (*) Department of Internal Medicine, University of Pittsburgh Medical Center, 200 Lothrop Street, MUH 826E, Pittsburgh, PA, USA e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_5, © Springer Science+Business Media, LLC 2011

47

48

D.J. McAdams

the house-officer, who is available at almost all times to the care needs of those patients in the hospital setting. It has been almost 10 years since the coining of the term “hospitalist” by Wachter and Goldman.1 Definitions have been reworked and adapted, but currently the Society of Hospital Medicine (SHM) has defined Hospitalists as, “physicians whose primary professional focus is the general medical care of hospitalized patients. Their activities include patient care, teaching, research, and leadership related to Hospital Medicine.”2 The drive and expansion of this newer field of medicine has been felt throughout the US and abroad, and its popularity continues to grow exponentially. During the past decade we have seen this new specialty form its own professional society, have dedicated journals, and inundate existing, wellrespected, traditionally general medicine journals with abundant evidence-based literature. SHM estimates indicate that there were about 2,000 hospitalists in 1998, 8,000 in 2003, and that number is expected to rise to as much as 30,000 by the end of 2010.2 Indeed, this is one of the only fields of medicine where there is a vast surplus of jobs comparative to physicians to fill them. The majority of hospitalists have been trained in internal medicine (89%), mostly general internal medicine (51%); there is a large subset of internists who have subspecialty training (38%), usually in pulmonary medicine, critical care medicine, or a combination of the two.3 The remainder are made up mainly of ­family practice physicians, pediatricians, and other subspecialists like infectious disease and cardiology. Currently, there is no formal training required to become a hospitalist. However it is indeed true that a significant amount of one’s training in internal medicine is geared toward care of the hospitalized patient. However, there are at least six hospitalist fellowships currently in existence with more being planned for the future. In many ways, these are designed like general internal medicine fellowships, gearing the physician towards further experience in education and clinical research while providing continued exposure to inpatients and their medical problems.

Models of Hospitalist Care Wachter has described four stages of hospital care that help to illustrate the driving forces behind hospitalist models.4 These stages help us to understand inpatient care structure, but they are not meant to be hierarchical nor do hospital systems sequentially pass through them. Rather, they are meant as a tool to help us understand that many external forces predicate how hospital care is provided. The first stage is the PCP model in which every PCP cares for his/her own patients admitted to the hospital. This has been the classic model of care in medicine. The second stage involves rotating coverage of hospitalized patients between members in a private practice. Each physician takes turns caring for those patients admitted. This model became popular as groups started to get larger and increase the number of patients in their practices. In the third stage, we see the emergence of a dedicated hospital physician who cares for inpatients. PCPs may hand over care to the hospital physician, but are

5  Acute Hospitalist Medicine and the Rapid Response System

49

not required to do so. In stage four, in contrast to the voluntary hospitalist stage, PCPs are required to hand over care to the inpatient physician. Every stage has its own associated advantages and disadvantages. For example, in stages three and four, the inpatient physician can provide continuous care to admitted patients while the PCP is free to spend more time in the office setting. However, this may lead to ­discontinuity of care due to multiple providers, or dissatisfaction in not being able to see one’s own doctor. Given the forces of healthcare today, many, if not most, hospital systems are at least in part relying on at least a voluntary hospitalist system of care, as described in stage three above. There are numerous hospitalist models of care in place today. In many ways, the models continually redefine themselves based on changes in hospitals themselves and changes in physician training. For example, recent restrictions in house staff work hours have necessitated that hospitals find alternative ways to cover patients. One type of model includes a private practice group employing a hospitalist to admit and care for their patients. A much more popular model is that in which a private practice group of inpatient physicians provide care to those patients admitted to the hospital. Typically, these hospitalist groups contract out to private practices or hospitals to care for their patients. These former models are popular with community facilities. Other models include those in which hospitals and health maintenance organizations hire their own inpatient physicians. Finally, many academic centers now have divisions or sections of Hospital Medicine. Academic hospitalists generally work in direct patient care less than private hospitalists, usually between 1 and 6 months per year. However, their time is usually supplemented by activities such as house staff training, academic research, and administrative duties.

Benefits of Hospitalist Systems As one might imagine, there are several areas of benefit inherent in having a dedicated physician caring for patients requiring hospitalization. This doctor is not limited or constrained by the problems imposed by having an office practice. As such, the hospitalist is available throughout the day or night to see patients immediately, meet with family members and loved ones, and respond to emergency ­situations. The hospitalist is also in a prime position to foster a culture of patient safety, primarily by participating in multidisciplinary teams.5 Additionally, by ­virtue of the fact that this doctor practices only in the hospital, over time he or she becomes more attuned to developing and maintaining the necessary skills to manage acute inpatient medical issues. Hospital medicine is a relatively young field, but the body of evidence in the literature showing benefits of this new system is growing rapidly. Published data demonstrates that utilizing hospitalists decreases total costs per case and patient’s length of stay6; preserve patient satisfaction despite not having direct PCP involvement in care6; trend towards a decrease in short-term mortality7; provide benefit in end-of-life care8; and improve resident education.9 There is also data to suggest that

50

D.J. McAdams

some of these changes, particularly length of stay and cost per case, are derived only when experienced hospitalists are present in a program or after a program has been established for some time.7 This is of growing concern since the recent explosion of popularity in hospital medicine has left many slots open for inexperienced hospitalists, and since some programs are designed to be transient in nature and are filled by recent graduates from residency. Nonetheless, benefits derived from ­hospitalist use are widely present, and certainly this concern will diminish over the next decade as the number of providers begins to equilibrate with the number of available employable positions. A large number of hospitalist literature and journals now exist. Interestingly, almost every journal devoted to hospital medicine has a section in each publication focusing on quality improvement or patient safety. Again, the fact that these physicians are working within the hospital most of the time affords them the unique ability to police the system, recognize areas of improper or inefficient care management, and formulate and carry out care plans that have been proven to enhance inpatient care.

Hospitalists as Acute Providers Compared to hospital models from decades ago, sicker patients are admitted to and stay in the hospital. No longer are the majority of patients simply staying in facilities awaiting tests. Most have serious, volatile problems, and clearly their conditions can change at any time. One can, therefore, make an argument that round-the-clock care of patients by an in-house physician is much more beneficial that traditional outside overnight call coverage.10 By virtue of focused training in hospital medicine as well as advanced cardiac life-support (ACLS) techniques, the hospitalist is in a prime position to care for the inpatient in urgent and emergent situations. In general, adverse events follow a gradual clinical patient deterioration, and much of the time the signs of impending doom go unrecognized or are even ignored.11 While there is not much direct data to suggest a link between hospitalists and early recognition of deterioration yet, there is some suggestion that the omnipresence of the hospitalist allows for more prompt recognition of acute problems with patients and implementation of appropriate and directed care to prevent adverse outcomes.7,9 The hospitalist can work alone in this venue, but more commonly he or she works on a multidisciplinary team as part of a Rapid Response System (RRS). The concept of a “code team” is not new, and certainly many facilities rely on intensivists and intensive care unit (ICU) teams to provide emergency care. The newer trend is an attempt to make these teams more focused, and more rapid to respond to less severe crises, and avoid delays in care for suddenly critically ill patients. This has mostly been done by having a smaller number of well-trained staff respond to the emergency in a controlled fashion, rather than a larger number of junior staff with sub-optimal skills or experience. While the RRS is composed of

5  Acute Hospitalist Medicine and the Rapid Response System

51

many parts, the response teams (MET or RRTs) can be comprised of any number and type of care provider experienced in resuscitation skills. METs generally involve a combination of hospitalists (non-intensivist or intensivist), ICU nurses, respiratory ­technicians, anesthesiologists, and even emergency room physicians, whereas an RRT does not have a physician as a first responder. In this situation, a hospitalist is a convenient and effective physician who can respond should the RRT need physician backup. The use of a hospitalist system does not preclude the need for an RRT or MET. Rather, the two usually go hand in hand. The two main types of hospitalists, nonintensivists and the intensivists, both can be a part of this response team. The nonintensivist (those inpatient physicians trained in a primary specialty but without critical care training) often do not have major instruction beyond basic life-support and ACLS, and in particular may lack complex airway management skills. Often this non-intensivist hospitalist is the first responder to an urgent or emergent situation, is involved in calling the RRT to the bedside, and can certainly be involved as an integral member of the MET, including the “code leader.” However, the more skills-based aspects of code management, again mainly related to difficult airway problems or other procedures, may be reserved for the intensivist. In general, the major benefit of the non-intensivist hospitalist to the system is being in a position (geographically and intellectually) to foster a more rapid and astute recognition of the clinical deterioration of patients, and then set in motion the necessary elements to trigger the rapid response system to send a team to the aide of the patient. There continue to be clinical trials examining the usefulness of RRTs and METs. At least two studies (both were non-randomized and non-blinded) have shown some benefit to having an RRS, namely a decreased incidence of unanticipated ICU transfers, lower incidence of death without a DNR, a decreased incidence of and mortality for in-hospital cardiac arrests, and a reduction in overall hospital mortality.12,13 There has been at least one meta-analysis evaluating the effects of an RRS on clinical outcomes, and this study failed to find any consistent improvement in outcomes through the use of an RRS in the 13 studies included.14 Nevertheless, there may be some advantage from an RRS; however the extent of the advantage remains to be demonstrated.

Thoughts for the Future It is evident that the wave of the future in hospital care will almost universally involve the hospitalist. Yet given the rapid nature with which this assumption is occurring, steps will need to be taken to ensure that hospitalists are prepared for the situations they will encounter on a daily basis. Instituting steps to improve retention (incentive programs, reasonable shifts and work hours, etc.) will likely improve performance and care delivery in programs that employ hospitalists. There may also be changes in residency training programs to allow candidates interested in a career in hospital medicine an opportunity to obtain

52

D.J. McAdams

more experience in the care of the inpatient and in managing inpatient emergencies. While more fellowships for hospital medicine may continue to emerge, it remains to be seen if completing a fellowship will ever be required for those wishing to pursue a hospital-based position. Ultimately, it may become necessary to define the specific types of training required for hospital medicine, and this will almost certainly evolve around management of acute scenarios. Steps need to be taken in order to fully train hospitalists to deal with emergent events, particularly non-intensivists. In academic centers, residents are being given less and less exposure to urgent or emergent events and procedures. Interestingly, they are getting more controlled experience in the lab setting but much less bedside emergency situation experience. And since there is no formal training in Hospital Medicine that is required, much of these duties are falling onto the shoulders of already busy and short-staffed intensivists. Focused training for hospitalists participating in multidisciplinary teams (METs) or aiding a nurse-led team when called (RRT), may prove to be extremely beneficial for every aspect of patient care. While the former will almost certainly allow for better management of quality issues and patient safety, the latter will literally be the basis for provision of care during acute hospital emergencies.

References 1. Wachter RM, Goldman L. The emerging role of “hospitalists” in the American health care system. N Engl J Med. 1996;335:514–517. 2. Society of Hospital Medicine (SHM). ; 2010. Accessed January 2010. 3. Lindenauer PK, Pantilat SZ, Katz PP, Wachter RM. Hospitalists and the practice of inpatient medicine: results of a survey of the National Association of Inpatient Physicians. Ann Intern Med. 1999;130(4 Pt 2):343–349. 4. Wachter RM. An introduction to the hospitalist model. Ann Intern Med. 1999;130:338–342. 5. Shojania KG, Wald H, Gross R. Understanding medical error and improving patient safety in the inpatient setting. Med Clin North Am. 2002;86(4):847–867. 6. Wachter RM, Goldman L. The hospitalist movement 5 years later. JAMA. 2002;287(4): 487–494. 7. Meltzer D, Manning WG, Morrison J, et  al. Effects of physician experience on costs and outcomes on an academic general medicine service: results of a trial of hospitalists. Ann Intern Med. 2002;137:866–875. 8. Auerbach AD, Pantilat SZ. End-of-life care in a voluntary hospitalist model: effects on communication, processes of care, and patient systems. Am J Med. 2004;116:669–675. 9. Kulaga ME, Charney P, O’Mahony SP, et al. The positive impact of initiation of hospitalist clinician educators: resource utilization and medical resident education. J Gen Intern Med. 2004;19:293–301. 10. SHM Benchmarks Committee. Value added by hospitalists: hospital medicine programs add value through extraordinary availability (24/7). Hospitalist. 2004;8:19–22. 11. Buist M, Bernard S, Anderson J. Epidemiology and prevention of unexpected hospital deaths. Surg J R Coll Surg Edinb Irel. 2003;1(5):265–268. 12. Bristow PJ, Hillman KM, Chey T, et al. Rates of in-hospital arrests, deaths and intensive care admissions: the effect of a medical emergency team. Med J Aust. 2000;173:236–240.

5  Acute Hospitalist Medicine and the Rapid Response System

53

13. Buist MD, Moore GE, Bernard SA, Waxman BP, Anderson JN, Nguyen TV. Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. BMJ. 2002;324:1–6. 14. Ranji SR, Auerbach AD, Hurd CJ, O’Rourke K, Shojania KG. Effects of rapid response ­systems on clinical outcomes: systematic review and meta-analysis. J Hosp Med. 2007;2(6): 422–432.

Chapter 6

Medical Trainees and Patient Safety Stephen W. Lam and Arthas Flabouris

Keywords  Medical • Trainees • Patient • Safety

Healthcare, Healthcare Facilities and Medical Trainees As the next generation of medical practitioners, medical trainees form an important part of the medical profession. Even though they are trainees, they are a vital healthcare resource, contributing to patient care, research and academia. Their ability to contribute is profoundly influenced by their prior undergraduate academic education and supervised clinical experience. Progression through the postgraduate years is associated with a gradual reduction in clinical supervision as expertise is accumulated. Assessment is typically both formative and summative and continues until the trainee is considered safe to practice without further supervision and gains the relevant qualifications to do so. Postgraduate clinical training generally occurs in large healthcare facilities. Patient profiles and illness types (or “case mix”) determine the healthcare provision that is required from medical trainees. In turn, the learning environment for trainees is determined by the case mix to which they are exposed, along with other factors such as level and quality of supervision, resources and working conditions. Thus there is a “shared dependence,” where patients depend on trainees to provide safe and competent care, while the trainees rely on patients to provide a suitable learning environment from which to gain quality training and experience.

S.W. Lam (*) University of Adelaide, Intensive Care Unit, Royal Adelaide Hospital, North Tce, Adelaide, SA 5001, Australia e-mail: [email protected]

M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_6, © Springer Science+Business Media, LLC 2011

55

56

S.W. Lam and A. Flabouris

The Healthcare Environment Delivering care in a hospital setting has become increasingly complex and risky.1 The population is aging2,3 with increasing co-morbidities and changing disease demographics.4–6 Healthcare technology is becoming more complex7 and patients often present with chronic and sometimes terminal conditions. Meanwhile, hospitals seek to achieve cost efficiency through reducing acute hospital beds, streamlining inpatient care, staff reductions, and a greater emphasis on home care. Within that setting, education is less of a “core” activity and viewed as more of an “add-on.”8 The changing needs of patients is also reflected by an increase in demand for a new specialist in hospital medicine, or “hospitalist,” capable of providing competent institutional care in a team environment and handling both acute medical events9 and palliative care issues.7,10 The rapid growth in technological and scientific advances resulting from better understanding of complex disease processes has fueled the growth of medical specialization. The number of American Medical Association accredited specialties and subspecialties increased from 14 in 1927 to 41 by 1985, after which growth was exponential, reaching 124 by the year 2000.11 Highly technical “proceduralists” and specialists are now limiting their practices to specific diseases, or organs; treat certain populations (e.g., pediatricians); or operate in specific geographical areas (e.g., emergency departments). Because of the associated technical complexity and cost of such procedures, many of these services are restricted to academic and acute care medical facilities. As a result, there has been a decline in the number of medical practitioners devoted to comprehensive and whole individual care. Medical specialization has been criticized as being fragmented12 and confusing to patients and general practitioners alike, with risk to the perception of medicine as an ­integrated profession.13

Medical Trainees: The Undergraduate Years The quality of a trainee’s learning depends on what they already know.14 When exposed to clinical practice, they attempt to make sense of new experiences by using their existing knowledge. Thus, the primary role of medical schools is the education of medical students – preparing them with the necessary knowledge and skills for structured and supervised practice in acute care facilities. Increasingly, this role has had to compete with research and other non-teaching activities. In the 1990s, the medical curriculum was criticized for being too rigid, with overuse of didactic teaching methods and rote memorization.15 The emphasis of undergraduate training and examination has shifted from the didactic acquisition of academic knowledge towards a focus on patient-oriented knowledge and problem-based learning.7,16,17 The adoption of patient-based learning methods has been undertaken with a view to improving the link between undergraduate training and postgraduate provision of patient care.7,16–18 It also allows undergraduate training to evolve with

6  Medical Trainees and Patient Safety

57

changing patient needs on the wards and the development of teamwork and leadership skills through group learning. Recognition of the importance of practical skills assessment has lead to the use of such examination techniques as the Observed Structured Clinical Examination (OSCE). For the assessment of competency in critical care skills, a teaching methodology that incorporates structured clinical and multidisciplinary, problem-based instruction19 with that of OSCE and/or computer simulation-based assessment have been shown to be effective.20 While clinically oriented, these teaching methods employ simulated scenarios with minimal direct provision of patient care; hence, there is little opportunity for compromising patient safety.

Medical Trainees and Patient Safety: The First Few Years Medical issues requiring provision of care can be separated into two types: • Those associated with an already identified illness (e.g., provision of elective surgical procedure or drug treatment for a known problem) • Those associated with unanticipated acute medical problems or complications (e.g., undiagnosed illnesses, idiosyncratic drug reactions, iatrogenic complications arising from treatment and nosocomial infections acquired from the healthcare environment)

Provision of Care for Identified Illnesses The supervision of postgraduate trainees in their provision of patient care should be adjusted according to the trainees’ level of experience and assessed competency. Predictable illness clinical pathways can be used to oversee clinical performance, but the value of senior clinical oversight should not be ignored. 21 Such oversight can be useful in detecting missed diagnoses as well as providing educational feedback upon performance.

Provision of Care for Medical Incidences Because of the frequent and routine nature of many aspects of general ward care, such as prescribing intravenous fluids, many healthcare institutions utilize the most junior members of the healthcare team as the first point of contact to address routine issues. In dealing with these, most junior trainees remain unsupervised and receive little feedback22 unless an adverse event occurs. This is despite the fact that

58

S.W. Lam and A. Flabouris

supervision has been recognized as the “single most important element upon which graduate educational training is dependent.”23,24 The high frequency of minor medical issues arising in general wards has also created the need for 24-h “on-call” medical officers to deal with them. On such shifts, medical trainees are often given a wide range of smaller, less focused tasks. Several tasks for multiple patients may be allocated to the same individual simultaneously, often from different areas of the hospital. The priority of such tasks may vary for reasons ranging from the level of perceived urgency of the patient’s clinical condition to the need to meet required timeframes and deadlines (e.g., awaiting transfer to the operating room). As such, medical trainees are often faced with the need to triage priorities and handle important tasks with multiple distracting issues under significant time pressures. “On-call” shifts are typically long and extend outside normal working hours. The mode of presentation of medical emergencies is frequently subtle or nonspecific; and their early recognition and correct management is crucial to patient safety and outcome.9,21,25–28 Subtle indicators of a more severe underlying process can be easily overlooked among the burden of routine tasks. A recent international survey of patients who suffered cardiac arrests, death, or unanticipated intensive care unit admissions in hospitals revealed that significant physiological abnormal findings were present in many patients prior to those events.29 For some patients in this study, medical staff had reviewed them, thus highlighting the possible preventability of such adverse events.29 Other studies have documented patients with abnormal and/or inadequately attended clinical findings who subsequently experienced potentially preventable adverse events.28,30 Inappropriate working conditions can also hinder a trainee’s ability to correctly identify and separate warning signs of impending disaster from more minor complaints and issues, and provide appropriate care in a timely manner where required.31–35 Postgraduate medical education in acute care facilities is tailored toward clinical expertise in select medical domains through a specialized, structured curriculum, which is not consistent with a more whole-patient approach. Medical trainees undertaking specialist training therefore often lack sufficient skills to meet patient needs,17,36–38 particularly with respect to the recognition and management of medical incidences, which may present initially in a subtle manner. Training in the basic aspects of recognizing and caring for patients with critical illness is often lacking, not only in the undergraduate years39,40 but often in the postgraduate period as well,41,42 and may remain in a poor state after completion of the chosen medical specialty and among specialty supervisors.43

Improving Patient Safety in Institutions with Medical Trainees Responses to acute medical emergencies should be immediate, organized, predetermined and involve a team of appropriately trained and resourced clinical staff. A good example is a trauma team response.44,45 The organization of trauma management has

6  Medical Trainees and Patient Safety

59

resulted in a significant reduction in preventable deaths.46,47 However, for inpatients with acute medical emergencies, the most junior doctors are usually left to recognize and manage such emergencies on their own. Not only may they lack the required critical care skills, but they also lack the crucial skills of being able to communicate, coordinate, and organize a team response. Often monitoring and procedural equipment are not available and senior assistance may be remote or not provided in a timely fashion. This is especially so in acute care facilities, where response to acute ward medical emergencies may be limited to a team that responds only to cardiac arrests. As demonstrated in this book, a team that responds to acute medical emergencies other than for in-hospital cardiac arrests is a concept that is becoming increasingly popular.48–51 Ideally, medical trainees should be trained in basic and advanced resuscitation skills, no matter what their primary specialty training. However, the expectation that all such trainees would be able to regularly perform or practice those skills is hard to sustain. It is more important that trainees be instructed in the early recognition of at-risk patients,9,21,25–28 and then perhaps integrate themselves within a formal hospital medical emergency response team.

Postgraduate Training and Specialization Complicated patients need practitioners who are able to manage undifferentiated illness, often along with multiple specialist teams and/or other generalists. Focus on one area of practice with specialist training invariably leads to a lack of knowledge and experience in other areas. The multi-factorial causes of adverse events in the Quality in Australian Health Care Study showed that technical competency, problem-solving ability, communication, performance, and system design all contributed to the quality of medical care and thus should be considered integral components of the postgraduate education curriculum.52 Thus, in this age of increasing specialization it is crucial that postgraduate training maintains a more balanced approach to acute care. At the very least, medical trainees should be taught to recognize warning signs that may herald a greater emergency and be empowered to refer as appropriate or seek other critical care involvement during times of medical crises. Training time spent within a critical care-type environment for all sub-specialization postgraduate training programs may provides trainees with a greater knowledge of acute illness.53 Support for the concept of the hospitalist has grown as a result of issues of patient safety and a drive for lower inpatient costs.54,55 The rise of the hospitalist mirrors that of intensive care and emergency medicine, but which is occurring in the general wards. Hospitalists have become the preferred providers of postgraduate medical education among medical trainees in some countries.56 Training and retention of basic and advanced life-support skills requires a multidisciplinary, coordinated, and integrated team approach. Such training is best

60

S.W. Lam and A. Flabouris

served through clinical exposure within dedicated critical care units, simulation technology and skills laboratories.57 Training that also involves instruction in triage, emergency planning and preparation, team leadership, teamwork and team organization during emergency response should be included. Consideration for such training should begin in the undergraduate years. A system of regular accreditation is an essential component. Practitioners who have been disciplined by state medical-licensing boards are three times as likely to have displayed unprofessional behavior while in medical school.58 It is increasingly recognized that the issue of “professionalism” should be introduced to undergraduate education.59,60

Summary With changing hospital patient demographics and rapidly advancing healthcare technology, it is becoming increasingly important for healthcare systems to evolve to meet their new challenges. Medical trainees, as a vital healthcare resource, provide both elective and emergency medical care within acute healthcare facilities. Postgraduate training and medical team structure often place junior trainees at the forefront of identifying and responding to the deteriorating patient. This requires them to deal with issues ranging from the trivial to the more complicated and often subtle presentations of acute medical emergencies. Their ability to recognize these signs and trigger an appropriate response is crucial to minimizing serious adverse events for such patients. For medical trainees to safely and efficiently fulfill their roles in emergent and elective patient care, undergraduate and postgraduate training will need to provide them with the appropriate skills, environment, balance between specialization and general medicine and appropriate supervision.

References 1. Kohn LT, Corrigan JM, Donaldson MS, Molla S. To Err Is Human: Building a Safer Health System. Washington, DC: National Academy Press; 2000. 2. Population Division, Department of Economic and Social Affairs, United Nations Secretariat. World population prospects: the 2002 revision. Highlights. New York: United Nations. ; 2003 Accessed 27.04.09. 3. World Health Organization. The World Health Report 1998: Life in the 21st Century – A Vision for All. Geneva: WHO; 1998. 4. van Weel C. Chronic diseases in general practice: the longitudinal dimension. Eur J Gen Pract. 1996;2:17–21. 5. van Weel C, Michels J. Dying, not old age, to blame for costs of health care. Lancet. 1997;350(9085):1159–1160.

6  Medical Trainees and Patient Safety

61

6. Resnick NM, Marcantonio ER. How should clinical care of the aged differ? Lancet. 1997;350(9085):1157–1158. 7. Chantler C. The role and education of doctors in the delivery of health care. Lancet. 1999;353(9159):1178–1181. 8. Dent AW, Crotty B, Cuddihy HL, et al. Learning opportunities for Australian prevocational hospital doctors: exposure, perceived quality and desired methods of learning. Med J Aust. 2006;184(9):436–440. 9. McQuillan P, Pilkington S, Allan A, et  al. Confidential inquiry into quality of care before admission to intensive care. BMJ. 1998;316(7148):1853–1858. 10. McCahill LE, Dunn GP, Mosenthal AC, Milch RA, Krouse RS. Palliation as a core surgical principle: Part 1. J Am Coll Surg. 2004;1999(1):149–160. 11. Donini-Lenhoff FG, Hedrick HL. Growth of specialization in graduate medical education. JAMA. 2000;284(10):1284–1289. 12. Grumbach K. Primary care in the United States – the best of times, the worst of times. N Engl J Med. 1999;341(36):2008–2010. 13. Martini CJM. Graduate medical education in the changing environment of medicine. JAMA. 1992;268(9):1097–1105. 14. Dyrbye LN, Harris I, Rohren CH. Early clinical experiences from students’ perspectives: a qualitative study. Acad Med. 2007;82(10):979–998. 15. Christakis NA. The similarity and frequency of proposals to reform U.S. medical education: constant concerns. JAMA. 1995;274(9):706–711. 16. Howe A, Campion P, Searle J, Smith H. New perspectives – approaches to medical education at four new U.K. medical schools. BMJ. 2004;329(7461):327–331. 17. Jones R, Higgs R, Angelis C, Prideaux D. Changing face of medical curricula. Lancet. 2001;357(9257):699–704. 18. Dornan T, Bundy C. What can experience add to early medical education? Consensus survey. BMJ. 2004;329(7470):834–840. 19. Hill D, Stalley P, Pennington D, Besser M, McCarthy W. Competency-based learning in ­traumatology. Am J Surg. 1997;173(2):136–140. 20. Rogers PL, Jacob H, Rashwan AS, Pinsky MR. Quantifying learning in medical students ­during a critical care medicine elective: a comparison of three evaluation instruments. Crit Care Med. 2001;29(6):1268–1273. 21. Reilly BM. Physical examination in the care of medical inpatients: an observational study. Lancet. 2003;362(9390):1100–1105. 22. Lack CS, Cartmill JA. Working with registrars: a qualitative study of interns’ perceptions and experiences. Med J Aust. 2005;182(2):70–72. 23. Australian Curriculum Framework for Junior Doctors. . Accessed 30.04.09. 24. Iglehart JK. Revisiting duty-hour limits – IOM recommendations for patient safety and ­resident education. N Engl J Med. 2008;359(25):2633–2635. 25. Hillman K, Bristow PJ, Chey T, et  al. Antecedents to hospital deaths. Intern Med J. 2001;31:343–348. 26. Franklin C, Mathew J. Developing strategies to prevent inhospital cardiac arrest. Analyzing responses of physicians and nurses in the hours before the event. Crit Care Med. 1994;22(2):244–247. 27. Bedell SE, Deitz DC, Leeman D, Delbanco TL. Incidence and characteristics of preventable iatrogenic cardiac arrest. JAMA. 1991;265(21):2815–2820. 28. Schein RMH, Hazday N, Pena M, Ruben BH, Sprung CL. Clinical antecedents to in-hospital cardiopulmonary arrest. Chest. 1990;98(6):1388–1392. 29. Kause J, Smith G, Prytherch D, Parr M, Flabouris A, and for the Intensive Care Society (UK), and Australian and New Zealand Intensive Care Society Clinical Trials Group ACADEMIA Study Investigators. A comparison of antecedents to cardiac arrests, deaths, and emergency intensive care admissions in Australia and New Zealand, and the United Kingdom – the ACADEMIA study. Resuscitation. 2004;62(3):275–282.

62

S.W. Lam and A. Flabouris

30. Garrad C, Young D. Suboptimal care of patients before admission to intensive care. BMJ. 1998;316(7148):1841–1842. 31. Homes G. Junior doctors’ working hours: an unhealthy tradition? [editorial]. Med J Aust. 1998;168(12):587–588. 32. Olson LG, Ambrogetti A. Working harder – working dangerously? Fatigue and performance in hospitals. Med J Aust. 1998;168(12):614–616. 33. Sexton JB, Thomas EJ, Helmreich RL. Error, stress, and teamwork in medicine and aviation: cross-sectional surveys. BMJ. 2000;320(7237):745–749. 34. Lockley SW, Cronin JW, Evans EE, et al. Effect of reducing interns’ weekly work hours on sleep and attentional failures. N Engl J Med. 2004;351(18):1829–1837. 35. Landrigan CP, Rothschild JM, Cronin JW, et  al. Effect of reducing interns’ work hours on serious medical errors in intensive care units. N Engl J Med. 2004;351(18):1838–1848. 36. Smith GB, Poplett N. Knowledge of aspects of acute care in trainee doctors. Postgrad Med J. 2002;78(920):335–338. 37. Meek T. New house officers’ knowledge of resuscitation, fluid balance and analgesia. Anaesthesia. 2000;55:1128. 38. Goldacre MJ, Lambert T, Evans J, Turner G. Preregistration house officers’ views on whether their experience at medical school prepared them well for their jobs: national questionnaire survey. BMJ. 2003;326(7397):1019–1022. 39. Harrison GA, Hillman KM, Fulde GWO, Jacques TC. The need for undergraduate education in critical care (results of a questionnaire to year 6 medical undergraduates, University of New South Wales and recommendations on a curriculum in critical care.). Anaesth Intensive Care. 1999;27(1):53–58. 40. Buchman TG, Dellinger RP, Raphaely RC, Todres ID. Undergraduate education in critical care medicine. Crit Care Med. 1992;20(11):1588–1603. 41. Gillard JH, Dent TH, Jolly BC, Wallis DA, Hicks BH. CPR and the RCP. Training of students and doctors in UK medical schools. J R Coll Physicians Lond. 1993;27:412–417. 42. Redmond AD. Training in resuscitation. Arch Emerg Med. 1987;4:205–206. 43. Thwaites BC, Shankar S, Niblett D, Saunders J. Can consultants resuscitate? J R Coll Physicians Lond. 1992;26:265–267. 44. West JG, Williams MJ, Trunkey DD, Wolferth CC. Trauma systems. Current status – future challenges. JAMA. 1988;259(24):3597–3600. 45. Pagliarello G, Dempster A, Wesson D. The integrated trauma program: a model for cooperative trauma triage. J Trauma. 1992;33(2):198–204. 46. Shackford SR, Hollingworth-Fridlund P, Cooper GF, Eastman AB. The effect of regionalization upon the quality of trauma care as assessed by concurrent audit before and after institution of a trauma system: a preliminary report. J Trauma. 1986;26(9):812–820. 4 7. Draaisma JM, de Haan AF, Goris RJ. Preventable trauma deaths in the Netherlands – a ­prospective multicenter study. J Trauma. 1989;29(11):1552–1557. 48. Lee A, Bishop G, Hillman KM, Daffurn K. The medical emergency team. Anaesth Intensive Care. 1995;23:183–186. 49. Stenhouse C, Coates S, Tivey M, Allsop P, Parker T. Prospective evaluation of a Modified Early Warning Score to aid detection of patients developing critical illness on a surgical ward. Br J Anaesth. 2000;179(6):663P. 50. Kerridge RK, Saul WP. The medical emergency team, evidence-based medicine and ethics. Med J Aust. 2003;179(6):313–315. 51. Goldhill OR, Worthing L, Mulcahy A, Tarling M, Sumner A. The patient-at-risk team: ­identifying and managing seriously ill ward patients. Anaesthesia. 1999;54:853–860. 52. Wilson RM, Runciman WB, Gibberd RW, et al. The quality in Australian health care study. Med J Aust. 1995;163:458–471. 53. Zhu JN, Weiland TJ, Taylor DM, Dent AW. An observational study of emergency department intern activities. Med J Aust. 2008;188(9):514–519. 54. Wachter RM, Goldman L. The emerging role of hospitalists in the American health care system. N Engl J Med. 1996;335(7):514–517.

6  Medical Trainees and Patient Safety

63

55. Wachter RM. Hospitalists in the United States – mission accomplished or work in progress? N Engl J Med. 2004;350(19):1935–1936. 56. Hauer KE, Wachter RM, McCulloch CE, Woo GA, Auerbach AD. Effects of hospitalist attending physicians on trainee satisfaction with teaching and with internal medicine rotations. Arch Intern Med. 2004;164(16):1866–1871. 57. Riley RH, Grauze AM, Chinnery C, Horley RA, Trewhella NH. Three years of “CASMS:” the world’s busiest medical simulation centre. Med J Aust. 2003;179(11/12):626–630. 58. Papadakis MA, Teherani A, Banach MA, et  al. Disciplinary action by medical boards and prior behavior in medical school. N Engl J Med. 2005;353(25):2673–2682. 59. Association of American Medical Colleges. Learning objectives for medical student education – guidelines for medical schools: Report I of the Medical School Objectives Project. Acad Med. 1999;74(1):13–18. 60. Cruess SR, Cruess RL. Professionalism must be taught. BMJ. 1997;315(7123):1674–1677.

Chapter 7

Rapid Response Systems: A Review of the Evidence Bradford D. Winters and Julius C. Pham

Keywords  Rapid • Response • Systems • Review • Evidence

Introduction Long before the development of the Rapid Response System (RRS) concept, nurses and physicians have been aware that patients rarely deteriorate suddenly. Once practitioners started to ask questions such as “What is happening in the minutes, hours and even days before a patient deteriorates to a cardiorespiratory arrest?” or “Is there something we can do to intervene?” the steps towards creating a new patient safety and quality initiative began. Several studies sought to identify the antecedents to these events and define what might be the most ­predictive signs and symptoms to watch for. The results of these inquires were then linked to an intervention where a team can be summoned based on these changes in a patient’s condition to attempt to halt the progression to a more severe situation or even an arrest. This idea has matured from the early Medical Emergency Teams (METs) piloted in Australia and the US to more comprehensive and diverse systems that include Rapid Response Teams (RRTs), Critical Care Outreach Teams (CCOTs) and other strategies to bring additional resources to the bedside of a patient developing a critical illness. While this strategy makes intuitive sense and has strong face validity, is it supported by the evidence? Do RRSs improve patient safety and quality of care and are they cost-effective? In this chapter we will review the published literature for RRSs to try and address these questions.

B.D. Winters (*) Departments of Anesthesiology, Critical Care Medicine and Surgery, The Johns Hopkins University School of Medicine, Meyer 297 600N Wolfe Street, Baltimore, MD 21298, USA e-mail: [email protected]

M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_7, © Springer Science+Business Media, LLC 2011

65

66

B.D. Winters and J.C. Pham

Evaluating the Evidence Evidence may come from a wide range of sources and vary significantly in quality and quantity. When we weigh the evidence for an intervention, we must be sure to consider these factors. While a thorough review of evidence-based medicine is out of the scope of this chapter, we will address some guiding principles. First, we should always seek the highest possible quality of evidence when designing studies and reviewing the data to inform our decision‑making. Evidence may be categorized based on level of quality (see Table 7.1).1 At the very top (Level 1) are systematic reviews and meta-analyses of randomized controlled ­trials (RCTs) that exhibit a high degree of homogeneity. These are followed by individual RCTs with narrow confidence intervals. Level 2 studies include historical or concurrent cohort trials. Level 3 includes case-controlled studies. Level 4 evidence consists primarily of case series and case studies and Level 5 data is based on expert opinion, physiological principles, bench‑top basic science data or experience. Level 1 trials are able to establish “cause and effect” while studies that do not actively control the intervention in their design can only establish an association. Level 1 evidence should always be the goal; however, it may not always be available to answer a particular question or may be impractical to carry out in a research setting. There are many situations where the “best” attainable evidence may only be Level 2 quality or lower, leaving us to rely on this best available evidence to guide our decisions. While we should not be dogmatic on insisting on Level 1 or even Level 2 data when it is not practical or possible, we must be very critical in our appraisals, especially if the intervention carries great cost or risk. How to balance quantity of evidence with quality is more problematic. Do numerous lower‑quality studies trump a single high‑quality study with conflicting results? How many lower‑quality studies does it take? What if the aggregate number of patients in the lower quality study(s) is vastly larger than the higher quality study(s)? Finally, how do we reconcile or own biases towards the results? Clearly, the answers are difficult, especially when study quality varies significantly. It is no wonder that different reviewers often come to very different conclusions. These issues will become important as we review the RRS literature. Table 7.1  Levels of evidence Evidence level Types of included studies Level 1a Meta–analysis and systematic reviews of RCTS Level 1b RCTs (single and multiple institution) Level 2a Cluster randomized trials, concurrent cohort controlled trials, step–wise trials Level 2b Historically (before/after) controlled trials Level 3 Case controlled studies Level 4 Case studies, case series without controls Level 5 Expert opinion, experience, pathophysiological reasoning, basic science data

7  Rapid Response Systems: A Review of the Evidence

67

Identifying the Deteriorating Patient, the RRS Afferent Limb Several groups became interested in the time period leading up to a cardiorespiratory arrest in the late 1980s and the early 1990s. While these studies do not address the question of whether RRSs are effective, they lay the foundation for why we have developed RRSs and deserve some consideration, since we still are asking them two decades later: “How do we best predict who is experiencing a serious deterioration and is in need of ‘rescue’?” Early chart reviews of patients who had experienced a cardiorespiratory arrest showed clear physiologic abnormalities in variables such as heart rate, respiratory rate and mental status and were also found to be present in patients who went on to arrest.2–19 These signs and symptoms were often present for many hours, if not days,3, 6–8,12 and often went unrecognized even when health care professionals visited the patients.7 Most agreed that tachycardia, bradycardia, tachypnea, oxygen desaturation, low systolic blood pressure, respiratory distress and mental status changes were the clearest signs. While difficult to quantify, concern or worry on the part of nursing staff was also considered predictive of progression to arrest. These retrospective results led to the widespread use of these physiological limits and “worry or concern” being used as “alert” or “activation” criteria for RRSs. This has been recently validated as effective.18 The sensitivity and specificity of physiological limits for preventing arrest is unclear. Several authors have tried to determine if better use can be made of these discretely measured physiological triggers. Some have tried to assign numerical values to degrees of abnormality, which is then summed to create a composite score. Many different “alert” or “activation” systems have been reported; some have been reported as successful,20–34 but others have poor inter‑rater reliability on assigning and calculating these scores.35 This has been noted in other scoring systems.36 Even individual physiological triggers used in RRSs have inter‑rater variability. On average though, simpler scores have been found to work better but not necessarily better than specific vital signs35,37 Bell et al.38 found similar results when comparing restricted to extended criteria. Extended criteria were found to have low specificity and would likely lead to numerous false alarms that could easily overburden an RRS. These authors, however, cautioned that restrictive criteria carried the risk of higher rates of failure to rescue. Others39–41 have also noted this tendency for low sensitivity and low positive predictive value, but one39 did find high specificity for a combination of heart rate, respiratory rate, systolic blood pressure and change in mental status. Attempts to modify the cutoffs for these triggers did not improve the results. A second consideration is the frequency with which patients are monitored. Even if physiological limits are predictive of deterioration, patients in general wards are infrequently monitored, allowing for significant deteriorations to occur before they may even have the chance to be recognized. Buist et al noted that the median time for the occurrence of vital sign changes in general ward patients before their arrest was about 6‑7 h.7 If vital signs are only taken every 8 h, which is common in hospitals in the US, there are significant opportunities to miss a serious problem.

68

B.D. Winters and J.C. Pham

As noted before, this is compounded by the frequent lack of realization that the signs are worrisome. If intermittently collected physiologic vital sign criteria (individual or aggregated into a score) may be less than adequate to optimize opportunities for intervention, can we develop electronic continuous systems, like those used in the ICU, and successfully apply them to general ward patients? Several barriers exist, including the need for patient mobility, high‑fidelity uninterrupted monitoring for these mobile patients and, most especially, an acceptable signal to noise ratio so that all patients’ deteriorations are recognized with minimal false alarms. Many technologies are rapidly developing to address the need for wireless high fidelity monitors. Controlling the signal noise problem is much more difficult. Attempts to address this problem using algorithms to analyze the data streams42–44 and create more reliable “indexed” values44 are in their early stages but offer potential. Improving how we monitor patients on general wards is a crucial element if we are to realize the full potential of RRSs.

The Efferent Limb: The Responding Team Unlike the afferent limb, there is less evidence evaluating the team that responds to a deteriorating patient. This efferent limb may be staffed in myriad ways, depending on the human resources available. Many programs use METs, which include physicians as well as nurses, respiratory therapists and others on the team, while RRTs are nurse‑led teams without physician staffing. While most of these clinicians are based out of the ICU or are critical care‑trained, other staffing models include hospitalists using the code team in a dual role, responding to both45 or even using the primary service as the responders although their response must be immediate when activation criteria are met. To date, there are no head‑to‑head comparisons of these different models of team staffing and there is no evidence to support the superiority of one model over another. In absence of clear data for guidance, choosing a staffing model for an RRS still depends on local resources and culture. Understanding and improving team function has become the subject of interest in a variety of fields, including RRSs. Jones et al have suggested that commonly encountered RRS situations, such as respiratory distress, deserve standardization and protocols, which alone has been shown to improve the quality of care.46 Simulation is a powerful tool for understanding how teams function, identifying opportunities for improvement and for developing and testing protocols for care in urgent/emergent situations such as those frequently encountered by RRS teams. DeVita et al at the University of Pittsburgh have been particularly active in using this strategy to assess RRS team performance. They have shown that team performance can be significantly enhanced through simulator education and training.47 Wallin et al have found similar results especially in the effect on skills. However, while they found improved performance, they also found that this type of education did not help improve attitudes towards “safe” teamwork.48 Education, in general, has been frequently cited as being essential to RRS success.46, 49, 50 Although performance

7  Rapid Response Systems: A Review of the Evidence

69

is essential, there is also much value in cultivating the “culture of safety.” However, to expect that simulation education alone can change culture is shortsighted. Changing culture and attitudes requires a multi‑pronged approach, including simulation.

The Rapid Response System: Is It Effective? The effectiveness of RRSs has been examined using a variety of outcome measurements including in‑hospital mortality, unanticipated ICU admission, incidence of cardiorespiratory arrest, ICU mortality, and hospital, and ICU length of stay (LOS). Recently, research has also evaluated the impact of the RRS on the institution of “not for resuscitation” (NFR) status and the timely goal‑directed resuscitation and administration of broad spectrum antibiotics in patients with sepsis. Outcome measures, like mortality, often are more attractive methodologically, and are often preferably sought by hospital administrators and regulatory agencies. Process measures are often more difficult to get a handle on but are potentially even more important. Rapid response systems seem to have benefit, especially for the outcomes in hospital mortality and cardiorespiratory arrest. However, the data is not homogeneous. Currently only two studies using a randomized methodology exist51,52 and only one examined more than one outcome (the MERIT study). This large, multi‑center study51 found no change in unexpected deaths, cardiac arrests or unplanned admissions to the intensive care unit (ICU). A smaller single institution study also found a significant improvement in mortality.52 Both studies were cluster randomized. Priestley’s study52 also performed a second historical control methodology that corroborated their results; and while not equivalent to blind randomization, such a study would be nearly impossible to achieve with an intervention like RRSs. Even cluster‑randomized methodologies are difficult and controlling, for the Hawthorne effect is extremely problematic. In fact one could argue that a Hawthorne‑like effect is needed to successfully implement a new system as opposed to introducing a new drug. This was one of the problems cited as a possible reason for the negative results of the MERIT study, which showed that many of the calls in the control hospitals demonstrated MET‑type activity rather than cardiac arrest calls.51 All hospitals in the study could not be blinded and were aware that they were being measured. This explanation is underscored by the fact that both the control and the RRS hospitals all significantly improved their mortality and arrest rates as compared to their baseline pre‑study values by roughly the same amounts negating any measurable differences between groups. Additionally, a very recent in‑depth re‑examination of the MERIT study data found that 50% of the time, when the standard cardiac arrest team was called in the “no‑MET” control hospitals, the call was for a patient not experiencing a cardiac or respiratory arrest. This suggests that control hospitals were, in fact, engaging in MET activities but through their cardiac arrest teams. Additionally, under‑utilization of the Medical Emergency Teams in the MET hospitals even when clear activation

70

B.D. Winters and J.C. Pham

criteria were present in patients and a wide variability across the individual hospitals in the number of MET calls and outcomes affects the ability to interpret the original study’s data and draw conclusions. Given this information, Chen et al approached the results from a different perspective, analyzing the data to determine if early intervention made a difference whether it was by a Medical Emergency Team or a cardiac arrest team. They found that early intervention (by any team, control or MET hospital) did result in a significant reduction in all cardiac arrests as well as unexpected cardiac arrests. This significant result was also found to exist for unexpected hospital deaths.53 This suggests that despite the original MERIT study results, a more in-depth look at the data demonstrated positive results that support the RRS and MET concept of early recognition and response to the patient’s bedside. Given these considerations, the MERIT study’s original data, when aggregated with the Priestley study, the point estimate, still, is favorable (0.76) but the confidence interval is wide (0.39–1.48), making it non‑significant as noted by both the systematic review of Winters et al54 and the Cochrane Systematic Review of randomized trials.55 The systematic review of Ranji et al treated the Priestley study separately as a time‑interrupted trial and did not aggregate the data.56 Unfortunately, the data re‑analysis by Chen et al cannot be subjected to meta‑analysis in combination with the Priestley data since the numerator and denominators are defined differently as this would likely lead to an even more favorable result. All of the other studies that have examined mortality and incidence of arrest in adults are nonrandomized but the results are more favorable and consistent. While data quality prevents including most studies in an aggregate analysis, meta‑analysis of the studies providing adequate numerator and denominator data57–68 shows a significant reduction in cardio‑respiratory arrest [OR = 0.625 (CI 95% = 0.502–0.777)] and hospital mortality [OR = 0.886 (CI 95% = 0.771–0.994)] (see Figs. 7.1 and 7.2). These represent 37.5% and 11.4% reductions, respectively. Most of the other non‑randomized studies, including those done in children, also show significant reductions in these outcomes, especially the incidence of arrest.69–73 These benefits also seem to be sustained over time. Buist et al have reported data over a 6‑year timeframe that shows sustained reductions in cardio‑respiratory arrest.74 This result is echoed by Jones et al, who found a sustained reduction in the incidence of cardio‑respiratory arrest over a 4‑year period. Additionally, over the 4  years, surgical patients had reductions in hospital mortality with the RRS program, with two of those years being statistically significant although this did not occur with medical patients. It is unclear why medical patients did not receive the same mortality benefit.59,60 Perhaps their problems are different from surgical patients and RRSs are more effective at intervening with the deteriorations that kill surgical patients. It is also interesting that the effect on arrest rates seems more powerful than the effect on mortality here and in other studies. This, along with the differential effect seen by Jones’s group, may suggest that while RRSs can prevent arrests, their ability to prevent eventual death is more limited, especially in certain patient groups. Many patients admitted to hospital have very severe or terminal illnesses and are unlikely to survive their hospital stay. Rapid response systems would not be expected to affect this, but trying to control it is nearly impossible.

7  Rapid Response Systems: A Review of the Evidence

71

Fig. 7.1  Aggregate analysis of effect on the incidence of cardio–respiratory arrest: nonrandomized studies

Bristow Buist Bellomo-03 Kenward Bellomo-04 Bellomo-07 Garcea Dacey Jolley Chan Combined .4

1

RR of death

2

Fig. 7.2  Aggregate analysis of RRS effect on in–hospital mortality: nonrandomized studies

Jones et al also examined the effect an RRS has on the long‑term mortality of the individual surgical patients finding a benefit extending well beyond hospital discharge.75 Again, early effective interventions in the types of problems surgical patients tend to have may explain this extended benefit as well as the in‑hospital benefit.

72

B.D. Winters and J.C. Pham

The effect of RRSs on another commonly examined outcome, unanticipated ICU admission, is quite variable. The MERIT study is the only randomized trial to evaluate for this and found no significant change.51 Two more recent non‑randomized studies found decreases in ICU admissions,67,76 while two found no change.68,77 Earlier studies58, 63 found statistically significant reductions but one61 found a statistically significant increase. This variability is not surprising, given the effect that individual hospital systems, resources and culture have on which patients get admitted to the ICU and under what circumstance. Changes in unanticipated ICU admissions should not be viewed as good versus bad but rather in the context of the institution’s culture. When admission rates go up, but mortality and arrest rates go down, it is inappropriate to consider this a failure. Two studies have examined RRSs in combination with other patient safety initiatives and reported favorable results.78,79 Both studies found reductions in mortality, but that component attributable to the RRS is unknown. One of the studies79 also found a reduction in the cardio‑respiratory arrest rate, which was most likely attributable to the RRS. The IHI 100,000 Lives Campaign likewise reports mortality reduction among its participating hospitals80 but uses a six‑plank program of which RRSs are only one component. Additionally, not all hospitals implemented all planks and the IHI cannot attribute their observed reductions to any specific plank. Other outcomes, such as ICU mortality and LOS, either have too few reports and/or unclear denominators, making interpretation difficult. Like ICU admission, it is unclear whether LOS should go up or down with an RRS, since patients who are rescued may have much longer LOS than those who die. Additionally, there are many local factors that affect LOS, making generalizability of the data difficult. The effect of RRSs on the process of care is an area for great potential that has only been explored in a limited fashion. Sebat et al studied septic and hypovolemic shock patients over a 5‑year period and found that having an RRS significantly improved such process measures as institution of appropriate goal‑directed fluid therapy. These improvements were associated with reductions in mortality especially for those with septic shock.81 While it has not been studied in a rigorous manner yet, many papers have alluded to the number of patients appropriately made NFR after an RRS activation. Perhaps the need to activate the RRS underscores the patient’s severe condition and creates a forum in which these discussions may begin. We have personally experienced this in our own RRS program on several occasions. A post hoc analysis of the MERIT study database,82 as well as a separate pilot study,83 have begun to more formally explore how RRSs may benefit this process of care. Other outcomes that RRSs may help with include nursing satisfaction, nurse retention, education and patient and family satisfaction. Unfortunately, much of the evidence in support of these is anecdotal. Where family RRS activation is available, families report tremendous support of the system. Nurses are generally quite positive, with some finding the idea of stopping such programs abhorrent.84–86 Hopefully, in the future, survey and resource studies of these measures will provide higher quality evidence.

7  Rapid Response Systems: A Review of the Evidence

73

Finally, we consider whether RRSs are cost‑effective. Unfortunately, there are almost no studies directly addressing whether RRSs have financial benefits through the implementation of an actual program. One study87 that did do so used a modified RRS benefits calculator originally created by William Ward,88 estimating a cost savings of US $170,000. However, as Mr. Ward pointed out in an editorial, their modifications of the calculator to emphasize cost saving misses an important point. Costs are mostly fixed. Keeping a patient on the general ward and avoiding an ICU admission does not allow you to reduce costs. You still need the staff to cover the ICU bed. The bed still “costs” money. The hospital is just not generating any revenue from it because it is now empty. Cutting staff to save costs only hampers your ability to provide quality. Generating revenue through improved processes and quality that allows you to care for more patients efficiently and more safely with fewer complications (like an arrest), is where the real financial benefit from RRSs can come from. As Ellen Lasser noted, “Cost savings will not be achieved by cutting corners, but rather by eliminating processes that lead to poor outcomes and ultimately high costs.”89 As RRSs seek to do just that, we should see evidence emerge that they are in fact, cost‑effective.

Summary Overall, the balance of evidence indicates that RRSs are effective. While the largest randomized study was underpowered and inconclusive, a smaller randomized one was positive. Numerous non‑randomized studies have demonstrated benefit individually and in aggregate. This is especially true for the ability of RRSs to reduce the incidence of cardio‑respiratory arrest. The situation is not as strong for their ability to reduce mortality, although the trend is quite favorable. It is likely that surgical patients and those experiencing shock may experience a greater benefit from RRSs, although this does not mean that RRSs should be reserved only for those patients. This would be logistically difficult in many hospitals and may be viewed as unethical by many, since the evidence is limited in this regard. The ability to study RRSs and their effectiveness is somewhat limited by the nature of the intervention. Their implementation cannot be blinded, individual patient randomization is nearly impossible, and confounders are difficult to control. Additionally, their widespread implementation into many hospitals in the US, Australia and the UK makes future randomization extremely difficult. Other places, like Europe or Asia, may be able to generate additional randomized studies, but non‑randomized cohort studies may be the highest quality data that can be achieved. Given these restrictions, we must rely on this level of evidence as it is. It is hoped that as further studies are published, additional attention will be given to outcomes such as nursing and patient/family satisfaction, staff retention, culture change, safety climate and especially process measures. All of these measures deserve attention as we seek to create safer systems for our patients.

74

B.D. Winters and J.C. Pham

References 1. Straus SE, Richardson WS, Glasziou P, Haynes RB. Evidence‑Based Medicine. 3rd ed. Edinburgh: Churchhill Livingstone; 2005:169. 2. Sax FL, Charlson ME. Medical patients at high risk for catastrophic deterioration. Crit Care Med. 1987;15(5):510–515. 3. Schein RM, Hazday N, Pena M, Robin BH, Sprung CL. Clinical antecedents to in‑hospital cardiopulmonary arrest. Chest. 1990;98(6):1388–1392. 4. Bedell SE, Deitz DC, Leeman D, Delbanco TL. Incidence and characteristics of preventable iatrogenic cardiac arrests. JAMA. 1991;265(21):2815–2820. 5. Daffurn K, Lee A, Hillman KM, Bishop GF, Bauman A. Do nurses know when to summon emergency assistance? Intensive Crit Care Nurs. 1994;10(2):115–120. 6. Smith AF, Wood J. Can some in‑hospital cardio‑respiratory arrests be prevented? A prospective survey. Resuscitation. 1998;37(3):133–137. 7. Buist MD, Jarmolowski E, Burton PR, Bernard SA, Waxman BP, Anderson J. Recognising clinical instability in hospital patients before cardiac arrest or unplanned admission to intensive care. A pilot study in a tertiary‑care hospital. Med J Aust. 1999;171(1):22–25. 8. Hillman KM, Bristow PJ, Chey T, Daffurn K, et al. Antecedents to hospital deaths. Intern Med J. 2001;31(6):343–348. 9. Hodgetts TJ, Kenward G, Vlachonikolis IG, et al. Incidence, location and reasons for avoidable in‑hospital cardiac arrest in a district general hospital. Resuscitation. 2002;54(2):115–123. 10. Kause J, Smith G, Prytherch D, et al. A comparison of antecedents to cardiac arrests, deaths and emergency intensive care admissions in Australia, New Zealand, and the United Kingdom ‑ the ACADEMIA study. Resuscitation. 2004;62(3):275–282. 11. Nurmi J, Harjola VP, Nolan J, Castren M. Observations and warning signs prior to cardiac arrest. Should a medical emergency team intervene earlier? Acta Anaesthesiol Scand. 2005;49(3):702–706. 12. Hillman K, Bristow PJ, Chey T, et al. Duration of life‑threatening antecedents prior to intensive care admission. Intensive Care Med. 2002;28:1629–1634. 13. Franklin C, Mathew J. Developing strategies to prevent in‑hospital cardiac arrest: analyzing responses of physicians and nurses in the hours before the event. Crit Care Med. 1994;22(2):244–247. 14. McGloin H, Adam SK, Singer M. Unexpected deaths and referrals to intensive care of patients on general wards. Are some cases potentially avoidable? J R Coll Physicians Lond. 1999;33(37):255–259. 15. Goldhill DR, White SA, Sumner A. Physiological values and procedures in the 24 h before ICU admissions from the ward. Anaesthesia. 1999;54(6):529–534. 16. Goldhill DR, Worthington L, Mulcahy A, Tarling M, Sumner A. The patient‑at‑risk team: identifying and managing seriously ill ward patients. Anaesthesia. 1999;54(2):853–860. 17. Buist M, Bernard S, Nguyen TV, Moore G, Anderson J. Association between clinically abnormal observations and subsequent in‑hospital mortality: a prospective study. Resuscitation. 2004;62(2):137–141. 18. Santiano N, Young L, Hillman K, et al. Analysis of medical emergency team calls comparing subjective to “objective” call criteria. Resuscitation. 2009;80(1):44–49. 19. Hodgetts TJ, Kenward G, Vlachonikolis IG, Payne S, Castle N. The identification of risk ­factors for cardiac arrest and formulation of activation criteria to alert a medical emergency team. Resuscitation. 2002;54(2):125–131. 20. Stenhouse C, Coates S, Tivey M, Allsop P, Parker T. Prospective evaluation of a Modified Early Warning Score to aid earlier detection of patients developing critical illness on a general surgical ward. Br J Anaesth. 2000;84:663P. 21. Lee A, Bishop G, Hillman K, Daffurn K. The medical emergency team. Anaesth Intensive Care. 1995;23:183–186.

7  Rapid Response Systems: A Review of the Evidence

75

22. Morgan RJM, Williams F, Wright MM. An early warning scoring system for detecting developing critical illness. Clin Intensive Care. 1997;8:100. 23. Goldhill DR, McNarry AF. Physiological abnormalities in early warning scores are related to mortality in adult in‑patients. Br J Anaesth. 2004;92(6):882–884. 24. Goldhill DR, McNarry AF, Mandersloot G, McGinley A. A physiologically‑based early warning score for ward patients: the association between score and outcome. Anaesthesia. 2005;60(6):547–553. 25. Sharpley JT, Holden JC. Introducing an early warning scoring system in a district general hospital. Nurs Crit Care. 2004;9:98–103. 26. Gardner‑Thorpe J, Love N, Wrightson J, Walsh S, Keeling N. The value of Modified Early Warning Score (MEWS) in surgical in‑patients: a prospective observational study. Ann R Coll Surg Engl. 2006;88:571–575. 27. Harrison GA, Jacques T, McLaws ML, Kilborn G. Combinations of early signs of critical illness predict in‑hospital death ‑ the SOCCER study (signs of critical conditions and emergency responses). Resuscitation. 2006;71:327–334. 28. Jacques T, Harrison GA, McLaws ML, Kilborn G. Signs of critical conditions and emergency responses (SOCCER): a model for predicting adverse events in the inpatient setting. Resuscitation. 2006;69:175–183. 29. Subbe CP, Hibbs R, Williams E, Rutherford P, Gemmel L. ASSIST: a screening tool for critically ill patients on general medical wards. Intensive Care Med. 2002;28(Suppl 1):S21. 30. Subbe CP, Kruger M, Rutherford P, Gemmel L. Validation of a modified Early Warning Score in medical admissions. Q J Med. 2001;94(10):521–526. 31. Subbe CP, Davies RG, Williams E, Rutherford P, Gemmell L. Effect of introducing the Modified Early Warning score on clinical outcomes, cardio‑pulmonary arrests and intensive care utilisation in acute medical admissions. Anaesthesia. 2003;58(8):797–802. 32. Haines C, Perrott M, Weir P. Promoting care for acutely ill children. Development and evaluation of a pediatric early warning tool. Intensive Crit Care Nurs. 2006;22(2):73–81. 33. Monaghan A. Detecting and managing deterioration in children. Paediatr Nurs. 2005;17(1):32–35. 34. Duncan H, Hutchison J, Parshuram CS. The pediatric early warning system score: a severity of illness score to predict urgent medical need in hospitalised children. J Crit Care. 2006;21(13):271–279. 35. Subbe CP, Gao H, Harrison DA. Reproducibility of physiological track‑and‑trigger warning systems for identifying at‑risk patients on the ward. Intensive Care Med. 2007;33(4): 619–624. 36. Polderman K, Christiaans HMT, Wester JP, Spijkstra JJ, Girbes ARJ. Intra‑observer variability in APACHE II scoring. Intensive Care Med. 2001;27(9):1550–1552. 37. Gao H, McDonnell A, Harrison DA, Moore T, et  al. Systematic review and evaluation of physiological track and trigger warning systems for identifying at‑risk patients on the ward. Intensive Care Med. 2007;33(4):667–679. 38. Bell MB, Konrad D, Granath F, Ekbom A, Martling CR. Prevalence and sensitivity of MET‑criteria in a Scandinavian university hospital. Resuscitation. 2006;70(1):66–73. 39. Cretikos M, Chen J, Hillman K, Bellomo R, Finfer S, Flabouris A. MERIT study investigators. The objective medical emergency team activation criteria: a case‑control study. Resuscitation. 2007;73(1):62–72. 40. Smith GB, Prytherch DR, Schmidt PE, Featherstone PI, Higgins B. A review, and performance evaluation, of single‑parameter “track and trigger” systems. Resuscitation. 2008;79(1):11–21. 41. Smith GB, Prytherch DR, Schmidt PE, Featherstone PI. Review and performance evaluation of aggregate weighted “track and trigger” systems. Resuscitation. 2008;77(2):170–179. 42. Smith GB, Prytherch DR, Schmidt P, Featherstone PI, et  al. Hospital‑wide physiological surveillance. A new approach to the early identification and management of the sick patient. Resuscitation. 2006;71(1):19–28.

76

B.D. Winters and J.C. Pham

43. Watkinson PJ, Barber VS, Price JD, Hann A, et al. A randomised controlled trial of the effect of continuous electronic physiological monitoring on the adverse event rate in high risk medical and surgical patients. Anaesthesia. 2006;61(11):1031–1039. 44. Tarassenko L, Hann A, Young D. Integrated monitoring and analysis for early warning of patient deterioration. Br J Anaesth. 2006;97(1):64–68. 45. Hunt EA, Zimmer KP, Rinke ML, et al. Transition from a traditional code team to a medical emergency team and categorization of cardiopulmonary arrests in a children’s center. Arch Pediatr Adolesc Med. 2008;162(2):117–122. 46. Jones D, Bates S, Warrillow S, Goldsmith D, et al. Effect of an education programme on the utilization of a medical emergency team in a teaching hospital. Intern Med J. 2006;36(4):231–236. 47. DeVita MA, Schaefer J, Lutz J, Wang H, Dongilli T. Improving medical emergency team (MET) performance using a novel curriculum and a computerized human patient simulator. Qual Saf Health Care. 2005;14(5):326–331. 48. Wallin CJ, Meurling L, Hedman L, Hedegård J, Felländer‑Tsai L. Target‑focused medical emergency team training using a human patient simulator: effects on behaviour and attitude. Med Educ. 2007;41(2):173–180. 49. Cretikos MA, Chen A, Hillman KM, Bellomo R, Finfer S, Flabouris A. MERIT Study Investigators. The effectiveness of implementation of the medical emergency team (MET) system and factors associated with use during the MERIT study. Crit Care Resusc. 2007;9(2):205–212. 50. Buist M, Bellomo R. MET. The emergency medical team or the medical education team? Crit Care Resusc. 2004;6:88–91. 51. MERIT Study Investigators. Introduction of the medical emergency team (MET) system: a cluster‑randomised controlled trial. Lancet. 2005;365(9477):2091–2097. 52. Priestley G, Watson W, Rashidian A, et  al. Introducing critical care outreach: a ward‑­ randomised trial of phased introduction in a general hospital. Intensive Care Med. 2004; 30(7):1398–1404. 53. Chen J, Bellomo R, Flabouris A, Hillman K, Finfer S. MERIT Study Investigators for the Simpson Centre and the ANZICS Clinical Trials Group. The relationship between early emergency team calls and serious adverse events. Crit Care Med. 2009;37(1):148–153. 54. Winters BD, Pham JC, Hunt EA, Guallar E, Berenholtz S. Pronovost PJl. Rapid response systems: a systematic review. Crit Care Med. 2007;35(5):1238–1243. 55. McGaughey J, Alderdice F, Fowler R, Kapila A, et al. Outreach and Early Warning System (EWS) for the prevention of intensive care admission and death of critically ill adult patients on general hospital wards. Cochrane Database Syst Rev. 2007;18:CD00552. 56. Ranji SR, Auerbach AD, Hurd CJ, O’Rourke K, Shojania KG. Effects of rapid response systems on clinical outcomes: systematic review and meta‑analysis. J Hosp Med. 2007; 2(6):422–432. 57. Bellomo R, Goldsmith D, Uchino S, et al. A prospective before‑and‑after trial of a medical emergency team. Med J Aust. 2003;179(6):283–287. 58. Bellomo R, Goldsmith D, Uchino S, et  al. Prospective controlled trial of effect of medical emergency team on postoperative morbidity and mortality rates. Crit Care Med. 2004;32(4):916–921. 59. Jones D, Bellomo R, Bates S, et al. Long‑term effect of a medical emergency team on cardiac arrests in a teaching hospital. Crit Care. 2005;9(6):R808–815. 60. Jones D, Opdam H, Egi M, et al. Long‑term effect of a medical emergency team on mortality in a teaching hospital. Resuscitation. 2007;74(2):235–241. 61. Buist MD, Moore GE, Bernard SA, Waxman BP, Anderson JN, Nguyen TV. Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. BMJ. 2002;324(7334):387–390. 62. Chan PS, Khalid A, Longmore LS, Berg RA, Kosiborod M, Spertus JA. Hospital‑wide code rates and mortality before and after implementation of a rapid response team. JAMA. 2008;300(21):2506–2513.

7  Rapid Response Systems: A Review of the Evidence

77

63. Bristow PJ, Hillman KM, Chey T, et al. Rates of in‑hospital arrests, deaths and intensive care admissions: the effect of a medical emergency team. Med J Aust. 2000;173(5):236–240. 64. Kenwood G, Castle N, Hodgetts T, Shaikh L. Evaluation of a medical emergency team one year after implementation. Resuscitation. 2004;61(3):257–263. 65. DeVita MA, Braithwaite RS, Mahidhara R, Stuart S, Foraida M, Simmons RL. Use of medical emergency team responses to reduce hospital cardiopulmonary arrests. Qual Saf Health Care. 2004;13(4):251–254. 66. Jolley J, Bendyk H, Holaday B, Lombardozzi KAK, Harmon C. Rapid response teams: do they make a difference? Dimens Crit Care Nurs. 2007;26(6):253–260. 67. Dacey MJ, Mirza ER, Wilcox V, et al. The effect of a rapid response team on major clinical outcome measures in a community hospital. Crit Care Med. 2007;35(9):2076–2082. 68. Garcea G, Thomasset S, McClelland L, Leslie A, Berry DP. Impact of a critical care outreach team on critical care readmissions and mortality. Acta Anaesthesiol Scand. 2004;48(9):1096–1100. 69. Tibballs J, Kinney S, Duke T, Oakely E, Hennessy M. Reduction of pediatric in‑patient cardiac arrest and death with a medical emergency team: preliminary results. Arch Dis Child. 2005;90(11):1148–1152. 70. Tibballs J, Kinney S. Reduction of hospital mortality and of preventable cardiac arrest and death on introduction of a pediatric medical emergency team. Pediatr Crit Care Med. 2009;10(3):306–312. 71. Brilli RJ, Gibson R, Luria JW, et al. Implementation of a medical emergency team in a large pediatric teaching hospital prevents respiratory and cardiopulmonary arrests outside the intensive care unit. Pediatr Crit Care Med. 2007;8(3):236–246. 72. Sharek PJ, Parast M, Leong K, et  al. Effect of a rapid response team on hospital‑wide ­mortality and code rates outside the ICU in a children’s hospital. JAMA. 2007;298(19): 2267–2274. 73. Zenker P, Schlesinger A, Hauck M, et al. Implementation and impact of a rapid response team in a children’s hospital. Joint Comm J Qual Saf. 2007;33(7):418–425. 74. Buist M, Harrison J, Abaloz E, Van Dyke S. Six‑year audit of cardiac arrests and medical emergency team calls in an Australian teaching hospital. BMJ. 2007;335(7631):1210–1212. 75. Jones D, Egi M, Bellomo R, Goldsmith D. Effect of the medical emergency team on long‑term mortality following major surgery. Crit Care. 2007;11(1):R12. 76. Ball C, Kirkby M, Williams S. Effect of the critical care outreach team on patient survival to discharge from hospital and re‑admission to critical care: non‑randomised population based study. BMJ. 2003;327(7422):1014–1016. 77. Leary T, Ridley S. Impact of an outreach team on re‑admissions to a critical care unit. Anaesthesia. 2003;58(4):328–332. 78. Mailey J, Digiovine B, Baillod D, Gnam G, Jordan J, Rubinfeld I. Reducing hospital standardized mortality rate with early interventions. J Trauma Nurs. 2006;13(4):178–182. 79. Tolchin S, Brush R, Lange P, Bates P, Garbo JJ. Eliminating preventable death at Ascension Health. Joint Comm J Qual Saf. 2007;33(3):145–154. 80. 100  K Lives Campaign. http://www.ihi.org/IHI/Programs/Campaign/Campaign; 2009 Accessed 05.01.09. 81. Sebat F, Musthafa AA, Johnson D, et  al. Effect of a rapid response system for patients in shock on time to treatment and mortality during 5  years. Crit Care Med. 2007;35(11): 2568–2575. 82. Chen J, Flabouris A, Bellomo R, Hillman K, Finfer S. MERIT Study Investigators for the Simpson Centre and the ANZICS Clinical Trials Group. The Medical Emergency Team System and not‑for‑resuscitation orders: results from the MERIT study. Resuscitation. 2008;79(3):391–397. 83. Jones DA, McIntyre T, Baldwin I, Mercer I, Kattula A, Bellomo R. The medical emergency team and end‑of‑life care: a pilot study. Crit Care Resusc. 2007;9(2):151–156. 84. Jones D, Baldwin I, McIntyre T, et al. Nurse’s attitudes to a medical emergency team service in a teaching hospital. Qual Saf Health Care. 2006;15(6):427–432.

78

B.D. Winters and J.C. Pham

85. Galhotra S, Scholle CC, Dew MA, Mininni NC, Clermont G, DeVita MA. Medical emergency teams: a strategy for improving patient care and nursing work environments. J Adv Nurs. 2006;55(2):180–187. 86. Salamonson Y, van Heere B, Everett B, Davison P. Voices from the floor: nurses’ perceptions of the medical emergency team. Intensive Crit Care Nurs. 2006;22(3):138–143. 87. Thomas K, Vanoyen Force M, Rasmussen D, Dodd D, Whildin S. Rapid response team: ­challenges, solutions, benefits. Crit Care Nurse. 2007;27(1):20–27. 88. Ward WJ. The business case for implementing rapid response teams [power point presentation]. http://www.ihi.org/IHI/Topics/CriticalCare/IntensiveCare/Tools Business Case for Implementing RRTs Presentation; 2009 Accessed 05.01.09. 89. Lanser E. Outcomes and performance measurement. Redefining how healthcare is strategized and delivered. Healthc Exec. 1999;14(4):20–24.

Chapter 8

Healthcare Systems and Their (Lack of ) Integration Ken Hillman, Jeffrey Braithwaite, and Jack Chen

Keywords  Healthcare • systems • integration European monks established acute hospitals as an adjunct to their monasteries to care for Christian soldiers on their way to and from the crusades. The hospitals gradually separated from the monasteries to become more autonomous in both their locations and functions. Although separated from the monasteries geographically, they often remained formally attached to religious orders; in some instances, this relationship continues to this day. The study of medicine soon became part of some European universities, such as Padua and Florence in Italy, and then gradually to other universities across Europe. The teaching of medicine remained largely centered on public hospitals. Even to this day, public hospitals play a major role in undergraduate teaching. However, the practice, as opposed to the teaching of medicine was, until the middle of the twentieth century, based on physicians treating fee-paying patients, usually in their own homes. While medical practitioners would earn their living treating individual fee-­paying patients, they remained attached to public hospitals in order to teach medical ­students and supervise junior doctors in training. These apprentices would be responsible for the majority of care in public hospitals. The hospitals were largely charitable. The clinician would visit the public hospital for several hours each week to attend the poor.

K. Hillman (*) The Simpson Centre for Health Services Research, Critical Care Services, Liverpool Hospital, The University of New South Wales, Locked Mailbag 7103, Liverpool BC, NSW 1871, Australia and The Australian Institute of Health Innovation, The University of New South Wales, Sydney, NSW 2052, Australia e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_8, © Springer Science+Business Media, LLC 2011

79

80

K. Hillman et al.

Teaching and research were performed on these patients. Everyone seemingly gained. The sick received free care; medical students and junior medical staff learned the art of medicine; and the senior clinicians’ conscience and sense of virtuousness were satisfied.1 Because the hospital depended on the physician giving his time for free, the hospital function was structured around his convenience. For example, each senior doctor had access to his own ward and nursing staff. A hierarchy of junior medical staff in training was attached to each specialist. Surgeons had their own operating rooms and recovery areas. The legacy of these arrangements is that hospital systems have been established around the needs of specialist doctors, not the patients.2–4 This is reinforced by medical training that is largely centered around the doctor/patient relationship. Doctors are taught about individual patient signs, symptoms, history, diagnosis and treatment. Students today may learn more about public health and communication skills than in the past, but usually not much about how to establish and work within systems built around the needs of the patient. The hospital system designed for the hospital’s senior physicians probably worked reasonably well until approximately the middle of the last century. Up until then, hospital treatment was based on a small number of operations and procedures, a few primitive investigative tools, and a small range of therapeutic medicines. Hospitals were often places to manage the sick with bed rest and convalescence. An explosion of medical knowledge occurred beginning in the 1950s. Complex surgery, such as cardiac valve replacement, transplant surgery and cancer surgery became commonplace. Great advances occurred in anesthesia, which made the performance of these procedures possible. Physicians developed a wide range of drugs, including antibiotics, chemotherapy and antiarrhythmics. As a result of complex surgery, sophisticated postoperative care was necessary. The speciality of intensive care was born. The skills used to support patients in the operating and recovery rooms (such as artificial ventilation and administration of powerful drugs such as vasopressors) enabled patients, who would have died otherwise to survive, including those with poliomyelitis and severe infections. However, to this day, patients usually remain under an individual clinician who “owns” them. In larger institutions, nursing staff, together with a hierarchy of junior medical staff, often still manage the day-to-day care of the patients. The management of patients through a hospital is often inefficient, dislocated and disorganized.4 In part, this is due to high levels of complexity. The problem of delivering safe care within an antiquated system is exacerbated by the increase in medical specialization combined with the changing nature of hospital patients. Patients are now older with multiple co-morbidities.5 There is an increasing mismatch between these complex at-risk patients with multi-system problems and physicians who have specialized training in a limited area; (e.g. neurology); in one geographical area (e.g. emergency medicine); one population type (e.g. rehabilitation and pediatrics); or in procedures (e.g. surgeons). For example, a patient may be treated by a renowned plastic surgeon supported by excellent nursing staff and trainee doctors. But, let us imagine that,

8  Healthcare Systems and Their (Lack of) Integration

81

despite a high level of care, the patient bleeds postoperatively. Hypovolemic shock is not a common occurrence on a plastic surgery ward. The patient becomes tachycardic and hypotensive for some time. Nursing staff inform the most junior medical staff, who, in turn, report through their hierarchy. The patient may also have co-morbidities, such as ischemic heart disease, hypertension, non-insulin dependent diabetes and chronic lung disease. The hypotension causes myocardial and renal ischemia. A physician is called who orders an EKG that shows ischemia. Aspirin and beta-blockers are ordered. As a result, the bleeding becomes worse and the tachycardic response is blunted. We argue that the system of care fails because it is a system designed to perform elective surgical procedures and assumes that patients will have an uncomplicated postoperative course and that physicians and nurses will be able to recognize and treat deviations in the course of the patients’ illness. The attending doctors and nurses are not trained in the basic principles of how to recognize seriously ill patients, let  alone on how to deliver advanced resuscitation. Furthermore, the system’s care co-ordinating function is missing. In-hospital patients with cross-specialty problems (for example, medical disease in a surgical patient, or surgical illness in a medical patient) are treated using a complex system of referral. The treating medical teams refer problems that are outside their own areas of expertise to other colleagues. This works well for situations that are not life-threatening but, for the increasing population of at-risk patients in acute hospitals, the lack of systems that work at the interfaces between specialists, professions and departments may contribute to excessive morbidity and mortality.6 Seriously ill patients require the urgent attention of specialists such as those trained in intensive care medicine. While intensive care specialists are competent at recognizing and managing events such as ischemia and hypoxia, they have, until recently, practiced within their own four walls.5–8 Thus, the intensive care unit (ICU) is simply another silo in the hospital and, if a seriously ill patient is fortunate enough to receive their care within this domain, they have a good chance of being well managed. However, if the hospital has no system for recognizing at-risk patients, their outcome may be poor. The usual referral process takes time. Ischemia and hypoxia, even in relatively minor degrees, need urgent recognition and intervention. When the patient behaves in a predictable fashion, the silos in healthcare usually work well. Individual doctors, and the areas in which they work, are trained to provide excellent standards of care. However, patients are increasingly falling through gaps between these silos. Many patients die potentially preventable deaths in acute hospitals.9,10 These deaths represent the “tip of the iceberg” of harms that befall patients due to system failures in hospital. For every potentially preventable death, there are many patients who suffer serious adverse events. Up to 80% of hospital cardiac arrests are preceded by at least 8  h of slow deterioration 11; and there are a similar number of potentially preventable antecedents that precede ­hospital deaths 12 and unexpected admissions to the ICU.13 New ways of integrating care around patient needs are being explored and tested. For example, North America has been using “hospitalists” for several years.14

82

K. Hillman et al.

They have a wide range of expertise but concentrate more on acute medicine, especially in the context of older patients with co-morbidities. The hospitalist also understands about continuity and co-ordination of patient care, managing the patient’s in-hospital course and arranging a seamless transition back to a community setting.14 This is one of the ways of counteracting the lack of integration in acute hospital systems. There have also been other patient-centered systems that work on a wholeof-hospital basis – so-called Rapid Response Systems (RRS). They have been designed to recognize seriously ill patients in terms of vital signs and observational abnormalities. There are many forms of RRSs,15–17 the details of which are described in other chapters in this book. While the RRSs were initially established to identify and respond to seriously ill patients, the challenge of implementing a hospital-wide system has highlighted the lack of integration in acute hospitals. In a cluster randomized study evaluating the Medical Emergency Team (MET) system,18 it was found that in the MET hospitals less than half of all patients who suffered an adverse event had their vital signs measured within 24 h of the event. Moreover, only half of all patients who had criteria that defined serious illness had an emergency call made. Obviously, a system based on identifying the seriously ill will not be effective if the criteria defining critical illness are not measured frequently enough. Nor will it function if, after a patient has met the criteria, staff do not activate a call to trigger a response. Silo-based care in acute hospitals has existed for many centuries and because it is so entrenched, it has become difficult to implement whole-of-hospital systems. In order to overcome the lack of integration and to effectively implement a RRS, there are a minimum of five essential components that must be simultaneously instituted (See Table 8.1). Table 8.1  Essential components for implementing a Rapid Response System Calling criteria

Rapid response

Evaluation of the system

Education Ownership

Predefined levels of vital sign and observational abnormalities need to be agreed on and displayed across the organization for the recognition of seriously ill and at-risk patients An around-the-clock response must be available with at least one person skilled in all aspects of advanced resuscitation Agreed data must be collected in order to define the extent of the problem and track the effectiveness of the implementation strategy. The data must be packaged and widely distributed in order to drive effect implementation All staff in the organization must be aware of the system and how to activate it The system must be supported by senior management and clinicians in a formal way in order to ensure effective implementation and sustainability

8  Healthcare Systems and Their (Lack of) Integration

83

Identification of the Seriously Ill At-Risk Patient Successful identification is based on: • Abnormal vital sign readings ‑ pulse rate, respiratory rate and blood pressure. • Abnormal observations, such as airway obstruction, seizures and a decreased level of consciousness. • Other measurements, such as oxygen saturation and urine output can also be included. • The so-called calling criteria must also include “staff concern.” In a mature RRS, staff concern accounts for almost 50% of all emergency calls.

Response to the Seriously Ill Patient • If a patient meets any one of the predefined criteria, a rapid response is mandatory. • The response must always include at least one person who is skilled in all aspects of advanced resuscitation. • It is acknowledged that a list of calling criteria and a response alone will not bring about a “safe” system. The availability of an appropriate response around the clock requires education of existing staff.

Education • All hospital staff need to be aware that an around-the-clock system exists for the identification of, and response to, the seriously ill. They must also be familiar with how to activate the system. This would be the responsibility of individual hospitals and would include strategies such as the calling criteria being included on the back of all hospital name badges; laminated versions of the criteria being made available in all general wards and public places; and explaining the system as part of all hospital staff orientation programs. • There is a general acceptance that the level of skills and knowledge of junior medical, nursing and other ward staff regarding acute medicine, recognizing signs of the deteriorating patient and simple treatment measures is less than satisfactory and education programs should aim to correct this. • The aim of any RRS is to have a least one person available around the clock who has the necessary skills and knowledge to attend to seriously ill at-risk patients in every acute hospital. It must be emphasized that this is a minimum standard, not an optional consideration. It is pointless to identify seriously ill patients if

84

K. Hillman et al.

there are no staff members trained in handling emergencies. Specific education programs, ideally based in specifically designed skills laboratories may be necessary.

Evaluation • Evaluation19 is critical as part of the implementation process and to ensure sustainability. • The evaluation data must be valid and reliable as well as inexpensive to collect. They must accurately reflect how well the system is working and how and where it can be improved. Evaluation must be owned and operated by each institution. • Evaluation of the local system by local staff is the single most important factor in transferring practices elsewhere and sustainability. • The four most important outcomes for evaluating a health system (and specifically the RRS) are: –– –– –– ––

Potentially preventable deaths Potentially preventable cardiac arrests Potentially preventable intensive care admissions Potentially preventable is defined by the event being preceded by one of the calling criteria where there has been no response. –– Number of rapid response calls per 1,000 admissions. This is the single most accurate determinant of effectiveness and sustainability. A system that is just beginning has between three and seven calls per 1,000 admissions and a mature, more effective system has between 40 and 50 calls per 1,000 admissions. • The outcome indicators must be attractively packaged and regularly fed back to all levels of the institution.

Support • Such a system will not work without high-level administrative as well as widespread clinical support. • The Institute of Healthcare Improvement (IHI) in the United States as well as Canadian and Scandinavian experiences suggests that making hospitals safe by providing RRSs for at-risk patients was the single most important factor and therefore the highest priority for saving lives. In Canada and some Scandanavian countries, the most senior executive in the health region/institution was given the responsibility for implementation and sustainability of the RRS. Leadership seems therefore essential support for a successful system.

8  Healthcare Systems and Their (Lack of) Integration

85

• Usually, a specific overarching regional and institutional committee is established to both establish and maintain the system. The committee reviews whether the strategies for the identification of the at-risk patient are being effectively implemented. They also ensure that widespread educational strategies are in place; and, importantly, regularly review the outcome indicators and adjust the system as appropriate. • While the organization framework for the system could be embedded within the existing hospital structure, the system itself requires a separate structure and brief and it should be directly accountable to the institutional senior executive. For example, data for evaluation may already be collected within the organization but successful implementation depends on recognizing the challenges of combining all existing functions and levels into a cohesive system with separate responsibility and accountability. The process of implementing an RRS in an organization influences the culture20 of that organization.21 For example, it facilitates the organization to develop a template for other hospital-wide patient-centered initiatives. It encourages hospital administration and clinicians to work together in establishing other patient safety programs. The process empowers staff who are delivering healthcare directly to patients to have the confidence not only to observe and measure patient vital signs and observations, but also to take action based on those measurements. The pioneers of intensive care created an environment in which the specialty could be nurtured. From within ICUs there developed educational strategies for training specialists, as well as a place to consolidate the legitimacy of the specialty and to refine the knowledge and skills necessary to deliver a high level of care. Now that the specialty of critical care medicine has been secured within the walls of the ICU, the next major advance in intensive care medicine is to contribute to the creation of systems to either prevent admission to ICU or optimize the outcome of those we manage in the ICU. As early as 1974, Peter Safar,22 one of the pioneers in intensive care, stated that “The most sophisticated intensive care often becomes unnecessarily expensive terminal care when the pre-ICU system fails,” The development of patient-centered systems to improve safety has been a major step forward in increasing hospital integration.

References 1. Marsh WJ. Physicians, practitioners and fees. BMJ. 1878;1:197–198. 2. Hillman K. The changing role of acute‑care hospitals. Med J Aust. 1999;170:325–328. 3. Braithwaite J, Vining RF, Lazarus L. The boundaryless hospital. Aust NZ J Med. 1994;24(5):565–571. 4. Braithwaite J, Lazarus L, Vining RF, Soar J. Hospitals: to the next millennium. Int J Health Plann Manage. 1995;10(2):87–98. 5. Hillman K, Parr M, Flabouris A, Bishop G, Stewart A. Redefining in-hospital resuscitation: the concept of the medical emergency team. Resuscitation. 2001;48(2):105–110. 6. Hillman K. Critical care without walls. Curr Opin Crit Care. 2002;8(6):594–599.

86

K. Hillman et al.

7. Hillman K. A hospital-wide system for managing the seriously ill. Minerva Anestesiol. 1999;65(6):346–347. 8. Hillman K. Expanding intensive care medicine beyond the intensive care unit. Crit Care. 2004;8(1):9–10. 9. Wilson RMcL, Runciman WB, Gibbert RW, et al. The quality in Australian health care study. Med J Aust. 1995;163:458–471. 10. Brennan TA, Leape LL, Laird N, et  al. Incidence of adverse events and negligence in ­hospitalized patients: results of the Harvard Medical Practice Study 1. N Engl J Med. 1991;324(6):370–376. 11. Schein RM, Hazday N, Pena M, Ruben BH, Sprung CL. Clinical antecedents to in-hospital cardiopulmonary arrest. Chest. 1990;98(6):1388–1392. 12. Hillman KM, Bristow PJ, Chey T, et  al. Antecedents to hospital deaths. Intern Med J. 2001;31(6):343–348. 13. Bristow PJ, Hillman KM, Chey T, et al. Rates of in-hospital arrests, deaths and intensive care admissions: the effect of a medical emergency team. Med J Aust. 2000;173:236–240. 14. Wachter RM, Goldman L. The emerging role of “hospitalists” in the American health care system. N Engl J Med. 1996;335(7):514–517. 15. Goldhill DR, Worthington L, Mulcahy A, Tarling M, Sumner A. The patient‑at‑risk team: Identifying and managing seriously ill ward patients. Anaesthesia. 1999;54(9):853–860. 16. Bright B, Walker W, Bion J. Clinical review: outreach ‑ a strategy for improving the care of the acutely ill hospitalized patient. Crit Care. 2004;8(1):33–40. 17. Coombs M, Dillon A. Crossing boundaries, re‑defining care: the role of the critical care ­outreach team. J Clin Nurs. 2002;11(3):387–393. 18. MERIT Study Investigators. Introduction of the medical emergency team (MET) system: a cluster-randomised controlled trial. Lancet. 2005;365(9477):2091–2097. 19. Braithwaite J, Westbrook MT, Tranglia JF, et al. Area health systems changing in support of patient safety? A multi-methods evolution of education, attitude and practice. Int J Health Care Qual Assur. 2007;20(7):585–601. 20. Braithwaite J, Greenfield D, Westbrook MT. Contrasting and converging perspectives on organisational culture and climate. In: Braithwaite J, Hyde P, Pope C, eds. Culture, Climate and Teams in Health Care Organisations. London: Polygrave Macmillan; 2009. 21. Berwick DM, Calkins DR, McCannon CJ, Hackbarth AD. The 100, 000 Campaign: setting a goal and a deadline for improving health care quality. JAMA. 2006;295(3):324–327. 22. Safar P. Critical care medicine-quo vadis. Crit Care Med. 1974;2(1):1–5.

Chapter 9

Creating Process and Policy Change in Healthcare Stuart F. Reynolds and Bernard Lawless

Keywords  Rapid • Response • Systems • Public policy • Hospital • Innovation

Introduction The word “policy” is often bantered about and it is assumed that everyone has or uses a fundamental definition for policy and hence, policy development. In broad terms, policy can refer simply to a “plan of action” or “statement of aims or goals.” However defined, it is then difficult to interpret who is involved in policy planning and what are the key functions of policy development, policy implementation and subsequent evaluation. To better understand some of these nuances, William Jenkins defines policy as a set of incremental decisions taken by a political figure or group regarding the prioritization of goals and the means to achieve these goals. James Anderson describes public policy as a purposive course of action aimed at dealing with a problem identified by the government. These begin to capture some of the key elements of public policymaking and the inherent link to government as a key player in policy development and implementation.

Changing Healthcare Policy Policy to support large-scale implementation has not historically been rooted in scientific evidence and frequently not well supported by anecdotal evidence.1 Using evidence for policymaking, particularly in the healthcare sector, has been weak, especially when compared to other industries. Health and social policies generally apply to a breadth of communities or individuals through a variety of disparate mechanisms that can subsequently make programs difficult to evaluate S.F. Reynolds (*) Associate Clinical Professor of Critical Care Medicine, University of Alberta Hospital, 3C1.12 WMC, 8440 112 Street, Edmonton, Alberta, Canada e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_9, © Springer Science+Business Media, LLC 2011

87

88

S.F. Reynolds and B.Lawless

with scientific rigor.2 Even focused healthcare policies seem to rarely be evaluated, owing to difficulties monitoring multiple activities, the challenges of observing for disparate effects, or timelines that do not allow for observing meaningful change. Also, the “politics” of policymaking would have that politicians or decision makers are less interested in science or health impact and more interested in financial implications or views of particular groups and communities.3 The process of ­“evidence policy making” is not enhanced by the fact that structural barriers to communication exist between researchers and decision makers. This can include differences of priorities, language, means of communication, interpretation of ­findings and definition of final products of research. The policy cycle makes up the regular business of government. It strives to be incremental, continuous and, hopefully, systematic. Despite what may appear to be a good model for prospective planning, the creation of new public policy is often rooted in a response to a disastrous situation. The outbreak of SARS in Ontario in 2003 demonstrated that as a provincial resource, critical care was not well coordinated on a systems level, or even within individual institutions. Even though SARS primarily affected only Ontario’s largest urban area, Toronto, the effect on hospital services and the impact on the availability of critical care resources were across the entire province. A relatively small number of infective patients (70) had a major impact on the province-wide critical care system. Within Toronto, this resulted in the closure of hospital emergency departments, cancellation of elective surgical procedures as well as halting the normal transfer patterns of patients between ­hospitals. Hospitals outside Toronto lost their usual referral destination for patients requiring tertiary and quaternary care. The impact that a poorly coordinated and functioning critical care system could have on many other processes within hospital systems became glaringly apparent. However, SARS was not the only precipitating event to give support to Ontario’s critical care strategy. It became obvious that the addition of critical care capacity could not be the sole answer. A well-functioning and integrated critical care strategy was required to mitigate the demand on surgical and critical care services imposed by the demographic pressure of an aging population. Furthermore the ongoing specter of a pandemic increased the awareness that a provincial critical care strategy was required. With strong support from the Ministry of Health and Long-Term Care and, in collaboration with healthcare decision makers and critical care stakeholders, a sevenpronged strategy was implemented to achieve three core goals: 1 . Improve access to critical care resources 2. Improve the quality of care provided in critical care areas across the province 3. Improve system integration so that intensive care units did not function in silos, but moved to be able to leverage critical care as a resource across the province (see Fig. 9.1) Key to the strategy was the development of a provincial rapid response system, known as the Critical Care Response Team Project (CCRT). It was readily apparent that CCRT’s were an innovation well situated to meet the three core goals (see Fig. 9.2).

9  Creating Process and Policy Change in Healthcare

Fig. 9.1  Critical care strategy

Fig. 9.2  Critical Care Response Team Project

89

90

S.F. Reynolds and B.Lawless Be not the first by who the new is tried, nor the last to lay the old aside. Alexander Pope, an essay on criticism

Healthcare innovations are not simple, and Pope’s quote presages that there is a dynamic sociologic interplay between healthcare innovation and healthcare culture. This complex interplay thus makes implementation and evaluation inherently ­complex. Rapid response systems (RRS) are a complex healthcare innovation that result in significant process and culture change. It presents research challenges. There is an inherent complexity to randomizing policies to whole populations or communities: achieving agreement of measurement and analysis, maintaining internal integrity, timing interventions with data collection and the inherent political nature of the evaluation process are unique to the policy experimental setting.3 Thus many policies and innovations need to be evaluated from a social science paradigm, rather than the traditional medical evaluation ­paradigm.4 A social science paradigm enriches the evaluation by explaining why certain innovations are readily adopted and successful, while others are rejected or do not perform, evaluation from a social science perspective answers the question of why a policy, process or program works or does not work. Rogers,5 in his book Diffusion of Innovation, defined an innovation as an idea, a practice that is perceived as new by an individual or an organization. The diffusion of the innovation is the process whereby the innovation is communicated through members of a social system. To obtain widespread acceptance, an innovation needs to demonstrate a relative advantage. This requires that the new idea must offer significant improvements from current practice. Also, innovations with minimal complexity are more apt to be accepted. Finally, when people see the results of an innovation, the observability increases the likelihood that the innovation will be both utilized and accepted. Current data suggests that patients are at risk of adverse events, including ­cardiac arrest and death while residing in hospitals. Most of these adverse events are predictable, and therefore preventable; often through simple interventions.6,7 The relative advantage of an RRS is that it provides a systematic mechanism to both identify and respond to patients at risk of clinical deterioration. This is significantly better than traditional medical practice, where the patient is required to significantly deteriorate prior to obtaining critical care expertise.8,9 The RRS concept is not a complex notion, but rather a common sense innovation; simply put ‑ early detection and action will be simpler and more effective than delayed detection and resuscitation. A landmark trial in the care of patients with severe sepsis and septic shock has given credence to this concept.10 Finally, an RRS is not restricted to the four walls of an ICU; instead, the impact is readily and consistently observed throughout the hospital. If an innovation appears to have sufficient merit on the basis of relative advantage and complexity, the degree to which it is adopted depends on the characteristics of those within the social system.11 The diffusion of innovation theory identifies characteristics of those who, within an organization, are apt to support change or innovation. Early support can be expected from those within the organization themselves are innovators. Early adopters of innnovation are usually well integrated within the

9  Creating Process and Policy Change in Healthcare

91

organization. These individuals are often the opinion leaders and perceived as role models for peers. Early majority adopters are characterized by frequent interactions with peers, but are unlikely to be opinion leaders and tend to deliberate before adopting new ideas. Late majority adopters are characterized as skeptical and cautious, and their motivations for change comes from pressure from peers and economic sources. The last group to adopt, or most likely to not adopt, are unfortunately referred to as ­“laggards.” Ryan and Gross characterized this group as isolated, with a point of reference in the past, suspicious of innovation and not considered opinion leaders. An effective communication strategy will increase the number of people willing to adopt an innovation. Stakeholder input will be helpful with initial design and subsequent acceptance of an innovation. Communication during implementation that clearly defines goals and responsibilities with opinion leaders, team members, end-users and administration is imperative. From an RRS perspective, it is important to seek out the opinion leaders and innovators within the hospital system. They are often, but not always, found in leadership positions within the hospital structure. Acceptance by leadership will serve to build a foundation of system-wide acceptance and ensure that there is alignment of RRS goals with organizational values; thereby improving the chance of RRS success.12 Rapid response systems have the opportunity to be a readily accepted innovation on the basis of need alone. However, as Dodgson and Bessant13 describe, the success of an innovation relies on the recipient of the innovation doing something with the resource provided. This explains the relationship between the uptake of service (adoption) of RRS and reduction in cardiac arrests and other adverse events.14,15 The question needs to move from “Does the innovation work?” to “Why does it work and in what context?” This is referred to as a realist evaluation.16,17 Policy creation can be the result of needs identified after stresses on a system unveil disparity. This was the case with the critical care strategy in Ontario. It was developed in response to SARs, and the anticipated demographic demand. As stated previously, policy evaluation is fraught with significant confounders; traditional medical evaluation techniques that focus on clinical outcomes are of limited utility. Many of the benefits and outcomes of RRS policy will not be realized during early days. The first focus during policy implementation should therefore be on diffusion of the innovation, creating an environment that will allow the innovation to thrive. Therefore, program acceptance or utilization is a reasonable first outcome measure. It is folly to focus on clinical outcomes early in the establishment of an RRS policy as demonstrated in the MERIT trial,15 whereby the failure of MET hospitals to significantly reduce clinical outcomes was attributed to the ineffectiveness of METs, rather than the incomplete diffusion of the MET concept. It is imperative that RRSs be continuously evaluated. The very presence of the program will generate a new equilibrium within the system. Continuous program evaluation and modification is required for feedback to improve RRS service and improve system-wide function.

92

S.F. Reynolds and B.Lawless

The success of an RRS policy is dependent on the diffusion of the RRS innovation via a structured and audited implementation process across multiple hospitals. This structure must promote local leadership and, at the same time, provide central coordination and support.

References 1. Letherman S, Sutherland K. Designing national quality reforms: a framework for action. Int J Qual Health Care. 2007;19(6):334–340. 2. Shimkhada R, Peabody JW, Quimbo SA, Solon O. The quality improvement demonstration study: an example of evidence-based policy-making in practice. Health Res Policy Syst. 2008;6:5. 3. Davies P. Policy evaluation in the United Kingdom. In: KDI International Policy Evaluation Forum, Seoul, Korea; 2004. 4. Berwick D. The science of improvement. JAMA. 2008;299(10):1182–1184. 5. Rogers EM. Diffusion of Innovations. 4th ed. New York: The Free Press; 1995. 6. McGloin H, Adam S, Singer M. The quality of pre-ICU care influences outcome of patients admitted from the ward. Clin Intensive Care. 1997;8:104. 7. McQuillan P, Pilkington S, Allan A, et  al. Confidential inquiry into quality of care before admission to intensive care. BMJ. 1998;316(7148):1853–1858. 8. Schein RM, Hazday N, Pena M, Ruben BH, Sprung CL. Clinical antecedents to in-hospital cardiopulmonary arrest. Chest. 1990;98(6):1388–1392. 9. Buist MD, Jarmolowski E, Burton PR, Bernard SA, Waxman BP, Anderson J. Recognizing clinical instability in hospital patients before cardiac arrest or unplanned admission to intensive e care. A pilot study in a tertiary-care hospital. Med J Aust. 1999;171(1):22–25. 10. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med. 2001;345(19):1368–1377. 11. Tyan B, Gross NC. The diffusion of hybrid seed corn in two Iowa communities. Rural Sociol. 1943;8:15–24. 12. Soumerai S, McLaughlin TJ, Gurwitz JH, et  al. Effect of local medical opinion leaders on quality of care for acute myocardial infarction. JAMA. 1998;279(17):1358–1363. 13. Dodgson M, Bessant J. Effective Innovation Policy: A New Approach. New York: Routledge; 1996. 14. Jones D, Bellomo R, Bates S, et al. Long-term effect of a medical emergency team on cardiac arrests in a teaching hospital. Crit Care. 2005;9:R808–R815. 15. MERIT Study Investigators. Introduction of the medical emergency team (MET) system: a cluster randomised controlled trial. Lancet. 2005;365(9477):2091–2097. 16. Reynolds S. Sustainability of RRS presentation. 17. Pawson R, Tilley N. Realistic Evaluation. London: Sage; 1997.

Chapter 10

The Challenge of Predicting In-Hospital Cardiac Arrests and Deaths Michael Buist

Keywords  Predicting • Patient • Hospital • Crisis • Cardiac • Deaths

Introduction In this chapter, we first explore the similarities and differences between the current hospital crisis of iatrogenic patient deaths, which are now the fourth most common cause of death in UK,1 and the sixth most common in US,2 and the theories that have been used to explain and manage organizational crises that occur in other industries. We then critically examine the studies to date that attempt to predict the in-hospital patient management crises. Finally, we examine the place of hard defenses such as electronic monitoring and alert systems to protect patients from the healthcare system. In the longer run, there needs to be a significant and fundamental change to the “soft defenses,” such as the training of all our frontline healthcare workers, so that such potential patient crises are predicted and managed earlier to prevent iatrogenic morbidity and mortality.

Organizational Crisis Theory: Hazards, Defenses and Latent Conditions James Reason, in his book “Managing the Risks of Organizational Accidents,” states that organizational accidents, as opposed to individual accidents, are predictable events.3 An individual accident is one in which a person or group of people

M. Buist (*) Rural Clinical School, University of Tasmania, Private Bag 3513, Burnie, TAS 7320, Australia e-mail: [email protected]

M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_10, © Springer Science+Business Media, LLC 2011

93

94

M. Buist

makes an individual slip, lapse, or error of judgment with the net result being an adverse outcome either to the person or the people who erred, or to the person or people in the immediate vicinity. As such there is usually a relatively tight, simple explanation for cause and effect in an individual accident. For, example if one makes the error of judgment to drive a motorcar on the wrong side of the road, there is a high likelihood of an accident, which will involve the person who made the error of judgment, along with any other bystanders. On the other hand, organizational accidents have “multiple causes involving many people at different levels of an organization.”4 These events, while usually infrequent, are often catastrophic. Analyses of such organizational accidents often reveal that the defenses an organization has to prevent such catastrophes are breached by a unique series of sequential hazards that play out in an environment of latent conditions. There is always a tension within an organization to balance resource allocation for production and profit generation as opposed to the implementation, maintenance and updating of defenses to protect the organization from crisis. Resource allocation for production of profit is a core tenant of a commercial organization. It is a process that has easily measured endpoints with relatively simple relationships between resource allocation and production. On the other hand, resource allocation for organizational defenses has no such relationship and the benefits of such defenses are difficult to measure. As such, if an organization has little exposure to hazards that may cause a crisis, it can be difficult to allocate resources to defenses, in the face of societal or financial drivers to maximize production. This tension thus creates the landscape, or latent, conditions that may predispose an organization to crisis.4 The defenses within an organization can be simply categorized into either “hard” or “soft.” Hard defenses are physical barriers, where no human discretion can apply. Soft defenses relate to laws, rules, policies, procedures, guidelines and often, as a last resort, “common sense.” These soft defenses are human constructs, however, and as such their implementation, utilization, analysis, improvement or otherwise and even avoidance can occur at an individual operator level. Furthermore, operator interpretation and implementation or lack thereof inevitably becomes an organizational issue often dependent on where in a particular organization the “tension” between production and protection sits.4

Iatrogenic Patient Death: Individual or Organizational Accident? Thus, we turn to the crisis in safety of healthcare. Up until this point, we have used the terms “crisis,” “catastrophe” and “accident” to mean the same thing: a sudden overwhelming event with considerable damage to those involved. For the purposes of the remainder of this chapter, we will confine ourselves to a definition of crisis as an unexpected, iatrogenic in-hospital death (Box 1). The fact that in an epidemiological, societal, political, and medico-legal sense these deaths constitute a crisis has been made in other chapters.

10  The Challenge of Predicting In-Hospital Cardiac Arrests and Deaths

95

BOX 1 A 47-year old, previously completely well male underwent a semi-elective thoracotomy for an empyema. The surgical procedure and anesthetic were uneventful. The patient returned to the ward at 15:00 with a heart rate of 130 bpm. Otherwise his observations were unremarkable. The surgical registrar was concerned about the heart rate and the patient’s inability to pass urine postoperatively. She instructed the intern to insert a urinary catheter if the patient failed to pass urine by 18:00. At 18:00 there was no urine output, the heart rate was 140 bpm. Despite the intern’s insistence, the patient refused to have a urinary catheter inserted. Otherwise the patient’s condition was stable. The day intern handed over the patient in a verbal report to the night resident medical officer (RMO) at 22:00. The night RMO was summoned urgently to see the patient at 23:30 when the patient’s blood pressure dropped to 85/60 mmHg. The heart rate was now 150 bpm. The RMO assessed that the patient was hypovolemic and administered 2  l of intravenous fluid, and ordered a blood transfusion. With this intervention, the blood pressure improved and the RMO went about his other tasks. There were no further observations on the patient until 02:30 when the blood pressure was observed to be 75/55 mmHg. The RMO again responded promptly and commenced further fluid resuscitation. Again there was a transient improvement in the patient’s condition. At about 04:00 the RMO was concerned enough about the patient to telephone the on-call surgical registrar (offsite, on-call due to financial restraints). The RMO explained the patient’s condition to the registrar. The surgical registrar was concerned and stated that he would come in early at 07:00 to review the patient prior to the commencement of his operating list. At 05:30, the patient lost consciousness, and the nursing staff put out a cardiac arrest call. Despite the best efforts of the anesthetic registrar and the ICU registrar, the patient could not be resuscitated and he was declared deceased at 06:00.

When a patient enters a hospital, he enters a system where he will be exposed to a variety of hazards, which in turn have numerous defenses in place to prevent an adverse patient outcome. Operations, anesthesia, medical interventions and procedures, drugs and fluids and even oxygen therapy constitute the hazards. Some hard defenses exist in anesthesia, whereby the administration of hypoxic gas mixtures is physically prevented, otherwise most other defenses in the general hospital ward environment are soft. These soft defenses include treatment policies and procedures, manual alarm systems, and ad hoc hierarchical and lateral human checking systems. Soft defenses are very reliant on the training and education that healthcare workers receive. Superimposed on these layers of hazards and defenses that the patient is confronted with, there are the latent conditions that exist most obviously within the patient, but more insidiously within the hospital as an organization. A patient’s past medical history, family history, social history-associated co-morbidities, drug regimen,

96

M. Buist

and allergies, largely constitute their latent conditions. These conditions, and their relation to the current presenting complaint that brings the patient into the hospital system, is territory that individual healthcare workers are usually extremely well trained in and familiar with. Hospital latent conditions are not so explicit, particularly to the patient or the frontline healthcare worker. They are made up of a complex matrix of production imperatives, such as the financial operating environment, political and societal imperatives, medico-legal and insurance concerns, compliance issues imposed by various regulatory bodies (often with associated financial incentives or disincentives) and workforce and work-practice issues. In the acute care hospital, the distinction between individual and organizational accidents is blurred. First, the crisis of iatrogenic patient death is insidious. Although epidemiologically this is a crisis that may constitute an epidemic, to the individual practitioner or even hospital, these unfortunate events may not appear as part of an epidemic, largely because, at an individual level, these events occur relatively infrequently, over a long timeframe. For example, The Quality in Australian Healthcare Study looked at a random sample of 14,179 admissions to 28 hospitals in two states of Australia, in 1992, and documented 112 deaths (0.79%), including 109 cases where the adverse event caused greater than 50% disability (0.77%).5 Nearly 70% of the deaths, and 58% of the cases of significant disability were considered to have a high degree of preventability. Thus, for the individual clinicians, treating departments and units, and even the 28 study hospitals themselves, their actual experience of these outcomes over the year would be minimal (one or two cases). Secondly, the defenses that hospitals have in place to protect patients have not significantly changed over recent decades. In particular, there are few, if any, “hard” defenses, and the “soft” defenses are overly reliant on the skills and abilities of the frontline healthcare workers, principally the junior doctor and nurse. In particular, at least in Australia and UK, several studies indicate that the medical undergraduate syllabus does not provide graduates with the basic knowledge, skills, and judgment to manage acute life-threatening emergencies.6–9 These studies identified deficiencies in cognitive abilities, procedural skills, and communication. Further analysis of the causative factors associated with the adverse events in The Quality in Australian Health Care Study found that cognitive failure was a factor in 57% of these adverse events.10 In this analysis, cognitive failure included such errors as failure to synthesize, decide and act on available information; failure to request or arrange an investigation, procedure or consultation; lack of care or attention; failure to attend; misapplication of, or failure to apply, a rule, or use of a bad or inadequate rule.10 In a two-hospital study from UK that looked at 100 sequential admissions to the intensive care unit (ICU) from ward areas, it was found that 54 had suboptimal care on the ward prior to transfer.11 This group of patients had a mortality rate of 56%. Some of the sub-optimal treatment factors included failure to seek advice, lack of knowledge, failure to appreciate clinical urgency, and lack of supervision. Undergraduate and postgraduate curricula have been slow to embrace a patient safety culture.12,13 The hospital organizational response to the issue of adverse events and iatrogenic deaths has generally been to attempt to document and audit incidence, reinforce the traditional hierarchal referral model of care, and to incorporate a plethora of written policies and procedures into the clinical

10  The Challenge of Predicting In-Hospital Cardiac Arrests and Deaths

97

e­ nvironment with little in the way of sustained organizational attempts to “close the loop.” In the acute hospital setting, the frequent turnover of workers through frontline care delivery positions and the expectation that the hospital is a “training” setting may reduce the organizational ability to “see” such events and to retain corporate memory of them, let alone to have the sophisticated procedures in place to undertake root cause analysis and organizational learning to take place. Finally, the hospital environment is a complex and dynamic matrix of political, administrative, financial, work place and workforce variables that interact to provide patient care.14 This effect overwhelms the fact that one could probably argue that the hazards that a patient may encounter have at worst changed little or at best diminished somewhat thanks to better operative, peri-operative techniques and safer drugs.

Can We Predict Hospital Iatrogenic Death? Implicit in the prediction of iatrogenic hospital death is the fact that there would need to be a number of easily identifiable simple clinical markers or factors that predict death. Unfortunately, there are no such markers or factors. In an attempt to look for such markers and risk factors, three types of study have been used. First, the large retrospective epidemiological case note review studies used to determine incidence and outcome from hospital adverse events have shed some light on factors that may predispose a patient to iatrogenic hospital death.1,5,10,15–18 The Harvard Medical Practice Study (HMPS) and the QAHCS both performed a separate analysis of the documented adverse events, by an iterative process with expert reviewers, to look for causative factors, degrees of preventability and with HMPS-associated negligence. The HMPS found that age, operative status, and negligence were associated with poor outcomes (death and permanent disability) from adverse events with associated high degrees of preventability. The HMPS documented that patients over 65 years of age had double the risk of an adverse event than patients between 16 and 44.15,16 It also estimated that 51.3% of the deaths from adverse events were caused by negligence. In a re-examination of the 2,351 adverse events from the QAHCS, 34.6% of the adverse events were categorized as “a complication of, or the failure in the technical performance of an indicated procedure or operation.”10 However, of more significance, 81.8% of events were associated with human error and cognitive failure as discussed above.10 The QAHCS also found that “delay,” both diagnostic and treatment was associated with 20% of adverse events and that 86–90% of these events were assessed to be highly preventable.10 A second methodology is the case or case-control study, which retrospectively examines features of care received by patients who had an unexpected in-hospital death, or other high risk event (in hospital cardiac arrest or unplanned ICU admission).A retrospective case note review at the Jackson Memorial Medical Center in Florida, documented, over a 4-month period in 1987, 64 consecutive ­in-hospital cardiac arrests in the general ward areas.19 Of these, 54 (84%), had

98

M. Buist

documented observations of clinical deterioration or new complaints within 8 h of the arrest. In a similar study performed at the Cook County Hospital in Illinois, 150 cardiac arrests were observed in the medical wards over a 20-month period from 1990 to 1991.20 In 99 of these (64%), a nurse or physician documented deterioration in the patient’s condition within 6 h of the cardiac arrest. The hospital mortality rate of the 150 cardiac arrests was 91%. Again, in a 28-week period at the Manchester Royal Infirmary, reported in 1999, 47 cardiac arrest calls in the general ward areas were analyzed.21 Twenty-four (51%) had premonitory signs prior to the cardiac arrest call. Similarly, in a study in a tertiary care hospital in metropolitan Melbourne over the 1997 calendar year, there was a median period of documented clinical instability of 6.5 h (range 0–432 h) prior to either cardiac arrest call or intensive care unit referral among 122 in-hospital patients.22 This was despite the fact that over the period of instability, these patients were reviewed, on average, twice by junior medical staff.22 In a case-control study of 118 consecutive cardiac arrests in 1999 performed at Selly Oak Hospital in Birmingham, multivariate analysis identified abnormal breathing, abnormal pulse and abnormal systolic blood pressure in the hours prior to the cardiac arrest as being positively associated with the event.23 More simply, Goldhill and Summer made the observation, across a group of UK hospitals, that admission to the intensive care unit from the general ward areas, as compared to ICU admission from the operating room or emergency department resulted in significantly higher mortality.24 Furthermore, in a study of 7,190 ICU admissions across 24 UK hospitals, the actual length of stay in a general hospital ward was an independent predictor of hospital mortality.25 This study documented a hospital mortality rate of 67.2% for patients who were on the ward for more than 15 days. An analysis of 1,047 general ward patients in a UK Trust Hospital that were assessed by an intensive care outreach team in 2005 found an association between all physiological variables except temperature and heart rate with hospital mortality.26 Looking at 321 in-hospital cardiac arrests in a single Finnish Hospital between 1993 and 2002 found that documented abnormal clinical observations, an initial rhythm at arrest that was nonshockable, unwitnessed arrest, and delay in resuscitation team response were all independent risk factors for mortality at hospital discharge.27 There are only two multi-center case or case control studies. In the ACADEMIA study undertaken jointly by the UK Intensive Care Society and the Australia and New Zealand Intensive Care Society clinical trials group,28 data was collected on the incidence of serious physiological abnormalities that preceded in-hospital death, cardiac arrest or unanticipated ICU admission over 3 consecutive days in 90 hospitals in the three study countries. Over the 3 days of the study, there were 638 such events, of which 60% had a total of 1,032 serious physiological abnormalities prior to the index event. In the second case control study, Cretikos29 conducted a nested case control study in seven Australian public hospital during the MERIT study.30 There were 450 cases as defined by ICU/HDU admission, cardiac arrest and death in hospital patients without a “do not resuscitate” order and 520 sex-, age- and locationmatched controls. In this analysis, various abnormal observations for activation of

10  The Challenge of Predicting In-Hospital Cardiac Arrests and Deaths

99

the medical emergency team were applied to determine the criteria with highest sensitivity and specificity and hence positive predictive value for a case event. The most predictable criteria was a combination of a heart rate greater than 140 bpm, respiratory rate greater than 36/min, systolic blood pressure of less than 90 mmHg, and a greater than two-point reduction in the Glasgow Coma Scale (sensitivity of 49% and specificity of 94% and a positive predictive value of only 10%.) Thirdly, there are prospective cohort studies. These studies look at a population of patients over time and try to determine risk factors from the patients who have sentinel events. A New South Wales (NSW) study of 50,942 acute-care admissions to three hospitals, performed over a 6-month study period in 1996, documented the antecedents of 778 deaths.31 Of these, only 66 were classified as unexpected in that they did not have a “do not resuscitate” order or were preceded by a cardiac arrest or intensive care unit admission. In the 8 h prior to the deaths of these patients, 50% had documented severe abnormalities in the observation charts or concerns noted in the nursing or medical record. Furthermore, 33% of these patients had abnormal observations, or concerns noted up to 48 h prior to their deaths. The most common abnormal observations were hypotension (systolic blood pressure <90 mmHg) and tachypnea (respiratory rate >36/min). In another study, Bellomo et al followed 1,125 patients admitted for more than 48  h (excluding day case surgical admissions) to the surgical units at the Austin Hospital over a 6-month period in 1999.32 They documented 414 serious adverse events including 80 patient deaths (7.1%). This study also identified increased age as a risk factor for death from adverse events. The mortality rate for patients aged more than 75 years who underwent unscheduled surgery was 20%. The major limitation of this study was that the definition of serious adverse events included postoperative complications such as sepsis, pulmonary edema and acute myocardial infarction, events that although serious and adverse, may not have been preventable. Similarly, a study of 6,303 patients33 admitted to Dandenong Hospital in Melbourne, over a 33-week period in 1999, 564 (8.9%) experienced a total of 1,598 abnormal observations. The two commonest abnormal observations were desaturation to less than 90% (51% of all events), and hypotension (17.3%). One hundred and forty-six patients died during the study. When the abnormal observations were considered simultaneously in a multiple linear logistic regression model, the following events were found to be significant predictors of mortality: decrease of consciousness, loss of consciousness, hypotension, respiratory rate <6/min, oxygen saturation <90%, and tachypnea >30/min. The presence of any one of the six events was associated with a 6.8-fold (95% CI: 2.7–17.1) increase in the risk of mortality (Table 10.1). A cross-sectional survey undertaken at the Royal London Hospital came to a similar conclusion.34 On a single day, the following data were collected from 433 adult nonobstetric inpatients: respiratory rate, heart rate, systolic blood pressure, temperature, oxygen saturation, level of consciousness, and urine output for catheterized patients. Mortality status at hospital discharge was then determined. Logistic regression modeling identified level of consciousness, heart rate, age, systolic blood pressure and respiratory rate as important variables in predicting

100

M. Buist Table 10.1  Risk of mortality: independent predictors Event Odds ratio and 95% CI Decrease of consciousness 6.4 (2.6–15.7) Hypotension 2.5 (1.6–4.1) Loss of consciousness 6.4 (2.9–13.6) Bradypnea 14.4 (2.6–80.0) SaO2 <90% 2.4 (1.6–4.1) Tachypnea 7.2 (3.9–13.2)

outcome. Patients receiving a lower level of care than desirable also had an increased mortality (P < 0.01). Other risk factors that have been studied as predictive patient severe adverse event include age,35 staff lack of awareness36 or, in contrast, that the staff at the bedside were worried about a patient.37 Based on these and other studies,38–43 various scoring systems have been developed with the aim of identifying patients who are at risk of having a serious adverse event. These systems fall into two major categories: simple single parameter systems and aggregate weighted systems. The most common example of the single parameter system is the activation criteria for the medical emergency team. In this system, if any of the single parameter observations broach the call criteria threshold, then the team should be called to the patient. In contrast, the aggregate weighted systems have a score assigned to each level of abnormality for the various physiological observations. The emergency response team is activated based on the total score that a patient gets across all the individual weighted physiological observations scores. These systems have all been reviewed and critically evaluated.44–49 The ideal system should be able to identify the majority of patients at risk of a sentinel event (i.e., high sensitivity) and at the same time not result in activation of an emergency response team to patients who, although they fulfill activation criteria, will not subsequently have a sentinel event (i.e., high specificity). Unfortunately, all of these systems have relatively low sensitivity, specificity and positive predictive value. This is not surprising given that the overall incidence of the type of sentinel event of interest (ICU/HDU admission, unexpected in-hospital death and cardiac arrest) occurs with an incidence of 5% or less and that some of these events will be genuinely sudden and unpredictable. All of these studies, which sought to identify factors that may predict unexpected iatrogenic hospital death (see Table  10.2 for summary), have major limitations, which limit the generalizability of the findings to some extent. All of the data in these studies have had to be collected manually from existing (paper) records. If the variables of interest are not documented, they will not be included in the analysis. Furthermore, all of the studies have collected data at some point, or, at best, over a limited timeframe, thus allowing for observation bias and Hawthorne effect.50 The time delay from data collection and publication in most of the above studies is years, during which time the hospital latent conditions and patient case mix may have altered significantly; with the advent of electronic data collection of patient observations, a number of these limitations have been eliminated or at least ameliorated.51–53 The main advantage of electronic observation data capture is the ability of such

10  The Challenge of Predicting In-Hospital Cardiac Arrests and Deaths

101

Table 10.2  Risk factors for patient crisis The adverse event epidemiological studies   Increased age   Negligence   Operative procedure   Human error and cognitive failure   Diagnostic and treatment delay The retrospective case control studies   Hypotension   Tachypnea   Documentation of concern   Abnormal pulse   Conscious level The prospective studies   Hypotension   Bradypnea   Tachypnea   Oxygen desaturation   Conscious level   Receiving a lower level of care for illness state

systems to analyze large numbers of patient observations and generate specific patient population risk factors or rules to identify particular at-risk patients. In the aggregated multi-parameter weighted systems, an additional benefit of electronic data entry is the assurance that at the very least, the patient score has been correctly calculated.54 Finally, with the data held electronically, the ability to automatically communicate to the appropriate healthcare providers becomes available.55

Prevention of Futile Clinical Cycles with Hard Defenses Given the data concerning the incidence and risk factors for unexpected in-hospital deaths, a wide range of interventions have been proposed to prevent them. These include the medical emergency team (MET),30,56–58 outreach teams,59–61 the “intensive care unit without walls,”62,63 the “hospitalist,”64,65 the British Early Warning System,66–68 and various education systems for both under- and postgraduates.69,70 However, all of these strategies constitute soft defenses. As such, success or otherwise is dependent on human cognitive processes and abilities. An alternative methodology to address the issue of the “patient crisis” is to examine the processes surrounding and leading up to such an event. This is sometimes described as a clinical audit, and may be reported as a root cause analysis when it examines both circumstances and latent conditions. A case report of a death in hospital is presented in Box  1. During the entire period of the patient’s ­postoperative course, there was no Medical Emergency Team Call or consultation with the treating surgeon or on-call intensive care specialist. This death occurred in

102

M. Buist

a hospital where the MET had been in operation for over 4 years and where a full-time nurse-educator was employed to ensure optimal compliance, education and MET utilization. The issues that this death raised were crystallized in the letter of complaint that the family wrote to the state coroner that asked, among other things: 1. Why didn’t the resident doctor contact the surgeon who operated on ….., during the night if there were signs of distress, complications or a deterioration in his condition? 2. Why didn’t the resident doctor contact the registrar? Was there a registrar on duty during the night? 3. Why wasn’t an MET call put into place after 11.00 p.m. when …… blood pressure fell and remained low? What criteria or symptoms presented by a patient instigate the MET process? The expert witness appointed by the coroner made the obvious conclusion: Another important observation is that the patient fulfilled the criteria for activation of the MET for at least 14 h. I understand that XXXX Hospital had an MET at the time of the patient’s death and that these MET criteria were widely advertised and known throughout the hospital. I also understand that these criteria were attached to the back of the hospital medical officer’s ID card and thus easily available in case of doubt when faced with a sick patient. The timely activation of the hospital MET might have saved the patient’s life.

However, the real question remained: Why didn’t the extremely competent and experienced medical and nursing staff involved with this man’s care not call for some sort of help, even if it was not the MET?71 Why did the soft defenses fail? Some answers came from the detailed debriefing that took place with the involved staff. First, because the patient was discharged from recovery with a heart rate of 130 bpm, the junior medical staff assumed that the patient was “okay” from both the consultant surgeon and anesthetist’s point of view. They assumed that if the operating team were in any way unhappy with the patient’s condition, then the patient would have been transferred to the intensive care unit postoperatively. In fact, both the surgeon and anesthetist were unaware that the patient was discharged from recovery with an elevated heart rate that mandated an MET call. Secondly, despite the patient’s heart rate, he “looked okay.” In particular, the patient was ­sitting up having a cup of tea and talking to relatives early that evening. Thirdly, the nursing staff was further reassured by the fact that the junior medical staff attended promptly to their concerns about the patient and seemed to be managing the situation appropriately. Finally, it was a very busy night for all concerned. With the benefit of hindsight, everyone involved would have put out an MET call. The traditional hierarchal referral model of care is the manner in which most “western hospitals” manage these and other acute medical scenarios in the general ward ­setting. In this model, when a patient deteriorates, the bedside nurse invariably documents the deterioration and then has to make a decision on whether or not to communicate this information and to whom. If this communication occurs, it is usually to the most junior member of the treating medical team, often a junior ­doctor with less than 12 months’ experience. To receive this communication, the said doctor needs to be available, not distracted by other tasks, and then undertake the steps of reading an alpha numeric

10  The Challenge of Predicting In-Hospital Cardiac Arrests and Deaths

103

pager, and make a telephone call back to the nurse (assuming that the line is not busy). The degree to which this vital first information transfer step succeeds is very much dependent on the communication skills of the nurse and the junior doctor. Based on this information and a large number of hospital latent conditions existent at the time, the junior doctor will show up and make an assessment of the patient. Despite this assessment, in most instances, the junior doctor will lack the skills, confidence and experience to manage the situation.6–9 As such, the communication step and associated processes will be repeated with the medical team’s next most senior doctor, generally a speciality trainee registrar or fellow. This person, because of his position in the hospital hierarchy, is generally not in a position to stop what he is doing to attend the patient. This type of “medical middle-manager” in the hospital hierarchy is usually busy in the operating room or the emergency department, doing ward rounds, seeing outpatients or attending to his own education. Although generally more experienced, particularly in his chosen field of specialization, even at this level the registrar/fellow may suffer from the same inadequacies as the junior.72 However, at this level, there is at some point some actions (often by telephone orders) undertaken at the behest of the registrar/fellow. Some of these actions include patient assessment, investigation and management. In some instances, there are further referrals to sub-speciality units, and, at some point, the handing over of care to other teams of on-call doctors. At some point the consultant is usually contacted about the patient’s situation. This traditional hierarchal referral model of care, while arguably appropriate for more chronic outpatient medical conditions, is not well suited for the acutely ill patient in the general ward setting. For it to be successful in this setting, all of the following steps need to be taken: 1 . Timely response of all staff in a well-coordinated sequence 2. The correct diagnosis 3. The correct assessment of the patient’s severity is appropriately communicated 4. The appropriate actions are taken 5. Documentation of these steps occurs 6. The response is documented When one considers these “clinical futile cycles”73 and the complexities under which all of these “soft defenses” must operate, it is little wonder that there are not more unexpected iatrogenic in-patient deaths. Often the only factors that prevent greater mortality are the patient’s physiological resilience and the state of hypervigilance of a junior medical or nursing staff member, which has been shown to be an inadequate defense against organizational accidents.

Communication Technology as a Hard Defense With the advances in communication technology that exist and the fact that there are easily identifiable risk factors for unexpected iatrogenic patient death (Table 10.2), there is a need for a patient-centric system that communicates ­accurate

104

M. Buist

Table 10.3  Automated, escalated alert interventions for Box 1 Abnormal Alert Time observation status Alert recipient 15:00 Heart rate: 130 bpm 2 Receiving surgical wardJunior medical officer 18:00 Heart rate: 140 bpm 3 Surgical registrar 23:30

Heart rate:150 bpm SBP: <90 mmHg

4

METTreating surgeon and ICU specialist

Action required Patient reviewed within 3 hReturn to operating room? Urgent patient review within 1 h Urgent resuscitation

patient abnormal physiological data to the most appropriate personnel in a timely fashion. Such a system would require electronic entry of simple bedside observational data; it could then process these data to determine appropriate severity and communicate this to the most appropriate doctors and nurses. This would remove many of the subjective, human cognitive factors that so often either fail or simply don’t occur in the acute hospital setting when our patients become acutely unwell. For example, with respect to the case history in Box  1, such a system could interact with the treating clinical staff in the manner as outlined in Table 10.3. The alerts are graded according to the severity of the clinical observations, and allow for alert escalation if the appropriate management does not occur. Alert escalation can be configured individually to alert other more senior staff of “failed alerts” or to automatically page the MET/cardiac arrest teams when appropriate. Theoretically, such a system should significantly reduce important errors of cognition, delays in diagnosis and treatment, and as such reduce morbidity and mortality from adverse events, in particular cardiac arrest.

References 1. Adam D. The counting house. Nature. 2002;417(6892):898. 2. Kohn LT, Corrigan JM, Donaldson MS, eds. To Err Is Human: Building a Safer Health System. Washington: National Academy Press; 1999. 3. Reason J. Managing the Risks of Organizational Accidents. Aldershot: Ashgate Publishing Limited; 1998. 4. Hazards RJ. Defence and Losses in Managing the Risks of Organizational Accidents. Aldershot: Ashgate Publishing Limited; 1998. 5. Wilson R, Runciman W, Gibberd R, Bernadette T, Newby L, John D. The quality in Australian health care study. Med J Aust. 1995;163:458–471. 6. Harrison GA, Hillman KM, Fulde GWO, Jacques TC. The need for undergraduate education in critical care. Anaesth Intensive Care. 1999;27:53–58. 7. Taylor D. Undergraduate procedural skills training in Victoria: is it adequate? Med J Aust. 1997;171:22–25. 8. Buist M, Jarmolovski E, Burton P, McGrath B, Waxman B, Meek R. Can Interns manage clinical instability in hospital patients? A survey of recent graduates. Focus Health Prof Educ. 2001;3(3):20–28.

10  The Challenge of Predicting In-Hospital Cardiac Arrests and Deaths

105

9. Cooper N. Medicine did not teach me what I really needed to know. BMJ. 2003;327(suppl):S190. 10. Wilson R, Harrison B, Gibberd R, Hamilton J. An analysis of the causes of adverse events from the quality in Australian health care study. Med J Aust. 1999;170:411–415. 11. McQuillan P, Pilkington S, Allan A. Confidential enquiry into the quality of care before intensive care unit admission. BMJ. 1998;316:1853–1858. 12. Mcdonald R. New initiative to improve undergraduate teaching in acute care. Student BMJ. 2004;12:68. 13. Stevens DP. Finding safety in medical education. Qual Saf Health Care. 2002;11:109–110. 14. Buist M, Bellomo R. MET: the medical emergency team or the medical education team? Crit Care Resusc. 2004;6:83–91. 15. Brennan TA, Leape LL, Laird N, et al. Incidence of adverse events and negligence in hospitalised patients: results of the Harvard Medical Practice Study I. N Engl J Med. 1991;324:370–376. 16. Leape LL, Brennan TA, Laird N, et al. The nature of adverse events in hospitalised patients: results of the Harvard Medical Practice Study II. N Engl J Med. 1991;324:377–384. 17. Thomas EJ, Studdert DM, Burstin HR, et al. Incidence and types of adverse events and negligent care in Utah and Colorado. Med Care. 2000;38(3):261–271. 18. Davis P, Lay-Yee R, Briant R, Wasan A, Scott A, Schug S. Adverse events in New Zealand public hospitals: occurrence and impact. NZ Med J. 2002;115(1167):203–205. 19. Schein RM, Hazdat N, Pena M, et  al. Clinical antecedents to in-hospital cardiopulmonary arrest. Chest. 1990;98(6):1388–1392. 20. Franklin C, Mathew J. Developing strategies to prevent in-hospital cardiac arrest: analyzing responses of physicians and nurses in the hours before the event. Crit Care Med. 1994;22:244–247. 21. Smith AF, Wood J. Can some in-hospital cardio-respiratory arrests be prevented? A prospective survey. Resuscitation. 1998;37(3):133–137. 22. Buist MD, Jarmolowski E, Burton PR, Bernard SA, Waxman BP, Anderson J. Recognising clinical instability in hospital patients before cardiac arrest or unplanned admission to Intensive Care. A pilot study in a tertiary-care hospital. Med J Aust. 1999;171:22–25. 23. Hodgetts T, Kenward G, Ioannis V, Payne S, Castle N. The identification of risk factors for cardiac arrest and formulation of activation criteria to alert a medical emergency team. Resuscitation. 2002;54:125–131. 24. Goldhill DR, Sumner A. Outcome of intensive care patients in a group of British intensive care units. Crit Care Med. 1998;28(8):1337–1345. 25. Goldhill DR, McNarry AF, Hadjianastassiou VG, Tekkis PP. The longer patients are in hospital before intensive care admission the higher their mortality. Intensive Care Med. 2004;30(10):1908–1913. 26. Goldhill DR, McNarry AF, Mandersloot G, McGinley A. A physiologically-based early warning score for ward patients: the association between score and outcome. Anaesthesia. 2005;60:547–553. 27. Skrifvars MB, Nurmi J, Ikola K, Saarinen K, Castran M. Reduced survival following resuscitation in patients with documented clinically abnormal observations prior to in-hospital cardiac arrest. Resuscitation. 2006;70:215–222. 28. Kause J, Smith G, Prytherch D, Parr M, Flabouris A, Hillman K. A comparison of antecedents to cardiac arrests, deaths and emergency intensive care admissions in Australia and New Zealand, and the United Kingdom: the ACADEMIA study. Resuscitation. 2004;62(3):275–282. 29. Cretikos M, Chen J, Hillman K, Bellomo R, Finfer S, Flabouris A. The objective medical emergency team activation criteria: a case-control study. Resuscitation. 2007;73:62–72. 30. MERIT study investigators. Introduction of the medical emergency team (MET) system: a cluster randomised controlled study. Lancet. 2005;365:2091–2097. 31. Hillman KM, Bristow PJ, Chey T, et  al. Antecedents to hospital deaths. Intern Med J. 2001;31(6):343–348.

106

M. Buist

32. Bellomo R, Goldsmith D, Russell S, Uchino S. Postoperative serious adverse events in a teaching hospital: a prospective study. Med J Aust. 2002;176:216–218. 33. Buist M, Nguyen T, Moore G, Bernard S, Anderson J. Association between clinically abnormal bedside observations and subsequent in hospital mortality: a prospective study. Resuscitation. 2004;62:137–141. 34. Goldhill DR, McNarry AF. Physiological abnormalities in early warning scores are related to mortality in adult inpatients. Br J Anaesth. 2004;92(6):882–884. 35. Smith GB, Prytherch DR, Schmidt PE, et al. Should age be included as a component of track and trigger systems used to identify sick adult patients? Resuscitation. 2008;78:109–115. 36. Fuhrmann L, Lippert A, Perner A, Ostergaard D. Incidence, staff awareness and mortality of patients at risk on general wards. Resuscitation. 2008;77:325–330. 37. Santiano N, Young L, Hillman K, et al. Analysis of Medical Emergency Team calls comparing subjective to “objective” call criteria. Resuscitation. 2009;80:44–49. 38. Jacques T, Harrison G, McLaws M, Kilborn G. Signs of critical conditions and emergency responses (SOCCER): a model for predicting adverse events in the inpatient setting. Resuscitation. 2006;69:175–183. 39. Kellet J, Deane B, Gleeson M. Derivation and validation of a score based on hypotension, oxygen saturation, low temperature, ECG changes and loss of independence (HOTEL) that predicts early mortality between 15 min and 24 h after admission to an acute medical unit. Resuscitation. 2008;78:52–58. 40. Kellett J, Deane B. The simple clinical score predicts mortality for 30 days after admission to an acute medical unit. Q J Med. 2006;99:771–781. 41. Paterson R, MacLeod DC, Thetford D, et al. Prediction of in-hospital mortality and length of stay using an early warning scoring system: clinical audit. Clin Med. 2006;6:281–284. 42. Groarke JD, Gallagher J, Stack J, et al. Use of an admission early warning score to predict patient morbidity and mortality and treatment success. Emerg Med J. 2008;25:803–806. 43. Bell MB, Konrad D, Granath F, Ekbom A, Martling C. Prevalence and sensitivity of METcriteria in a Scandinavian University Hospital. Resuscitation. 2006;70:66–73. 44. Daly K, Esmonde L, Goldhill DR, Parry GJ, Rashidian A, Subbe CP. Systematic review and evaluation of physiological track and trigger warning systems for identifying at-risk patients on the ward. Intensive Care Med. 2007;33:667–679. 45. Smith GB, Prytherch DR, Schmidt PE, Featherstone PI. Review and performance evaluation of aggregate weighted “track and trigger” systems. Resuscitation. 2008;77:170–179. 46. Smith GB, Prytherch DR, Schmidt PE, Featherstone PI, Higgins B. A review, and performance evaluation of single-parameter “track and trigger” systems. Resuscitation. 2008;79:11–21. 47. Peberdy MA, Cretikos M, Abella BS, et  al. Recommended guidelines for monitoring, ­reporting, and conducting research on medical emergency team, outreach and rapid response systems: an Utstein-style scientific statement. Circulation 2007;116(21):2481–500. 48. Young D, Griffiths J. Clinical trials of monitoring in anesthesia, critical care and acute ward care: a review. Br J Anaesth. 2006;97(1):39–45. 49. Subbe CP, Gao H, Harrison DA. Reproducibility of physiological track-and-trigger warning systems for identifying at-risk patients on the ward. Intensive Care Med. 2007;33:619–624. 50. Campbell JP, Maxey UA. The Hawthorne effect: implications for pre-hospital research. Ann Emerg Med. 1995;26:590–594. 51. Mohammed MA, Hayton R, Clements G, Smith G, Prytherch D. Improving accuracy and efficiency of early warning scores in acute care. Br J Nurs. 2009;18(1):18–24. 52. Smith GB, Prytherch DR, Schmidt P, et al. Hospital-wide physiological surveillance: a new approach to the early identification and management of the sick patient. Resuscitation. 2006;71:19–28. 53. Hravna KM, Edwards L, Clontz A, Valenta C, DeVita M, Pinsky M. Defining the incidence of cardiorespiratory instability in patients in step-down units using an electronic integrated monitoring system. Arch Intern Med. 2008;168(12):1300–1308. 54. Prytherch DR, Smith GB, Schmidt P, et al. Calculating early warning scores: a classroom comparison of pen and paper and hand-held computer methods. Resuscitation. 2006;70:173–178.

10  The Challenge of Predicting In-Hospital Cardiac Arrests and Deaths

107

55. Jones S, Eddelstone J, Buist M, et al. Getting a clinical response for the unstable in-hospital medical patient: The use of bedside electronic capture of observations and automated clinical alerts to improve compliance with an NHS Trust Early Warning Score (EWS) protocol. Abstract book International Forum on Quality and Safety. Berlin; 2009. 56. Bristow PJ, Hillman KM, Chey T, et al. Rates of in-hospital arrests, deaths and intensive care admissions: the effect of a medical emergency team. Med J Aust. 2000;173:236–240. 57. Buist MD, Moore GE, Bernard SA, Waxman BP, Anderson JN, Nguyen TV. Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. BMJ. 2002;324:1–6. 58. Bellomo R, Goldsmith D, Uchino S, et al. A prospective before-and-after trial of a medical emergency team. Med J Aust. 2003;179(6):283–287. 59. Goldhill D, Mcginley A. Outreach critical care. Anaesthesia. 2002;57(2):183. 60. Goldhill DR. Preventing surgical deaths: critical care and intensive care outreach services in the post-operative period. Br J Anaesth. 2005;95(1):88–94. 61. Garcea G, Thomasset S, McClelland L, Leslie A, Berry D. Impact of a critical care outreach team on critical care readmissions and mortality. Acta Anaesthesiol Scand. 2004;48(9):1096–1100. 62. Hillman K. Critical care without walls. Curr Opin Crit Care. 2002;8(6):594–599. 63. Hillman K. Expanding intensive care medicine beyond the intensive care unit. Crit Care. 2004;8(1):9–10. 64. Hillman K. Hospitals and hospitalists: an alternative view. Med J Aust. 2000;172(6):299. 65. Hillman K. The hospitalist: a US model ripe for importing? Med J Aust. 2003;178(2):54–55. 66. Goldhill DR. The critically ill: following your MEWS. QJM. 2001;94(10):507–510. 67. Subbe CP, Davies RG, Williams E, Rutherford P, Gemmell L. Effect of introducing the Modified Early Warning Score on clinical outcomes, cardio–pulmonary arrests and intensive care utilization in acute medical admissions. Anaesthesia. 2003;58(8):797–802. 68. Cuthbertson BH. Outreach critical care: cash for no questions. Br J Anaesth. 2003;90(1):4–6. 69. The Royal College of Surgeons of England and the Royal College of Surgeons of Australasia. Care of the Critically Ill Surgical Patient. London: Arnold; 2003. 70. Harrison J, Buist M, Flanagan B. A simulation-based MET training course for Physician Trainees at Southernhealth. Abstract book, the 9th National forum for pre-vocational medical education. Melbourne, October 24–26th; 2004. 71. Buist M. The rapid response team paradox: why doesn’t anyone call for help? Crit Care Med. 2008;36(2):634–636. 72. Harrison J, Flanagan B. What are PGY3’s worried about? Exploration of physician trainees concerns regarding their resuscitation skills. Abstract book, the 9th National forum for prevocational medical education. Melbourne, October 24–26th; 2004. 73. Buist M, Harrison J, Abaloz E, Van Dyk S. Six-year audit of cardiac arrests and medical emergency team calls in an outer metropolitan teaching hospital. BMJ. 2007;335(7631):1210–1212.

Chapter 11

The Meaning of Vital Signs John Kellett, Breda Deane, and Margaret Gleeson

Keywords  Meaning • Vital • Signs • Pulse • Rate

Introduction The four classic vital signs are respiratory rate, temperature, pulse rate and blood pressure. Although their measurement has been standard practice for over a century, there have been few attempts to quantify their clinical performance. Until recently, the largest study of respiratory rate was performed by Hutchinson in 18461 and the largest studies on fever remain those performed by Wunderlich in the nineteenth century.2 Amazingly, the ominous significance of low temperatures has only recently been appreciated,3,4 and the mortality risk associated with transient hypotension only reported for the first time in 2006.5 The “normal” range for a medical test has been traditionally defined as being within two standard deviations of the mean of the healthy population. However, this does not mean that all values outside the normal range are necessarily associated with illness or an increased risk of death. Conversely, values within the normal range may occasionally be associated with illness and death if the patient is sick. While occasionally healthy individuals can present with a persistent abnormality of one vital sign and remain well, it is most unusual for healthy patients to have two or more vital signs persistently outside the normal range. Yet, it was not until 1966 that the prognostic significance of the relationship between a high heart rate and low blood pressure (i.e., the Shock Index) was recognized.6 Pulse oximetry has been proposed as the fifth7,8 and altered mental status as the sixth vital sign.9 The brain is as sensitive and vital an organ as the immune (temperature), cardiac (pulse, blood pressure) and respiratory system (respiratory rate and oxygen saturation) for heralding serious illness. While florid dementia and delirium may be obvious, mild alteration of mental status is often not noticed,

J. Kellett (*) Department of Medicine, Nenagh Hospital, Nenagh, County Tipperary, Ireland e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_11, © Springer Science+Business Media, LLC 2011

109

110

J. Kellett et al.

thought to be a normal part of the ageing process or, in the intoxicated young, not taken seriously.10 The cardinal features of impaired cognition are inattention and altered level of consciousness (i.e., agitation, lethargy, stupor and coma) and can be quickly assessed.11 On the other hand, fluctuations in confusion or disorganized thinking require prolonged observation and may, therefore, not be appreciated during the initial emergency room encounter. From February 17, 2000, to January 29, 2004, all medical patients admitted to our hospital had their histories and physical data entered into a computerized database from which a clinical score that predicted mortality for up to 30 days after hospital admission was developed.12 This study was one of the largest ever undertaken of the vital signs of unselected acutely ill medical patients. Even so, the number of patients with abnormal vital signs was relatively small. For example, out of nearly 10,000 patients studied, only 77 patients had a fever over 39°C, only 22 a pulse below 40 beats per minute, and only 11 a respiratory rate below 12 breaths per minute. However, abnormalities of both oxygen saturation and respiratory rate were frequent, especially so in the elderly (see Table 11.1). Although most reports consider altered mental status to be largely confined to the elderly,10 in our experience it was more frequent in the young, doubtless a consequence of the high rate of admission of alcohol and drug intoxication to our hospital. Although such patients are disruptive and consume a considerable amount of time and attention none of them died while in hospital. On the other hand, any degree of altered mental status was associated with death in older patients.12

Pulse Rate Sir William Osler, while neither defining bradycardia nor tachycardia, remarked that a slow pulse was sometimes normal and that Napoleon had a pulse rate of only 40 beats per minute.13 The accepted limits for heart rate have long been set at between 60 and 100 beats per minute14; however, these recommendations are “traditional” and not based on any systematic clinical studies.15 Although the lower and upper limit of normal for the pulse rate of 60 and 100 beats per minute was “generally agreed” and endorsed by the New York Heart Association in 1953,14 they were never confirmed by any systematic clinical investigations. Only relatively recently have Spodick et al reported that the mean heart rate for men and women was 70–75 beats per minute, with two standard deviation limits of 43 and 93 beats per minute in men, and 52 and 94 beats per minute for women. There was no significant change from ages 50 and 80.15 Nevertheless it is still common practice for beta-blockers to be withheld, even in those with a recent myocardial infarction, if the pulse falls below 60 beats per minute. Moreover, sinus bradycardia occurs in normal children16 and adults,17,18 particularly during sleep, when rates of 30 beats per minute and pauses of up to 2 s are not uncommon.17,19 Bradycardia may also be seen in the absence of heart disease in the elderly,20 has no adverse effect on longevity21 and is of no prognostic significance in otherwise healthy subjects. We observed a

11  The Meaning of Vital Signs Table 11.1  Proportion of patients with vital sign abnormalities by age Age (years) ³14 and <48 ³48 and <68 ³68 and <78 ³78 (n = 2,449) (n = 2,462) (n = 2,422) (n = 2,631) (%) (%) (%) (%)   5.2   4.3   5.9   7.6 Blood pressure < 100 mmHg Heart rate <40 or >90 42.1 33.7 37.3 35.1 beats per minute Oxygen saturation < 95% 10.0 18.1 27.0 29.6 Respiratory rate > 22 16.0 19.9 27.3 27.6 breaths per minute Temperature < 35°C or   2.6   2.3   4.0   2.6 ≥39°C Shock Index ≥ 0.9 12.9   8.3   9.9 10.5 11.2   5.7   4.5   6.7 Altered mental status

111

All patients (all ages) (n = 9,964) (%)   5.8 37.0 21.3 22.8   2.6 10.4   6.9

mortality rate of 9.1% in the 22 patients who presented with a heart rate below 40 beats per minute. However, above this heart rate, mortality remains constant up to 90 beats per minute, after which it gradually increases to 6.6% at 120 beats per minute and 12.8% at greater than 130 beats per minute.

Blood Pressure Hypotension has been defined as a systolic blood pressure below 90 mmHg or a reduction of more than 40 mmHg from the patient’s usual pressure.22 The obvious problem with this definition is determining the patient’s “usual” blood pressure in an emergency. In a study of 815 healthy subjects, the two standard deviation lower limit of systolic blood pressure over 24 h was 102 mmHg in men and 95 mmHg in women; overnight this fell to 90 mmHg in men and 84 mmHg in women.23 Low blood pressure, therefore, is not a reliable indicator of severe illness and, on its own, does not indicate inadequate cardiac output, intravascular volume or peripheral perfusion unless other symptoms and signs are present. However, for “sick” patients requiring hospitalization, even a transient reduction of systolic blood pressure below 100  mmHg increases the risk of death.5 Ironically, once blood pressure drops below 100 mmHg, it is difficult to detect manually both by auscultation and palpation, and is unreliably detected by the electronic devices that are replacing the traditional mercury sphygmomanometer. Most of these automated devices rely on a piezo-electric crystal that measures arterial waveforms distal to a deflating cuff to obtain systolic and diastolic pressures. Systolic blood pressure is taken at the appearance of a pulsatile signal, while diastolic blood pressure is calculated using a proprietary formula.24 Many such devices have been found to be inaccurate at systolic blood pressures below 120  mmHg.25 Therefore, all low blood pressure readings (i.e., systolic blood pressure below 120  mmHg) should be carefully

112

J. Kellett et al.

v­ erified. Indeed, once the pulse can no longer be felt at the wrist, the systolic blood pressure probably cannot be measured accurately noninvasively.26 Although several severity-of-illness scores3,27–31 consider hypertension a risk factor, we found no level of high blood pressure to be associated with an increased risk of death. There was an association, however, between hypertension and mortality if mental status was altered (see Fig. 11.1). This is in keeping with the report by Ikeda et al that an intracranial lesion was more likely if patients with impaired consciousness had a systolic blood pressure over 170  mmHg.32 Moreover, the Cushing response of bradycardia with hypertension is a well-recognized clinical manifestation of increased intra-cranial pressure.33

The Shock Index The Shock Index – the ratio between pulse rate and systolic blood pressure (HR/ SBP) – was first described by Allgower and Buri in 19676 and has a normal range from 0.5 to 0.7 in healthy adults. It is elevated in acute hypovolemia and left ventricular dysfunction and correlates with left ventricular end-diastolic pressure.34–36 Hypotension and tachycardia are common in metabolic brain dysfunction caused by, for example, intoxication, endocrine disease and sepsis.37 The Shock Index has 60% Normal mental status

30-day mortality (%)

50%

Altered mental status

40%

30%

20%

10%

0

9

21

20

>=

9 20

0

to

19

9 19

0

to

9

18

17 18

0

to

9 16 17

0

to

9 16

0

to

15

9 15

0

to

14

9 14

0

to

9

13

12 13

0

to

9 11 12

0

to

9 10 11

0

to

99 10

0

to

89

to 90

to

80

<8

0

0%

Systolic blood pressure (mmHg) Fig. 11.1  30-day mortality by systolic blood pressure of patients with normal and altered mental status. Data from 9,964 consecutive unselected acute medical patients admitted to Nenagh Hospital (for Methods, etc., see 12)

11  The Meaning of Vital Signs

113

been used to predict the outcome of ectopic pregnancy,38 traumatic injury,39 sepsis, gastrointestinal hemorrhage40 and pulmonary embolus.41 Persistent elevation of the Shock Index over 1.0 for several hours following trauma or acute circulatory failure has been related to a poor outcome.42 While a pulse rate below 100 beats per minute with a systolic blood pressure over 110 mmHg and a respiratory rate of 16 breaths per minute indicates a blood loss of less than 750 ml, a pulse rate over 140 beats per minute in association with a systolic blood pressure below 90  mmHg and a respiratory rate over 26 breaths per minute indicates Class IV shock, or a blood loss of over 2,000 ml.43 We found mortality increased from 3.9% below a Shock Index of 0.9–34.2% over a Shock Index of 1.5. Once the pulse rate exceeds systolic blood pressure (i.e., Shock Index > 1.0), the odds ratio for 30-day mortality was 4.0 (p < 0.0001), and the overall risk of death is 14.5% (see Fig. 11.2).

Temperature More than 120 years ago, Wunderlich’s million observations of 25,000 subjects defined the “normal” human temperature as 37°C. Unfortunately, recent tests suggest that his thermometers may have been calibrated as much as 1.4–2.2°C 40%

2500 30-day mortality

35%

Number of patients 2000

25%

1500

20% 1000

15%

Number of patients

30-day mortality (%)

30%

10% 500 5% 0%

0

0.

0-

0.

09

19

0.

10

. -0

20

0.

29

. -0

30

0.

39

. -0

40

0.

49

. -0

50

0.

59

. -0

60

0.

69

. -0

70

0.

79

. -0

80

0.

89

. -0

99

. -0

90

0.

00

1.

09

. -0

10

1.

19

. -1

20

1.

29

. -1

30

1.

39

. -1

40

1.

50

49

. -1

. =1

>

Shock Index Fig. 11.2  30-day mortality by Shock Index (i.e., pulse rate in beats per minute divided by systolic blood pressure in mmHg). Data from 9,964 consecutive unselected acute medical patients admitted to Nenagh Hospital (for Methods, etc., see 12)

114

J. Kellett et al.

higher than today’s instruments.44 More recently, Mackowiak et  al reported that 36.8 ± 0.4°C was the normal range of oral temperature, and the 37.2 and 37.7°C were the upper limits of normal morning and evening temperature, respectively.2 Body temperature varies at different parts of the body. In 2002 Sund-Levander et al defined a wide normal range of oral temperature from 33.2 to 38.2°C.45 Tympanic temperature, however, had a narrower normal range from 35.4 to 37.8°C. Since the time of Florence Nightingale, doctors and nurses have feared fever and responded to it energetically. As a result even tiny rises in temperature can result in immediate treatment, which often includes the inappropriate use of antibiotics. Such aggressive and, albeit inappropriate treatment notwithstanding, we noted very little increase in the risk of mortality until fever exceeded 39°C. In contrast, a low temperature was much more ominous. We found the lowest mortality rate (3.5%) was associated with a temperature between 36.5 and 36.9°C – below this temperature mortality increased exponentially with each half-degree fall in temperature to reach 6.1% between 35.0 and 35.4°C. Once the temperature fell below 35°C, 30-day mortality rapidly rose to 12% or more (see Fig. 11.3).

20%

3500 30-day mortality

18%

Number of patients

3000

2500

14% 12%

2000

10% 1500

8% 6%

1000

Number of patients

30-day mortality (%)

16%

4% 500

2% 0%

<3 4. 0 34 .0 -3 4. 34 4 .5 -3 35 4.9 .0 -3 35 5.4 .5 -3 36 5.9 .0 -3 6. 4 36 .5 -3 6. 37 9 .0 -3 37 7.4 .5 -3 7. 38 9 .0 -3 38 8.4 .5 -3 8. 39 9 .0 -3 39 9.4 .5 -3 9. 9

0

Temperature (˚C) Fig. 11.3  30-day mortality by tympanic temperature (°C). Data from 9,964 consecutive unselected acute medical patients admitted to Nenagh Hospital (for Methods, etc., see 12)

11  The Meaning of Vital Signs

115

Respiratory Rate While only 2.5% of our patients had a temperature abnormality, more than a quarter of them had either hypoxia and/or tachypnea. Respiratory rate is a powerful predictor of disease severity and of a poor outcome.46 It has been called the vexatious clinical sign because it requires patience and diligence to measure accurately and cannot be performed reliably by a machine.47 Although many authors have defined the “normal” respiratory rate to be from 8 to 18 breaths per minute, there are few studies to support their conclusions. In the studies that are available, the mean respiratory rate is between 16 and 20.5 breaths per minute and only one of these studies was performed on emergency patients.48 The largest published study was probably that of Hutchinson in 1846,1 who reported a mean respiratory rate in 1,714 adult males of 20.2 with a range of 6–40 breaths per minute – patients with pneumonia had a mean rate of 28 breaths per minute.49 In 1959, Mead measured the respiratory rate in 75 adults seated at a public gathering and found the mean to be 16 with a range between 11 and 26.50 In 1982, McFadden found the normal respiratory rate in 82 hospitalized elderly adults to be 20.5 breaths per minute.51 In 1988, Hooker et al48 reported the “normal” respiratory rates in 110 emergency department patients as 20.1 ± 4.0 breaths per minute – almost identical to our findings.12 For any given level of alveolar ventilation, there is an optimum respiratory rate at which the muscular work of breathing is minimal.52,53 In normal subjects, this “theoretical” optimum frequency was predicted by Otis et al to be 15 breaths per minute.52 However, when diseases such as pneumonia modify the elastic and airway properties of the lung, the respiratory frequency at which minimum work is performed for the required level of alveolar ventilation changes.54

Oximetry Although developed in 1974,55 many newly qualified doctors are still unaware of the normal range of oxygen saturation and do not know how to investigate or adequately manage patients with low values.56 Patients cannot by an act of will lower their oxygen saturation below 95% at sea level and mortality rises dramatically once oxygen saturation falls below 90% (see Fig.  11.4). Oximetry requires little skill or training and rapidly measures both oxygen saturation and pulse rate – if neither is detectable it is likely that the patient has poor peripheral circulation and is either suffering from hypovolemia or heart failure. In addition, the response of both oxygen saturation and pulse rate to exercise and change in posture can be quickly assessed. A rise in heart rate of more than 30 beats per minute on standing indicates the presence of hypovolemia,57 while a fall in oxygen saturation with exercise indicates that serious lung or heart disease is likely,58–61 especially in the elderly.58

116

J. Kellett et al. 35%

4500 30-day mortality Number of patients

30%

4000

3000 20%

2500

15%

2000 1500

10%

Number of patients

30-day mortality (%)

3500 25%

1000 5%

500

-7 70

<7

72 1 -7 3 74 -7 5 76 -7 7 78 -7 9 80 -8 1 82 -8 3 84 -8 5 86 -8 7 88 -8 9 90 -9 1 92 -9 3 94 -9 5 96 -9 7 >= 98

0

0

0%

Oxygen saturation (%) Fig.  11.4  30-day mortality by oxygen saturation (%). Data from 9,964 consecutive unselected acute medical patients admitted to Nenagh Hospital (for Methods, etc., see 12)

Age and Vital Signs The proportion of patients with an abnormal pulse rate, temperature and blood pressure was more or less constant across all age groups, while an abnormal respiratory rate and/or low oxygen saturation was present in 10–15% of younger patients, and increased to nearly 30% in older patients (see Table 11.1). The mortality associated with all vital sign abnormalities is greatly influenced by age. None of our patients younger than 48 years with a blood pressure below 100 mmHg died, compared with 24% of those 78 years or older. With the possible exception of a pulse rate over 140 beats per minute, no vital sign abnormality predicted death in younger patients (see Table  11.2). Since 59% of young patients had at least one abnormal vital sign, compared with 46% of those over 78 years of age, it cannot be argued that younger patients were “less sick” than older ones. Similar dramatic age-related increases in mortality associated with all vital signs abnormalities have been observed in hospitalized patients elsewhere.62 This may be attributable to the natural physiology of illness in the elderly, or it may be that older patients require more rapid correction of vital sign abnormalities than is currently provided.

Table 11.2  Mortality associated with vital signs according to age Age (Years) ³14 and <48 (%) ³48 and <68 (%) Systolic blood pressure   >100 mmHg 0.2 1.9   >80 and £100 mmHg 0.0 3.2   ³70 and £80 mmHg 0.0 20.0   <70 mmHg 0.0 50.0 Pulse rate   <40 beats per minute 0.0 0.0   ³40 and <90 beats per 0.1 1.2 minute   ³90 and <130 beats per 0.1 3.8 minute   ³130 beats per minute 2.3 2.9 Oxygen saturation (%)   ³95% 0.2 1.4   <95% and ³90% 0.0 3.9   <90% 0.0 8.8 Respiratory rate   <12 breaths per minute 0.0 0.0   ³12 and <22 breaths per 0.1 1.8 minute 0.3 3.2   ³22 and <30 breaths per minute   ³30 breaths per minute 0.0 4.3 Temperature   ³35 and <39°C 0.2 2.0   <35°C 0.0 4.9   ³39°C 0.0 12.5 ³78 (%) 8.5 20.1 30.0 66.7 16.7 7.7 11.3 35.7 7.6 11.6 21.5 0.0 7.4 14.4 23.3 9.3 15.7 33.3

³68 and <78 (%) 5.7 14.2 22.2 33.3 11.1 5.1 7.9 12.5 4.8 7.4 16.8 50.0 5.2 8.7 13.7 5.8 27.3 28.6

4.4 13.6 13.0

13.8

7.8

9.1 3.5

3.3 7.2 15.8

12.8

5.7

9.1 3.7

4.1 10.7 23.4 44.4

All patients

(continued)

11  The Meaning of Vital Signs 117

Shock Index (Pulse/SBP)   Shock Index < 0.9   Shock Index ³ 0.9 Altered mental status   Altered   Normal Overall mortality

Table 11.2  (continued)

1.7 6.4 4.3 1.9 2.1

0.7 0.1 0.2

³48 and <68 (%)

0.1 1.0

Age (Years) ³14 and <48 (%)

18.3 5.8 6.3

5.5 14.2

³68 and <78 (%)

17.5 9.1 9.6

7.9 24.4

³78 (%)

8.3 4.4 4.6

3.9 11.3

All patients

118 J. Kellett et al.

11  The Meaning of Vital Signs

119

Combining Vital Signs In 1997, Morgan et al first reported an early warning score (EWS) that combined a variety of vital sign and mental status abnormalities to help detect acutely ill patients who were seriously ill and likely to deteriorate.63 Since then several ­modifications of this EWS have been developed and adopted into clinical practice. These scores were developed empirically from expert opinion that selected the standard vital signs and changes in mental status and allocated scores to them ­without any prospective validation.4 As a result, the discriminative performance of these scores is poor, with the area under their receiver operator curve (AUROC) ranging from 64 to 80%.64 Some scores, derived from data analysis, have reported far higher discrimination with AUROC of 85–90%.12,28,65 However, these were developed and/or validated in-patient populations that may not be representative of patients elsewhere.66 The major concern, therefore, with all EWS systems is that although they may perform satisfactorily in the local environments for which they were developed, it is not known if they have universal applicability to other patient  populations. Also, although combining vital signs to calculate EWS can identify the “patient-at-risk” on initial clinical assessment, their relative complexity and susceptibility to calculation error limits their use as monitoring tools for the ­detection of clinical deterioration.67 This issue has been addressed by the ­introduction of electronic data entry and decision support systems.68 Recently a computerized ­system has used a neural network to combine changes in heart rate, blood pressure, respiratory rate and oxygen saturation to determine a single ­integrated monitoring index that successfully detects early “preclinical” signs of cardio-respiratory instability.69

Summary Although measured routinely for many years, there is still uncertainty as to the “normal” values of many of the classic vital signs, and certainly considerable doubt as to what change in vital signs identifies the patient-at-risk and the deteriorating patient. Some vital sign abnormalities are more common than are others – abnormal temperature and low blood pressure are relatively rare, whereas hypoxia, increased respiration and abnormal heart rates are common. Serious illness is more likely to be present when there is a combination of abnormal vital signs rather than a single abnormality. Furthermore, changes in vital signs cannot be fully interpreted unless the age and other clinical features, such as mental status of the patient, are known. It may well be that relative changes in vital signs to each other over time, rather than absolute values, will prove to be the most powerful indictors of clinical deterioration. Vital signs are the simplest and probably the most important information gathered on hospitalized patients, yet they are rarely stored anywhere except as a paper chart at the foot of the bed. Recently, however, several clinicians have begun to develop large electronic databases of vital signs and related clinical data, the analy-

120

J. Kellett et al.

sis of which will allow greater insights into the physiology of acute illness.68,70 Such research, after almost a century of neglect, will ensure that the continuous monitoring and competent interpretation of vital signs remains a pivotal part of medical care. Acknowledgement  The authors would like to acknowledge the help of all the medical, nursing and administrative staff of Nenagh Hospital in the collection of data that made this paper possible. In particular, we would like to thank Mrs. Marie Kennedy for her meticulous help.

References 1. Hutchinson J. On the capacity of the lungs, and on the respiratory functions with a view to establishing a precise and easy method of detecting disease by the spirometer. Med Chir Trans. 1846;29:137–252. 2. Mackowiak PA, Wasserman SS, Levine MM. A critical appraisal of 98.6°F, the upper limit of the normal body temperature and other legacies of Carl Reinhold August Wunderlich. JAMA. 1992;268(12):1578–1580. 3. Subbe CP, Kruger M, Rutherford P, Gemmel L. Validation of a modified early warning score in medical admissions. QJM. 2001;94(10):521–526. 4. Duckitt RW, Buxton-Thomas R, Walker J, et al. Worthing physiological scoring system: derivation and validation of a physiological early-warning system for medical admissions. An observational, population-based single-centre study. Br J Anaesth. 2007;98(6):769–774. 5. Jones AE, Yiannibas V, Johnson C, Kline J. Emergency department hypotension predicts sudden unexpected in-hospital mortality. Chest. 2006;130(4):941–946. 6. Allgower M, Buri C. Shock index. Dtsch Med Wochenschr. 1967;46:1–10. 7. Mower WR, Myers G, Nicklin EL, Kearin KT, Baraff LJ, Sachs C. Pulse oximetry as a fifth vital sign in emergency geriatric assessment. Acad Emerg Med. 1998;5(9):858–865. 8. Neff T. Routine oximetry. A fifth vital sign? Chest. 1988;94(2):227. 9. Flaherty JH, Rudolph J, Shay K, et al. Delirium is a serious and under-recognized problem: why assessment of mental status should be the sixth vital sign. J Am Med Dir Assoc. 2007;8(5):273–275. 10. Meagher DJ. Delirium: optimising management. BMJ. 2001;322(7279):144–149. 11. Inouye SK, van Dyck CH, Alessi CA, Balkin S, Siegal AP, Horwitz RI. Clarifying confusion: the confusion assessment method: a new method for detecting delirium. Ann Int Med. 1990;113(12):941–948. 12. Kellett J, Deane B. The simple clinical score predicts mortality for 30 days after admission to an acute medical unit. QJM. 2006;99(11):771–781. 13. Osler W. Neuroses of the heart. In: Osler W, ed. The Principles and Practice of Medicine. Edinburgh: Young J Pentland; 1901:759. 14. Criteria Committee of the New York Heart Association. Nomenclature and Criteria for Diagnosis of Diseases of the Heart. 5th ed. New York: New York Heart Association; 1953. 15. Spodick DH, Raju P, Bishop RL, Rifkin RD. Operational definition of normal sinus heart rate. Am J Cardiol. 1992;69(14):1245–1246. 16. Scott O, Williams GJ, Fiddler GI. Results of 24-hour ambulatory monitoring of electrocardiogram in 131 healthy boys aged 10 to 13 years. Br Heart J. 1980;44:304–308. 17. Brodsky M, Wu D, Denes P, Kanakis C, Rosen KM. Arrhythmias documented by 24-hour continuous electrocardiographic monitoring in 50 male medical students without apparent heart disease. Am J Cardiol. 1977;39(3):390–395. 18. Bjerregaard P. Mean 24-hour heart rate, minimal heart rate, and pauses in healthy subjects 40–79 years of age. Eur Heart J. 1983;4(1):44–51.

11  The Meaning of Vital Signs

121

19. Hilgard J, Ezri MD, Denes P. Significance of ventricular pauses of three seconds or more detected on 24-hour Holter recordings. Am J Cardiol. 1985;55(8):1005–1008. 20. Agruss NS, Rosin EY, Adolph RJ, Fowler NO. Significance of chronic sinus bradycardia in elderly people. Circulation. 1972;46(5):924–930. 21. Tresch DD, Fleg JL. Unexplained sinus bradycardia: clinical significance and long-term prognosis in apparently healthy persons older than 40 years. Am J Cardiol. 1986;58(10):1009–1013. 22. Bone RC, Balk RA, Cerra FB, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest. 1992;101(6):1644–1656. 23. O’Brien E, Murphy J, Tyndall A, et al. Twenty-four-hour ambulatory blood pressure in men and women aged 17 to 80 years: the Allied Irish Bank Study. J Hypertens. 1991;9(4):355–360. 24. Wonka F. Oscillometric blood pressure measurement: description of the method used. Z Kardiol. 1996;85:1–7. 25. Beaubien ER, Card CM, Card SE, Biem HJ, Wilson TW. Accuracy of the Dinamap 1846 XT automated blood pressure monitor. J Hum Hypertens. 2002;16(9):647–652. 26. Deakin CD, Low JL. Accuracy of the advanced trauma support guidelines for predicting systolic blood pressure using carotid, femoral and radial pulses: observational study. BMJ. 2000;321(7262):673–674. 27. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med. 1985;13(10):818–829. 28. Olsson T, Lind L. Comparison of the rapid emergency medicine score and APACHE II in non-surgical emergency department patients. Acad Emerg Med. 2003;10(10):1040–1048. 29. Rhee KJ, Fisher CJ, Willitis NH. The rapid acute physiology score. Am J Emerg Med. 1987;5(4):278–286. 30. Le Gall J-R, Lemeshow S, Saulnier F. A new simplified acute physiology score ­(SAPS II) based on a European/North American multicenter study. JAMA. 1993;270(24):2957–2963. 31. Lemeshow S, Teres D, Klar J, Avrunin JS, Gehlbach SH, Rapport J. Mortality probability models (MPM II) based on an international cohort of intensive care patients. JAMA. 1993;270(20):2478–2486. 32. Ikeda M, Matsunaga T, Irabu N, Yoshida S. Using vital signs to diagnose impaired consciousness: cross-sectional observational study. BMJ. 2002;325(7378):800–805. 33. Guyton AC, Hall JE. Textbook of Medical Physiology. 9th ed. Philadelphia: W.B. Saunders; 1996:209–220. 34. Rady MY, Nightingale P, Little RA, Edwards JD. Shock index: a re-evaluation in acute respiratory failure. Resuscitation. 1992;23(3):227–234. 35. Rady MY. The role of central venous oximetry, lactic acid concentration and shock index in the evaluation of clinical shock: a review. Resuscitation. 1992;24(1):55–60. 36. Rady MY, Rivers EP, Martin GB, et al. Continuous central venous oximetry and shock index in the emergency department: use in the evaluation of clinical shock. Am J Emerg Med. 1992;10(6):538–541. 37. Victor M, Ropper AH. Coma and related disorders of consciousness. In: Adams and Victor’s Principles of Neurology. 7th ed. New York: McGraw-Hill; 2001:366–389. 38. Birkhahn RH, Gaeta TJ, Bei R, Bove JJ. Shock index in the first trimester of pregnancy and its relationship to ruptured ectopic pregnancy. Acad Emerg Med. 2002;9(2):115–119. 39. King RW, Plewa MC, Buderer NM, Knotts FB. Shock index as a marker for significant injury in trauma patients. Acad Emerg Med. 1996;3(11):1041–1045. 40. Rady MY, Smithline HA, Blake H, Nowak R, Rivers E. A comparison of the shock index and conventional vital signs to identify acute critical illness in the emergency department. Ann Emerg Med. 1994;24(4):685–690. 41. Kline JA, Nelson RD, Jackson RE, Courtney DM. Criteria for the safe use of d-dimer testing in emergency department patients with suspected pulmonary embolism: a multicenter US study. Ann Emerg Med. 2002;39(2):144–152.

122

J. Kellett et al.

42. Ostern HJ, Trentz O, Hemplemann G, Trentz OA, Sturm J. Cardiorespiratory and metabolic patterns in multiple trauma patients. Resuscitation. 1979;7(3–4):169–183. 43. Zimmerman JL. In: Taylor RW, Delliner RP, Farmer JC, eds. Fundamental Critical Care Support. 3rd ed. Society of Critical Care Medicine Des Plaines: Illinois; 2001:4–9. 44. Mackowiak PA. Concepts of fever. Arch Intern Med. 1998;158(7):1870–1881. 45. Sund-Levander M, Forsberg C, Wahren LK. Normal oral, rectal, tympanic and axillary body temperature in adult men and women: a systematic literature review. Scand J Caring Sci. 2002;16(2):122–128. 46. Cretikos M, Chen J, Hillman K, et al. The objective medical emergency team activation criteria: a case-control study. Resuscitation. 2007;73(1):62–72. 47. Lovett PB, Buchwald JM, Sturmann K, Bijur P. The vexatious vital: neither clinical measurements by nurses nor electronic monitor provides accurate measurements of respiratory rate in triage. Ann Emerg Med. 2005;45(1):68–76. 48. Hooker EA, O’Brien DJ, Danzel DF, Barefoot JAC, Brown JE. Respiratory rates in emergency department patients. J Emerg Med. 1989;7(12):129–132. 49. Hutchinson J. Thorax. In: Todd RB, ed. Cyclopaedia of Anatomy and Physiology, vol. IV. London: Longman, Brown, Green, Congmans and Roberts; 1849:1079–1087. 50. Mead J. Control of respiratory frequency. J Appl Physiol. 1960;15(3):325–336. 51. McFadden JP, Price RC, Eastwood HD, Briggs RS. Raised respiratory rate in elderly patients: a valuable physical sign. BMJ. 1982;284(6316):626–627. 52. Otis AB, Fenn WO, Rahn H. Mechanics of breathing in man. J Appl Physiol. 1950;2(11):592–607. 53. McIlroy MB, Marshall R, Christie RV. The work of breathing in normal subjects. Clin Sci. 1954;13:127–136. 54. Marshall R, Christie RV. The visco-elastic properties of the lungs in acute pneumonia. Clin Sci. 1954;13:403–408. 55. Sinex JE. Pulse oximetry: principles and limitations. Am J Emerg Med. 1999;17(1):59–67. 56. Smith GB, Poplett N. Knowledge of aspects of acute care in trainee doctors. Postgrad Med J. 2002;78(920):335–338. 57. McGee S, Abernethy WB, Simel DL. Is this patient hypovolemic? JAMA. 1999;281(11):1022–1029. 58. Pilling J, Cutaia M. Ambulatory oximetry monitoring in-patients with severe COPD: a preliminary study. Chest. 1999;116(2):314–321. 59. Schenkel NS, Burdet L, de Muralt B, Fitting JW. Oxygen saturation during daily activities in chronic obstructive pulmonary disease. Eur Respir J. 1996;9(12):2584–2589. 60. Rao V, Todd TRJ, Kuus PTA, Buth KJ, Pearson FG. Exercise oximetry versus spirometry in the assessment of risk prior to lung resection. Ann Thorac Surg. 1995;60(3):603–608. 61. Warner L, Bartlett KA, Charles SA, O’Brien LM. Use of pulse oximetry with exercise in the diagnosis of PCP. Int Conf AIDS. 1991;7:232. 62. Smith GB, Prytherch DR, Schmidt PE, et al. Should age be included as a component of track and trigger systems used to identify sick adult patients? Resuscitation. 2008;78(2):109–115. 63. Morgan RJ, Williams F, Wright MM. An early warning scoring system for detecting developing critical illness. Clin Intens Care. 1997;8:100. 64. Smith GB, Prytherch DR, Schmidt PE, Featherstone PI, Higgins B. A review and performance evaluation of single–parameter “track and trigger” systems. Resuscitation. 2008;79(1):11–21. 65. Kellett J, Deane B, Gleeson M. Derivation and validation of a score based on Hypotension, Oxygen saturation, low Temperature, ECG changes and Loss of independence (HOTEL) that predicts early mortality between 15 min and 24 h after admission to an acute medical unit. Resuscitation. 2008;78(1):52–58. 66. Goodacre S, Turner J, Nicholl J. Prediction of mortality among emergency medical admissions. Emerg Med J. 2006;23(5):372–375. 67. Prytherch DR, Smith GB, Schmidt P, et al. Calculating early warning scores – a classroom comparison of pen and paper and hand-held computer methods. Resuscitation. 2006;70(2):173–178.

11  The Meaning of Vital Signs

123

68. Smith GB, Prytherch DR, Schmidt P, et al. Hospital-wide physiological surveillance – a new approach to the early identification and management of the sick patient. Resuscitation. 2006;71(1):19–28. 69. Hravnak M, Edwards L, Clontz A, Valenta C, DeVita MA, Pinsky MR. Defining the incidence of cardiorespiratory instability in patients in step-down units using an electronic integrated monitoring system. Arch Intern Med. 2008;168(12):1300–1308. 70. Rooney T, Moloney ED, Bennett K, O’Riordan D, Silke B. Impact of an acute medical admission unit on hospital mortality: a 5–year prospective study. QJM. 2008;101(6):457–465.

Chapter 12

Matching Illness Severity with Level of Care Gary B. Smith and Juliane Kause

Keywords  Matching care • Levels • Illness “…My ward was now divided into three rooms; and, under favor of the matron, had managed to sort out the patients in such a way that I had what I called my “duty room,” my ‘pleasure room,’ and my ‘pathetic room,’ and worked for each in a different way. One, I visited with a dressing tray full of rollers, plasters, and pins; another, with books, flowers, games, and gossip; a third, with teapots, lullabies, consolation and sometimes, a shroud…” – Louisa May Alcott, 18631

Matching the level of care to the severity of a patient’s illness seems fundamental to the provision of quality health care. However, until recently, admission to hospital has generally been an unsystematic process, with sick patients being admitted directly from home, an outpatient clinic, or the emergency department (ED) to general wards wherein the level of monitoring and clinical supervision varied considerably. Only a few, very sick, patients are admitted directly from the ED to a high dependency unit (HDU) or an intensive care unit (ICU) for higher levels of monitoring or care. In times of high HDU/ICU occupancy, even sick patients are triaged to lower care areas.2 The placement of patients on general wards has often been based on the type of disease, rather than their care needs or severity of illness. As a result, patients with different illness severity have often been treated together. Medical and nursing care was usually specialty-based, with most urgent clinical care being provided by trainee staff working in hierarchical structures. Patients were admitted under specialists who were experts in their field but who often lacked the ability to detect and manage critical illness arising from conditions outside their specialty. There was often little planning of the patients’ in-hospital stay, and patient flow through

G.B. Smith (*) Department of Cirtical care, Queen Alexandra Hospital, Portsmouth, PO6 3LY, UK e-mail: [email protected]

M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_12, © Springer Science+Business Media, LLC 2011

125

126

G.B. Smith and J. Kause

hospital areas often bore little relation to the patient’s level of acuity. While these systems served some patients well, others were disadvantaged and often received suboptimal care.3–10 Today, reduced hospital bed numbers and increased reliance on outpatient surgery mean that only the sickest patients are hospitalized. In general, patients are older and more dependent than in previous decades; the incidence of co-existing morbidity is higher, and patient management has become more complex, placing greater pressure on health care staff. Developments in nursing, medical, surgical, and anesthetic care mean that complex therapies and investigations that were previously unavailable or deemed too risky are now commonplace. In effect, the hospital has become the “intensive care unit of the community.”

Evidence of Incorrect Placement of Patients In an ideal world, the sickest patients would be admitted to an area that provided the greatest supervision and the highest level of organ support and nursing care. While this is often the case, it is clear that many patients are incorrectly placed for their level of acuity.11–17 For example, in a 2-week survey of medical and surgical wards in a UK hospital, Leeson-Payne et al recorded 111 “HDU days,” representing 57 patients.11 Similarly, Crosby and Rees showed that 6.8% of patients in surgical ward beds and 50.8% of patients in ICU beds would have been more appropriately cared for in an HDU.12 Over 60% of patients in a group of Welsh HDUs were also incorrectly placed; the majority of them were well enough to be cared for on a general ward.13 A study of 8,040 ICU admissions in the United States demonstrated that 76.8% of patients admitted simply for monitoring were reported to have a 10% chance of receiving active ICU treatment during their stays; only 4.4% actually received it.14 These data suggest that although ICU and HDU beds are scarce, lowacuity patients are often placed there inappropriately. This is important, as an inability to admit a sick patient to an ICU when necessary, or even just a delay in admitting, can lead to a poor outcome.18–21 Improvements in these processes have occurred, due to the introduction of recommendations for admission and discharge criteria for critical care units and the associated levels of care.22–28 The mismatch of patient needs and care capabilities is not limited to those who are sufficiently sick to warrant an HDU or ICU bed. For instance, medical patients “lodged” on surgical wards because of a shortage of medical beds may fail to receive care that is appropriate to their disease or level of acuity if staff are unfamiliar with the disease process or its treatment. Recent research29,30 shows that patient flow through the hospital may have an important effect on outcome. A large review of ICU patients demonstrated that emergency patients admitted to ICU from general wards had a higher mortality than those admitted directly from the ED.29 Similarly, elective and emergency high-risk surgical patients admitted to critical care following initial postoperative care on a standard ward had high mortalities compared to those admitted directly from the operating room complex.30 Clearly,

12  Matching Illness Severity with Level of Care

127

some of these differences may be due to variations in case-mix in the two groups, but it is likely that process errors also have an effect.

Definitions of Levels of Care The disparity between patient severity of illness and the location of their care has encouraged the UK health service to define levels of care (levels 0–3) for hospitalized patients.26 Using this system, patients are allocated to a level of care according to their clinical need; location and prevailing nurse-to-patient ratio are not considered.26 The system uses four levels of care: Level 0:  Patients whose needs can be met through normal ward care in an acute hospital. Such patients might require intravenous medication and their vital signs measured less frequently than every 4 h. Level 1:  Patients at risk of their condition deteriorating, or those recently relocated from higher levels of care, whose needs can be met on an acute care ward with additional support from the critical care team. Such patients might require vital sign monitoring at a minimum every 4  h, continuous oxygen therapy (FiO2 < 0.5), respiratory physiotherapy to treat or prevent respiratory failure, boluses of intravenous fluids, or advanced techniques such as epidural analgesia or patientcontrolled analgesia. Level 2:  Patients requiring more detailed observation or intervention, including support for a single failing organ system or postoperative care, and those stepping down from higher levels of care. Such patients might require pre-operative physiological optimization, a minimum of hourly vital signs measurements or continuous oxygen therapy by face mask with a FiO2 ³ 0.5. Level 3:  Patients requiring advanced respiratory support alone or basic respiratory support together with support of at least two organ systems. This level includes all complex patients requiring support for multi-organ failure. More details of therapy appropriate for the different levels of care are provided in the original document.26 It is important to be aware that the UK classification system for levels of care shares terminology with the Society of Critical Care Medicine’s recommendations for categorizing intensive care units, with the potential for confusion.24 Nevertheless, the former provides a starting point from which levels of care can be matched to patient severity of illness. Patient movement between these levels of care has been portrayed as linear,31 but the true pattern of movement is unknown. In some cases, the speed of physiological deterioration can be dramatic and sudden as compensatory mechanisms fail. Occasionally, patients will suffer acute deterioration and a “false arrest,” where the arrest or code team is called to the patient but no cardiopulmonary arrest has occurred. These patients should be treated as high risk, as up to 30% of them subsequently die in the hospital.32–34 When in doubt, it is probably wise to opt for a higher level of care, as it is much easier to step down care if the patient is later found to be stable or improving.

128

G.B. Smith and J. Kause

Identifying a Patient’s Level of Illness Matching clinical resources to a patient’s severity of illness requires an accurate evaluation of the patient’s condition and a recognition that the patient is in need of enhanced monitoring and care. Signs of illness reflect the interaction between the patient’s physiological reserve (i.e. age, prior health), physiological deterioration (particularly respiratory rate, heart rate, SaO2, and level of consciousness), and the underlying clinical condition. The level of treatment being received by the patient is also important to consider, as this will influence physiological values; for example, consider the impact of supplementary oxygen on SaO2. In general, the clinical signs of acute illness are similar whatever the underlying process, as they usually reflect failing cardiovascular, respiratory, and neurological systems. It is rare for a single symptom, clinical sign, vital sign measurement, or laboratory investigation to be pathognomonic of a specific clinical condition. Consequently, sensitive methods of identifying patients at risk of deterioration are difficult to develop. Although numerous studies have identified heart rate, blood pressure, respiratory rate, and conscious level abnormalities to be associated with subsequent critical events,35–41 suggestions that they have predictive value must be questioned at present, as not all important vital signs are, or can be, recorded continuously in general ward areas. The process of evaluating a patient’s level of sickness, or determining that a patient is deteriorating, now relies heavily upon the use of “track-and-trigger” systems incorporating vital sign measurements.42–46 The use of these systems is intuitive, but their sensitivity, specificity, and accuracy are only just being evaluated in large studies.47–50 Several different types of “track-and-trigger” systems exist47,48,51 and preference varies internationally. In the UK, most hospitals use an aggregate weighted “track-and-trigger” system that allocates points in a weighted manner, based on the derangement of patients’ vital signs variables from an arbitrarily agreed “normal” range.51 The sum of the allocated points is known as the early warning score [EWS], the value of which is used to direct care, e.g. an increase in vital signs monitoring, the involvement of more experienced staff or the calling of a rapid response team (e.g. outreach or medical emergency team). In Australia and the US, there is a preference for using a set of predetermined, largely objective, “calling criteria.” The original of these comprised a range of specific conditions (e.g. pulmonary edema), physiological/pathological abnormalities (e.g. pulse rate < 40 or > 120 beats per minute) and the loose criterion “any time urgent help is required.”52 The occurrence of one or more of these is used as an indicator to call for a rapid response team. Since their original description, the original calling criteria have been modified by many others, often with only subtle changes to the physiological trigger points.48 Generally, the inclusion of calling criteria variables and their trigger points are solely based on expert clinical opinion and intuition. While the use of a physiologically-based “track-and-trigger” system is appealing as a method of identifying the need for increased monitoring and care, it is possible that a more subjective approach, based loosely on staff experience and expertise may also be effective.50,53–55

12  Matching Illness Severity with Level of Care

129

The usefulness of “track-and-trigger” systems depends upon the frequent and regular measurement of vital signs. However, the measurement and charting of vital signs is often poor, with resultant gaps in data.38,39,56–63 Nurses may also be unaware of patients’ abnormal vital signs.64 The reason for these failures is multifactorial.65–67 Fortunately, the use of “track-and-trigger” systems can itself increase the frequency of vital sign monitoring.68,69 Future research needs to consider how the performance of physiologically based “track-and-trigger” systems can be enhanced by the inclusion of factors such as the results of routine investigations,70,71 symptoms, diagnosis, and therapy.72 If “track-and-trigger” systems’ can be incorporated into patient monitoring systems that include so-called “smart alarms,” decision-support and neural network technology, they may prove useful in detecting subtle trends sufficiently early for critical illness to be averted.73,74 Such monitoring systems could also contact response teams directly using wireless or mobile phone technology. Integrated monitoring and analysis for early warning of patient deterioration are discussed in detail in Chap. 12.

Response to Acute Illness Even when medical staff are alerted to a patient’s abnormal physiology, there is often delay in attending the patient or in referral for higher levels of care.3,38,39,75,76 The use of algorithms that dictate specific actions and response times are helpful, but not perfect.75,76 Delayed treatment on wards can result in poor outcomes.77,78

Knowledge and Experience of Ward Staff Patients should expect to be treated in areas where the knowledge and clinical expertise of doctors, nurses, and physiotherapists is appropriate for their condition. However, at a time when hospitalized patients’ acuity is increasing, deficiencies in the acute care knowledge and skills of medical and nursing staff continue to be identified.59,61,65,79–84 The “deskilling” of ward staff appears to be an undesired byproduct of the opening of ICUs and HDUs. Previously, when overall hospital acuity levels were low, patients with complex medical problems and invasive monitoring could easily be admitted to these new, high-care areas. However, over time, hospital activity levels have risen, patients have become sicker, and ICUs and HDUs have become full. Ill patients are now discharged earlier to the general wards, where some clinical skills have not been maintained85,86 (since patients of the highest acuity have been removed to the ICUs) and as a result, staff often lack confidence when dealing with acute care problems.87 The discovery that trainee doctors and ward nurses often lack the skills necessary to detect critical illness and manage sick patients is worrisome, as in many healthcare

130

G.B. Smith and J. Kause

systems, they are the first to assess and treat patients. Of particular concern are reports that medical school training provides poor preparation for doctors’ early careers in clinical medicine and fails to teach essential aspects of applied physiology and acute care.88-90 Furthermore, the common textbooks used by medical students and trainee doctors to learn how to examine patients rarely offer advice on how to assess the acutely ill patient.91 In the UK, these shortfalls have led to the development of core competencies in acute care for a range of clinical staff 92,93 and a series of courses to teach ward staff how to recognize and manage critical illness outside of critical care areas.94–98 The Society of Critical Care Medicine has also introduced a course, directed at non-intensivists, which focuses on managing sick patients in the first 24 h of critical illness when more direct critical care expertise is unavailable.99 Training for members of rapid response teams is also now emerging.100–102

Potential Impact of Staffing Levels and Patient Flow on Outcomes Hospital staffing tends to be at its lowest during nighttime hours and on weekends, potentially making it difficult to match levels of care to patient acuity. Admission to a general medical ward after 17:00103 or to the hospital on weekends104 is associated with increased mortality. Patients who are admitted to ICU from general wards have a greater mortality than those admitted directly from ED.29 Patients who are discharged from ICUs to general wards at night have an increased risk of in-hospital death compared to those discharged during the day and to those discharged to HDUs.105,106 If hospitals were well staffed at all times, this difference might be eliminated. Often those who need to be readmitted to ICU are shown to have residual organ failure at the time of ICU discharge.107 Ward nurses are often uncertain about the dependency of patients received from ICU; better communication from ICU staff might improve this.108 Greater registered nurse staffing is associated with the reduction in the rates of pneumonia, shock, cardiac arrest, and death.109 Mortality in the ICU is also increased at times of lower staffing levels,110 perhaps suggesting poor matching of care with demand.

New Approaches to Matching Care with Patient Severity of Illness Many of the reported deficiencies in acute care relate to inadequate management of the patient’s airway, breathing, and circulation, or aspects of patient monitoring.3–5,35–39 These areas of practice are extremely familiar to specialists working in critical care, anesthesia, and emergency medicine, but less so to the clinicians under whom patients are traditionally admitted. Consequently, many new models of care delivery attempt to support the primary admitting team with the skills of resuscitation specialists. These

12  Matching Illness Severity with Level of Care

131

advances can be categorized into new patient admission processes, early emergency department treatment, new general medicine specialists, rapid response teams, and better decisions about limitation of care and resuscitation.

New Patient Admission Processes In many UK hospitals, emergency patients are now rarely admitted directly to a general ward without a degree of in-hospital triage. As before, some of this occurs from the emergency department; however, new admission wards – medical and surgical assessment units – have been created for the rapid triage for patients referred by primary physicians.111 These units perform two major functions; they monitor and observe patients, usually up to 72 h, and act as a single location for all acute admissions until their required level of care is evaluated. The single location provides rapid access to senior medical staff, diagnostics, and urgent treatment, acting as a central focus for on-call medical, nursing, and physiotherapy staff, in contrast to the traditional system in which staff and patients were dispersed throughout the hospital. In some centers, the opening of assessment units has been accompanied by the appointment of senior clinicians in acute medicine.112 There is emerging evidence that assessment units save lives and improve efficiency.113,114 Other developments that have accompanied the introduction of medical assessment units and surgical assessment units are the concepts of post-admission ward rounds and, perhaps more appropriately, multiple ward rounds during a single day.115

Early Treatment of Patients in the Emergency Department Many acutely ill patients enter the hospital via the emergency department and are obviously in need of immediate ICU-type interventions. It makes little sense to defer these interventions until ICU admission, which may be delayed by organizational factors. Emergency departments usually have the resources to intervene with techniques, such as invasive cardiovascular monitoring, non-invasive cardiac output measurement, and oximetry, which facilitate the use of early goal-directed therapy.116 Early therapy in the emergency department reverses physiological deterioration,117 and although several publications from other clinical settings demonstrate that late application of goal-directed therapy in critically ill patients is not beneficial,118,119 early therapy appears capable of improving patient survival.116

New General Medicine Specialists In the US, cost pressures, increased patient acuity, shortened hospital stays, and time pressures on primary physicians led to the development of the “hospitalist,” a new type of generalist physician who cares for patients with a wide range of

132

G.B. Smith and J. Kause

organ dysfunctions.120 In this book, the role of the hospitalist is described in detail in Chap. 5. The hospitalist provides the immediate, cross-specialty clinical care that has so often been lacking in the hospital from community-based, primary care physicians. The majority of hospitalists are specialists in general internal medicine.121 While their primary professional focus is general inpatient care, 80% of hospitalists also care for these patients when they are admitted to critical care units.122 Hospitalists also have teaching, research, and administrative responsibilities, and where they have been introduced they seem to have been beneficial.120,123 In the UK, many hospital physicians have specialized to the point where their involvement with the “on-call” admission of unselected, acutely ill medical patients has become intermittent and a minor part of their work. Many also state that they would never participate in hands-on emergency care.115 In response, some hospitals have appointed acute- care physicians who often work within medical assessment units.112 Others have introduced systems in which the on-call physician is relieved of all conflicting duties while on call,115 or have adopted the concept of “physician-of-the-week,” in which a senior clinician’s schedule for the week is dedicated exclusively to the care of medical emergencies.115,124

Rapid Response and Medical Emergency Teams In most hospitals, the only nonspecific acute care team is the cardiac arrest team. Although such teams appear to improve the rates of survival for patients after cardiac arrest in circumstances where no team had previously existed,125,126 there is evidence that most victims survive due to the actions of staff before the team arrives.127 Cardiac arrest has low rates of survival.128,129 Consequently, some hospitals in Australia and the United States have introduced Medical Emergency Teams whose role is much broader than, but includes, care of the patient in cardiac arrest.46,130–138 These teams are an attempt to deliver urgently needed experienced critical care skills to the patient, thereby improving the balance between supply and demand. In the UK, a similar system of preemptive ward care, based predominantly on individual members or teams of nursing staff was introduced in 2000.139–146 Models of outreach services are, perhaps, less prescriptive than those described for the MET. For instance, they may range from a single critical care consultant nurse147 to a 24-h, 7-day-a-week multi-professional team. Some are based upon existing acute pain relief teams148 while others may manage only postoperative patients.149 A disadvantage of nurse-based outreach teams is that although they may use patient group directives to enable them to administer fluids and oxygen without consultation with doctors, their pharmacological interventional armamentarium is limited. Rapid response teams such as Medical Emergency or Outreach teams are covered in detail elsewhere in this book.

12  Matching Illness Severity with Level of Care

133

Better Decisions About Limitation of Care and Resuscitation If a hospital patient is not expected to live, alterations are often made in the level of care required. Although a do-not-resuscitate decision does not exclude other active care, for some patients the focus of care will shift to palliation of symptoms. However, there is evidence that the resuscitation status of many patients referred for active therapy has often not been considered, or that the decision made was inappropriate.100 Even when patients have clear evidence of severe physiological deterioration, which could be anticipated to lead to cardiac arrest or death, decisions about resuscitation status are uncommon.39 This may be due to a reluctance of staff to engage in difficult do-not-resuscitate discussions with patients or their relatives, or because their knowledge of such policies is poor.150 Internationally, there are varying attitudes with regard to the training and practice of ethical aspects of resuscitation, including consulting patients about the decision.151–153 The recording of do-not-resuscitate decisions also varies.152,153 Improved treatment limitation decision making and better care of the dying patient are likely to improve patient care, reduce unnecessary and futile cardiopulmonary resuscitation attempts, and make resource utilization more rational.154 Medical Emergency Teams may have a role in this process.138,155

Summary Existing models of acute care do not necessarily match care provision with patient severity of illness and may lead to substandard care and medical error. New methods of care delivery, in which the work of primary clinicians is supported by the skills of resuscitation specialists, provide opportunities to improve this situation. These advances involve medical and surgical assessment units, hospitalists, consultants in acute care, early emergency department treatment, medical emergency and outreach teams, and better decisions about limitation of care and resuscitation.

References 1. Alcott LM. Hospital Sketches. Boston: James Redpath Publishers; 1863. 2. Strauss MJ, LoGerfo JP, Yeltatzie JA, Temkin N, Hudson LD. Rationing of intensive care unit services. An everyday occurrence. JAMA. 1986;255:1143–1146. 3. McQuillan PJ, Pilkington S, Allan A, et al. Confidential inquiry into quality of care before admission to intensive care. BMJ. 1998;316:1853–1858. 4. Neale G. Risk management in the care of medical emergencies after referral to hospital. J R Coll Physicians Lond. 1998;32:125–129. 5. McGloin H, Adam S, Singer M. Unexpected deaths and referrals to intensive care of patients on general wards. Are some cases potentially avoidable? J R Coll Physicians Lond. 1999;33:255–259.

134

G.B. Smith and J. Kause

6. Vincent C, Neale G, Woloshynowych M. Adverse events in British hospitals: preliminary retrospective record review. BMJ. 2001;322:517–519. 7. Brennan TA, Leape LL, Laird NM, et  al. Incidence of adverse events and negligence in hospitalized patients: results of the Harvard Medical Practice Study I. N Engl J Med. 1991;324:370–376. 8. Kohn LT, Corrigan JM, Donaldson MS, eds. To Err Is Human: Building a Safer Health System. Washington, DC: National Academies Press; 2000. 9. RMcL W, Runciman WB, Gibberd RW, Harrison BT, Newby L, Hamilton JD. The quality in Australian health care study. Med J Aust. 1995;163:458–471. 10. Seward E, Greig E, Preston S, et al. Confidential study of deaths after emergency medical admission: issues relating to quality of care. Clin Med. 2003;3:425–434. 11. Leeson-Payne CG, Aitkenhead AR. A prospective study to assess the demand for a high dependency unit. Anaesthesia. 1995;50:383–387. 12. Crosby DL, Rees GA. Provision of postoperative care in UK hospitals. Ann R Coll Surg Engl. 1994;76:14–18. 13. Donnelly P, Sandifer QD, O’Brien D, Thomas EA. A pilot study of the use of clinical guidelines to determine appropriateness of patient placement on intensive and high dependency care units. J Public Health Med. 1995;17:305–310. 14. Zimmerman JE, Wagner DP, Knaus WA, Williams JF, Kolakowski D, Draper EA. The use of risk predictions to identify candidates for intermediate care units. Implications for intensive care utilization and cost. JAMA. 1995;108:490–499. 15. Bodenham AR, Knappett P, Cohen A, Bensley D, Fryers P. Facilities and usage of general intensive care in Yorkshire. A need for high-dependency units. Clin Intensive Care. 1995;6:260–265. 16. Ryan DW, Bayly PJ, Weldon OG, Jingree M. A prospective two-month audit of the lack of provision of a high-dependency unit and its impact on intensive care. Anaesthesia. 1997;52:265–270. 17. Kilpatrick A, Ridley S, Plenderleith L. A changing role for intensive therapy: is there a case for high-dependency care? Anaesthesia. 1994;49:666–670. 18. Frisho-Lima P, Gurman G, Schapira A, Porath A. Rationing critical care – what happens to patients who are not admitted? Theor Surg. 1994;9:208–211. 19. Simchen E, Sprung CL, Galai N, et al. Survival of critically ill patients hospitalized in and out of intensive care. Crit Care Med. 2007;35:449–457. 20. Metcalfe A, Slogget A, McPherson K. Mortality among appropriately referred patients refused admission to intensive care units. Lancet. 1997;350:7–12. 21. Young MP, Gooder VJ, McBride K, James B, Fisher ES. Inpatient transfers to the intensive care unit: delays are associated with increased mortality and morbidity. J Gen Intern Med. 2003;18:77–83. 22. Society of Critical Care Medicine Task Force of the American College of Critical Care Medicine, Society of Critical Care Medicine. Guidelines on intensive care unit admission, discharge, and triage. Crit Care Med. 1999;27:633–638. 23. Nasraway S, Cohen IL, Dennis RC, et al. Guidelines on admission and discharge for adult intermediate care units. Crit Care Med. 1998;26:607–610. 24. Haupt MT, Bekes CE, Brilli RJ, et  al. Guidelines on critical care services and personnel: recommendations based on a system of categorization of three levels of care. Crit Care Med. 2003;31:2677–2683. 25. Department of Health. Guidelines on Admission to and Discharge from Intensive Care and High Dependency Units. London: HMSO; 1996. 26. Intensive Care Society. Levels of Critical Care for Adult Patients. London: Intensive Care Society; 2002. Modified 2008. 27. Smith G, Nielsen M. ABC of intensive care: criteria for admission. BMJ. 1999;318:1544–1547. 28. European Society of Intensive Care Task Force. Guidelines for utilisation of intensive care units. Intensive Care Med. 1994;20:163–164.

12  Matching Illness Severity with Level of Care

135

29. Goldhill DR, Sumner A. Outcome of intensive care patients in a group of British intensive care units. Crit Care Med. 1998;26:1337–1345. 30. Jhanji S, Thomas B, Ely A, Watson D, Hinds CJ, Pearse RM. Mortality and utilisation of critical care resources among high-risk surgical patients in a large NHS trust. Anaesthesia. 2008;63:695–700. 31. Intensive Care Society. Guidelines for the Introduction of Outreach Services. London: Intensive Care Society; 2002. Modified 2003. 32. Cashman JN. In-hospital cardiac arrest: what happens to the false arrests? Resuscitation. 2002;53:271–276. 33. Hein A, Thorén A, Herlitz J. Characteristics and outcome of false cardiac arrests in hospital. Resuscitation. 2006;69:191–197. 34. Kenward G, Bradburn S, Steeds R. False cardiac arrests: the right time to turn away? Postgrad Med J. 2007;83:344–347. 35. Franklin C, Mathew J. Developing strategies to prevent in-hospital cardiac arrest: analyzing responses of physicians and nurses in the hours before the event. Crit Care Med. 1994;22:244–247. 36. Schein RM, Hazday N, Pena M, Ruben BH, Sprung CL. Clinical antecedents to in-hospital cardiopulmonary arrest. Chest. 1990;98:1388–1392. 37. Hodgetts TJ, Kenward G, Vlachonikolis IG, Payne S, Castle N. The identification of risk factors for cardiac arrest and formulation of activation criteria to alert a medical emergency team. Resuscitation. 2002;54:125–131. 38. Hillman K, Bristow PJ, Chey T, et al. Duration of life-threatening antecedents prior to intensive care admission. Intensive Care Med. 2002;28:1629–1634. 39. Kause J, Smith GB, Hillman K, Prytherch D, Parr M, Flabouras A. A comparison of antecedents to cardiac arrests, deaths and emergency intensive care admissions in Australia and New Zealand, and the United Kingdom – the ACADEMIA study. Resuscitation. 2004;62:275–282. 40. Smith GB, Prytherch DR, Schmidt PE, et al. Should age be included as a component of trackand-trigger systems used to identify sick adult patients? Resuscitation. 2008;78:109–115. 41. Jones AE, Yiannibas V, Johnson C, Kline JA. Emergency department hypotension predicts sudden unexpected in-hospital mortality: a prospective cohort study. Chest. 2006;130:941–946. 42. Morgan RJ, Williams F, Wright MM. An early warning scoring system for detecting developing critical illness. Clin Intensive Care. 1997;8:100. 43. Stenhouse C, Coates S, Tivey M, Allsop P, Parker T. Prospective evaluation of a modified Early Warning Score to aid earlier detection of patients developing critical illness on a surgical ward. Br J Anaesth. 2000;84:663. 44. Goldhill DR, Worthington L, Mulcahy A, Tarling M, Sumner A. The patient-at-risk team: identifying and managing seriously ill ward patients. Anaesthesia. 1999;54:853–860. 45. Subbe CP, Hibbs R, Williams E, Rutherford P, Gemmel L. ASSIST: a screening tool for the critically ill patients on general medical wards. Intensive Care Med. 2002;28:S21. 46. Hourihan F, Bishop G, Hillman K, Daffurn K, Lee A. The Medical Emergency Team: a new strategy to identify and intervene in high-risk patients. Clin Intensive Care. 1995;6:269–272. 47. Smith GB, Prytherch DR, Schmidt PE, Featherstone PI. Review and performance evaluation of aggregate weighted “track-and-trigger” systems. Resuscitation. 2008;77:170–179. 48. Smith GB, Prytherch DR, Schmidt PE, Featherstone PI, Higgins B. A review, and performance evaluation, of single-parameter “track-and-trigger” systems. Resuscitation. 2008;79:11–21. 49. Gao H, McDonnell A, Harrison DA, et al. Systematic review and evaluation of physiological track-and-trigger warning systems for identifying at-risk patients on the ward. Intensive Care Med. 2007;33:667–679. 50. Santiano N, Young L, Hillman K, et al. Analysis of medical emergency team calls comparing subjective to “objective” call criteria. Resuscitation. 2009;80:44–49. 51. Department of Health and NHS Modernisation Agency. The National Outreach Report. London: Department of Health; 2003.

136

G.B. Smith and J. Kause

52. Lee A, Bishop G, Hillman KM, Daffurn K. The medical emergency team. Anaesth Intensive Care. 1995;23:183–186. 53. Cioffi J. Recognition of patients who require emergency assistance: a descriptive study. Heart Lung. 2000;29:262–268. 54. Cioffi J. Nurses’ experiences of making decisions to call emergency assistance to their patients. J Adv Nurs. 2000;2:108–114. 55. Boockvar K, Brodie HD, Lachs M. Nursing assistants detect behavior changes in nursing home residents that precede acute illness: development and validation of an illness warning instrument. J Am Geriatr Soc. 2000;48:1086–1091. 56. Chellel A, Fraser J, Fender V, et al. Nursing observations on ward patients at risk of critical illness. Nurs Times. 2002;98:36–39. 57. Kenward G, Hodgetts T, Castle N. Time to put the R back in TPR. Nurs Times. 2001;97:32–33. 58. Subbe CP, Williams EM, Gemmell LW. Are medical emergency teams picking up enough patients with increased respiratory rate? Crit Care Med. 2004;32:1983–1984. 59. National Confidential Enquiry into Patient Outcomes and Death. An Acute Problem. London: National Confidential Enquiry into Patient Outcome and Death; 2005. 60. National Institute for Health and Clinical Excellence. Acutely Ill Patients in Hospital: Recognition of and Response to Acute Illness in Adults in Hospital. London: National Institute for Health and Clinical Excellence; 2007. NICE Clinical Guideline No 50. 61. National Patient Safety Agency. Safer Care for the Acutely Ill Patient: Learning from Serious Incidents. London: NPSA; 2007. 62. Chen J, Hillman K, Bellomo R, Flabouris A, Finfer S, Cretikos M. The impact of introducing medical emergency team system on the documentations of vital signs. Resuscitation. 2009;80:35–43. 63. Armstrong B, Walthall H, Clancy M, Mullee M, Simpson H. Recording of vital signs in a district general hospital emergency department. Emerg Med J. 2008;25:799–802. 64. Fuhrmann L, Lippert A, Perner A, Ostergaard D. Incidence, staff awareness and mortality of patients at risk on general wards. Resuscitation. 2008;77:325–330. 65. National Patient Safety Agency. Recognising and Responding Appropriately to Early Signs of Deterioration in Hospitalised Patients. London: NPSA; 2007. 66. Andrews T, Waterman H. Packaging: a grounded theory of how to report physiological deterioration effectively. J Adv Nurs. 2005;52:473–481. 67. Hogan J. Why don’t nurses monitor the respiratory rates of patients? Br J Nurs. 2006;15:489–492. 68. McBride J, Knight D, Piper J, Smith GB. Long-term effect of introducing an early warning score on respiratory rate charting on general wards. Resuscitation. 2005;65:41–44. 69. Odell M, Rechner IJ, Kapila A, et al. The effect of a critical care outreach service and an early warning scoring system on respiratory rate recording on the general wards. Resuscitation. 2007;74:470–475. 70. Prytherch DR, Sirl JS, Weaver PC, Schmidt P, Higgins B, Sutton GL. Towards a national clinical minimum data set for general surgery. Br J Surg. 2003;90:1300–1305. 71. Prytherch DR, Sirl JS, Schmidt P, Featherstone PI, Weaver PC, Smith GB. The use of routine laboratory data to predict in-hospital death in medical admissions. Resuscitation. 2005;66:203–207. 72. Cullen DJ, Nemeskal AR, AM Zaslavsky. Intermediate TISS: a new Therapeutic Intervention Scoring System for non-ICU patients. Crit Care Med. 1994;22:1406–1411. 73. Smith GB, Prytherch DR, Schmidt P, et al. Hospital-wide physiological surveillance – a new approach to the early identification and management of the sick patient. Resuscitation. 2006;71:19–28. 74. Tarassenko L, Hann A, Young D. Integrated monitoring and analysis for early warning of patient deterioration. Br J Anaesth. 2006;97:64–68. 75. Day BA. Early warning system scores and response times: an audit. Nurs Crit Care. 2003;8:156–164.

12  Matching Illness Severity with Level of Care

137

76. Carberry M. Implementing the modified early warning system: our experiences. Nurs Crit Care. 2002;7:220–226. 77. Lundberg JS, Perl TM, Wiblin T, et al. Septic shock: an analysis of outcomes for patients with onset on hospital wards versus intensive care units. Crit Care Med. 1998;26:1020–1024. 78. Chen J, Bellomo R, Flabouris A, Hillman K, Finfer S. The relationship between early emergency team calls and serious adverse events. Crit Care Med. 2009;37:148–153. 79. Smith GB, Poplett N. Knowledge of aspects of acute care in trainee doctors. Postgrad Med J. 2002;78:335–358. 80. Howell M. Pulse oximetry: an audit of nursing and medical staff understanding. Br J Nurs. 2002;11:191–197. 81. Wood I, Douglas J, Priest H. Education and training for acute care delivery: a needs analysis. Nurs Crit Care. 2004;9:159–166. 82. Higginson R, Lewis C, De D, Jones B. The need for critical care nurse education at preregistration level. Br J Nurs. 2004;13:1326–1328. 83. McGaughey J. Acute care teaching in the undergraduate nursing curriculum. Nurs Crit Care. 2009;14:11–16. 84. Meek T. New house officers’ knowledge of resuscitation, fluid balance and analgesia. Anaesthesia. 2000;55:1127–1143. 85. Ingelby S, Eddleston J, Naylor S. Evaluating knowledge in acute illness: Critical Care Educational Project. Crit Care. 2003;7:S119. 86. Quniton S, Higgins Y. Critical care training needs analysis for ward staff. Crit Care. 2003;7:S119–S120. 87. Featherstone P, Smith GB, Linnell M, Easton S, Osgood VM. Impact of a one-day interprofessional course (ALERT™) on attitudes and confidence in managing critically ill adult patients. Resuscitation. 2005;65:329–336. 88. Cooper N. Medical training did not teach me what I really needed to know. BMJ [Career Focus]. 2003;327:190s. 89. Goldacre MJ, Lambert T, Evans J, Turner G. Preregistration house officers’ views on whether their experience at medical school prepared them well for their jobs: national questionnaire survey. BMJ. 2003;326:1011–1012. 90. Harrison GA, Hillman KM, Fulde GW, Jacques TC. The need for undergraduate education in critical care. (Results of a questionnaire to year 6 medical undergraduates, University of New South Wales and recommendations on a curriculum in critical care). Anaesth Intensive Care. 1999;27:53–58. 91. Cook CJ, Smith GB. Do textbooks of clinical examination contain information regarding the assessment of critically ill patients? Resuscitation. 2004;60:129–136. 92. Perkins GD, Barrett H, Bullock I, et al. The Acute Care Undergraduate Teaching (ACUTE) Initiative: consensus development of core competencies in acute care for undergraduates in the United Kingdom. Intensive Care Med. 2005;31:1627–1633. 93. Department of Health. Competencies for Recognising and Responding to Acutely Ill Patients in Hospital. London: Department of Health; 2009. 94. Smith GB, Osgood VM, Crane S. ALERT™ – a multiprofessional training course in the care of the acutely ill adult patient. Resuscitation. 2002;52:281–286. 95. Anderson ID. Care of the critically ill surgical patient courses at the Royal College of Surgeons. Br J Hosp Med. 1997;57:274–275. 96. White RJ, Garrioch MA. Time to train all doctors to look after seriously ill patients – CCrISP and IMPACT. Scott Med J. 2002;47:127. 97. O’Riordan B, Gray K, McArthur–Rouse F. Implementing a critical care course for ward nurses. Nurs Stand. 2003;17:41–44. 98. Smith CM, Perkins GD, Bullock I, Bion JF. Undergraduate training in the care of the acutely ill patient: a literature review. Intensive Care Med. 2007;33:901–907. 99. Dellinger RP. Fundamental critical care support: another merit badge or more? Crit Care Med. 1996;24:556.

138

G.B. Smith and J. Kause

100. Parr MJ, Hadfield JH, Flabouris A, Bishop G, Hillman K. The Medical Emergency Team: 12-month analysis of reasons for activation, immediate outcome, and not-for-resuscitation orders. Resuscitation. 2001;50:39–44. 101. DeVita MA, Schaefer J, Lutz J, Wang H, Dongilli T. Improving medical emergency team (MET) performance using a novel curriculum and a computerized human patient simulator. Qual Saf Health Care. 2005;14:326–331. 102. Johnson AL. Creative education for rapid response team implementation. J Contin Educ Nurs. 2009;40:38–42. 103. Hillson SD, Rich EC, Dowd B, Luxenberg MG. Call nights and patient care: effects on inpatients at one teaching hospital. J Gen Intern Med. 1992;7:645. 104. Bell CM, Redelmeier DA. Mortality among patients admitted to hospitals on weekends as compared with weekdays. N Engl J Med. 2001;345:663–668. 105. Beck DH, McQuillan PJ, Smith GB. Waiting for the break of dawn? The effects of discharge time, discharge scores, and discharge facility on hospital mortality after intensive care. Intensive Care Med. 2002;28:1287–1293. 106. Goldfrad C, Rowan K. Consequences of discharges from intensive care at night. Lancet. 2000;355:1138–1142. 107. Moreno R, Miranda DR, Matos R, Fevereiro T. Mortality after discharge from intensive care: the impact of organ system failure and nursing workload use at discharge. Intensive Care Med. 2001;27:999–1004. 108. Whittaker J, Ball C. Discharge from intensive care: a view from the ward. Intensive Crit Care Nurs. 2000;16:135–143. 109. Needleman J, Buerhaus P, Mattke S, Stewart M, Zelevinsky K. Nurse-staffing levels and the quality of care in hospitals. N Engl J Med. 2002;346:1715–1722. 110. Tarnow-Mordi WO, Hau C, Warden A, Shearer AJ. Hospital mortality in relation to staff workload: a 4-year study in an adult intensive-care unit. Lancet. 2000;356:185–189. 111. Cooke MW, Higgins J, Kidd P. Use of emergency observation and assessment wards: a systematic literature review. Emerg Med J. 2003;20:138–142. 112. Armitage M, Raza T. A consultant physician in acute medicine: the Bournemouth Model for managing increasing numbers of medical emergency admissions. Clin Med. 2002;2:331–333. 113. Rooney T, Moloney ED, Bennet K, O’Riordan D, Silke B. Impact of an acute medical admission unit on hospital mortality: a 5-year prospective study. QJM. 2008;101:457–465. 114. Maloney ED, Smith D, Bennett K, O’Riordan D, Silke B. Impact of an acute medical admission unit on length of hospital stay, and emergency department “wait times”. QJM. 2005;98:283–289. 115. Mather HM, Connor H. Coping with pressures in acute medicine – the second RCP consultant questionnaire survey. J R Coll Physicians Lond. 2000;34:371–373. 116. Rivers E, Nguyen B, Havstad S, et al. Early Goal-Directed Therapy Collaborative Group: early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med. 2001;345:1368–1377. 117. Nguyen HB, Rivers EP, Havstad S, et al. Critical care in the emergency department: a physiologic assessment and outcome evaluation. Acad Emerg Med. 2000;7:1354–1361. 118. Hayes MA, Timmins AC, Yau EHS, Palazzo M, Hinds CJ, Watson D. Elevation of systemic oxygen delivery in the treatment of critically ill patients. N Engl J Med. 1994;330:1717–1722. 119. Gattinoni L, Brazzi L, Pelosi P, et al. A trial of goal-oriented hemodynamic therapy in critically ill patients. SvO2 Collaborative Group. N Engl J Med. 1995;333:1025–1032. 120. Wachter RM, Goldman L. The hospitalist movement 5 years later. JAMA. 2002;287:487–494. 121. Hoff TH, Whitcomb WF, Williams K, Nelson JR, Cheesman RA. Characteristics and work experiences of hospitalists in the United States. Arch Intern Med. 2001;161:851–858. 122. Lindenauer PK, Pantilat SZ, Katz PP, Wachter RM. Hospitalists and the practice of in– patient medicine: results of a survey of the National Association of Inpatient Physicians. Ann Intern Med. 1999;130:343–349.

12  Matching Illness Severity with Level of Care

139

123. Baudendistel TE, Wachter RM. The evolution of the hospitalist movement in the USA. Clin Med. 2002;2:327–330. 124. Worth R, Youngs G. Consultant physician of the week: a solution to the bed crisis. J R Coll Physicians Lond. 1996;30:211–212. 125. Sandroni C, Ferro G, Santangelo S, et al. In-hospital cardiac arrest: survival depends mainly on the effectiveness of the emergency response. Resuscitation. 2004;62:291–297. 126. Henderson SO, Ballesteros D. Evaluation of a hospital-wide resuscitation team: does it increase survival for in-hospital cardio-pulmonary arrest? Resuscitation. 2001;48:111–116. 127. Soar J, McKay U. A revised role for the hospital cardiac arrest team? Resuscitation. 1998;38:145–149. 128. Gwinnutt CL, Columb M, Harris R. Outcome after cardiac arrest in adults in UK hospitals: effect of the 1997 guidelines. Resuscitation. 2000;47:125–135. 129. Peberdy MA, Kaye W, Ornato JP, et  al. Cardio-pulmonary resuscitation of adults in the hospital: a report of 14, 720 cardiac arrests from the National Registry of Cardiopulmonary Resuscitation. Resuscitation. 2003;58:297–308. 130. DeVita MA, Bellomo R, Hillman K, et  al. Findings of the first consensus conference on medical emergency teams. Cril Care Med. 2006;34:2463–2478. 131. Buist MD, Moore GE, Bernard SA, Waxman BP, Anderson JN, Nguyen TV. Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. BMJ. 2002;324:1–6. 132. DeVita MA, Braithwaite RS, Mahidhara R, Stuart S, Foraida M, Simmons RL. Medical Emergency Response Improvement Team (MERIT). Use of medical emergency team responses to reduce hospital cardiopulmonary arrests. Qual Saf Health Care. 2004;13:251–254. 133. Bellomo R, Goldsmith D, Uchino S, et al. A prospective before-and-after trial of a medical emergency team. Med J Aust. 2003;179:283–287. 134. Braithwaite RS, DeVita MA, Mahidhara R, Simmons RL, Stuart S, Foraida M. Medical Emergency Response Improvement Team (MERIT). Use of medical emergency team (MET) responses to detect medical errors. Qual Saf Health Care. 2004;13:255–259. 135. Bristow PJ, Hillman KM, Chey T, et al. Rates of in-hospital arrests, deaths, and intensive care admissions: the effect of a medical emergency team. Med J Aust. 2000;173:236–240. 136. Bellomo R, Goldsmith D, Uchino S, et al. Prospective controlled trial of effect of medical emergency team on postoperative morbidity and mortality rates. Crit Care Med. 2004;32:916–921. 137. Hillman K, Chen J, Cretikos M, et al. Introduction of the medical emergency team (MET) system: a cluster-randomised controlled trial. Lancet. 2005;365:2091–2097. 138. Jones DA, McIntyre T, Baldwin I, Mercer I, Kattula A, Bellomo R. The medical emergency team and end-of-life care: a pilot study. Crit Care Resusc. 2007;9:151–156. 139. Department of Health. Comprehensive Critical Care. A Review of Adult Critical Care Services. London: Department of Health; 2000. 140. Anderson K, Atkinson D, McBride J, Moorse S, Smith S. Setting up an outreach team in the U.K. Crit Care Nurs Eur. 2002;2:8–12. 141. Ball C, Kirkby M, Williams S. Effect of the critical care outreach team on patient survival to discharge from hospital and readmission to critical care: non-randomised population based study. BMJ. 2003;327:1014–1016. 142. Garcea G, Thomasset S, McClelland L, Leslie A, Berry DP. Impact of a critical care outreach team on critical care readmissions and mortality. Acta Anaesthesiol Scand. 2004;48:1096–1100. 143. Pittard AJ. Out of our reach? Assessing the impact of introducing a critical care outreach service. Anaesthesia. 2003;58:882–885. 144. Story DA, Shelton AC, Poustie SJ, Colin-Thome NJ, McNicol PL. The effect of critical care outreach on postoperative serious adverse events. Anaesthesia. 2004;59:762–766. 145. Priestley G, Watson W, Rashidian A, et  al. Introducing Critical Care Outreach: a wardrandomised trial of phased introduction in a general hospital. Intensive Care Med. 2004;30:1398–1404.

140

G.B. Smith and J. Kause

146. McDonnell A, Esmonde L, Morgan R, et al. The provision of critical care outreach services in England: findings from a national survey. J Crit Care. 2007;22:212–218. 147. Jones P. Consultant nurses and their potential impact upon health care delivery. Clin Med. 2002;2:39–40. 148. Counsell DJ. The acute pain service: a model for outreach critical care. Anaesthesia. 2001;56:925–926. 149. Bamgbade OA. The peri-operative care team: a model for outreach critical care. Anaesthesia. 2002;57:1028–1044. 150. Smith GB, Poplett N, Williams D. Staff awareness of a “do not attempt resuscitation” policy in a district general hospital. Resuscitation. 2005;65:159–163. 151. Mohr M, Bahr J, Kettler D, Andres J. Ethics and law in resuscitation. Resuscitation. 2002;54:99–102. 152. Baskett PJ, Lim A. The varying ethical attitudes towards resuscitation in Europe. Resuscitation. 2004;62:267–273. 153. Sidhu NS, Dunkley ME, Egan MJ. “Not-for-resuscitation” orders in Australian public hospitals: policies, standardised order forms and patient information leaflets. Med J Aust. 2007;186:72–75. 154. Ellershaw J, Ward C. Care of the dying patient: the last hours or days of life. BMJ. 2003;326:30–34. 155. Chen J, Flabouris A, Bellomo R, Hillman K, Finfer S. The Medical Emergency Team System and not-for-resuscitation orders: results from the MERIT study. Resuscitation. 2008;79:391–397.

Chapter 13

Causes of Failure to Rescue Marilyn Hravnak, Andrea Schmid, Lora Ott, and Michael R. Pinsky

Keywords  Causes • Failure • Rescue

Introduction Failure-to-rescue (FTR) is a term first coined by Silber in 1992.1 The term has been used increasingly as a measure of hospital quality of care, and has been named by the Agency of Health Care Quality as one of 20 patient safety indicators.2 The term is defined as the probability of death after a complication or adverse occurrence within a particular hospital.3 Both patient-level factors as well as hospital-level factors and characteristics can influence a facility’s FTR rate. The risk of FTR is lowered when complications and adverse occurrences are recognized quickly and treated aggressively.3 Likewise, the risk of FTR increases when deleterious changes in patient status occur and steps to reverse the change are not taken.4,5 The FTR literature relates several models, which define “complications” in different ways.3,6–8 The most comprehensive list of complications in the hospitalized patient used in the traditional FTR analysis first developed by Silber includes 14 categories3 (see Table 13.1). The occurrence of the complication, however, does not necessarily imply wrongdoing, and although some complications may be hospital-acquired, others may be present as admission co-morbidities. Nevertheless, the premise is that failure to identify these complications quickly and treat them aggressively in order to prevent death during hospitalization results in FTR.4 Thus it is the “rescue capability” of a hospital in recognizing and then responding to complications and adverse occurrence that denotes quality.9

M. Hravnak (*), School of Nursing, University of Pittsburgh, 3500 Victoria Street, 336 Victoria Building, Pittsburgh, PA 15261, USA e-mail: [email protected]

M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_13, © Springer Science+Business Media, LLC 2011

141

142

M. Hravnak et al.

Table  13.1  Complications developed by hospitalized patients which can lead to death and failure-rescue. Terminal sequelae common to all complications is severe hypotension, cardiorespiratory instability and death Categories of complications Examples and sequelae Cardiac Arrhythmias, cardiac arrest, infarction, congestive heart failure Respiratory Pneumonia, pneumothorax, bronchospasm, respiratory compromise, aspiration pneumonia Hypotension Shock, hypovolemia Neurologic Stroke, transient ischemic attack, seizure, psychosis, coma Deep vein thrombosis Pulmonary embolus, arterial clot, phlebitis Internal organ damage, return to surgery Infection Deep wound infection, sepsis Gangrene Amputation Gastrointestinal bleeding Blood loss Peritonitis, intestinal obstruction Renal dysfunction Hepatitis Pancreatitis Decubitus ulcers Orthopedic complications and compartment syndromes Adapted from Silber et al.3

There may be a number of opportunities between the initial development of a complication and eventual death resulting from that complication wherein the complication can be identified and treated. It is intuitively obvious that the earlier the complication is recognized and acted upon, the less the negative impact that the complication will have on patient outcome. Examples from Table 13.1 serve as illustrations. Obviously the common end-result for complications that result in death is the eventual development of hemodynamic instability manifested by abnormalities in vital signs and compromise of end-organ perfusion and associated metabolic derangement due to tissue hypoperfusion. At present, vital-sign monitoring and surveillance of easily discernable behavioral changes are therefore the key to first detecting instability, identifying it, and then applying appropriate interventions in order to prevent FTR.10 For the purposes of this discussion, vital signs include not only the basic intermittently measured blood pressure (BP), temperature, heart rate (HR) and respiratory rate (RR), but also noninvasively measured pulse oximetry (SpO2). In fact, all these vital signs can be measured continuously using readily available bedside monitoring devices. Medical Emergency Teams (METs) can only have an impact as an FTR intervention if the important initial steps of detection and identification are first enacted.11,12 There are several different models that denote the factors and characteristics that influence FTR.9,10The most simplistic approach is probably that suggested by Silber, which specifies that causes of FTR can be attributed to patient-level factors and hospital- or system-level factors.

13  Causes of Failure to Rescue

143

Causes of FTR: Patient-Level Factors Patient-level factors contributing to FTR can be static or dynamic. The traditional static patient-level factors that contribute to death due to complications are age, race, gender, procedure and the presence of co-morbidities such as obesity, smoking, intravenous drug use or previous myocardial infarction, diabetes and hypertension.1 Thus, known demographic risk factors, the complexity of the patient’s health problem, and the presence of co-morbidities are important attributes in targeting those patients who are most at risk of developing complications and therefore in need of a higher level of surveillance.13 These static patient-level factors, long known to contribute to the mortality of hospitalized patients, are incorporated in a variety of illness-acuity scoring systems. Dynamic patient-level factors contributing to FTR are those variable signs of physiologic deterioration that precede cardiac arrest, death or emergency intensive care unit (ICU) admission. These approaches have been much less successful in both identifying and quantifying the dynamic monitoring variables and their thresholds that precede such events14,15 in a sensitive and specific manner. Such information is essential to detect and identify instability in order to make a call for help and/ or intervene. However, the efficacy or lack thereof of METs are totally dependent upon the mechanisms to both “track,” meaning to detect instability using periodic observation of the dynamic physiologic signals and “trigger,” meaning to identify instability using predetermined threshold criteria of the tracked information to place the call for help. Dynamic physiologic data is traditionally measured via the periodic observation of vital signs. Bell et  al.16 demonstrated that patients who exhibited even one ­episode of single-parameter vital-sign abnormality during hospitalization had a higher 30-day mortality rate (25%) as compared to patients who did not (3.5%). However, it has proven extremely difficult to determine which vital-sign parameters and which threshold values can demonstrate MET efficacy when being used for track-and-trigger functions, and can do so in a generalizable manner. DeVita et al. showed that using some predefined numeric trigger thresholds for HR, RR, BP and SpO2 were superior to an open-ended clinician judgment call alone.17 However, which parameters and which threshold triggers for these vital signs remain debatable. There are many reports published using single-parameter track-and-trigger criteria, but there is considerable variation among the parameters and thresholds.18 In response to the limitations of multiple single parameters, the so-called Early Warning Scores (EWS) have been developed. These scores assign increasing levels of abnormality to develop a single index score, which in turn triggers a threshold.19,20 Smith et  al. conducted a systematic review of 33 unique aggregate-weighted ­track-and-trigger systems.21 Although most were poor at discriminating between survivors and non-survivors, 12 had area-under-the-receiver-operating curves (AUROC) between 0.700 and 0.799, with the top four methods incorporating age, and the top two methods incorporating temperature. However, none had AUROC ³ 0.8, which would indicate good discrimination. In their current iterations, EWS track-and-trigger

144

M. Hravnak et al.

systems still require intermittent clinician presence at the bedside to gather the data and develop the score by hand, with personal digital assistant (PDA) calculation at the bedside22 or through linkage of the PDA or electronic medical record to a local area network and centralized evaluation23 to make a call decision. Possibly adding additional parameters such as age24 or laboratory values25 could increase the accuracy of EWS but would, in turn, add more complexity. Interfacing such scoring systems to the electronic medical record with automated alerts might be one approach to simplifying the calculation mechanisms. In another systematic review18 Gao et  al. evaluated 36 papers identifying 25 distinct track-and-trigger methods – both single parameter and aggregate scoring mechanisms. All track-and-trigger mechanisms utilized HR, RR, systolic BP and level of consciousness, but varied over other parameters utilized. Although specificity and negative predictive values were high (median values of 89.5 [64.2–95.7] and 94.3 [89.5–97.0] respectively), sensitivities and positive predictive values were low (median values of 43.3 [25.4–69.2] and 36.7 [29.3–43.8] respectively), indicating that the majority of the chosen thresholds were placed too high in an effort to reduce false calls and MET overload. None of the track-and-triggers evaluated in the review had achieved the requirement of a Level 1 clinical decision rule, i.e. validated for use in a wide variety of settings with confidence that it can change behavior and improve patient outcomes.26 In addition to the difficulty in identifying which vital sign parameters and thresholds – either singly or in combination – should serve as activation criteria, the frequency with which the surveillance occurs presents another barrier. Galhotra et al. demonstrated that even mature rapid response systems with track-and-trigger mechanisms based upon intermittent patient evaluation still miss potentially avoidable cardiac arrests.27 Parameters and triggers aside, this can, in part, be explained by the fact that instability, when it occurs, is not a static phenomenon. Rather, it is characterized by patient’s cyclically moving above and below abnormality thresholds as their compensatory mechanisms first correct and then fail to correct the physiologic cause for instability. An example is shown in Fig. 13.1. This patient first demonstrates hypoxemia with low SpO2. In compensation, RR and HR increase to improve oxygen uptake and transport, which is effective – SpO2 improves for a brief period, and RR and HR lowers. Shortly however, the cycle repeats itself, with four cycles of deterioration, compensation and improvement over a 5-h period that increases in intensity and duration. Intermittent observation and recording of vital signs for this patient between cycles would have missed the instability. In a newer approach, Integrated Monitoring Systems (IMS) can “marry” weighted aggregate scoring with continuous monitoring. Such systems can ­continuously evaluate multiple physiologic parameters for subtle departure from normality.28 In one example, we evaluated the ability of an IMS (Visensia™, OBS Medical, Carmel, IN) to detect instability on a single stepdown unit. This IMS used the standard noninvasive hardwired monitoring variables of HR, RR, BP and SpO2 to develop a single neural networked signal. In the first phase, the IMS was ­connected but the index values were not visible to the clinicians. In patients for

13  Causes of Failure to Rescue

145

160 HR

120 80 40 1 40

2

3

4

5

6

7

8

9

10

11

12

13

14

15

2

3

4

5

6

7

8

9

10

11

12

13

14

15

2

3

4

5

6

7

8

9

10

11

12

13

14

15

2

3

4

5

6

7

8

9

10

11

12

13

14

15

RR

30 20 10

SpO2

0 1 100 95 90 85 80 1 160 BP

120 80 40 1

Fig. 13.1  Illustration of the cyclic nature of typical cardiorespiratory instability. Oxygen saturation (SpO2) begins to trend downward beginning at 02:30, with gradual rise in respiratory rate (RR). At shortly after 10:00, abrupt drop in SpO2 triggers compensatory mechanisms of tachypnea and tachycardia, resulting in transiently improved oxygen saturation. At 12:30 the cycles begin again. At the fourth cycle at 14:30 hypertension occurs as well. Although the patient is unstable for 5 h beginning at 10:00, intermittent observation not timed to the vital sign nadirs and apices would have not recognized the patient as being across track-and-trigger threshold

whom an MET was called, the IMS detected instability with a mean advance time of 6.3 h.29 In a later phase, index values were available to the staff, who responded to audible alerts when the IMS crossed an instability threshold. In this phase, both the incidence and duration of patient instability was reduced, presumably due to earlier instability recognition and treatment.30 Thus, use of an IMS that aggregates and weights physiologic data within a continuous monitoring environment, coupled with an alerting system and algorithm for clinician action, has shown promise in permitting clinicians to reduce patient instability. The future in assessing dynamic patient-level factors contributing to FTR may lie in perfecting track-and-trigger systems that utilize wireless technology to ­continuously and non-invasively monitor patients for early subtle departures from normality, and utilize aggregate data to develop dynamic index scores that can alert the clinician to attend to the patient for further evaluation and early intervention. Detection will be further improved with the ability to measure metabolic markers of tissue wellness non-invasively and continuously.31,32 Nevertheless, improvements in detection of the dynamic patient-level factors contributing to FTR can only be fully realized when system-level factors as indicated below are also addressed, as improved detection can only positively impact on patient outcomes when coupled with appropriate recognition and timely therapy.33

146

M. Hravnak et al.

Causes of FTR: Hospital- or System-Level Factors Acute-care hospitals are highly complex environments in which many factors contribute to the achievement of quality patient outcomes. Care delivered within such environments is an interdependent and interactive process that is influenced by nurses, physicians, support staff and the infrastructure and technology designed to support patient care. Providing clinically effective care has been defined as doing the right thing in the right way for the right patient at the right time.34 Silber and other colleagues who introduced FTR as an outcome measurement found it to be influenced by characteristics other than patient acuity, such as whether anesthesia care was provided by a board-certified anesthesiologist, or the level of registered nurse staffing.35 Their assumption was that the development of a complication is most strongly influenced by patient-level characteristics, but that once a complication develops, provider and organizational characteristics are more strongly linked to patient survival.1,35 Hospital characteristics originally determined by Silber as contributory to FTR are level of technology, number of beds, average daily census, total number of physicians, physician specialties and board certification, nurse-topatient ratio, nurse-to-bed ratio.1 Additional system-level factors that are described by others, especially within the nursing literature, are system-factors that inhibit recognition of instability and are a call for help, such as inadequate nurse staffing or education.36–38 Using the same surgical patient population as Silber and colleagues, Aiken and colleagues reported that, after adjusting for patient and hospital characteristics, an assignment of one additional patient per nurse was associated with a 7% increase in the likelihood of FTR.7 Both Aiken et al. and Needleman et al.7,9 found significant differences in FTR rates that were directly related to nurse education and the number of patients assigned. Senior nursing staff, certified nurses and lower nurse/patient ratios were all associated with lower FTR rates. Common features of system-level factors hinge on the ability of the healthcare facility to provide both adequate personnel as defined by numbers, education and training, and experience, as well as sufficient resources to permit opportunities for evaluative encounters, process the evaluative information accurately and implement the correct interventions in a timely manner. Linking those common system-level factors and the role of nursing to FTR, Aiken’s conceptual model demonstrates the relationship between patient outcomes to nursing outcomes: burnout, retention and nurse-rated quality of care.7 Nurseassessed quality of care was directly related to effective organizational support provided to the nursing staff. Nurse job satisfaction and retention of experienced nurses improves patient outcomes. Magnet and magnet-like hospitals experiencing adequate nursing resources, nurse autonomy and improved nurse-physician relations reported higher nurse satisfaction, retention and lower FTR.7,39 Nurses rated lower quality of care in the presence of low organizational support independent of staffing levels. Aiken’s work has emphasized the role of the clinical staff to ­improving patient outcomes.

13  Causes of Failure to Rescue

147

A variety of other system-level factors contributing to FTR have also been reported. Since it is assumed that patient instability should be a relatively constant event rate over a 24-h day, whereas staffing and surveillance levels vary, the common finding of diurnal variation of MET activation strongly suggests that system ­factors relate to delayed MET activation and ultimately FTR. Presently, the interaction of these factors and FTR are not well understood. Several studies have reported a pattern characterized by more MET activations during the day shift, compared with the night shift, as well as clustering of MET activations at times associated with routine observations and nursing handover.40–42 One explanation for this is that non-clinical or support staff availability is typically greater during the day shift, which would increase the number of people supporting the registered nurse and visits to patient rooms. Accordingly, it is possible that we are “seeing” the patient with more vigilance during the day. A corollary to this may be that patients are equally unstable during the night hours, but that this instability is missed because of fewer patient encounters by the staff. Staff fatigue may be another contributing factor.43 A model that incorporates both Aiken’s conceptual model for hospital-level ­factors, Silber’s patient and hospital characteristics, and our own observations regarding the static and dynamic patient level factors for FTR is illustrated in Fig. 13.2. Such a model may help to guide the understanding of the complexity of the interrelationships impacting FTR.

Medical Staff Number and Qualifications Hospital-Level Factors

Level of Technology Communication

Nurse: Patient Ratios, Staffing, Skill Mix Static Patient Characteristics and Demographics age, gender, race, procedure, co-morbidities; admission severity of illness Patient-Level Factors Dynamic Physiologic Data single parameters trends of change

Surveillance of Dynamic Patient Characteristics/ Early Detection of Complication Type of monitoring Delay to MET event call

Organizational Support for Nursing Care Resource Adequacy, Nurse autonomy, Nurse control, Nurse/MD relations

Nurse Outcome Fatigue, Burnout, Retention

Process of Care/ Activation of MET

Patient Outcome/ FTR

Fig.  13.2  Conceptual model for causes of failure to rescue incorporating medical emergency team availability

148

M. Hravnak et al.

Summary FTR is a problem that is due to a complex chain of both static and dynamic patient-level factors as well as system-level factors and there is interplay and likely even synergism between factors. Evaluation of the impact of METs and their variants have provided both positive44–48 and ambivalent49–51 results. As the push to minimize hospitalizations for the treatment of acute illness increases and the time of acute care hospitalization decreases with greater pressure to move more potentially unstable patients out of intensive monitoring units into less-well-monitored sites, the risk of FTR will only increase. The often-observed inconsistency in the effectiveness of the MET approach across different healthcare systems might be explained by the many other patient and system-level factors that also impact on FTR and are beyond the scope of the MET. Developing mechanisms to improve the acquisition of continuous, sensitive and specific physiological data to detect instability and support a call for help and eliminating system barriers that inhibit recognition of, and response to, instability, such as staff availability, education and experience,36 are necessary for a successful comprehensive approach to FTR reduction.

References 1. Silber JH, Williams SV, Krakauer H, Schwartz JS. Hospital and patient characteristics associated with death after surgery. A study of adverse occurrence and failure to rescue. Med Care. 1992;30(7):615–629. 2. Agency for Healthcare Research and Quality. Patient Safety Indicators. AHRQ Quality Indicators. Rockville, MD: Agency for Healthcare Research and Quality; 2006. 3. Silber JH, Romano PS, Rosen AK, Wang Y, Even-Shoshan O, Volpp KG. Failure-to-rescue: comparing definitions to measure quality care. Med Care. 2007;45(10):918–925. 4. Schmid A, Hoffman L, Happ MB, Wolf GA, DeVita M. Failure to rescue: a literature review. J Nurs Adm. 2007;37(4):188–198. 5. Aiken LH, Clarke SP, Sloane DM, Lake ET, Cheney T. Effects of hospital care environment and patient mortality and nurse outcomes. J Nurs Adm. 2008;38(5):223–229. 6. Needleman J, Buerhaus P, Mattke S. Nurse-staffing levels and the quality of care in hospitals. N Eng J Med. 2002;346(22):1715-1722. 7. Aiken LH, Clarke SP, Sloane DM, Sochalski J, Silber JH. Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction. JAMA. 2002;288(16):1987–1993. 8. National Quality Forum (NQF). National Voluntary Consensus Standards for NursingSensitive Care: An Initial Performance Measure Set. Washington, DC: National Quality Forum; 2004. 9. Needleman J, Buerhaus PI. Failure to rescue: comparing definitions to measure quality of care. Med Care. 2007;45(10):913–915. 10. Aiken LH, Clarke SP, Chueng RB, Sloan DM, Silber JH. Educational levels of hospital nurses and surgical patient mortality. JAMA. 2003;290(12):1617–1623. 11. Devita MA, Bellomo R, Hillman K, et al. Findings of the first consensus conference on medical emergency teams. Crit Care Med. 2006;34(9):2463–2478. 12. DeVita MA, Smith GB. Rapid response systems: is it the team or the system that is working? Crit Care Med. 2007;35(9):2218–2219.

13  Causes of Failure to Rescue

149

13. Talsma A, Bahl V, Campbell DA. Exploratory analyses of the “failure to rescue” measure: evaluation through medical record review. J Nurs Care Qual. 2008;23(3):202–210. 14. Cuthbertson BH, Boroujerdi M, McKie L, Aucott A, Prescott G. Can physiological variables and early warning scoring systems allow early recognition of the deteriorating surgical patient? Crit Care Med. 2007;35(2):402–409. 15. Bobay KL, Fiorelli KL, Anderson AJ. Failure to rescue: a preliminary study of patient-level factors. J Nurs Care Qual. 2008;23(8):211–215. 16. Bell MB, Konrad D, Granath F, Ekbom A, Martline CR. Prevalence and sensitivity of METcriteria in a Scandinavian University Hospital. Resuscitation. 2006;70(1):66–73. 17. DeVita MA, Braithwaite RS, Mahidhara R, Stuart S, Foraida M, Simmons RL. Use of medical emergency team responses to reduce cardiopulmonary arrests. Qual Saf Health Care. 2004;13(4):251–254. 18. Gao H, McDonnell A, Harrison DA, et al. Systematic review and evaluation of physiological track-and-trigger warning systems for identifying at-risk patients on the ward. Intensive Care Med. 2007;33(4):667–679. 19. Subbe CP, Kruger M, Rutherford P, Gemmel L. Validation of a modified Early Warning Score in medical admissions. QJM. 2001;94(10):521–526. 20. Subbe CP, Gao H, Harrison DA. Reproducibility of physiological track-and-trigger warning systems for identifying at-risk patients on the ward. Intensive Care Med. 2007;33(4):619–624. 21. Smith GB, Prytherch DR, Schmidt PE, Featherstone PI. Review and performance evaluation of aggregate weighted “track and triggertrack-and-trigger” systems. Resuscitation. 2008;77(2):170–179. 22. Prytherch DR, Smith GB, Schmidt P, et al. Calculating early warning scores – a classroom comparison of pen and paper and hand-held computer methods. Resuscitation. 2006;70(2):173–178. 23. Smith GB, Prytherch DR, Schmidt P, et al. Hospital-wide surveillance – a new approach to the early identification and management of the sick patient. Resuscitation. 2006;71(1):19–28. 24. Smith GB, Prytherch DR, Schmidt PE, et  al. Should age be included as a component of  track-and-trigger systems used to identify sick adult patients? Resuscitation. 2008;78(2):109–115. 25. Prytherch DR, Sirl JS, Schmidt P, Featherstone PI, Weaver PC, Smith GB. The use of routine laboratory data to predict in-hospital death in medical admissions. Resuscitation. 2005;66(2):203–207. 26. McGinn TG, Guyatt GH, Wyer PC, Naylor CD, Stiell IG, Richardson WS. Users’ guides to the medical literature XXII: how to use articles about clinical decision rules. JAMA. 2000;284(1):79–84. 27. Galhotra D, DeVita MA, Simmons RL, Dew MA, Members of the Medical Emergency Response Improvement Team (MERIT) Committee. Mature rapid response system and potentially avoidable cardiopulmonary arrests in hospital. Qual Saf Health Care. 2007;16(4):260–265. 28. Tarassenko L, Hann A, Young D. Integrated monitoring and analysis for early warning of patient deterioration. Br J Anaesth. 2006;97(1):64–68. 29. Hravnak M, Edwards L, Clontz A, Valenta C, DeVita M, Pinsky M. Defining the incidence of cardio-respiratory instability in patients in step-down units using an electronic integrated monitoring system. Arch Intern Med. 2008;168(12):1300–1308. 30. Hravnak M, DeVita MA, Edwards L, Clontz A, Valenta C, Pinsky MR. Cardio-respiratory instability before and after implementing an integrated monitoring system. Am J Respir Crit Care Med. 2008;177:A842. 31. Hadian M, Pinsky MR. Functional hemodynamic monitoring. Curr Opin Crit Care. 2007;13(3):318–323. 32. Puyana JC, Pinsky MR. Searching for non-invasive markers of tissue hypoxia. Crit Care. 2007;11(1):116.

150

M. Hravnak et al.

33. Pinsky MR. Hemodynamic evaluation and monitoring in the ICU. Chest. 2007;132(6): 2020–2029. 34. Royal College of Nursing. Clinical Effectiveness: Royal College of Nursing Guide. London: Royal College of Nursing; 1996. 35. Silber JH, Rosenbaum PR, Schwartz JS, Ross RN, Williams SV. Evaluation of the complication rate as a measure of quality of care in coronary artery bypass graft surgery. JAMA. 1995;274(4):317–323. 36. Hinshaw AS. Navigating the perfect storm: balancing a culture of safety with workforce ­challenges. Nurs Res. 2008;57(1):S4–10. 37. Rafferty AM, Clarke SP, Coles J, et  al. Outcomes of variation in hospital nurse staffing in English hospitals: cross-sectional analysis of survey data and discharge records. Int J Nurs Stud. 2007;44(2):175–182. 38. Seago JA, Williamson A, Atwood C. Longitudinal analyses of nurse staffing and patient ­outcomes: more about failure to rescue. J Nurs Adm. 2006;36(1):13–21. 39. Aiken LH, Sloane DM, Lake ET, Sochalski J, Weber AL. Organization and outcomes of ­in-patient AIDS care. Med Care. 1999;37(8):760–772. 40. Jones D, Bellomo R, Bates S, et al. Patient monitoring and the timing of cardiac arrests and medical emergency team calls in a teaching hospital. Intensive Care Med. 2006;32(9):1352–1356. 41. Galhotra S, DeVita MA, Simmons RL, Schmid A, Members of the Medical Emergency Response Improvement Team (MERIT) Committee. Impact of patient monitoring on the diurnal pattern of medical emergency team activation. Crit Care Med. 2006;34(6):1700–1706. 42. Schmid A. Frequency and pattern of medical emergency team activation among medical cardiology patient care units. Crit Care Nurs Q. 2007;30(1):81-84. 43. Manojlovich M, Talsma A. Identifying nursing processes to reduce failure to rescue. J Nurs Adm. 2007;37(11):504–509. 44. Jones D, Bellomo R, Bates S, et al. Long-term effect of a medical emergency team on cardiac arrests in a teaching hospital. Crit Care. 2005;9(6):R808–R815. 45. Jones D, Egi M, Bellomo R, Goldsmith D. Effect of the medical emergency team on long-term mortality following major surgery. Crit Care. 2007;11(1):R12. 46. Bellomo R, Goldsmith D, Uchino S, et al. Prospective controlled trial of effect of medical emergency team on postoperative morbidity and mortality rates. Crit Care Med. 2004;32(4):916–921. 47. Buist M, Harrison J, Abaloz E, Van Dyke S. Six-year audit of cardiac arrests and medical emergency team calls in an Australian outer metropolitan teaching hospital. BMJ. 2007;335(7631):1210–1212. 48. Chen J, Bellomo R, Flabouris A, Hillman K, Finfer S, The MERIT Study Investigators for the Simpson Centre and the ANZICS Clinical Trials Group. The relationship between early emergency team calls and serious adverse events. Crit Care Med. 2009;37:148–153. 49. MERIT Study Investigators. Introduction of the medical emergency team (MET) system: a cluster-randomized controlled trial. Lancet. 2005;365(9477):2091–2097. 50. Pham WBD, Hunt EA JC, Guallar E, Berenholtz S, Pronovost PJ. Rapid response systems: a systematic review. Crit Care Med. 2007;35(5):1238–1243. 51. Chan PS, Khalid A, Longmore LS, Berg RA, Kosiborod M, Spertus JA. Hospital-wide code rates and mortality before and after implementation of a rapid response team. JAMA. 2008;300(21):2506–2513.

Part II

Creating an RRS

Chapter 14

Impact of Hospital Size and Location on Feasibility of RRS Daryl Jones and Rinaldo Bellomo

Keywords  Hospital • Size • Location • Feasibility • Rapid • Response • System Change is not made without inconvenience, even from worse to better. Samuel Johnson We do not experience and thus we have no measure of the disasters we prevent. J.K. Galbraith.

Introduction Modern hospitals are complex institutions that treat increasingly unwell patients with multiple co-morbidities. The aim of such institutions is obviously to improve the outcome of the patients they treat. Unfortunately, this goal is not always achieved and some patients are inadvertently harmed. Studies in the U.S. conducted more than 30 years ago reported that patients admitted to hospital suffer serious adverse events unrelated to their admission diagnosis or underlying medical condition.1,2 Subsequently, similar findings have been reported from Australia,3 Canada,4 and New Zealand.5,6 The frequency and nature of serious adverse events and unexpected deaths is likely to be affected by factors such as the number of patients treated by the institution, the general health status of the patients, the nature of the services provided e.g. trauma versus cardiac surgery versus elective day surgery as well as the presence of quality ­improvement initiatives and hospital governance mechanisms. Thus, universityaffiliated teaching hospitals are likely to experience a greater burden of serious adverse events or unexpected deaths than smaller regional hospitals.

D. Jones (*) Department of Intensive Care, Austin Hospital, Austin Health, Heidelberg 3084, VIC, Australia e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_14, © Springer Science+Business Media, LLC 2011

153

154

D. Jones and R. Bellomo

On the other hand, reports from the U.S. suggest that the risk of operative death is related to the total number of procedures that the hospital performs each year.7 Thus, smaller or medium-sized regional hospitals may have a higher rate of postoperative complications for any given procedure, even if the total number of events is lower. Hence, serious adverse events and unexpected deaths are likely to be a ubiquitous phenomenon that all hospitals must somehow aim to prevent. Many hospitals have implemented Rapid Response Systems (RRS) to identify, review and treat acutely ill ward patients. This chapter outlines the various approaches for the early warning systems that have been employed to prevent serious adverse events and unexpected deaths, and how RRSs can be implemented in different locations and in hospitals of different size.

Antecedents to Serious Adverse Events and Cardiac Arrests, and Criteria for RRS Activation The logic behind the conception of early warning systems to identify deteriorating patients must be understood. In brief, a number of studies have demonstrated that serious adverse events and unexpected deaths are preceded by a period of physiological instability8–10 which manifest in derangements of commonly measured observations and vital signs. These derangements in physiology are often present for some time before deterioration occurs, thus allowing time for appropriate intervention. Accordingly, criteria for the activation of the Rapid Response Team (RRT) are typically based on acute changes in heart rate, respiratory rate, blood ­pressure, conscious state, urine output, and oxygen saturation derived from pulse oximetry.11 In our institution, 82% of the Medical Emergency Team (MET) reviews are initiated by a nurse11. It is perhaps not surprising that analysis of the timing of 2,568 MET reviews that occurred in the 3.5 years after the introduction of the MET ­service at our institution revealed that activation of the MET is more likely around periods of routine nursing observation and nursing shift handovers (Fig. 14.1). These findings emphasize the need for creating simple criteria for activating review of an unwell ward patient, regardless of the personnel that comprise the team that ­performs the review.

Models, Location and Size The structure and personnel comprising the team that reviews acutely unwell ward patients must by necessity vary between hospitals according to local resources, patient acuity, and the frequency of serious adverse events and cardiac arrests. In well-resourced, university-affiliated teaching hospitals, the high degree of patient acuity and complexity will necessitate the requirement for an Intensive

Number of MET calls per half hour

14  Impact of Hospital Size and Location on Feasibility of RRS

155

∗

100 90

∗

∗

80 70 60 50 40 30 20 10 0 1

3

5

7

9

11

13

15

17

19

21

23

Time of day (hours)

Fig. 14.1  Graph illustrating number of MET calls made per half hour over a 24-h period in relation to aspects of daily nursing and medical routine for 2,568 episodes of MET review. Arrows demonstrate periods of nursing handover (�), beginning and end of daily medical shift (¯), as well as periods of routine nursing observations (Ý). The dotted line represents the average number of MET calls made per half-hourly interval. Statistically significant (p < 0.05) levels of increased activity are also indicated (*)

Care-based team. In less well-resourced hospitals or district general hospitals servicing patients of lower acuity, alternative models may be adopted (Table 14.1). In addition to variations in the personnel comprising the team, the goals and objectives of the team may differ between models. Ultimately, the aim of the system is to “improve recognition and response to changes in a patient’s condition”.12 Available evidence suggests that there is considerable heterogeneity in the composition of the response teams throughout the world12. It may also be necessary to customize the threshold for the calling criteria that are used to activate the team. In the U.S. there is wide variation in team composition. Some centers utilize a nurse-led RRT,13–16 while others have implemented a physician-led MET.16,17 In Australia and New Zealand, the typical model is the MET, which is usually led by an intensive care fellow (registrar) and nurse.18,19 More recently, many hospitals in Australia have also introduced intensive care liaison nurses who assist ward nurses in “trouble-shooting” issues regarding aspects of nursing care and equipment related to high-acuity and complex ward patients.20,21

Teaching Hospitals and Academic Medical Centers RRSs in teaching hospitals are frequently METs composed of intensive care based staff. In our institution, the MET is composed of an intensive care fellow and nurse, as well as the admitting medical care fellow of the day.11,19

156

D. Jones and R. Bellomo

Table 14.1  Summary of various models of emergency teams for reviewing acutely unwell hospital patients Description of team Personnel Roles and objectives Advanced medical Intensive care-based Intensive care fellow resuscitation medical emergency Intensive care nurse Safe transfer to critical care team Internal medicine fellow environment if needed e.g., university teaching Respiratory care Formulate and coordinate hospital11,19 practitioner ongoing management plan for patients remaining on the ward Dual level medical emergency Level 1 Identification of patients team26 requiring intensive care Internal medical fellow and fellow review hospital medical officer e.g., secondary referral center Treatment and follow-up of with limited critical care acutely unwell ward patient personnel not requiring intensive care review or admission Level 2 Activated at the discretion of Intensive care fellow and nurse ward staff or following review by medical fellow Emergency department based Emergency department hospital Resuscitation conducted by MET emergency department medical officer e.g., district general hospital27 (Consultant and/or registrar doctor Ongoing management carried attend if available) out by ward doctors or visiting medical practitioner Intensive care liaison Intensive care nurse Review complex patients prior nurse20,21 to MET criteria developing Follow-up of patients e.g., district general discharged from the ICU hospital, in conjunction Consultation service with MET

Teaching hospitals typically provide very frequent MET review for “at-risk” ward patients. Reported studies in large teaching hospitals show an MET calling rate of between 25.822 and 71.323 calls per 1,000 hospital admissions. This equates to MET reviews in 2.6–7.1% of all hospital admissions. In the 3.5 years after the introduction of the MET service at the Austin Hospital, there were 2,568 review episodes (average of 734 calls per year). The distribution of these calls was relatively even throughout the week (Fig. 14.2), indicating that the MET is an important mechanism for managing unwell ward patients in the periods not staffed by the parent unit doctors. Such information also makes it clear that a model created to service a teaching hospital must deliver the service with uniformity 24 h a day, 7 days a week. Although larger institutions have a greater need for an MET service resulting in a greater demand on resources, they are more likely to have a capacity to obtain more resources, making it possible to meet such demands.

14  Impact of Hospital Size and Location on Feasibility of RRS

157

% of overall MET calls

16 14 12 10 8 6 4 2 0 mon

tues

wed

thurs

fri

sat

sun

Day of the week Fig. 14.2  Graph illustrating the distribution throughout the week for the 2,568 episodes of MET review that occurred in the 3.5 years following its introduction to the Austin Hospital

A major role of the MET in a university teaching hospital is to carry out resuscitation of the patient, and to decide the location of their ongoing management after the MET review. If the patient is to remain on the ward, an ongoing management plan is communicated with the medical fellow and the parent unit caring for the patient. Each institution needs to develop protocols for intensive care medical handover of ward patients requiring MET review, as well as protocols for the management of patients receiving multiple MET reviews during a single admission episode. In a University teaching hospital, the MET system may reduce the incidence of unplanned intensive care admissions.11 In other institutions with less critical care personnel and resources, the MET may actually facilitate the process of intensive care referral and admission. A recent review suggested that 13 of the 14 controlled studies that report a benefit in patient outcome associated with the introduction of an RRS, included a physician-led MET in the efferent arm. This may suggest that medical presence in the efferent arm treatment can also affect outcomes of RRT review.24

Secondary Referral Centers Several different models of review have been adopted for patients fulfilling early warning system criteria in secondary referral centers. The MET may be implemented to supersede the existing cardiac arrest team25 so that it reviews all medical emergencies in the hospital. In this model, the criteria for system activation are expanded to include MET criteria similar to those previously described. This approach is an effective way to meet the resource challenge by simply re-deploying those resources to intervene at an earlier time in the evolution of critical illness. As most hospitals have a cardiac arrest team, this is an easy initial way of allocating

158

D. Jones and R. Bellomo

the necessary resources to provide an MET service. In centers with limited critical care personnel, the MET can be divided into two levels or tiers (Table 14.1). The first tier (“MET review – medical”) involves review by the medical fellow or the parent unit staff for patients who fulfill MET call criteria, but are not critically unwell. The second tier (“MET review – intensive care”) is activated at the ­discretion of the nurse initiating the call, or following initial review by the medical fellow.26 It is conceivable that the implementation of an MET could be initially restricted to a limited number of wards. In this model, wards with the highest incidence of cardiac arrests and serious adverse events could be targeted to obtain maximum impact, with minimal outlay of resources. An alternative model is to use a different form of Rapid Response Team (RRT). Thus, the team leader may be an intensive care nurse who is specifically trained to perform the initial review of the unwell ward patient.13,14,16 Involvement of ICU medical staff is then at the discretion of the nurse team leader. In all of these models, one of the aims of the review process is to improve the process of identification and referral for patients who require intensive care management. If a cardiac arrest team is re-deployed to provide an RRT service, it is likely that its workload will increase as it attends more patients. This may require subsequent adjustments in resources. In addition, because the demands made by attendance to acute patient care under more complex circumstances require a wider array of interventions and knowledge, it may require a particular kind of nursing and medical expertise and training. These will have to be assessed in each institution on the basis of patient characteristics and acuity.

District General Hospitals For institutions with very limited or no critical care facilities, the RRT can be comprised of emergency department staff, who review the ward patient and then communicate with the patient’s visiting medical practitioner.27 This system is appropriate for hospitals in which there are no dedicated ward medical staff, and in which the overall number of MET calls is not excessive. Daly and ­co-workers27 reported on the implementation of such a model at the Swan District Hospital (Western Australia). Over a 12-month period, there were 68 reviews in 63 patients. The system reduced the time delay for recognition of a life ­threatening incident, and improved the process of communication with the visiting medical practitioner. This model required training of the emergency department staff in the skills of advanced resuscitation. The experience of these investigators provides proof of concept that an effective RRT service can be provided in a small ­hospital. It also emphasizes the need to use resources that are already available by re-engineering their use and also underlines the need to provide adequate training.

14  Impact of Hospital Size and Location on Feasibility of RRS

159

Small City Hospitals with an Intensive Care Unit Increasingly, RRTs are being implemented into smaller, privately operated ­hospitals in Australia. In 2004, we implemented an MET service in a small ­private city hospital in Melbourne containing a seven-bed ICU with 24-h cover by an ­in-house ICU fellow. The hospital has 120 beds, and is located adjacent to a university teaching hospital with a well-established MET service. The hospital services a mixed population of surgical patients (including open heart surgery) and medical patients (mostly cardiology and oncology). It does not have an emergency ­department, and patient care outside of the ICU is provided by visiting specialists. In response to the occurrence of cardiac arrests (approximately one every month) and other serious adverse events, the medical advisory committee, in conjunction with the ICU staff, introduced an MET using the available resources. The ICU fellow and an ICU nurse became the MET. The hospital’s nursing staff were ­educated to the benefits of the MET and the calling criteria were made known and available throughout the hospital. There was a rapid uptake of the system and over a 6-month period, the number of MET calls became stable at approximately ten per month. The service has proved sustainable and a preliminary data review that over a 6-month period there were only two cardiac arrests and an estimated six patients probably had their lives saved by the availability of the MET. As the number of events is small, it is not possible to demonstrate a statistically significant reduction in cardiac arrests. However, the benefits of the MET service are already visible to nursing and visiting medical staff and the system is already fully supported by both groups of stakeholders. Although the system is not perfect, and may require additional resources as well as better auditing, it is recognized that it delivers a much better level of care than was previously available to acutely ill patients on the hospitals wards.

Summary Despite the best efforts of hospital medical and nursing staff, serious adverse events and unexpected deaths are an unfortunate aspect of medicine in the modern-day ­hospital. Although the overall burden of such events may be higher for teaching ­hospitals, all medical institutions can (and should) develop a system for the ­identification and management of unwell ward patients, and they are likely to benefit from its ­introduction. This system should be tailored to meet the burden of events and to incorporate the most appropriately trained personnel available within the hospital. The fact that a somewhat imperfect system may initially be deployed should not be a justification for inaction. Even an imperfect early ­intervention system is likely to be better than what is normally available in most institutions. The need for ongoing auditing and modification of the system cannot be overemphasized.

160

D. Jones and R. Bellomo

References 1. McGlynn EA, Asch SM, Adams J, et al. The quality of health care delivery to adults in the United States. N Engl J Med. 2003;348:2635–2645. 2. Brennan TA, Leape LL, Laird NM, et al. Incidence of adverse events and negligence in hospitalized patients. Results of the Harvard Medical Practice Study I. N Engl J Med. 1991;324:370–376. 3. Wilson RM, Runciman WB, Gibberd RW, et al. The quality in Australian health care study. Med J Aust. 1995;163:458–471. 4. Baker GR, Norton PG, Flintoft V, et al. The Canadian Adverse Events Study: the incidence of adverse events among hospital patients in Canada. CMAJ. 2004;170(11):1678–1686. 5. Davis P, Lay-Yee R, Briant R, Ali W, Scott A, Schug S. Adverse events in New Zealand public hospitals I: occurrence and impact. N Z Med J. 2002;115(1167):U271. 6. Davis P, Lay-Yee R, Briant R, Scott A. Adverse events in New Zealand public hospitals II: preventability and clinical context. N Z Med J. 2003;116(1183):U624. 7. Birkmyer JD, Siewers AE, Finlayson EVA, et al. Hospital volume and surgical mortality in the United States. N Engl J Med. 2002;346:1128–1137. 8. Buist MD, Jarmaolowski E, Burton PR, et  al. Recognising clinical instability in hospital patients before cardiac arrests or unplanned admissions to intensive care. Med J Aust. 1999;171:22–25. 9. Franklin C, Mathew J. Developing strategies to prevent in-hospital cardiac arrest: analyzing responses of physicians and nurses in the hours before the event. Crit Care Med. 1994;22:244–247. 10. Shein RMH, Hazday N, Pena M, et al. Clinical antecedents to in-hospital cardiopulmonary arrests. Chest. 1990;98:1388–1392. 11. Bellomo R, Goldsmith D, Uchino S, et al. A prospective before-and-after trial of a medical emergency team. Med J Aust. 2003;179:283–287. 12. Øvretveit J, Suffoletto J-A. Improving rapid response systems: progress, issues, and future directions. Jt Comm J Qual Patient Saf. 2007;33:512–519. 13. Chan PS, Khalid A, Longmore LS, Berg RA, Kosiborod M, Spertus JA. Hospital-wide code rates and mortality before and after implementation of a rapid response team. JAMA. 2008;300:2506–2513. 14. Benson L, Mitchell C, Link M, Carlson G, Fisher J. Using an advanced practice nurse model for a rapid response team. Jt Comm J Qual Patient Saf. 2007;33:512–519. 15. Steel AC, Reynolds SF. The growth of rapid response systems. Jt Comm J Qual Patient Saf. 2008;34:489–495. 16. Wood KA, Ranji SR, Ide B, Dracup K. Rapid response systems in adult academic medical centers. Jt Comm J Qual Patient Saf. 2009;35:475–482. 17. DeVita MA, Braithwaite S, Mahidhara R, et al. Use of medical emergency team responses to reduce hospital cardiopulmonary arrests. Qual Saf Health Care. 2004;13:251–254. 18. Jacques T, Harrison GA, McLaws ML. Attitudes towards and evaluation of medical emergency teams: a survey of trainees in intensive care medicine. Anaesth Intensive Care. 2008;36:90–95. 19. Bellomo R, Goldsmith D, Uchino S, et al. Prospective controlled trial of effect of medical emergency team postoperative morbidity and mortality rates. Crit Care Med. 2004;32:916–921. 20. Green A. ICU liaison nurse clinical marker project. Aust Nurs J. 2004;11:27–30. 21. Chaboyer W, Foster MM, Foster M, Kendal E. The intensive care unit liaison nurse: towards a clear role description. Intensive Crit Care Nurs. 2004;20:77–86. 22. Jones D, Mitra B, Barbetti J, Choate K, Leong T, Bellomo R. Increasing the use of an existing medical emergency team in a teaching hospital. Anaesth Intensive Care. 2006;34:731–735. 23. Santiano N, Young L, Hillman K, et al. Analysis of medical emergency team calls comparing subjective to “objective” call criteria. Resuscitation. 2009;80(1):44–49.

14  Impact of Hospital Size and Location on Feasibility of RRS

161

24. Jones D, Bellomo R, DeVita M. Effectiveness of the Medical Emergency Team: the ­importance of dose. Crit Care. 2009;13(5):313. 25. Lee A, Bishop G, Hillman KM, Daffurn K. The Medical Emergency Team. Anaesth Intensive Care. 1995;23:183–186. 26. Casamento AJ, Dunlop C, Jones DA, Duke G. Improving the documentation of medical emergency team reviews. Crit Care Resusc. 2008;10:24–29. 27. Daly FF, Sidney KL, Fatovich DM. The Medical Emergency Team (MET): a model for the district general hospital. Aust N Z J Med. 1998;28:795–798.

Chapter 15

Barriers to the Implementation of RRS Michael A. DeVita and Ken Hillman

Keywords  Potential • Sociological • Political • Barriers • Medical • Emergency • Team • Implementation

Introduction While there is an abundance of literature on the success or otherwise of simple medical interventions such as a new drug or procedure, there is little in the way of evaluating system implementation. The Rapid Response System (RRS) requires implementation across the whole organization – especially close cooperation between senior clinicians and administration.1 This chapter will discuss potential obstacles to, as well as possible enhancement strategies for, the implementation of an RRS across an organization. The major barriers and strategies for overcoming them are noted in Table 15.1.

Sources of Obstacles and Inertia The Medical Emergency Team (MET) was first described in 1995,1 but investigators are still attempting to quantify the types and magnitude of the benefits. A cohort comparison study involving three hospitals demonstrated a reduction in case-mix-adjusted rates of unanticipated admissions to the intensive care unit (ICU).2 Another MET evaluation study demonstrated a significant reduction in the incidence of, and mortality from, unexpected hospital cardiac arrests.3 A further prospective before-and-after trial demonstrated an impressive reduction of

M.A. DeVita (*) Critical Care Medicine University of Pihsbury Medical School, 15213, Pittsburgh, PA, USA e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_15, © Springer Science+Business Media, LLC 2011

163

164

M.A. DeVita and K. Hillman

Table 15.1  Barriers to MET implementation and methods to overcome them Barrier Suggested approach Multidisciplinary event reviews of care antecedent to a Failure to view errors as crisis products of the system rather than individual mistakes Lack of data that METs are Review current data; run-focused trial; multidisciplinary life-saving crisis event reviews Professional silos Multidisciplinary event reviews; teach “system” of care Professional control Emphasize METs to support, not supplant primary team’s coverage; return patients to primary team immediately after event Educational system Emphasize benefit of better supervision of trainees by crisis team responders; track outcomes, delays in current system Financial Utilize current resources to staff MET response; identify frequency of avoiding ICU admission; identify mortality benefit to offset cost

in-hospital cardiac arrests, deaths following cardiac arrest, and overall in-hospital mortality after the introduction of an MET.4 Although these and other studies are preliminary in the sense that they are not randomized prospective placebo-­ controlled clinical trials, they nevertheless provide considerable support for the concept of a planned system response to crises that would reliably rescue patients as they deteriorate. A larger-cluster, randomized study has demonstrated a reduction in mortality in the hospitals with an RRS using an MET as responders5 as well as insight into factors such as the inadequacies of manual vital-sign recording and the difficulty changing the culture of an organization.6–7 While many hospitals across North America, Australasia and Europe have now implemented such Rapid Response Systems (RRSs), why did it take so long, and why don’t all hospitals have such a fundamental and soundly based system to rapidly detect and respond to at-risk and deteriorating patients? The barriers to the introduction of an RRS have a number of cultural issues across organizations that are difficult to discern and to overcome, and they have not been well studied or described. The first barrier is the failure to accept that almost all errors are a predictable result of the system of care that permits them to occur. While healthcare delivery may work well in providing individual patient clinical care, few systems function across existing healthcare “silos” – the professional groupings such as nursing or internal medicine physicians, or geographic groupings like emergency rooms or intensive care units (ICUs). Quality work within silos tends to be relatively easy to foster because members of a group tend to have common incentives and disincentives. Hospital-wide systems, however, are more difficult to effectively introduce and maintain in part because the framework for interdisciplinary systems’ improvement is relatively new to healthcare organizations, and because the members of diverse groups may have conflicting incentives. This perspective sometimes fosters the blaming of an individual rather

15  Barriers to the Implementation of RRS

165

than looking at it from a systematic perspective. Overcoming the perception that errors are not systematic is essential to creating an effective crisis response system. The RRS is a hospital-wide patient safety system. It assumes that errors will be made, and provides an important (and potentially life-saving) mechanism for the system to recover from failure and prevent the deterioration of a patient’s condition, irrespective of whether the deterioration was due to an error of omission or commission. The RRS requires interdisciplinary resources and teamwork. It presupposes that the system views these events as relatively common and preventable, as well as worth preventing and worth the associated costs. In other words, the hospital system of care must prioritize patient safety and view errors as a problem within the system, instead of an individual error. A second barrier may be the general reluctance of health to take up new interventions even when there is an evidence base as there now has been with the RRS.2–8 None of these studies focus on other factors that may impact on the success or otherwise of implementing an organization-wide system, such as the culture of the organization. Some might argue that a conventional placebo-controlled trial for METs is impossible because withholding early intervention in half of the recruited seriously ill patients may be unethical. Thus, no trial will truly be controlled or be a “pure placebo.” Furthermore, the MERIT study showed (among other things) that it is difficult to control an intervention that spans an entire healthcare system with individual hospitals as the randomized unit. In this study, the investigators found that both intervention and “control” hospitals had similarly varying degrees of MET intervention6. A third barrier is the fact that existence of entrenched professional silos within hospitals can create barriers to collaboration. Most hospital clinicians were trained in a system in which only their specialty is taught. This creates cultural and intellectual isolation as well as increasing incompetence in areas not related to their own practice. Specialist clinicians become increasingly knowledgeable and confident in their own silos, but remain relatively ignorant of current practice outside that silo or of interactions between the professional silos. The intellectual and role isolationism sets up a system of ownership, competition, and egocentrism, and is perhaps the foundation for blame when things go wrong. Similarly, the “healthcare team,” consisting of other healthcare workers, is often deficient because they are rarely trained together and sometimes do not cooperate in system-improvement activities. Better models for teamwork exist in sports or in the military. For example, the aviation industry sees itself as a global team continually striving to improve effectiveness and reduce error. A team learns and practices together before working as a team; a prime example is in sports, where there is a long history of effective training to improve effectiveness and reduce error among those competing together as part of a team. In contrast, members of the healthcare professions tend to view themselves first as a physician, or nurse, rather than a team member. This cognitive “set” can prevent individuals from taking actions that are within their capability but outside the traditional boundaries of their profession.

166

M.A. DeVita and K. Hillman

Another fundamental problem is that healthcare education – while outstanding in training diagnosis and therapeutics – is deficient in teaching the “system” of care. Instead, it focuses on diseases, diagnoses, and treatments, as well as procedural skills such as setting up a ventilator, inserting a central line, or changing a dressing. Training programs traditionally do not emphasize the “healthcare system”: how a hospital works, including the hospital hierarchy; the roles and responsibilities of various staff; the interactions within the system; and the informatics infrastructure. Implementing systematic change is often left to health graduates to learn while on the job. This sets up system “blindness,” where members of the healthcare professions may not trust the environment within which they are working, which leads them to set up their own methods for “getting around the system to get things done right.” With this mindset, the system is the problem, not the solution. Medicine is relatively resistant to change. For example, it took trauma systems 10 years before they demonstrated a decrease in mortality.9–11 But there is also little acknowledgment or understanding of the complexity of healthcare and therefore little understanding of implementation strategies for any new process. The identification of facilitators and impeders is, for the most part, through personal experience, and only for those “trying to get something done.”12–15

Foundations for System Change Some recent social changes are laying the foundation for in-hospital transformations. First, recent publications and scandals over potentially preventable deaths in the healthcare system have highlighted the frequency of errors and the harm that they cause.16 To Err Is Human,17 a book published by the Institute of Medicine in the United States, previously highlighted the under-publicized poor safety record of the US healthcare system. As a result, society now expects and is demanding more safety from the healthcare system. Federal policy is beginning to shift, as constituencies demand safer care and greater accountability. This change is occurring globally. National safety bodies oversee strict standards for drugs and devices, but currently there is little in the way of evaluating and imposing standards around health systems. In the past decade the Joint Council on Accreditation of Healthcare Organizations (JCAHO) in the United States has introduced tough new safety system audits.18 Their initiatives coincided with a tragic death in a hospital that had very recently been audited and accredited by the organization, which led some to conclude that the auditing system itself needed repair.19–20 The US Food and Drug Administration has altered its reporting mechanism and its methodology for notification of important drug error concerns. They now observe for sources of medication error and put pressure on manufacturers to alter packaging, labeling claims, and marketing approaches to prevent systematic sources of clinical errors and harm. Healthcare marketing strategies have also changed. Healthcare buyers are working together to get the best value instead of the best cost. For example, the Leapfrog

15  Barriers to the Implementation of RRS

167

group in the United States has defined system standards and care goals that have prompted providers to alter their approach to care delivery, marketing, and data collection.21–23 Senior healthcare officials are now attentive to safety as an important indicator of quality of care within a hospital. Thus a number of agencies in the US, Australia, and Europe are showing interest in the concept of RRS. For example, the Institute for Healthcare Improvement is providing courses on METs, and JCAHO has insisted that each accredited hospital have a system to identify and respond to the seriously ill. Both the federal and state governments in Australia are currently sponsoring implementation of the RRS concept and are advocating the MET model. Because safety is a goal for all caregiver organizations, these forces are leading administrators and caregivers to recognize the possible benefit of the RRS for their institution and their patients. It would also demonstrate that they are practicing according to newly emerging practice patterns and safety initiatives.

Impediments Within the Hospital There are a number of impediments within the hospital that may challenge the implementation of an MET system. The first is cost. For at least the last decade in the US, there has been a huge focus on cutting costs. However, as noted above, there is now a shifting focus to safety. While cost is and will always remain an issue, the balance is changing from favoring cost considerations to a new concentration on quality. The RRS seems to increase cost because they appear to require new equipment and staffing. When an RRS is undertaken, one should expect a three- to fivefold rise in the frequency of calls to seriously ill patients,24 although the majority will not be cardiac arrest events.1–25 Based on those numbers, it is easy to predict the staffing required for emergency stabilization. Medical Emergency Teams often include one or more intensive care or emergency medicine physicians and nurses, an anesthesiologist or nurse anesthetist, and one or more respiratory therapists. In contrast, a Rapid Response Team (RRT), advocated by the Institute for Healthcare Improvement, does not utilize a physician, instead relying on senior critical care nurses to perform an assessment and triage. They may call in a physician as needed. This triaging system enables the hospital to expend the costliest staff for situations where they are most needed. In addition, other members of staff who are not part of the team will respond and attempt to help manage the crisis; this activity is usually in addition to the other responsibilities they have. This perception of added work is a barrier to implementation and may stop discussion before accurate estimates of cost and benefit can be analyzed. It is estimated that training every staff member of a hospital to deal with cardiac arrests would cost over $500,000 per survivor.26 On the other hand, an RRS aims to concentrate a small number of experts who respond to all hospital emergencies, rather than universal training for all staff in basic cardio-pulmonary resuscitation. In this sense, RRSs will save costs and improve care.

168

M.A. DeVita and K. Hillman

A second impediment is that crisis teams intervene for patients who are usually being cared for by other individuals. This raises two issues. The first is power: Who is in “control” of the patient’s care? If one group is already treating the patient, calling in a second group sets up a conflict regarding who is in charge, with the implied question of “who is better?” The second issue is based on historical perceptions that each team should give “total” care to its patients. In this model, calling for help may be perceived as a sign of weakness, perhaps both emotional and intellectual, implying that the caller is somehow not equipped to deal with the situation. These attitudes, which promote barriers between clinical services, need to be removed; working across traditional silos is required for successful implementation of an RRS. Because RRSs require cultural as well as behavioral change, hospital nursing, physician, and administration leadership is required to make the system work. These leaders are needed for political support and for the required funding needed for the program. Without hospital leadership forcefully advocating for improved care of patients in crisis, and finding the resources needed to demonstrate both the need for and benefit of this service, the project is unlikely to succeed. Senior colleagues in the allied health professions are key allies in culture change – in particular, senior nurses and physician leaders. Staff nurses and doctors working in the hospital will not participate fully in the project without support from their leadership. The work occurring at the crisis response requires a significant influx of nursing and physician support; often, the support arrives from other areas of the hospital. This shift of work responsibilities will be in addition to other responsibilities and so may be resisted. The leadership has to view the larger hospital perspective and be able to allocate resources that can both handle the added workload and have the skills necessary for the management of hospital crises. To protect these resources from added work, the leaders may balk at lending their support. Opting for the status quo – especially when it seems that individuals, rather than systems, are the cause of crisis events – is often the politically easier course to take. We have observed that the role perception of caregivers is a barrier to implementation and successful use of an MET or RRT. Professional differences may create unanticipated barriers to implementation of an RRS. For example, nursing staff have a culture of recording patient vital signs and then reporting it to medical colleagues rather than acting on the findings directly. This is due in part to a long history of not being empowered to translate concern into action. In this process, nurses who find a low blood pressure, will try to contact the physician responsible for that patient rather than immediately act. This may create unnecessary delays in responding to a change in patient status. Usually this is not a problem; however, when the finding is critical, delays can allow harm to occur. The exception to the rule of reporting along traditional lines to a responsible physician is when the patient is found without a pulse or respiration; then the nurse may act to activate a crisis response to deal with the patient. In contrast, an RRS empowers nurses to act immediately to bring a crisis response. Nurses (and physicians) may feel uncomfortable in participating in a new process that appears to change the traditional role of the nurse, and they may be reluctant to take on responsibilities that are not

15  Barriers to the Implementation of RRS

169

traditionally within their own boundaries. We have observed experienced ­individuals in our human simulator crisis team training course avoiding important tasks that are “not their job” – for example, no one may support respiration for a dyspneic patient because that is the job of anesthesia or critical care staff. Creating new processes for MET or RRT responses will challenge traditional roles and responsibilities, which is a potential barrier to the system’s implementation. Just as remaining mindful of one’s profession may prove a barrier to the correct response, being mindful that one’s performance may be criticized will alter behavior. Junior medical staff, who are trying to learn and impress their seniors, are likely to see calling for help as a sign of weakness. Junior doctors traditionally look after their own patients no matter how sick they are and call for help only when they have insight into their own inadequacies and the potential consequence of this for patients. The problem is that until one is knowledgeable, one can neither appreciate these possibilities and dangers nor recognize when he is “in trouble.” Medical trainees often do not possess the skills, knowledge, or experience necessary to recognize and resuscitate seriously ill patients.27,  28 Physicians, nurses, and others in the hospital have operated in hierarchical and separate teams, creating an important barrier that needs to be acknowledged and understood. A nurse may recognize trouble and ask for input from other, more experienced nurses; when the situation is deemed to be beyond the capabilities of the nursing chain, they will call a physician. That physician will respond based on his or her skills and priorities. When that person finds that the problem exceeds his skills, a second call is made, and a third or a fourth, until finally all the resources are assembled. This knowledge and skills ladder is hierarchical in nature and builds in delays in response. Even though there are well-recognized delays, caregivers may be reluctant to go outside the chain of command when a crisis occurs. For the RRS to work, that chain must be identified as a systematic barrier to rapid and effective care of a patient in crisis. This tacit statement that the current hierarchy is a source of error and harm could prove a barrier to accomplishing implementation of the system. Traditional medical and nursing education is also a potential impediment to the RRS, since it teaches that learning is best when one thinks through a problem on one’s own, and then learns from the successes and mistakes. Crisis-response teams optimally intervene rapidly and with appropriate expertise. The response may be so fast that trainees may not have the opportunity to think it through and decide for themselves what are the most important considerations and best course of action. There may be a belief that not allowing mistakes somehow impedes learning. Since many advocate that education is an important component of medical and nursing care, creating a situation where teachers believe learning is not possible will cause concern. The belief that an error results from an individual’s actions, and not because of a failure of the “system” to prevent error, is a major barrier to RRS implementation.29 Recognition that a faulty system permits error and harm to a patient is relatively advanced thinking in today’s medical world. The Morbidity and Mortality (M & M) conference structure often concentrates on individual errors in judgment

170

M.A. DeVita and K. Hillman

or skill rather than contextualizing it into “system thinking.” The failure to recognize that “a system that permits errors” is a faulty system creates a fundamental barrier to implementing METs or RRTs. It seems to be easier to blame an individual than a system in such traditional M & M conferences because of the obvious connection between the caregiver and the patient, and so the need for a “system fix” can remain unrecognized. As long as this perception of “problem individuals” persists, there will be a significant barrier to creating a new process that threatens established hierarchy and current practice patterns.

Strategies to Overcome Hurdles The authors have noted a number of strategies to overcome a variety of hurdles. Some strategies may operate optimally during the implementation phase and others are more appropriate for system maintenance. There is no data to support which barriers need to be overcome first, nor any to determine which strategies are most effective, easiest, or surest. Both authors have found that hospital “stories” of failures or near misses can become the basis for a forum where an alternative system approach can be discussed. There are two types of stories that tend to work. The first is the “cause célèbre,” in which some tragic event occurs, demanding analysis and action – for example, the wife of a staff physician who dies from an error involving an opioid overdose, after which careful analysis reveals that the death was due to both a life-threatening situation (the opioid dosing) but also to the hospital system’s failure to respond to the event. The second is a compendium of smaller stories – for example, analysis of a series of people who had adverse events in a 6-month interval all due to opioid adverse events can be a powerful motivator to action. While it is easy to attribute a single adverse event to a single faulty practitioner, analysis of many events together will demonstrate myriad causes. We glibly call this “following the parade of the cadavers” to highlight how tracking similar events can lead to observations that may not be evident when singular events are scrutinized. This method makes it clear that the system is faulty if so many mistakes can occur, each mistake by a different individual, at a different step in the process, and at different times. This makes the need for a system response more evident. It is in this context that the person promoting an MET or RRT response must propose a system that may prevent serious adverse events no matter what their causes. Successes with one type of crisis tend to lead to the recognition by others that the system may work well for other types of problems. In this way, the MET and RRT response becomes the system’s “goalkeeper” to rescue patients when other error-detection mechanisms have failed. A major selling point for the RRS is that they can prevent deaths and serious complications from myriad causes that result in patient deterioration. Using data is often effective for motivating change; data is impersonal and unbiased, and can track harm and the benefits of process change. Possibly the most

15  Barriers to the Implementation of RRS

171

important data to track is the frequency and duration of delays in care that occurred prior to a crisis event. Reviewing the 24 h prior to a crisis or cardiac arrest event for delays in delivery of appropriate treatment is possible once standards for response times and severity of illness are created. Crisis criteria enable reviewers to determine how long after the criteria were met that it takes to deliver the definitive treatment, or even get the responsible and capable person to the bedside of the patient in crisis.24,27 Analysts can then graph delays by frequency, duration, location, service, day of week, time of day, etc. These data provide a powerful tool to recognize system deficiencies and motivate process change. We believe that delays in delivering definitive treatments are the hallmarks of hospital care systems without an RRS. Continued data collection and analysis will demonstrate effectiveness of process change in removing delays in care from an institution. Both authors have encountered individuals who remain resistant to group efforts to solve medical crisis situations. Data from the RRS can, in itself, facilitate the implementation and maintenance process,30 and may be processed and targeted to every level of the organization. Specific patient details are provided for individual clinicians, while departments, divisions, and the hospital would review aggregated, de-identified data. Data includes details of MET calls, number of deaths, cardiac arrests, and unplanned ICU admissions in which MET criteria had been met but no call made; these are called “potentially preventable events.” A graphic depiction of duration and frequency of delays from the onset of a crisis situation (determined by satisfying crisis criteria) can help target areas that need extra educational effort. Another important indicator is the number of MET calls/1,000 hospital admissions as an increasing number of calls is strongly correlated with a significant reduction in deaths and cardiac arrests.5 Some clinicians or departments may not like having “unconsulted” doctors assume care of their patients, creating a political barrier to implementation of the RRS. One method to overcome this concern and facilitate the implementation of a hospital emergency system is to remind staff that calling for help in cases of complex acute medicine is no different from seeking consultations from colleagues in different specialties. Guarantees that the patient’s care will remain under the control of the “primary” caregivers after the crisis is resolved (or even during it) can also foster an environment where an MET or RRT may be implemented successfully. One author discovered that the director of a residency education program did nothing to promote use of an MET, but permitted the use of METs with the caveat that it was the responsibility of the MET response organizers to “make sure my residents get taught.” After the trainees reported decreased stress when caring for sick patients, and improved understanding of the management of suddenly critically ill patients – learned during observation of and working with the MET responders – the education director relented. Residents found that with METs they could learn in a context where they had support, and they could observe the “right way”; on the other hand, they did feel that losing the opportunity to learn by doing detracted from their educational program. The RRS can help to cost-effectively treat patients. For example, nursing infrastructure required to care for seriously ill patients on a general ward can detract

172

M.A. DeVita and K. Hillman

from other routine activities. The MET and RRT response not only provides timely and expert care at all times, but can decrease the burden on the rest of the staff having to care for the seriously ill in inappropriate environments. The average duration of an MET response is approximately 30 min.1 Thus a patient who is deteriorating can be assessed, treated, and, if necessary, triaged in short order. This allows the nursing unit routine to remain relatively undisturbed in spite of the crisis event. The RRS improves the safety of patients on general wards with patients in crisis. Both authors have observed what we call “domino codes,” where a second patient medical crisis occurs because staff either fail to deliver treatment or adequately monitor patients while they are coping with the first patient in crisis. We believe that an RRS decreases “domino codes” because they swiftly bring new critical-care resources to the unit (which is suddenly understaffed to meet the workload), and they either rapidly resolve the crisis or triage the patient elsewhere. Identifying these unit-based resource issues and recognizing the successes that occur after RRS implementation are great motivators to overcome pockets of resistance. Adding or increasing MET or RRT responses means increased work for the response team, because each response brings critical care workers from other areas to treat a single patient. It may seem a daunting task to marshal the resources to take on the task of responding to all patient medical crisis events. Both authors’ hospitals offer a service similar to the traditional cardiac arrest team: that is, resuscitate first, discuss after, and return the patient to the care of the primary doctor immediately after the crisis is resolved. By using the cardiac arrest team, no new resources need to be identified – the current resources are just taxed a bit more. Recognition that many emergency patients may go on to become cardiac arrest patients can help motivate responders to arrive early, before the heart stops. Early calls improve outcome and decrease the effort needed to restore homeostasis. Critical care admissions may be avoided, decreasing the downstream work of the ICU staff. On balance, responding to crises early and effectively reduces workload for hospital staff. In hospitals, education is a continuous and essential activity needed to maintain quality care. To improve learning about the importance of an RRS and the behaviors needed to foster one, it helps to find opportunities to educate staff about the RRS. For example, new staff orientation should include a module on crisis management and appropriate use of MET or RRT capability. New staff will accept the process based on “accepted and expected” practice. In contrast, existing staff need to be re-educated in why a systematic approach makes more sense than ad hoc processes to build crisis intervention teams for each critical event. New systems of care must be perceived by existing staff to be both easier and more effective than current practice; otherwise, the new process will fail. RRS rules must be made simple and objective. Any “interpretable” rules will not be consistently followed; the MET or RRT response should be viewed as “one-stop-shopping” for management of any medical crisis. Buist et al have found that “caregivers ‘worried’ about a patient” is a common trigger for an MET response.3 Reliably rescuing staff members who have patient concerns will reinforce use of the response in the future. For managers of the RRS, positive and negative reinforcement can foster culture change. Congratulate those who “call for help”; tell them, their bosses, and their

15  Barriers to the Implementation of RRS

173

c­ olleagues how a life was saved. The University of Pittsburgh used e-mail for this feedback to effect culture change.24 Private notification of superiors about failure to trigger the MET or RRT response will demonstrate the impact of the failure. It is essential that RRS responders reinforce the call as well; it will do no good if superiors praise calling for help while the RRS responders criticize the same action. Every criticism of an RRS call must trigger re-education of those individuals. All members of the institution must view an RRS call as an act of heroism; it is putting patient care above ego. Other reinforcing educational strategies include placing team-triggering criteria in all parts of the hospital and on pocket cards for responders and staff, and notifying patients and families about the MET system to protect the patients. As noted previously, “calling for help” can be perceived as a sign of weakness, and this perception is promoted when the criteria for what constitutes a crisis are subjective or ambiguous. As such, the request is an indirect measure of competency: the person perceiving the patient in crisis defines crisis by his or her inability to manage the situation alone. To prevent this barrier, objective and readily recognized crisis criteria must be adopted. With objective criteria, the person who finds the crisis is merely notifying others that the crisis exists (following hospital policy), and this does not imply that the person’s ability to manage the situation is inadequate. Instead, the MET or RRT call becomes a mark of excellence in patient care and clinical judgment. Hospitals that have utilized crisis criteria have shown an increase in MET and RRT response frequency and a decrease in delays to treatment.3,24

Summary Implementing a Rapid Response System in a hospital will likely alter the culture of care and threaten the status quo. There are many potential psychological, emotional, sociological, and economic barriers to bringing a new system of care to a stable environment. Nevertheless, strong data indicate that such a system of care will decrease unexpected mortality in a variety of hospital settings. Therefore, the key question with which hospital leadership must grapple is how to implement the system, not whether to implement it. There is no strong data to define the particular barriers, nor how to overcome them. Instead, in this chapter we have proposed strategies that have been effective in our hospital environments and may benefit others as well.

References 1. Lee A, Bishop G, Hillman KM, Daffurn K. The medical emergency team. Anaesth Intensive Care. 1995;23:183–186. 2. Bristow PJ, Hillman KM, Chey T, et al. Rates of in-hospital arrests, deaths, and intensive care admissions: the effect of a medical emergency team. Med J Aust. 2000;173:236–240.

174

M.A. DeVita and K. Hillman

3. Buist MD, Moore GE, Bernard SA, Waxman BP, Anderson JN, Nguyen TV. Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. BMJ. 2002;324:387–390. 4. Bellomo R, Goldsmith D. Postoperative serious adverse events in a teaching hospital: a prospective study. Med J Aust. 2002;176:216–218. 5. Chen J, Flabouris A, Bellomo R, Hillman K, Finfer S, The MERIT Study Investigators in the Simpson Centre and the ANZICS Clinical Trials Group. The relationship between early emergency team calls and serious adverse events. Crit Care Med. 2009;37:148–153. 6. MERIT Study Investigators. Introduction of the medical emergency team: a cluster-randomised controlled trial. Lancet. 2005;365:2091–2097. 7. Chen J, Flabouris A, Bellomo R, Hillman K, Finfer S, The MERIT Study Investigators in the Simpson Centre and the ANZICS Clinical Trials Group. The impact of introducing medical emergency teams on the documentation of vital signs. Resuscitation. 2009;80:35–43. 8. Bellomo R, Goldsmith D, Uchino S, et al. Prospective controlled trial of effects of medical emergency team on postoperative morbidity and mortality rates. Crit Care Med. 2004;32:916–921. 9. Nathens AB, Jurkovich GJ, Cummings P, Rivara FP, Maier RV. The effect of organized systems of trauma care on motor vehicle crash mortality. JAMA. 2000;283:1990–1994. 10. Lecky F, Woodford M, Yates DW, The UK Trauma Audit Research Network. Trends in trauma care in England and Wales 1989–97. Lancet. 2000;355:1771–1775. 11. Mullins RJ, Veum-Stone J, Helfand M, Zimmer-Gembeck M, Trunkey D. Outcome of hospitalized patients after institution of a trauma system in an urban area. JAMA. 1994;27:1919–1924. 12. Plsek PE, Greenhalgh T. The challenge of complexity in health care. BMJ. 2001;323:625–628. 13. Eccles FR, Grimshaw J. Why does primary care need more implementation research? Fam Pract. 2001;18:353–355. 14. Dellinger RP. Fundamental critical care support: another merit badge or more? Crit Care Med. 1996;24:556–557. 15. Cook RI, Render M, Woods DD. Gaps in the continuity of care and progress on patient safety. BMJ. 2000;320:791–794. 16. Schneider EC, Lieberman T. Publicly disclosed information about the quality of health care: response of the U.S. public. Qual Saf Health Care. 2001;10:96–103. 17. Institute of Medicine. To Err is Human: Building a Safer Health System. In: Kohn LT, Corrigan JM, Donaldson MS, eds. Washington, DC: National Academy Press; 2000. 18. Beyea SC. Implications of the 2004 National Patient Safety Goals. AORN J. 2003;78:834–836. 19. Griffith JR, Knutzen SR, Alexander JA. Structural versus outcomes measures in hospitals: a comparison of Joint Commission and Medicare outcomes scores in hospitals. Qual Manag Health Care. 2002;10:29–38. 20. Landis NT. Government finds fault with hospital quality review by joint commission and states. Am J Health Syst Pharm. 1999;56:1699–1700. 21. Scanlon M. Computer physician order entry and the real world: we’re only humans. Jt Comm J Qual Saf. 2004;30:342–346. 22. Pugliese G. In search of safety: an interview with Gina Pugliese. Interview by Alison P. Smith. Nurs Econ. 2002;20:6–12. 23. Cors WK. Physician executives must leap with the frog. Accountability for safety and quality ultimately lie with the doctors in charge. Physician Exec. 2001;27:14–16. 24. Foraida M, DeVita M, Braithwaite RS, Stuart S, Brooks MM, Simmons RL. Improving the utilization of medical crisis teams (Condition C) at an urban tertiary care hospital. J Crit Care. 2003;18:87–94. 25. Hourihan F, Bishop G, Hillman KM, Daffurn K, Lee A. The medical emergency team: a new strategy to identify and intervene in high-risk patients. Clin Intensive Care. 1995;6:269–272. 26. Lee KH, Angus DC, Abramson NS. Cardio-pulmonary resuscitation: what cost to cheat death? Crit Care Med. 1996;24:2046–2052. 27. McQuillan P, Pilkington S, Alan A, et  al. Confidential inquiry into quality of care before admission to intensive care. BMJ. 1998;316:1853–1858.

15  Barriers to the Implementation of RRS

175

28. Goldhill DR, White SA, Sumner A. Physiological values and procedures in the 24 hours before ICU admission from the ward. Anaesthesia. 1999;54:529–534. 29. Reason J. Combating omission errors through task analysis and good reminders. Qual Saf Health Care. 2002;11:40–44. 30. Hillman K, Alexandrou E, Flabouris M, et al. Clinical outcome indicators in acute hospital medicine. Clin Intensive Care. 2000;11:89–94.

Chapter 16

An Overview of the Afferent Limb Gary B. Smith and David R. Prytherch

Keywords  Afferent limb • Rapid response systems

Introduction The afferent limb of a rapid response system (RRS) is responsible for monitoring the patient, detecting any deterioration and triggering a response (i.e. the efferent arm of an RRS).1 The provision of a fully functional, effective afferent limb would seem simple. However, there is good evidence that failures in the individual components of the afferent limb are common and frequently result in potentially avoidable, adverse clinical outcomes, such as cardiac arrest, death, unanticipated intensive care unit (ICU) admission and morbidity.2–24 Failures include inadequate frequency of observations, incomplete observation sets, lack of knowledge of meaning of abnormal values, and failure to call for assistance when required.2–24 For example, the UK National Confidential Enquiry into Patient Outcomes and Death (NCEPOD) report “An Acute Problem” found that in 439 acute medical emergency patients who died in the ICU after admission from general wards, notes seldom contained written requests regarding the type and frequency of physiological observations to be measured.4 Additionally, vital signs datasets were often incomplete. Pulse rate, blood pressure and temperature were the most frequently recorded variables and breathing rate the least.4 Instructions giving parameters that should trigger a patient review were rarely documented.4 In the Australian MERIT study (n = 5,899), 81% of patients without a documented “Do Not Attempt Resuscitation” (DNAR) order had an incomplete or absent record of heart rate, blood pressure and respiratory rate in the 15-min period before an unanticipated cardiac arrest, unplanned ICU admission, or an unexpected death.12 This study also showed that in

G.B. Smith (*) Department of Critical Care, Queen Alexandra Hospital, Portsmouth, PO6 3LY, UK e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_16, © Springer Science+Business Media, LLC 2011

177

178

G.B. Smith and D.R. Prytherch

hospitals where medical emergency teams (METs) were present, and there were documented physiological MET criteria present >15 min before the above adverse events, the team was called in only 41% of cases.12 Failures to call may be because of lack of recognition of patient deterioration, lack of knowledge of the escalation protocol, incorrect clinical judgement, a lack of confidence in escalating or worry on the part of the caller that they might receive criticism (perhaps based on previous experiences).24 The activation of rapid response teams also shows a circadian pattern, suggesting that processes for identifying patient deterioration may vary throughout the day.25–26 Monitored units (those with continuous monitoring of at least either pulse oximetry or electrocardiography, a central monitor station and alarms at both the central station and the patient room) have more activations than unmonitored units, but the pattern of activation is essentially identical.26 Staff may not adequately inform each other regarding patients’ care, particularly during handovers and transfers.5,6,27 Andrews has shown that quantifiable evidence is the most effective means of referring patients to doctors, and that using an early warning score (EWS), rather than reporting changes in individual vital signs, increases the chances of achieving good communication about a deteriorating patient.28

Improving the Function of the Afferent Limb Improving Regular Monitoring and Assessment There should be a clear, documented vital signs measurement plan for each patient containing unambiguous instructions regarding the variables to be measured and the frequency of measurement.2,29,30 This is probably most easily implemented by each hospital having an organization-wide policy that emphasizes the minimum uniform measurement plan for all patients. The plan should also take account of the patient’s diagnosis, co-morbidities, treatment plan and severity of illness, and should modify the frequency of measurement and level of care accordingly.2,29,30 There is a dearth of research about the frequency with which patients should have their vital signs measured;17,18 however, it makes sense that the frequency should adapt to changing clinical situations. In the UK, current recommendations suggest that the minimum frequency for observations should be 12-h and should include at least heart rate, breathing rate, systolic blood pressure, level of consciousness, SpO2 and temperature.29 However, some attendees at a recent Consensus Conference on Monitoring preferred at least every 6 h.2 The use of an early warning score (EWS) and the implementation of a Rapid Response Team have both been associated with an increase in the frequency of vital signs measurements.15,31,32 Alterations to the monitoring frequency, especially ­reductions, should be made by senior staff.2,29 Vital signs datasets should be complete every time, as physiological variables are inter-related, often by physiological compensatory mechanisms. To give context

16  An Overview of the Afferent Limb

179

to measurements of peripheral saturation measurements (SpO2), the percentage of oxygen concentration being delivered should also be recorded.

Ensuring Vital Signs Measurements Are Accurate In the developed world, most vital signs are now measured by machine. It is obvious that these machines need to be properly calibrated and well maintained. Staff using them should have the appropriate training and should attain the necessary competence in monitoring, measurement and interpretation.33 They should also know the limits of accuracy and applicability of any machine that they use. For example, factors such as nail varnish influence the accuracy of pulse oximeters;34–35 the size of blood pressure cuff in relation to the patient’s arm influences the accuracy of BP measurements.36 Measurements should be standardized as much as possible – patients should ideally be in the same position and, where a measurement is made in a particular limb, the same limb should always be used. Breathing rate is generally not measured by machine outside of critical care areas, but rather by observation. There is considerable evidence that breathing rate is measured poorly and subject to poor inter-observer reliability.4,9,15,19,21,22,31,37–41

Ensuring Vital Signs Measurements Are Accurately Recorded An important role of the vital signs chart is allowing the early recognition of patient deterioration. To do this, it must be immediately available, accurate, up-to-date and legible. These goals are often not met. It is common for many different versions of vital signs charts to exist within a single hospital. At least within a given institution, and perhaps within a whole health system, it would seem sensible to have a single format. There remains a requirement to define the optimal format for a vital signs chart, as their design can affect interpretation and, therefore, the detection of deterioration.42 If an EWS system is used, calculations and recording must also be accurate and legible. Some hospitals have introduced color-coded or color-banded vital signs charts that are believed to assist in the recognition of patient deterioration.43,44 These use different colors to represent different levels of physiological abnormality linked to “track-and-trigger systems” (see below) or ghosted EWS values on the chart area to assist in calculations.43

Systems for Identifying the Sick or Deteriorating Patient Abnormalities of heart rate, blood pressure, breathing rate and consciousness are  markers of impending critical events.11,14,45–51 It seems clinically intuitive that

180

G.B. Smith and D.R. Prytherch

c­ ombinations of abnormalities would have greater predictive value than individual measures, and that trends may contain further information. Many hospitals now use a set of predetermined “calling criteria” as indicators of the need to escalate monitoring or to call for more expert help, often in the form of a Rapid Response Team (RRT). These calling criteria, or “track-and-trigger” systems, can be categorized as single-parameter systems, multiple-parameter systems, aggregate weighted scoring systems or combination systems.52 The most commonly used are the aggregate weighted track-and-trigger systems (AWTTS), which predominate in the UK, and the single-parameter track-and-trigger systems (SPTTS) common in the US and Australia. A few institutions are reported to use a multiple parameter track-and-trigger system (MPTTS), which require two or more observational values to exceed preset limits. AWTTS offers a graded escalation of care, whereas SPTTS provides an all-or-nothing response (i.e. the calling of a RRT). Aggregate Weighted Track-and-Trigger Systems AWTTS allocate points in a weighted manner to reflect the derangement of physiology from predetermined “normal” ranges.51,53–56 The sum of the allocated points is known as the early warning score (EWS) and this is used to direct care, e.g. to increase vital signs monitoring, to involve more experienced staff, or to call a rapid response team. The constituent physiological variables typically include pulse rate, blood pressure, breathing rate and conscious level, but some also include other parameters. With few exceptions, the variables included, and their weightings, are based solely on clinical experience and intuition. There is wide range of unique, but very similar, AWTTSs in clinical use.57 There is no consistency regarding their physiological components, the majority differing only in minor variations in the weightings for physiological derangement and/or the cut-off points between physiological weighting bands. A typical AWTTS is shown in Table 16.1. Until recently, the majority of AWTTS have either not been validated or have only been assessed in small studies where patients have experienced a specific clinical outcome, e.g. intensive care unit (ICU) admission or visit by a rapid

Table 16.1  A typical aggregate-weighted track-and-trigger score Value 3 2 1 0 1 Pulse (bpm) 41–50 51–100 101–110 £40 Respiratory Rate (bpm) £8 9–14 15–20 Temperature (°C) £35.0 35.1–36 36.1–37.4 ³37.5 CNS level (using Alert Voice AVPU scale) Systolic BP (mmHg) £70 71–80 81–100 101–199

2 3 111–130 ³131 21–29 ³30 Pain ³200

Unresponsive

16  An Overview of the Afferent Limb

181

response team.58 The performance of some AWTTS has been studied in groups of unselected patients59–61 early on in their hospital stay. More recently, Smith et al have compared the performance of 33 unique AWTTS using a single, large vital signs dataset collected on admission to hospital from unselected acute medical admissions.57 This demonstrated AUROC values62 (±95% CI) ranging from 0.657 (0.636–0.678) to 0.782 (0.767–0.797) for mortality at hospital discharge. AWTTSs have the disadvantage of being relatively complex compared to SPTTSs and require complete observations sets to allow calculation of an EWS. Obtaining the EWS involves calculation (summing the individual weighted scores) and this often introduces error. Errors tend to increases with the degree of physiological abnormality.63,64 In AWTTS, a high or rising EWS is considered premonitory of an adverse outcome (e.g. unanticipated ICU admission, cardiac arrest, or death). Whilst some work has been done, further investigation is required to fully understand the relationship between EWS values and adverse outcomes. This is difficult because of the intermittent nature of most ward-based monitoring and the lack of completeness of the vital signs datasets.

Single Parameter Track-and-Trigger Systems The first SPTTS was described in 1995 by Lee.65 It consisted of specific physiological/pathological abnormalities (e.g. respiration rate <10 or >30/min), specific conditions (e.g. new arrhythmia), and the non-specific criterion “any time urgent help is required.” The occurrence of one or more of these criteria indicates that the clinician should call for help, usually from an RRT. Compared to AWTTSs, SPTTSs are simple and easy to use. There is now a variety of SPTTSs in which the calling criteria have been modified from those of Lee, though often only subtly.66,67 As with AWTTSs, the variables and their trigger points in SPTTSs are based on expert ­clinical opinion and intuition. The wide ranges of variables and trigger points used underline the arbitrary nature of the process by which they have been chosen. A typical SPTTS is shown in Table 16.2. Most publications that refer to the use of SPTTS do not report data about their sensitivity and specificity. Where data are presented, it usually only refers to those patients who triggered the call for an RRT and or suffered an adverse clinical outcome.58,68 Proper sensitivity and specificity calculations were not performed, as the studies contain no data regarding those patients who did not trigger an RRT response or had not developed a specific adverse outcome. Recently, Smith et  al used a single, large vital signs dataset to compare the performance of 39 unique SPTTS.69 There was marked variation in sensitivity (7.3–52.8%) and specificity (69.1–98.1%) for mortality at hospital discharge.69 In SPTTS, extreme physiological values are considered to predict an increased risk of an adverse outcome. As with AWTTS, further investigation is required to

182

G.B. Smith and D.R. Prytherch Table 16.2  Medical emergency team (MET) calling criteria Airway If threatened Breathing All respiratory arrests Respiratory rate <5 breaths/min Respiratory rate >36 breaths/min Circulation All cardiac arrests Pulse rate <40 beats/min Pulse rate >140 beats/min Systolic blood pressure <90 mmHg Neurology Sudden fall in level of consciousness (fall in GCS of >2 points) Repeated or extended seizures Other Any patient you are seriously worried about that does not fit the above criteria

understand the relationship between physiology and adverse outcomes. The intermittent nature of most ward-based monitoring and the lack of completeness of the vital signs datasets present researchers with challenges. Efficiency of Aggregate Weighted and Single Parameter Track-and-Trigger Systems The number of calls generated by any calling system has implications for resource utilization. The aforementioned differences in performance of both AWTTSs and SPTTSs lead to considerable differences in the number of calls generated by different systems. For example, for SPTTSs, Smith et al demonstrated a 14-fold difference in the number of calls generated (i.e. workload) when they applied different SPTTSs to the same dataset.69 The number of calls generated by an AWTTS will depend upon the particular EWS score chosen to trigger by any particular healthcare institution. Other Clinical Observations that May Be Used to Trigger Rapid Response Systems AWTTS and SPTTS can only work in the dimensions of the parameters that they include. There are many other signs and symptoms that may well be valid reasons to trigger an RRT that are not (usually) included in these systems (Table  16.3). Importantly, SPTTSs specifically include the “any time urgent help is required” as a trigger for calling for help (often termed “nurse worried”).70 This is not explicitly included in AWTTSs. However, whatever system is used to identify a sick or deteriorating patient, it is important not to ignore the clinical intuition of the observer.

16  An Overview of the Afferent Limb

183

Table 16.3  Examples of signs and symptoms that may well be valid reasons to trigger an RRT that are not (usually) included in AWTTSs and SPTTSs Pallor Diaphoresis Changes in patient behavior (e.g. unexplained agitation) Prolonged capillary refill time Airway obstruction Seizure Uncontrolled bleeding Changes in functional capacity (e.g. a new inability to stand, loss of movement [or weakness] of face or limb, changes in speech or loss of ability to speak) Reduced urine output Chest pain

Any such system must allow the observer to call for help based on their worry even when the system in use would not trigger at that level of physiological abnormality. Clinical staff should be encouraged to call for help if they are concerned about a patient’s condition, even if they are unable to identify the cause of their worry. In some hospitals, the patient’s family and visitors are encouraged to form part of the calling process for RRTs as their intimate knowledge of the patient gives them a particular ability to recognize deterioration or subtle changes in the patient.71,72 Family activation of the Rapid Response System is discussed further in Chapter 18. We consider family activation to be an important addition to an institution’s ability to find all deteriorating patients as soon as possible. The Value of Monitoring Systems for Improving Detection of Critical Events in Low-Risk Populations To date, there is insufficient data to recommend continuous monitoring for low-risk patients; however, logic would suggest that continuous monitoring would be required to identify every patient deterioration. Hravnack and colleagues utilized continuous monitoring to determine the frequency of critical deteriorations in a high-dependency neurosurgical unit.73 They found that only one in four patients had even a single abnormal vital sign, and only 5% of patients had a clinical deterioration to the point where the RRS should have been triggered. Even among those patients, few progressed to the point where the RRS was in fact triggered. From this, one may suggest that many “critical” events may be missed, and yet little harm occurs. Interestingly, only two patients suffered a cardiac arrest during the trial; both had been removed from the monitor because the patient was “doing well.” When using continuous monitoring technology, alert levels should be preset by staff, or, more importantly, directed toward specific staff who would need to respond. The advantages of continuous monitoring include: 1 . More data points permitting more accurate trending of patient status 2. Continuous data means critical events can be discovered immediately

184

G.B. Smith and D.R. Prytherch

3 . Enables immediate life-saving interventions that would be futile if delayed 4. Enables work flow adjustments by floor staff 5. Enables staffing adjustments. 6. Improved patient flow potentially lowering cost The downside of continuous monitoring is mostly related to increased work generated by needing to investigate or troubleshoot abnormal signals (actual abnormalities as well as artifactual “noise”). There may be an increase in unnecessary treatments, increased staff work, and, more importantly, increased fatigue to alarms. This can lead to the potentially dangerous occurrence of staff ignoring alarms. It is important to note that ~33% of patient found “dead-in-bed” events occur in unmonitored (considered to be the healthiest) patients. Monitoring might prevent these deaths.

Calling for Assistance There should be a universally known and understood, mandated, unambiguous, activation protocol for summoning a response to a deteriorating patient. The culture of the organization should be such that staff are never criticized for calling. This final step in the afferent limb – the alert to or notification mechanism – is a key link that can easily fail. Because of the need for speed and reliability, a preset calling system is preferred. A dedicated phone exchange, well advertised, is helpful in preventing failures. While potentially more costly, a redundant system may be best since the cost of failure is potentially a life lost. A dedicated telephone number can reduce operator delays. Responders may be notified both via belt pagers and overhead pagers to prevent notification failure due to “silent” spots relative to the overhead speakers, and “dead” spots relative to the pagers. Safest is a check to ensure that there are no areas of the hospital where neither system is in place. The message can be misheard or interpreted as well. For example “five” and “nine” sound similar on overhead speakers. Duplicate modes of communication overcome this failure, as does a carefully scripted call that helps provide context to the listener. For example, the alert announcement can be specified: an alert tone signal, type of alert event, building, floor, wing. Communication between patient caregivers and the responding team is the final critical element of the afferent limb. Hospitals should consider the use of standardized method of communication, such as RSVP or SBAR,74,75 as these may improve the communication of patient deterioration between staff.

The Role of Technology Existing vital signs monitoring techniques are being developed further. These improvements take advantage of advances in technology that make the monitoring systems light, wearable, mobile and wirelessly connected. These advances may make continuous monitoring and recording of vital signs available to more patients.

16  An Overview of the Afferent Limb

185

Too long an interval between intermittently sampled vital signs may mean that opportunities to detect deterioration are missed. The ubiquitous presence of computers in healthcare permits the accurate and faster calculation of EWS30,63,76 and could permit the use of more complex rules for calling criteria, e.g. trends. Additionally, there are new methods of synthesizing information from data.77 New sensors are also being developed.78,79 Modern communication and computer technology provide the ability for a ­sensor to directly contact the responder without the involvement of the local carer. greater assurance in both detecting deterioration and alerting responders using clear and reliable algorithms and auditing of the whole system, including detection, calling and response.

Summary A properly functioning afferent limb is essential to the workings of an effective Rapid Response System. However, considerable deficiencies exist in the three components of the afferent limb – the monitoring of patients, the detection of patient deterioration and the triggering of the efferent limb. Considerable improvements in the function of the afferent limb could be achieved by simple, obvious interventions such as regular, complete monitoring and assessment of patients and a monitoring plan for each patient. Organizations should consider the use of “track-and-trigger” systems to assist in the detection of patient deterioration and the use of a standardized method of communicating patient illness between staff. Technology is also likely to provide solutions to improving the function of the afferent limb.

References 1. DeVita MA, Bellomo R, Hillman K, et  al. Findings of the first consensus conference on ­medical emergency teams. Crit Care Med. 2006;34:2463–2478. 2. DeVita MA, Smith GB, Adams SK, et al. “Identifying the hospitalised patient in crisis” – a consensus conference on the afferent limb of Rapid Response Systems. Resuscitation. 2010;81:375–382. 3. McQuillan PJ, Pilkington S, Allan A, et al. Confidential inquiry into quality of care before admission to intensive care. BMJ. 1998;316:1853–1858. 4. National Confidential Enquiry into Patient Outcomes and Death. An Acute Problem? London: National Confidential Enquiry into Patient Outcome and Death; 2005. 5. National Patient Safety Agency. Safer Care for the Acutely Ill Patient: Learning from Serious Incidents. London: NPSA; 2007. 6. National Patient Safety Agency. Recognising and Responding Appropriately to Early Signs of Deterioration in Hospitalised Patients. London: NPSA; 2007. 7. Chellel A, Fraser J, Fender V, et al. Nursing observations on ward patients at risk of critical illness. Nurs Times. 2002;98:36–39. 8. Smith S, Fraser J, Plowright C, et  al. Nursing observations on ward patients – results of a five-year audit. Nurs Times. 2008;104:28–29.

186

G.B. Smith and D.R. Prytherch

9. Wheatley I. The nursing practice of taking level 1 patient observations. Intensive Crit Care Nurs. 2006;22:115–121. 10. Fuhrmann L, Lippert A, Perner A, Ostergaard D. Incidence, staff awareness and mortality of patients at risk on general wards. Resuscitation. 2008;77:325–330. 11. Hillman KM, Bristow PJ, Chey T, et  al. Duration of life-threatening antecedents prior to intensive care admission. Intensive Care Med. 2002;28:1629–1634. 12. Hillman K, Chen J, Cretikos M, et  al. Introduction of the medical emergency team (MET) system: a cluster-randomised controlled trial. Lancet. 2005;365:2091–2097. 13. Harrison GA, Jacques TC, Kilborn G, McLaws ML. The prevalence of recordings of the signs of critical conditions and emergency responses in hospital wards – the SOCCER study. Resuscitation. 2005;65:149–157. 14. Kause J, Smith G, Prytherch D, Parr M, Flabouris A, Hillman K. A comparison of antecedents to cardiac arrests, deaths and emergency intensive care admissions in Australia and New Zealand, and the United Kingdom – the ACADEMIA study. Resuscitation. 2004;62:275–282. 15. McBride J, Knight D, Piper J, Smith G. Long-term effect of introducing an early warning score on respiratory rate charting on general wards. Resuscitation. 2005;65:41–44. 16. Nurmi J, Harjola VP, Nolan J, Castrén M. Observations and warning signs prior to cardiac arrest. Should a medical emergency team intervene earlier? Acta Anaesthesiol Scand. 2005; 49:702–706. 17. Zeitz K, McCutcheon H. Observations and vital signs: ritual or vital for the monitoring of postoperative patients? Appl Nurs Res. 2006;19:204–211. 18. Lockwood C, Conroy-Hiller T, Page T. Vital signs. JBI Rep. 2004;2:207–230. 19. Hogan J. Why don’t nurses monitor the respiratory rates of patients? Br J Nurs. 2006;15:489–492. 20. Bristow PJ, Hillman KM, Chey T, et al. Rates of in-hospital arrests, deaths and intensive care admissions: the effect of a medical emergency team. Med J Aust. 2000;173:236–240. 21. Kenward G, Hodgetts T, Castle N. Time to put the R back in TPR. Nurs Times. 2001;97:32–33. 22. Subbe CP, Williams EM, Gemmell LW. Are medical emergency teams picking up enough patients with increased respiratory rate? Crit Care Med. 2004;32:1983–1984. 23. Armstrong B, Walthall H, Clancy M, Mullee M, Simpson H. Recording of vital signs in a district general hospital emergency department. Emerg Med J. 2008;25:799–802. 24. Buist M. The rapid response team paradox: why doesn’t anyone call for help? Crit Care Med. 2008;36:634–635. 25. Jones D, Bates S, Warrillow S. Circadian pattern of activation of the medical emergency team in a teaching hospital. Crit Care. 2005;9:R303–R306. 26. Galhotra S, DeVita MA, Simmons RL, Schmid A, et al. Impact of patient monitoring on the diurnal pattern of medical emergency team activation. Crit Care Med. 2006;34:1700–1706. 27. Day BA. Early warning system scores and response times: an audit. Nurs Crit Care. 2003;8:156–164. 28. Andrews T, Waterman H. Packaging: a grounded theory of how to report physiological deterioration effectively. J Adv Nurs. 2005;52:473–481. 29. National Institute for Health and Clinical Excellence. NICE Clinical Guideline 50. Acutely Ill Patients in Hospital: Recognition of and Response to Acute Illness in Adults in Hospital. London: National Institute for Health and Clinical Excellence; 2007. 30. Smith GB, Prytherch DR, Schmidt P, et al. Hospital-wide physiological surveillance – a new approach to the early identification and management of the sick patient. Resuscitation. 2006; 71:19–29. 31. Odell M, Rechner IJ, Kapila A, et al. The effect of a critical care outreach service and an early warning scoring system on respiratory rate recording on the general wards. Resuscitation. 2007;74:470–475. 32. Chen J, Bellomo R, Flabouris A, Hillman K, Finfer S. The impact of introducing medical emergency team systems on the documentations of vital signs. Resuscitation. 2009;80:35–43. 33. Department of Health. Competencies for Recognising and Responding to Acutely Ill Patients in Hospital. London: Department of Health; 2009.

16  An Overview of the Afferent Limb

187

34. Kruger PS, Longden PJ. A study of a hospital staff’s knowledge of pulse oximetry. Anaesth Intensive Care. 1997;25:38–41. 35. Stoneham MD, Saville GM, Wilson IH. Knowledge about pulse oximetry among medical and nursing staff. Lancet. 1994;344:1339–4132. 36. Beevers G, Lip GYH, O’Brien E. Blood pressure measurement. Part II. Conventional ­sphygmomanometry: technique of auscultatory blood pressure measurement. BMJ. 2001; 322:1043–1047. 37. Edwards SM, Murdin L. Respoiratory rate – an under-documented clinical assessment. Clin Med. 2001;1:85. 38. Helliwell VC, Hadfield JH, Gould T. Documentation of respiratory rate for acutely sick ­inpatients – an observational study. Intensive Care Med. 2002;28:S21. 39. Cretikos MA, Bellomo R, Hillman K, Chen J, Finfer S, Flabouris A. Respiratory rate: the neglected vital sign. Med J Aust. 2008;188:657–659. 40. Hudson A. Prevention of in-hospital cardiac arrests – first steps in improving patient care. Resuscitation. 2004;60:113–115. 41. Lim WS, Carty SM, et al. Respiratory rate measurement in adults – how reliable is it? Respir Med. 2002;96:31–33. 42. Chatterjee MT, Moon JC, Murphy R, McCrea D. The “OBS” chart: an evidence-based approach to re-design of the patient observation chart in a district general hospital setting. Postgrad Med J. 2005;81:663–666. 43. Oakey RJ, Slade V. Physiological observation track-and-trigger system. Nurs Stand. 2006;20:48–54. 44. Patient Safety First “How-to Guide” for Reducing Harm from Deterioration. http://www. patientsafetyfirst.nhs.uk. Accessed 24.05.09. 45. Buist M, Bernard S, Nguyen TV, Moore G, Anderson J. Association between clinically abnormal observations and subsequent in-hospital mortality: a prospective study. Resuscitation. 2004;62:137–141. 46. Lighthall GK, Markar S, Hsuing R. Abnormal vital signs are associated with an increased risk of critical events in U.S. veteran in-patients. Resuscitation. 2009;80:1264–1269. 47. Buist MD, Jarmolowski E, Burton PR, Bernard SA, Waxman BP, Anderson J. Recognising clinical instability in hospital patients before cardiac arrest or unplanned admission to intensive care. A pilot study in a tertiary-care hospital. Med J Aust. 1999;171:22–25. 48. Hillman KM, Bristow PJ, Chey T, et  al. Antecedents to hospital deaths. Intern Med J. 2001;31:343–348. 49. Goldhill DR, McNarry AF. Physiological abnormalities in early warning scores are related to mortality in adult in-patients. Br J Anaesth. 2004;92:882–884. 50. Jacques T, Harrison GA, McLaws M-L, Kilborn G. Signs of critical conditions and emergency responses (SOCCER): a model for predicting adverse events in the in-patient setting. Resuscitation. 2006;69:175–183. 51. Hodgetts TJ, Kenward G, Vlachonikolis IG, Payne S, Castle N. The identification of risk factors for cardiac arrest and formulation of activation criteria to alert a medical emergency team. Resuscitation. 2002;54:125–131. 52. Department of Health and NHS Modernisation Agency. The National Outreach Report. London: Department of Health; 2003. 53. Subbe CP, Kruger M, Rutherford P, Gemmel L. Validation of a modified Early Warning Score in medical admissions. QJM. 2001;94:521–526. 54. Morgan R, Williams F, Wright M. An Early Warning Scoring System for detecting developing critical illness. Clin Intensive Care. 1997;8:100. 55. Stenhouse C, Coates S, Tivey M, Allsop P, Parker T. Prospective evaluation of a modified Early Warning Score to aid earlier detection of patients developing critical illness on a general surgical ward. Br J Anaesth. 2000;84:663. 56. Goldhill DR, McNarry AF, Mandersloot G, McGinley A. A physiologically-based early warning score for ward patients: the association between score and outcome. Anaesthesia. 2005;60:547–553.

188

G.B. Smith and D.R. Prytherch

57. Smith GB, Prytherch DR, Schmidt PE, Featherstone PI. Review and performance evaluation of aggregate weighted “track-and-trigger” systems. Resuscitation. 2008;77:170–179. 58. Gao H, McDonnell A, Harrison DA, et al. Systematic review and evaluation of physiological track-and-trigger warning systems for identifying at-risk patients on the ward. Intensive Care Med. 2007;33:667679. 59. Subbe CP, Slater A, Menon D, Gemmell L. Validation of physiological scoring systems in the accident and emergency department. Emerg Med J. 2006;23:841–845. 60. Paterson R, MacLeod DC, Thetford D, et al. Prediction of in-hospital mortality and length of stay using an early warning scoring system: clinical audit. Clin Med. 2006;6:281–284. 61. Duckitt RW, Buxton-Thomas R, Walker J, et  al. Worthing physiological scoring system: derivation and validation of a physiological early-warning system for medical admissions. An observational, population-based single-centre study. Br J Anaesth. 2007;98:769–774. 62. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver-operating characteristic (ROC) curve. Radiology. 1982;143:29–36. 63. Mohammed MA, Hayton R, Clements G, Smith G, Prytherch D. Improving accuracy and efficiency of early warning scores in acute care. Br J Nurs. 2009;18:18–24. 64. Smith AF, Oakey RJ. Incidence and significance of errors in a patient “track-and-trigger” system during an epidemic of Legionnaires’ disease: retrospective case note analysis. Anaesthesia. 2006;61:222–228. 65. Lee A, Bishop G, Hillman KM, Daffurn K. The medical emergency team. Anaesth Intensive Care. 1995;23:183–186. 66. Goldhill DR, Worthington L, Mulcahy A, Tarlin M, Sumner A. The patient-at-risk team: identifying and managing seriously ill ward patients. Anaesthesia. 1999;54:853–860. 67. Hourihan F, Bishop G, Hillman K, Daffurn K, Lee A, et al. The medical emergency team: a new strategy to identify and intervene in high-risk patients. Clin Intensive Care. 1995;6:269–272. 68. Cretikos J, Chen K, Hillman R, Bellomo S, Finfer A. Flabouris and the MERIT study investigators. The objective medical emergency team activation criteria: a case-control study. Resuscitation. 2007;73:62–72. 69. Smith GB, Prytherch DR, Schmidt PE, Featherstone PI, Higgins B. A review and performance evaluation of single-parameter “track-and-trigger” systems. Resuscitation. 2008;79:11–21. 70. Santiano N, Young L, Hillman K, et al. Medical Emergency Team calls comparing subjective to “objective” call criteria. Resuscitation. 2009;80:44–49. 71. http://www.ihi.org/IHI/Topics/CriticalCare/IntensiveCare/Tools/ConditionHBrochure forPatientsandFamilies.htm; 2009 Accessed 24.05.09. 72. Ray EM, Smith R, Massie S, et al. Family alert: implementing direct family activation of a pediatric rapid response team. Jt Comm J Qual Patient Saf. 2009;35:575–580. 73. Hravnak M, Edwards L, Clontz A, Valenta C, DeVita MA, Pinsky MR. Defining the incidence of cardio-respiratory instability in patients in step-down units using an electronic integrated monitoring system. Arch Intern Med. 2008;168:1300–1308. 74. Featherstone P, Chalmers T, Smith GB. RSVP: a system for communication of deterioration in hospital patients. Br J Nurs. 2008;17:860–864. 75. Thomas CM, Bertram E, Johnson D. The SBAR communication technique: teaching nursing students professional communication skills. Nurse Educ. 2009;34:176–180. 76. Prytherch DR, Smith GB, Schmidt P, et al. Calculating early warning scores – a classroom comparison of pen and paper and hand-held computer methods. Resuscitation. 2006;70:173–178. 77. Tarassenko L, Hann A, Young D. Integrated monitoring and analysis for early warning of patient deterioration. Br J Anaesth. 2006;97:64–68. 78. Parry DT, Woodward L, Smith GB, et  al. Validation of a novel radar-based breathing rate measurement device in human volunteers. Resuscitation. 2008;77:S14–S15. 79. Jacobs JL, Eron L. Automated medical vigilance on the medical/surgical ward of an acutecare hospital. J Hosp Med. 2007;2:196–197.

Chapter 17

The Impact of Delayed RRS Activation Daryl Jones, Michael Haase, and Rinaldo Bellomo

Keywords  Impact • Delayed • Rapid • Response • System • Activation

Background: Principles of the Rapid Response System Modern hospitals service patients of increasing complexity and co-morbidity.1 Despite advances in technology and the best efforts of hospital staff, several studies have demonstrated that 6–17% of all hospital admissions are complicated by serious adverse events.2–8 These events are often unrelated to the patient’s underlying medical condition and in approximately 10% of cases they will result in permanent disability and even death.8 Other studies have shown that these events are not sudden or unpredictable. Instead, they are often preceded by signs of physiological instability that manifest as derangements in commonly measured vital signs.9–12 Most importantly, in many cases, the rate of deterioration is relatively slow, and occurs over 12–24 h.10 One of the most important tenets underlying the Rapid Response System (RRS) concept is that early intervention in the course of critical illness improves outcome. This principle has been seen for the early management of severe trauma,13 resuscitation of patients presenting to the emergency department with sepsis,14 as well as for thrombolysis in the management of myocardial infarction15 and selected cases of ischemic stroke.16 As stated by England and Bion, the principle of the RRS is to “take critical care expertise to the patient before, rather than after, multiple organ failure or cardiac arrest occurs”.17 This chapter reviews the evidence for the prevalence, consequences and causes of delayed RRS activation. In addition, we will suggest ways in which delay might be prevented. In the chapter, we will be describing our RRS, which has a physician-led

D. Jones (*) Department of Intensive Care, Austin Hospital, Studley Road, Heidelberg, VIC 3084, Australia e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_17, © Springer Science + Business Media, LLC 2011

189

190

D. Jones et al.

response team, or MET. Therefore, although we will be reporting evaluations of our MET-type RRS, our findings are equally applicable to ab RRS that has a non-physician-led Rapid Response Team (RRT).

How Often Is RRS Activation Delayed? There are a number of studies suggesting that activation of the RRS is delayed in a substantial proportion of cases. In one of the earliest comparative “before-andafter” studies of the Medical Emergency Team (MET), Buist and co-workers reported that there was a reluctance to call the MET because of adherence to the traditional model of calling the most junior member of medical staff first.18 Four studies of the MET at the Austin Hospital (Melbourne, Australia) provide objective evidence of the prevalence of delayed MET activation. An assessment of the 162 cardiac arrests that occurred in the 4 years following the introduction of the MET reported that 45 (28%) of cardiac arrest calls occurred shortly after an initial MET activation.19 This suggests that the MET call activation was excessively delayed and that there was insufficient time for MET intervention to prevent the cardiac arrest. An assessment of the role of the MET in 105 consecutive deaths at the same hospital revealed that five of 105 deaths did not have a “Do Not Resuscitate (DNR)” order at the time of death. Three of these five patients suffered a cardiac arrest and did not receive an MET call despite the presence of MET activation criteria.20 Two retrospective studies of patients subject to MET review identified a high incidence of delayed MET activation.21,22 These studies defined delayed MET activation as an interval of more than 30 min between the onset of MET criteria and subsequent MET review. They examined four “MET syndromes” that included altered conscious state, arrhythmia, respiratory distress, and hypotension. They reported that 24–39% of patients had a delayed activation of the MET, and that the median time of delay was between 5 and 13 h depending on the MET syndrome (Table 17.1).21,22 Finally, the Medical Early Response Intervention and Therapy (MERIT) study reported a high incidence of delayed or failed activation in patients who suffered one of the composite outcomes. Thus, documented MET criteria were present for more than 15 min in 30% of cardiac arrests, 51% of unplanned ICU admissions, and 50% of unexpected deaths.23 Table 17.1  Incidence and median time of delay for MET syndromes at the Austin hospital Percentage (%) of patients with delayed Median time of delay in patients with MET call MET syndrome delayed MET call (h) Altered conscious state 35 16 Arrhythmia 24 13 Respiratory distress 50 12 Hypotension 39  5

17  The Impact of Delayed RRS Activation

191

What Are the Consequences of Delayed MET Activation? Delay in the review of deteriorating patients admitted to hospital wards undermines one of the most important principles underlying the MET principle, that early intervention improves outcome. It is not surprising, therefore, that studies by Downey et al and Quach et  al suggest that delayed MET activation is associated with worse outcome. Thus, Downey and co-workers reported that the patients who experience delayed MET review for altered conscious state and arrhythmia experienced an increased mortality with an odds ratio (OR) of 3.1 (p = 0.005) compared with reviews in which delay did not occur (Fig. 17.1).21 Similarly, Quach and co-workers revealed that delayed MET/RRS review for patients with respiratory distress or hypotension was associated with an increased risk of death with an OR of 2.1 (p = 0.045) (Fig. 17.2).22

How Should Delayed MET/RRT Activation Be Classified? We propose that delayed MET/RRT activation should be classified as “afferent limb failure.” This should be further characterized as “complete” or “partial.” Complete afferent limb failure has occurred when a patient develops MET/RRT criteria, and the 1.0

P=0.023

Mortality

0.8

0.6

Delayed MET call

0.4

0.2 Non -delayed MET call 0.0 0.00

5.00

10.00

15.00

20.00

25.00

30.00

Days after MET call Fig. 17.1  Kaplan Meier survival curve showing increased mortality for 30 days for patient who had delayed versus non-delayed MET/RRS review for altered conscious state and arrhythmia

192

D. Jones et al. 1.0

P=0.049

Mortality

0.8

0.6

0.4

Delayed MET call

0.2

Non -delayed MET call

0.0 0.00

5.00

10.00

15.00

20.00

25.00

30.00

Days after MET call Fig. 17.2  Kaplan Meier survival curve showing increased mortality for 28 days for patient who had delayed versus timely MET/RRS review for respiratory distress or hypotension

MET/RRT is not activated, and the patient suffers a serious adverse event. Partial afferent limb failure has occurred when the patient develops MET/RRT criteria, and team activation is delayed and the patient still suffers a serious adverse event.

What Are the Causes of Delayed Response Activation? Existing evidence suggests that there are at least four major causes of delayed or failed response activation. The first of these is a fear by staff that they will be criticized either for calling the team or for a perception of inadequate management of the patient.24 Other studies suggest that staff may not be happy with the criteria for, or method of, MET/RRT activation. Thus, a study at a teaching hospital in Melbourne revealed that altering the limits of the calling criteria, reducing the number of people in the MET/RRT, and making the mode of activation more subtle resulted in a marked increase in calling rate.25 Foraida and co-workers reported that the introduction of objective calling criteria resulted in a near doubling of MET call activations in a university hospital in Pittsburgh.26 Awareness and acceptance of the RRS concept is also an important aspect of RRS activation rates. Casamento and co-workers reported that an education program for ward nurses resulted in increased call rates.27 Similarly, Buist and co-workers reported increased call rates associated with

17  The Impact of Delayed RRS Activation

193

a sustained education program of new and existing ward medical and nursing staff.28 A number of studies suggest that monitoring of acutely unwell patients may be inadequate to detect crisis early enough. Thus, several studies have shown that respiratory rate is infrequently monitored, despite the fact that an elevated respiratory rate is shown to predict increased mortality.29 An analysis of cardiac arrests at the Austin Hospital revealed that cardiac arrests were most common during the night, when RRS call rates were lowest, and that arrest rates were lowest in the evening when RRS call rates were highest.30 Finally, a number of studies suggest that ward staff may not appreciate the significance of a patient developing MET/RRS activation criteria. Many of the studies that preceded the development of the RRS concept revealed that the management of acutely unwell ward patients by ward doctors and nurses is frequently sub-optimal in the period leading up to cardiac arrests and unplanned ICU admissions.10,31 A questionnaire for nurses at our hospital revealed that nurses frequently use discretion when confronted with a patient who has developed RRS-activation criteria, and frequently do not call the MET/RRT.24

How Can Delayed RRS Activation Be Avoided? We believe that it is important to regularly review RRS activation rates and instances where there has been partial or complete afferent limb failure. Such cases should be regularly fed back to the appropriate medical or surgical units at peer review grand rounds or “mortality and morbidity” meetings. This should be performed in a non-confrontational and blame-free manner. In addition, regular focus groups should be conducted and questionnaires used to assess institution-specific barriers to RRS activation. The studies outlined above suggest that the reasons for failed activation may vary between hospitals. Members of the responder team (whether MET or RRT) should be constantly reminded that no member of staff should ever be criticized for activating the RRS, or for the management of the patient. Finally, there should be regular education and feedback should for all new and existing staff to reinforce or introduce the concept, background and principles of the RRS in identifying, reviewing and treating acutely unwell hospital ward patients.

References 1. Zajac JD. The public hospital of the future. Med J Aust. 2003;179:250–2. 2. Andrews LB, Stocking C, Krizek T, et al. An alternative strategy for studying adverse events in medical care. Lancet. 1997;349:309–13. 3. Baker GR, Norton PG, Flintoft V, et al. The Canadian Adverse Events Study: the incidence of adverse events among hospital patients in Canada. CMAJ. 2004;170:1678–86.

194

D. Jones et al.

4. Brennan TA, Leape LL. Adverse events, negligence in hospitalized patients: results from the Harvard Medical Practice Study. Perspect Health Risk Manage. 1991;11:2–8. 5. Davis P, Lay-Yee R, Briant R, Ali W, Scott A, Schug S. Adverse events in New Zealand public hospitals I: occurrence and impact. N Z Med J. 2002;115:U271. 6. Leape LL, Brennan TA, Laird N, et al. The nature of adverse events in hospitalized patients. Results of the Harvard Medical Practice Study II. N Engl J Med. 1991;324:377–84. 7. Vincent C, Neale G, Woloshynowych M. Adverse events in British hospitals: preliminary retrospective record review. BMJ. 2001;322:517–9. 8. Wilson RM, Runciman WB, Gibberd RW, Harrison BT, Newby L, Hamilton JD. The Quality in Australian Health Care Study. Med J Aust. 1995;163:458–71. 9. Bell MB, Konrad D, Granath F, Ekbom A, Martling CR. Prevalence and sensitivity of MET criteria in a Scandinavian University Hospital. Resuscitation. 2006;70:66–73. 10. Buist MD, Jarmolowski E, Burton PR, Bernard SA, Waxman BP, Anderson J. Recognising clinical instability in hospital patients before cardiac arrest or unplanned admission to intensive care. A pilot study in a tertiary-care hospital. Med J Aust. 1999;171:22–5. 11. Hodgetts TJ, Kenward G, Vlachonikolis IG, Payne S, Castle N. The identification of risk factors for cardiac arrest and formulation of activation criteria to alert a medical emergency team. Resuscitation. 2002;54:125–31. 12. Hodgetts TJ, Kenward G, Vlackonikolis I, et al. Incidence, location and reasons for avoidable in-hospital cardiac arrest in a district general hospital. Resuscitation. 2002;54:115–23. 13. Blow O, Magliore L, Claridge JA, Butler K, Young JS. The golden hour and the silver day: detection and correction of occult hypoperfusion within 24 hours improves outcome from major trauma. J Trauma. 1999;47:964–9. 14. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med. 2001;345:1368–77. 15. Fresco C, Carinci F, Maggioni AP, et al. Very early assessment of risk for in-hospital death among 11,483 patients with acute myocardial infarction. GISSI investigators. Am Heart J. 1999;138:1058–64. 16. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N Engl J Med. 1995;333:1581–7. 17. England K, Bion JF. Introduction of medical emergency teams in Australia and New Zealand: a multicentre study. Crit Care. 2008;12:151. 18. Buist MD, Moore GE, Bernard SA, Waxman BP, Anderson JN, Nguyen TV. Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. BMJ. 2002;324:387–90. 19. Jones, D, Opdam H, Egi M, et al. Long-term effect of a Medical Emergency Team on mortality in a teaching hospital. Resuscitation. 2007;74:235–41. 20. Jones D, McIntyre T, Baldwin I, Mercer I, Kattula A, Bellomo R. The Medical Emergency Team and End-of-Life care: A pilot study. Crit Care Resus. 2007;9:151–156. 21. Downey AW, Quach JL, Haase M, Haase-Fielitz A, Jones D, Bellomo R. Characteristics and outcomes of patients receiving a medical emergency team review for acute change in conscious state or arrhythmias. Crit Care Med. 2008;36:477–81. 22. Quach JL, Downey AW, Haase M, Haase-Fielitz A, Jones D, Bellomo R. Characteristics and outcomes of patients receiving a medical emergency team review for respiratory distress or hypotension. J Crit Care. 2008;23:325–31. 23. Hillman K, Chen J, Cretikos M, et al. Introduction of the medical emergency team system: a cluster-randomised controlled trial. Lancet. 2005;365(9477):2091–7. 24. Jones D, Baldwin I, McIntyre T, Nurses' attitudes to a medical emergency team service in a teaching hospital. Qual Saf Health Care. 2006;15:427–32. 25. Jones DA, Mitra B, Barbetti J, Choate K, Leong T, Bellomo R. Increasing the use of an existing medical emergency team in a teaching hospital. Anaesth Intensive Care. 2006;34:731–5. 26. Foraida MI, DeVita MA, Braithwaite RS, Stuart SA, Brooks MM, Simmons RL. Improving the utilization of medical crisis teams (Condition C) at an urban tertiary care hospital. J Crit Care. 2003;18:87–94.

17  The Impact of Delayed RRS Activation

195

27. Casamento AJ, Dunlop C, Jones DA, Duke G. Improving the documentation of medical emergency team reviews. Crit Care Resusc. 2008;10:29. 28. Buist M, Harrison J, Abaloz E, Van Dyke S. Six-year audit of cardiac arrests and medical emergency team calls in an Australian outer metropolitan teaching hospital. BMJ. 2007; 335(7631):1210–1212. 29. Cretikos MA et al. Respiratory rate: the neglected vital sign. Med J Aust. 2008;188(11): 657–659. 30. Jones D et  al. Patient monitoring and the timing of cardiac arrests and medical emergency team calls in a teaching hospital. Intensive Care Med. 2006;32(9):1352–1356. 31. McQuillan P et al. Confidential inquiry into quality of care before admission to intensive care. BMJ. 1998;316(7148):1853–1858.

Chapter 18

The Case for Family Activation of the RRS Helen Haskell

Keywords  Patient • Family • Activation • Rapid • Response • Systems

Introduction Simply put, patient and family activation of the rapid response system (RRS) allows patients or family members to trigger an institution’s RRS directly, without a nurse or physician as intermediary. While RRSs have been implemented in a growing number of hospitals in the U.S., many hospitals have been reluctant to include the feature of patient and family activation. Opponents of family activation of the RRS cite concerns that allowing patients to go over the heads of their caregivers might erode authority, damage relationships, or overload an already-stressed system of care. The most commonly cited concern has been that the program could result in calls for nonemergency situations, or calls that would be better handled by patient relations.1,2 This would create additional work in a healthcare system that may already be struggling to manage the workload. The case in favor of family activation has two major thrusts. The first is that it can provide a fail-safe mechanism to supplement professional caregivers should the caregivers be unable to do a patient assessment and determine if triggering criteria have been met. The second is far more personal and moral: it allows family members to have a measure of control over the destiny of their loved ones’ lives and care. The author of this chapter personally experienced the worst-case scenario of what can happen when hospital staff fails to recognize the signs of a deteriorating patient. Her 15-year-old son, Lewis Blackman, underwent elective surgery to correct a congenital chest defect, pectus excavatum. As part of his postoperative pain regimen, he was prescribed the NSAID analgesic ketorolac, known to be associated with gastrointestinal bleeding and perforation, among other side effects. Lewis

H. Haskell (*) Mothers Against Medical Error, 155 S. Bull St., Columbia, SC 29205, USA e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_18, © Springer Science+Business Media, LLC 2011

197

198

H. Haskell

appeared to be doing well until the third day after surgery (a Sunday), when he complained of sudden, excruciating epigastric pain. Nurses and residents ascribed his pain to constipation and, despite the family’s repeated requests for help, failed to act upon his increasingly unstable vital signs. After 30 h of progressive decline, Lewis died from infection and blood loss caused by a perforated duodenal ulcer, with no critical care assessment ever having been done.3 There was obvious serious regret that Lewis’s family knew or at least felt strongly that there was something seriously wrong, and had no recourse to help him when caregivers failed to respond. This kind of disempowerment arises from another era of medicine when families were less engaged, less informed, and physicians considered to be error-free. Cases such as Lewis’s have led to rising demands for a system allowing hospital patients to call for help directly. In an echo of Kerridge’s call for implementation of medical emergency teams on moral grounds, advocates of family activation argue that even if a family trigger cannot quantitatively be proven to improve outcomes, hospitals have an ethical obligation to provide patients and families with systems of emergency backup.4 It is argued that nurses cannot be available at every moment, and that friends and family members may more readily pick up on subtle early signs of deterioration because of their greater familiarity with the patient.5,6 In addition, many patients have experienced the situation, documented in the medical literature, in which nurses can be slow to call for help for reasons ranging from failure to recognize the signs of deterioration to fear of “looking stupid” if they should sound a false alarm.7,8

The Origins of Patient- and Family-Activated Rapid Response: Condition H Patient- and family-activated rapid response1 initiatives began to appear in American hospitals shortly after rapid response teams gained widespread attention as part of the Institute for Healthcare Improvement’s “Save 100,000 Lives” Campaign in 2004.9 The generally acknowledged pioneer in patient activation is Condition Help, or Condition H, a national initiative that began in May 2005 at Shadyside Hospital of the University of Pittsburgh Medical Center. UPMC Shadyside Hospital had recently implemented a Condition C (Medical Emergency Teams) program emulating that of their sister hospital, UPMC Presbyterian Hospital. There was a single difference between the two programs: at Presbyterian, anyone was allowed to trigger the RRS, while at Shadyside, only caregivers were allowed to do so due to concerns about potential for “inappropriate” calls. But that soon changed. Shadyside initiated the program after one of the hospital’s administrators heard patient safety advocate Sorrel King say that a family activation feature might have prevented the death of her hospitalized young daughter, Josie. A standard terminology has yet to be worked out for rapid response systems meant to be activated by persons other than hospital staff. Throughout this chapter, the terms “patient-activated,” “family-activated,” and “family trigger” are used interchangeably with the more accurate but cumbersome “patient- and family-activated” rapid response.

1 

18  The Case for Family Activation of the RRS

199

Created to complement existing cardiac arrest and medical emergency teams (Conditions A and C), Condition H is a separate patient-activated team that patients are advised to call if the patient or family feels there has been a breakdown in care, or if they feel the healthcare team is failing to address a noticeable change in the patient’s condition.5 Thus UPMC Shadyside not only added the ability for patients and families to call for help for medical crises, they also fostered family calls for other important, even if not life-threatening, breakdowns. This focus on the patient and expansion of the reasons a call might be made led to overwhelming approval by families and caregivers alike.10 Condition H was piloted at UPMC Shadyside with assistance from the Institute for Healthcare Improvement, the Robert Wood Johnson Foundation, and the Josie King Foundation. Because of the tools made available through the websites of those organizations, Shadyside’s Condition H has been a widely copied model for patient- and family-activated rapid response.11–15 Other hospitals have created their own versions, including some hospitals where the team is technically not patientactivated, but where patients are encouraged to ask a staff member to activate the rapid response team if they have a medical emergency.11,16 Notable examples of hospitals creating their own systems are the Family Activated Safety Team (FAST) program at Virginia Mason Medical Center in Seattle, seen as an example of Virginia Mason’s Toyota Production System methodology of “stopping the line;” and the Family Alert initiative of the North Carolina Children’s Hospital in Chapel Hill, NC, developed after analysis of earlier data showed “family concern” as the reason for activation of 8% of rapid response calls.17,18

Patient- and Family-Activated Rapid Response in Legislation, Accreditation, and Safety Organizations Some of the most important steps toward patient and family activation have originated outside the medical system. An early development came in June 2005, when the South Carolina legislature passed the Lewis Blackman Hospital Patient Safety Act in response to the death of Lewis Blackman, a South Carolina resident.19 Among other things, the law requires hospitals to create an independently accessible patient assistance system that provides timely aid any time of the day or night. In 2008, Massachusetts passed a law with a similar directive for hospitals to establish a “method for healthcare staff members, patients and families to request additional assistance directly from a specially trained individual if the patient’s condition appears to be deteriorating.”20 This method must also be ­available 24 h a day. It should be noted that neither of these laws requires a rapid response team or medical emergency team per se. They simply mandate that hospitals provide their patients with a mechanism to request emergency help. Perhaps partly for this reason, confusion over enforcement slowed effective implementation of this section of the Lewis Blackman Act. In 2008, the South Carolina Hospital Association passed a resolution urging hospitals to meet their obligations under the Lewis Blackman Act

200

H. Haskell

by adopting patient- and family-activated rapid response teams. Recent surveys show that implementation of family-activated RRSs is complete in almost 100% of hospitals in South Carolina.21 Other American hospitals began their family trigger programs in response to the Joint Commission for Accrediting Healthcare Organizations, whose 2008 Patient Safety Goal 16A included the expectation that accredited organizations would “empower staff, patients, and/or families to request additional assistance when they have a concern about the patient’s condition.” While this language was softened in later years to “encouraging” and “informing” families how to seek assistance, the Joint Commission’s stated goal of patient-centeredness remains an important impetus for programs of this sort.22–24 In addition, the “Save 100,000 Lives” and “Save 5 Million Lives” patient safety campaigns of the Institute for Healthcare Improvement provided a framework and resources enabling hospitals to undertake patient safety initiatives including familyactivated rapid response teams. The support and moral example of the Josie King Foundation also cannot be overestimated.12,25

Features of Patient- and Family-Activated Rapid Response Systems A survey of available materials yielded an abundance of online tools and resources, as well as articles in the popular press and trade publications.2,5,11–17,25–31 But while hospitals have been generous in sharing their experiences, formal publications are sparse. The following summaries are derived from a variety of electronic and print sources and rely heavily on a handful of relatively well-documented programs, primarily UPMC’s Condition H, Virginia Mason Medical Center’s FAST, and the North Carolina Children’s Hospital’s Family Alert.1,17,18,32,33 They suggest that, probably because of the cross-fertilization brought about by open sharing of resources, most family-activated RRSs have basic elements in common, despite considerable variation between facilities.

Administration and Design Organizers at Pittsburgh, Virginia Mason, and North Carolina Children’s Hospital report phasing in the family activation feature over a period of months, with a long planning period and a pilot on one or several units before rollout to the entire hospital. All three programs went to considerable lengths to obtain buy-in and input from hospital leadership, staff, and patients. Elements of program development included surveys, focus groups, and interviews with both patient and staff to allay concerns and develop best practices.15,17,26,32 At Virginia Mason, organizers

18  The Case for Family Activation of the RRS

201

developed and repeatedly reviewed a toolkit that included an algorithm, scripts, and process overview, with special attention to clarifying respondents’ roles and duties.17 At North Carolina Children’s Hospital, nurses worked with the communications department to develop the ­program’s posters and written material, and collaborated with the medical team to develop an assessment tool.26

Patient Education Initially, it was thought that patient and family activation would be a success if nurses were not overwhelmed with inappropriate calls from patients who might mistake the rapid response number for a service line. It quickly became evident, however, that the larger problem would be not overuse but reticence on the part of patients fearful of alienating hospital personnel.14,18 As a consequence, a common element in the promotion of family activation programs is introductory material designed to encourage patients to be comfortable with the idea of using them. In line with this thinking, patient education in most hospitals begins with the admitting nurse personally explaining the program, guided by a mock script or other prompt. The message is reinforced by written material such as brochures, posters, and telephone stickers prominently featuring the program’s telephone number.11,14,28,32 UPMC Shadyside and Children’s Hospital of Pittsburgh additionally show patients a video.1,27

Triggering Criteria Most hospitals report following the UPMC Shadyside example in advising patients to call the helpline if they notice a medical change that is not being addressed by the health team or if there is a breakdown in care or confusion over treatment.1,14 The North Carolina Children’s Hospital more broadly tells parents to trigger a Family Alert if their child’s condition is rapidly worsening, or if they feel they have a medical emergency. To avoid non-urgent calls to the team, North Carolina Children’s Hospital posts the patient relations department number in every room, and has nurses instruct families how to contact staff or patient relations for non-urgent needs.26

Screening On the assumption that patient and family calls will not always be for medical emergencies, screening is built into many family-activated systems. The degree of screening can range from instructing patients and families to ask a staff member to call for help,11,25 to asking them to contact the charge nurse and attending physician

202

H. Haskell

before calling the team,13 to encouraging the patient and family to activate the rapid response or medical emergency team directly.32 Many hospitals use the team itself as a form of screening, by having the family-activated team assess the patient before deciding whether to summon the “real” Medical Emergency Teams or RRT.17 UPMC Shadyside has a double level of triage, with the hospital operator screening calls to weed out non-medical problems, and the Condition H team then assessing patients to determine if escalation to the medical emergency team is warranted.5,15

Team Composition The makeup of family-activated teams is diverse. Many, if not most, familyactivated teams are nurse-led and do not have full Medical Emergency Teams or RRT capabilities. A call to the Family Alert at Children’s Hospital of North Carolina, however, activates a full Medical Emergency Tears: critical care physicians, critical care nurses, and a respiratory therapist.32 UPMC Shadyside’s Condition H team also includes a (an internal medicine house physician), but like many family-activated teams, it is composed largely of non-clinical personnel (a patient relations coordinator, an administrative nursing coordinator, and the unit nurse) and does not include critical care specialists. The number of people called to the bedside at different hospitals can vary from one to five, although Shadyside administrators note that overly large teams have the potential to frighten patients.11,14 At most hospitals the bedside nurse is treated as a member of the team.11,15,17

Follow-Up and Data Collection UPMC Shadyside led the way in developing an assessment process for its familyactivated program in addition to that for the traditional Medical Emergency Teams or RRT.5 Their two-page form, available online, has been adapted by several hospitals. It consists of one page of questions for team members, including the reason for activation, and a second page designed to assess patient satisfaction with the familyactivated system.13,14 In order to keep educational efforts from declining, North Carolina Children’s Hospital also conducts random in-person surveys to determine family awareness of the Family Alert system, and rewards staff whose units show high family awareness. Even so, family awareness after 276 surveys averaged only 27%, perhaps an indication of the difficulty of providing complete and lasting information in the midst of a stressful situation like a hospital admission.32

Gauging Success Most hospitals seem to consider their family trigger programs “successful” if they are accepted by staff and utilized by patients, even if infrequently. Published data, however, are inconclusive. UPMC and Virginia Mason have compiled information

18  The Case for Family Activation of the RRS

203

from several analyses of assessment surveys. While the numbers are minuscule (ranging from a total of 20 calls at Virginia Mason to 71 in the largest UPMC analysis), they tend to support anecdotal reports that family-activated calls are not usually true emergencies. In three analyses at UPMC hospitals, the greatest numbers of calls recorded were for pain management and medical management. While these are broad categories, accompanying observations imply that patient situations generally were not urgent, even though the calls were usually thought to be warranted.1,15,27 At the UPMC Children’s Hospital of Pittsburgh, the author noted that root cause analysis showed the underlying cause of all calls to be communication breakdown.1 At Virginia Mason, “delay in care” was the most common reason given for activation. Nevertheless, the FAST team did not activate the full medical emergency team in any of the system’s first 20 calls.17 This experience seems to echo that of a North Carolina hospital whose administrators reported that they stopped mobilizing the full team after finding that most family-activated calls were for pain management or communication problems rather than acutely deteriorating patients.11 North Carolina Children’s Hospital reports very low usage of the Family Alert system, with only two calls directly triggered by families in the first year. The number of staff-activated rapid response calls, however, increased by 50% following the implementation of family activation, from an average of 16 per 1,000 discharges to 24 per 1,000.26 This occurred in an environment in which cardiac arrest rates, ward mortality, and median duration of clinical instability had decreased dramatically following implementation of a staff-activated rapid response team 1 year earlier. Patient mortality did not decrease further following the advent of family activation.33 Taken together, these results indicate that, as one might expect, staff awareness has a more significant effect than family awareness on use of critical care resources. They also suggest that the continuous education and constant visual reminders required for a family activation plan may heighten staff awareness and give added incentive for staff to place the call themselves. Early surveys at UPMC and Virginia Mason showed high satisfaction with the patient activation program. Although the surveys apparently did not question patients about the events precipitating their calls, almost all patients agreed that their concerns were addressed by the team and that they would call the system again. Data also showed gradually increasing use of the system over time, presumably indicating greater awareness and acceptance on the part of both patients and staff. Nevertheless, usage remained low, at an average of around two calls a month.15,17,29

Summary Although there is little to no evidence regarding the effect on patient outcomes, the first few years of patient- and family-activated rapid response have served to dispel several preconceptions about the functioning of the system itself. First, widespread fears that patients would overwhelm RRSs with nonessential calls have not materialized.

204

H. Haskell

Patient education has instead proven be the greater challenge, with calls infrequent and hospitals reporting that constant vigilance is needed to maintain awareness. Second, findings from assessment surveys indicate that family-activated calls do not usually rise to the level of true patient crisis. This lends support to the concept that the larger role of family trigger programs is not just in rescuing deteriorating patients but also in giving both patients and hospitals a means of addressing less urgent but still problematic gaps in the system. Third, the growing acceptance of patient- and family-activated rapid response should be seen in the context of the growing movement of patientcentered care and patient engagement. Adherence to the spirit of patient-centered care has been an explicit motivation in the development of family activation programs at a number of hospitals, including UPMC and North Carolina Children’s Hospital. Administrators at those hospitals cite their family activation programs as a way to improve information flow and incorporate the family into the medical team. Within this framework, family activation is also perceived as changing culture by making staff more cognizant of the needs of patients and families.1,32 These concepts resonate with the public, who may find it mystifying that an emergency number for patients should be controversial.34 With its parent advocates and legislative initiatives, patient- and family-activated rapid response seems to be in a different category from many other patient safety initiatives, and may be seen as part of a growing impatience with the exigencies of an overtaxed system. The fact that two American states have seen fit to write emergency family response mechanisms into law suggests that patients view a hospital helpline not so much as a medical issue but as a right. From this perspective, the family trigger is less about improving mortality than about the moral imperative of giving patients a number to call when they are worried or afraid. It doesn’t matter how infrequently the system is triggered as long as it is there when needed.

References 1. Dean BS, Decker MJ, Hupp D, Urbach AH, Lewis E, Benes-Stickle J. Condition HELP: a pediatric rapid response team triggered by patients and parents. J Healthc Qual. 2008;30:28–31. 2. Giordano T. Patient-activated RRTs catch on around the U.S. The Hospitalist. March 2008. http://www.the-hospitalist.org/details/article/187663/Patientactivated_RRTs_Catch_On_ Around_the_U_S_.html. Accessed October 2009. 3. Monk J. How a hospital failed a boy who didn’t have to die. The State. June 16, 2002:A1, A8–A9. 4. Kerridge RK. The medical emergency team: no evidence to justify not implementing change. Med J Aust. 2000;173(5):228–229. 5. Greenhouse PK, Kuzminsky B, Martin SC, Merryman T. Calling a Condition H(elp). Am J Nurs. 2006;106(11):63–66. 6. King S. Josie’s Story. New York: Atlantic Monthly Press; 2009. 7. Odell M, Victor C, Oliver D. Nurses’ role in detecting deterioration in ward patients: systematic literature review. J Adv Nurs. 2009;65(10):1992–2006. 8. Tee A, Calzavacca P, Licari E, Goldsmith D, Bellomo R. Bench-to-bedside review: the Medical Emergency Teams syndrome – the challenges of researching and adopting medical

18  The Case for Family Activation of the RRS

205

emergency teams. Crit Care. 2008;12:205. Available at: http://ccforum/content/12/1/205. Accessed October 2009. 9. Berwick D, Calkins DR, McCannon CJ, Hackbarth AD. The 100,000 lives campaign: setting a goal and a deadline for improving health care quality. JAMA. 2006;295(3):324–327. 10. DeVita, M. Exec VP Medical Affairs, West Penn Allegheny Health System. Pers. comm. December 10, 2009. 11. Spath PL. Empowering families in emergencies. Hospitals and Health Networks Magazine. February 2, 2007. http://www.hhnmag.com/hhnmag_app/jsp/articledisplay. j s p ? d c r p a t h = H H N M AG / Article/data/02FEB2007/0702HHN_Online_ Spath&domain=HHNMAG. Accessed October 2009. 12. Josie King Foundation. Foundation programs: Condition Help (Condition H). http://www. josieking.org/page.cfm?pageID=18. Accessed October 2009. 13. Institute for Healthcare Improvement. IHI.org resources. Intensive care tools: rapid response teams. Condition H (Help) brochure for patients and families. http://www.ihi.org/IHI/Topics/ CriticalCare/IntensiveCare/Tools/ConditionHBrochureforPatientsandFamilies.htm. Accessed October 2009. 14. Robert Wood Johnson Foundation Publications and Research. Developing a patient/familyactivated rapid response team. Published June 4, 2008. http://www.rwjf.org/pr/product. jsp?id=30391. Accessed October 2009. 15. University of Pittsburgh Medical Center. UPMC quality and innovation: Condition H (Help). http://www.upmc.com/aboutupmc/qualityinnovation/centerforqualityimprovementandinnovation/pages/conditionh.aspx. Accessed October 2009. 16. Davis R. Medical teams swoop in at family’s behest. USA Today. January 29, 2007. 17. Noel E. F.A.S.T. Family-Activated Safety Team: a rapid response for patient and family concerns. http://www.wapatientsafety.org/downloads/Noel.pdf. Accessed October 2009. 18. Van Voorhis KT, Willis TS. Implementing a pediatric rapid response system to improve quality and patient safety. Pediatr Clin North Am. 2009;56:919–933. 19. Lewis Blackman Hospital Patient Safety Act of 2005. South Carolina Code of Laws § 44-73410 et seq. 20. Chapter 305 of the Laws of 2008, an act to promote cost containment, transparency, and efficiency in the delivery of healthcare. Requests for additional assistance for deteriorating patients. The General Laws of Massachusetts Part I, Title XVI, Chapter 111: Section 53F. 21. South Carolina Hospital Association. Summary rapid response team baseline and remeasurement analysis. 2009. Located at: South Carolina Hospital Association, Quality and Patient Safety, Columbia, SC. 22. The Joint Commission on Accreditation of Healthcare Organizations. 2008 National Patient Safety Goals, Hospital. 2008. 23. The Joint Commission on Accreditation of Healthcare Organizations. 2009 National Patient Safety Goals. The Joint Commission Accreditation Program: Hospitals. Pre-publication version, 2008. 24. The Joint Commission on Accreditation of Healthcare Organizations. 2010 Provision of care, treatment, and services. The Joint Commission Accreditation Program: Hospitals. Prepublication version, 2009. 25. Institute for Healthcare Improvement. 5 Million Lives Campaign Resources. Tools for hospitals working to implement and deploy rapid response teams. http://www.ihi.org/IHI/Programs/ Campaign/RapidResponseTeams.htm. Accessed October 2009. 26. Agency for Healthcare Research and Quality. Family-activated pediatric rapid response team increases calls from both families and staff, supports improvements in outcomes. AHRQ Health Care Innovations Exchange. Published September 30, 2009. http://www.innovations. ahrq.gov/content.aspx?id=2435. Accessed October 2009. 27. Agency for Healthcare Research and Quality. Patient- and family-activated response team averts potential problems and generates high levels of patient, family, and staff satisfaction. AHRQ Health Care Innovations Exchange. Published December 12, 2008, updated June 22, 2009. http://www.innovations.ahrq.gov/content.aspx?id=1759. Accessed October 2009.

206

H. Haskell

28. Institute for Healthcare Improvement. IHI.org resources. Intensive care tools: rapid response teams. Family Activated Safety Team (FAST) tools. http://www.ihi.org/IHI/Topics/ CriticalCare/IntensiveCare/Tools/FASTTools.htm. Accessed October 2009. 29. Kuzminsky, B. Condition Help: making the hospital a safe place for patients, the implementation story. http://www.rwjf.org/files/research/conditionhupmc.pdf. Accessed October 2009. 30. Landro L. Patients get power of fast response. Wall Street Journal. September 1, 2009. 31. Landro L. When a hospital lets families call for rapid-response help. Wall Street Journal Health Blog. August 31, 2009. http://blogs.wsj.com/health/2009/08/31/calling-for-rapidresponse-hospital-help-should-be-family-affair/. Accessed September 2009. 32. Ray EM, Smith R, Massie S, et al. Family alert: implementing direct family activation of a pediatric rapid response team. Jt Comm J Qual Patient Saf. 2009;35:575–580. 33. Hanson CC, Randolph GD, Erickson JA, et al. A reduction in cardiac arrest and duration of clinical instability after implementation of a pediatric rapid response system. Qual Saf Health Care. 2009;18:500–504. 34. Initiative gives families access to rapid response team. Healthcare Benchmarks Qual Improv. 2008;15(Patient Safety Alert Supplement 1):1–2.

Chapter 19

RRT: Nurse-Led RRSs Kathy D. Duncan and Christiane Levine

Keywords  Nurse • Rapid • Response • Teams • Recipe • Community • Hospitals Studies have shown, and clinicians are keenly aware, that subtle signs of deterioration can precede life-threatening events, and early identification and treatment of unstable patients may rescue them from progressing to serious instability or death.1–3 Many healthcare systems are now implementing Rapid Response Systems (RRSs) designed to provide for an immediate match between the needs of rapidly deteriorating patients with the knowledge, skills and resources to meet those needs. Whereas the detection of instability and placing the call for help represents the afferent (crisis detection) arm of the RRS, the response to the call represents the efferent arm (crisis response), and can take a variety of forms. In many hospitals and healthcare facilities, the efferent response to recognized instability is to deploy a specially trained team of professionals to immediately respond to the needs of deteriorating patients. Such teams are known as the Medical Emergency Teams (METs) or Rapid Response Teams (RRTs). MET usually denotes a physician-led “high capability” team (both in terms of numbers of responders and treatment options). MET competencies include (1) ability to prescribe therapy; (2) advanced airway management skills; (3) capability to establish central venous access; and (4) ability to begin an ICU-level of care at the bedside. RRT usually refers to a nurseled “intermediate capability” or “ramp-up” team with additional members, such as a respiratory therapist. RRT competencies include (1) rapid assessment of patient needs; (2) beginning basic care to stabilize the patient; (3) rapid triage of patients to a higher level of care if needed; and (4) ability to call in other resources to provide immediate ICU-level care on an expedited basis.4 As healthcare facilities

K.D. Duncan (*) Institute for Healthcare Improvement, Cambridge, MA, USA e-mail: [email protected]

M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems : Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_1, © Springer Science+Business Media, LLC 2011

207

208

K.D. Duncan and C. Levine

and systems embark upon developing, implementing and evaluating their efferent crisis response structure, the Institute for Healthcare Improvement (IHI) recommends that they consider several key elements5: • • • • • • • •

Engage senior leadership support Determine the best structure for the Rapid Response Teams (RRTs) Establish criteria for activating the Rapid Response Teams (RRTs) Establish a simple process for activating the Rapid Response Teams (RRTs) Provide education and training Use standardized tools Establish feedback mechanisms Measure effectiveness

However, one of the hardest elements to address prospectively in the development process is determining the best crisis response structure for individual facilities. As facilities design their rapid response system, a key component in making this decision is determining “who” is available and able to assist patients who are deteriorating in a timely matter. Many hospitals, namely, community and rural hospitals, generally do not have the option of providing a response team led by a physician-intensivist certified in critical-care medicine. Indeed, many community hospitals may not have physician coverage – either intensivists or hospitalists – in the facility around the clock. Instead, they must look within their current facility resources for personnel that can respond to the patient in a timely manner. When no physician is available, the development of the Rapid Response Teams (RRTs) requires not only delineation of specific team roles but also designation of a treatment leader, who may be a nurse with critical-care skills. This can be difficult because of the traditional professional roles in which health care workers have been constrained. There are hypothesized advantages and disadvantages to each model. For example, it is proposed that MET advantages may be the application of definitive care as quickly as possible, and “one-stop-shopping” for emergent services, while disadvantages may include that the MET requires highly trained (and costly) first responders, and may be a model that is intimidating to nurses, leading to delayed call. Proposed advantages to the RRT may be that it feels less intimidating to nurses resulting in earlier calls, and be less expensive, but can at least theoretically delay emergent care delivery.4 Therefore, examining the patient’s needs as well as the hospital’s available services and culture, may be important when determining which response team model is chosen. For example, Children’s Healthcare of Atlanta, a tertiary referral system with 474 staffed beds across three campuses, opted for the RRT model around findings from root cause analyses of patient deteriorations. In some instances, they had uncertainty of their own clinical assessment and wanted validation prior to calling. Their nurse-led team was created to provide staff with a “second look” and support when the staff might be hesitant to call a provider or escalate care. Thus the RRT model seemed to be the best fit to address instability in their institution. In this chapter, we will describe characteristics and logistics of a nurse-led RRT.

19  RRT: Nurse-Led RRSs

209

Identification of Hospital Resources The development of RRT personnel makeup and roles depends on several factors: 1. Availability of RRT members: It is crucial that the staff of the hospital can call for an RRT whenever needed – 24 h a day, 365 days a year. Small community or rural hospitals may have difficulty identifying available resources; they must look at several areas of the hospital that could provide resources to the RRT but currently do not. When the RRT is called, the need is immediate, so the team members must be able to stop whatever they are doing and respond to the call. If the team members – especially the leader – have to prioritize tasks and make a snap decision, they may make incorrect choices. For example, they may choose to complete their current activity, and not make the priority the unseen patient who has begun to deteriorate. Thus the goal of intervening early in the patients’ downhill spiral is thwarted before the response even starts. Therefore, RRT membership – in particular the designation of the team leader – must provide for immediate availability. 2. Accessibility to the RRT: Calling the RRT should be easy – one number, one call. Staff members will not call for the “subtle changes” or when they are “worried” if it is difficult to make the call. For example, if there are different numbers to call on the day shift or the weekend, it becomes a chore to call. Also, educating staff to request the RRT becomes more complex as the number of methods (phone numbers) to activate the crisis response increases. Simplicity and standardization are key. If the RRT is easy to reach, the staff is more likely to call at the first sign of instability. 3. Ability/skills of RRT members: To form a treatment plan, each team member must be able to assess the patient quickly and critically, perform his/her specified duties, and be confident in his/her decision-making skills. The team leader must not only be clinically competent in diagnosis and treatment of patients in crisis, but must also possess and be confident in the skills needed to lead a small group during a crisis. The team members must possess skills that match the tasks they are being asked to complete. For example, it would be neither prudent nor safe to delegate the role of airway manager to someone who is untrained, inexperienced, and unskilled. Nevertheless, non-physicians can perform advanced skills when appropriate training and credentialing is ensured.6

Nursing Leadership of RRTs With these factors in mind, an experienced critical-care nurse frequently fulfills the RRT leadership role. Intensive care units, emergency departments, and post-anesthesia recovery rooms offer nurses opportunities to develop vital skills, such as: • The ability to identify both overt and subtle signs of impending or present patient instability

210

K.D. Duncan and C. Levine

• The ability to accurately collect and recognize key laboratory data • The ability to quickly assess a variety of complex patients • The opportunity to implement evidenced-based protocols and observe immediate patient outcomes • The ability to quickly respond and effectively perform in critical patient situations • Confidence in ability and motivation by the urgency of the patient populations • The ability to work with physicians in consultation rather than at the bedside. Nurse-led RRTs have been successful in various hospital settings, from the small community hospital to the very large tertiary referral centers. Success can be described in a multitude of ways: a decrease in mortality rate; decrease in the number of cardiac arrests; decrease in the delay in transferring patients to a higher; more appropriate level of care; and arguably the most widespread of all staff satisfaction.

Support for the Nurse-Led Rapid Response Team An effective nurse-led RRT must contain the following components: • Communication tools • Specific protocols • Chain of command process.

Communication Tools A common barrier is the ability of the bedside nurse to communicate a concern to the patient’s physician, especially by telephone. Frequently the on-call physician may not be familiar with the patient, and the nurse will only relate the issue that is of concern. For example, the nurse may call and say “Mr. Smith’s temperature is 101.4.” The physician may ask for more information but without the entire picture of the patient, and this may result in an incorrect order of treatment, or in directing the nurse to “watch him and call if he gets worse.” It is imperative that a nurse-led RRT have a standardized, concise method of communicating with physicians by phone. This tool should be brief and include several aspects that construct a complete picture of the patient for the physician who is not in the room. For example, a tool utilizing the acronym SBAR (Situation, Background, Assessment, and Recommendation) allows the RRT nurse-leader to gather all the information and communicate all aspects to the physician when care decisions require escalation (Table 19.1). The facility will need to consistently review the trends noted in the calls and make adjustments to meet the needs of the patients.

19  RRT: Nurse-Led RRSs

211

Table 19.1  Tool for communicating the situation, background, assessment, and recommendation (SBAR) of a patient’s changing condition SBAR Before calling the physician: Tool for calling physicians Assess the patient Review the chart, Identify correct physician Read most recent progress notes Have available: chart, allergies, list, MAR, lab results S B A

Situation: State your name and unit. “I am calling about: (patient name and room number)” “The problem I am calling about is:” Background: State the admission diagnosis and date of admission. State the pertinent medical history Brief synopsis of treatment to date Assessment: Most recent vital signs BP:________ Pulse:________ Respirations:________ Any changes from prior assessments: Mental status: Resp. rate/quality Pain Skin color: Pulse/rhythm change Wound drainage Neurologic changes: BP: N/V, output

Temp:________

R Recommendation: Do you think we should: (state what you would like to see done) For example: Transfer the patient to ICU? Come to see the patient at this time? Ask for a consultant to see the patient now? Do you need any tests like CXR? ABG? EKG? CBC? N otification

Specific Protocols When an RRT response is under the direction of an experienced nurse, protocols may offer resources for immediate action in the absence or unavailability of a physician. Treatment protocols and algorithms that have the support of the physician staff may be utilized by the team; these protocols may include respiratory support and emergency medications. These protocols are facility-specific and must be approved by the appropriate departments and medical staff committees. The protocols should be highly detailed, as specific as the approval to start an IV and/or move to a higher level of care. Saint Francis Hospital in Memphis, Tennessee, has implemented several simple protocols that permit the RRT nurse-leader to order initial interventional and diagnostic procedures early in a call. Table 19.2 is an example of the procedure for the protocols, and Table  19.3 relates physicians’ orders that have been approved by the medical staff structure for use by the RRT nurse-leader.

212

K.D. Duncan and C. Levine

Table 19.2  Examples of nurse-led rapid response team protocols Stat arterial blood gasses:  This may be ordered by the Medical Response Team registered nurse for the patient with suspected pulmonary status deterioration. Patients who are anticoagulated with a continuous heparin infusion or within 24 h of thrombolytic or anti-platelet agent must have a physician’s order to obtain arterial blood gasses by arterial puncture. Pressure will be held until hemostasis occurs, at least 10–15 min or longer Stat electrocardiograms:   A 12-lead electrocardiogram (EKG) may be ordered by the Medical Response Team registered nurse for patients with prolonged severe chest pain and/or significant monitor changes. Chest Pain Emergency Department physician will read the EKG if no cardiologist assigned to patient. The primary physician must be notified of the 12-lead results Stat portable chest X-rays:  This may be ordered by the Medical Response Team registered nurse for the suspected displacement of invasive lines, endotracheal tube displacement, and for severe dyspnea or other signs of respiratory distress Noninvasive temporary pacing (transcutaneous pacing): Medical Response Team registered nurse may initiate noninvasive temporary pacing when the patient’s condition requires Oxygen therapy:  Oxygen via bi-nasal cannula at 2 L/min may be placed on any patient with chest pain or cardiac equivalent or for shortness of breath or increased work of breathing

The RRT at Missouri Baptist Hospital (St. Louis, MO) uses several Respiratory Care Protocols (Table 19.4), while other hospitals have competencies that enable a respiratory therapist to increase respiratory support and even intubate a patient if necessary when a physician is not available; the therapist is both trained and credentialed to perform the procedure. It is important that the RRT protocols be written such that they match available skills, and exclude procedures or tasks which cannot be performed by the team members designated. Thus the protocol development will guide the team development, and vice versa. The contribution of the current nurse caregiver is invaluable. Even though all RRT members may have critical care skills, they will be responding to a variety of sites such, as an orthopedic floor, post-operative unit, or the radiology department. They will rely on the current nurse-caregiver to provide important history and recent physical assessment findings. The current nurse-caregiver also has a rapport with the patient and family that the RRT does not possess. Chain of Command Process In a community and/or rural hospital where a physician in not a member of the crisis response team, or when there are limited physician resources in the hospital during the off-hours, there must be a clear, well-articulated chain of command process. If staff is unable to reach the patient’s physician, the RRT must be well versed in the facility’s physician chain of command. An important task for the nurse-led RRT and physician leadership is to develop and articulate a simple chain of command process for ready reference by RRT members. The RRT must be able to demonstrate the ability to use the established chain of command to obtain physician direction when needed. For example, intensivists, medical directors, or emergency

213 Table 19.3  Treatment protocols for nursing staff participating on a RRT Medical Response Team protocols Decreased level of consciousness Respiratory distress •  Pulse oximetry-oxygen to keep saturation •  Start saline lock above 92%. If history of COPD, get •  Pulse oximetry-oxygen to keep saturation oxygen saturation to 88–90% above 92%. If history of COPD, get •  Arterial blood gas, chest X-ray (for oxygen saturation to 88–90% respiratory distress), complete blood •  If hemodynamically unstablea or orthostatic count, complete metabolic panel hypotensive, give 250 mL bolus of normal •  Begin nebulized albuterol 2.5 mg,one-time saline treatment •  EKG • Cardiac monitor if in moderate or serious •  CBC, basic metabolic panel, urinalysis, distress arterial blood gas, accuchek Physician signature:______________________ •  Start saline lock Date:______________ Time:______________ •  Heart rate and oxygen saturation to be documented after albuterol treatment. Notify physician of HR above 140 or oxygen saturation <92% Physician signature:______________________ Date:______________ Time:______________ Chest pain •  Stat electrocardiogram and have read by critical-care physician immediately •  Place on cardiac monitor •  Start saline lock •  Pulse oximetry-oxygen to keep saturation above 92%. If history of COPD, get oxygen saturation to 88–90% Physician signature:__________________________ Date:______________ Time:______________ Gastrointestinal bleed

Seizures non-pediatric •  IV saline lock •  Pulse oximetry-oxygen to keep saturation above 92% if history of COPD, get oxygen saturation to 88–90% •  Place on cardiac monitor •  Accuchek •  Anti-convulsant drug level if applicable •  Pad bedrails Physician signature:__________________________ Date:______________ Time:______________ Additional orders

•  Start IV with normal saline with large Physician signature:______________________ bore # 18 or larger to keep vein open Date:______________ Time:______________ •  If patient is hemodynamically unstable,a start a second large bore IV and call critical-care physician to bedside •  Send STAT complete blood count, complete metabolic panel, protime, PTT, and type and screen •  If hemodynamically unstable,a change type and screen to type and cross match for 2 units packed red blood cells •  Electrocardiogram •  Pulse oximetry •  Place on cardiac monitor Physician signature:_______________________ Date: Time: Source: Saint Francis Hospital Memphis a Hemodynamically unstable: HR > 120 or systolic blood pressure < 90

214

K.D. Duncan and C. Levine

Table  19.4  Missouri Baptist Medical Center Respiratory Care Department oxygen therapy protocol 1. Purpose: To provide protocol-driven respiratory therapy to evaluate, treat, and monitor appropriate oxygen administration for all non-mechanically ventilated patients, with the intent of treating or preventing the symptoms and manifestations of hypoxia 2. Patient type: All patients currently receiving oxygen will be evaluated with the following exclusions: •  Patients receiving bleomycin treatments •  Patients with CO poisoning for first 12 h •  Acute myocardial infarction patients for first 72 h •  Premature infants •  Patients with specific orders not to change or titrate oxygen •  Patients using oxygen at home who have met home oxygen requirements established by doctor 3. Clinical area: All patient care areas 4. Equipment needed: Pulse oximeter 5. Guidelines: A. The following guidelines will be used in selecting an appropriate oxygen delivery device: 1. High-flow systems provide an adequate flow of oxygen to meet/exceed patients’ inspired flow rate needs 2. Low-flow systems will only provide flow of oxygen to supplement the patients’ inspired flow rate needs 3.  Criteria for use of a high-flow system:   a. Required FIO2 >0.45   b. Tidal volume <300 mL   c. Respiratory rate >25 breaths/min B. Types of low-flow devices: 1.  Cannula   a. Delivers FIO2: ~0.24–0.45   b. Most appropriate initial device for patients with chronic obstructive pulmonary disease 2.  Simple oxygen mask: Delivers FIO2: 0.40–0.50 3.  Non-rebreather mask: Delivers FIO2: 0.60–0.95+ C. Types of high-flow devices: 1.  Venturi mask: Delivers FIO2: 0.24–0.50 2.  High-flow aerosols: Delivers 0.24–0.95+

department physicians may be prospectively designated as resources for the RRT, along with a dedicated rapid contact mechanism so that consultation and escalation of care is not delayed.

Benefits of a Nurse-Led RRT Utilizing experienced nurses in the RRT leadership role may have several benefits. Nurses spend more time with patients than any other healthcare team member, and often have an instinctive and experiential ability to sense a patient’s deterioration,

19  RRT: Nurse-Led RRSs

215

even when they may not possess the fuller range of diagnostic or interventional skills of a physician. Experienced critical-care nurses also demonstrate a patient-centered focus: a steadfast determination to get the failing patient the treatment needed. In an acute care hospital environment, staff nurses are often overwhelmed with multiple complex patients. When a nurse-lead RRT is called to assess a patient, and the team leader is a colleague, the interaction between the RRT leader and the bedside nurse is that of peer-to-peer, with a sense of collaboration for the good of the patient. The call to an RRT should not trigger feelings of judgment as to what did or did not happen before the call – that someone may have missed a crucial element in the assessment, or that the nurse does not know what he or she is doing. The nurse leader – the right nurse leader – will keep the discussion and actions focused on what is best for the patient and refrain from judgment or criticisms. This attitude of mentoring also provides for the learning opportunities for the staff. The ability for bedside nurses to recognize the signs of patient deterioration and call for help are critical components to rescuing patients from harm. A crisis response team, regardless of its makeup, can only intervene when first called upon.7 When bedside nurses are uncertain of their skill set or of a patient’s questionable assessment, the nurse-led RRT can be a great resource. Many staff members are afraid to call the physician with a “gut feeling” or have limited experience to know how to frame their concerns about a patient. Perhaps they know that “something isn’t right” with the patient, but they want another set of eyes at the bedside. The nurse-led Rapid Response Team has had great success in mentoring less experienced staff into understanding the signs of deterioration, with the goal that the bedside nurse recognizes instability earlier and can call, based upon more subtle findings when faced with their next case. Less-experienced nurses may also be afraid to place a call to a physician, even if they have ample evidence to present about the patient of concern. The nurse-led RRT can also help the staff by organizing their thoughts into a structured framework (SBAR), and either calling the physician while the less experienced nurse listens and learns, or by staying with them while they call in the event when the physician asks them questions of which they are uncertain. It is important that the RRT not “take over” the care of the deteriorating patient. The role of the RRT is to provide needed resources emergently to prevent death. Staff nurses who have cared for this patient for a prolonged period will not learn from the event if they are pushed out of the way and the RRT takes over completely. The team led by an experienced nurse should be able to guide that nurse to assist with or observe the assessment, intervention, and communication with the physician. This spirit of peer-to-peer collaboration may not be as strong with a physician-led team (although some hospitals assign a critical-care nurse to include the floor nurse on the response team). An RRT nurse-leader who has a passion for mentoring other nurses to become more confident and comfortable in managing patients in crisis will contribute greatly to that bedside nurse’s development. It may also increase their willingness to call upon the RRT for future events.

216

K.D. Duncan and C. Levine

Another benefit is that many hospitals that may have demonstrated improvement in patient outcomes with the implementation of a crisis-response team have also observed a noticeable improvement in the relationship between ICU nursing staff and ward nursing staff. For example, the nurse-led RRT spends time on each call, assessing, implementing treatment plans, coaching, and observing outcomes. The team can rescue the patients from an acute event and often an ICU stay. In another example, Children’s Healthcare of Atlanta implemented its nurse-led team in December 2006, consisting of an ICU RN and Respiratory Therapist. Atlanta has seen their rate of preventable codes decrease, and has experienced a decrease in the number of patients being found in shock by the Rapid Response Teams (RRTs). Staff has verbalized great satisfaction with the program and the historical dissonance between the ICU staff and the General Care staff has also improved. In over 350 calls, the majority were for respiratory concerns, averaging 30 min for the call, with only 40% of patients requiring ICU admission (Evaluations Tables 19.5 and 19.6).

Nursing Leadership and Mentoring After the RRT Call Several models of nurse-led RRTs provide a follow-up visit after the initial RRT call is completed. The follow-up visit is made by the nurse 12–24 h after the initial visit and is an intentional redundancy that provides another safety net for the patient. During this visit, the nurse assesses the patient and reviews the events that have occurred since the RRT visit to ensure that the interventions were effective and that the current level of care is still appropriate for the patient. This visit also provides an opportunity for a debriefing with the staff currently assigned to the patient. The discussion and review of the patient during a less-urgent time provides a great opportunity for learning for staff members, and is a rare opportunity for collaboration of two professionals to discuss the care of their patient. These discussions build relationships between nursing staff from different areas of the hospital and offer an opportunity to learn from each other. This follow-up visit may be noted on the “call tool,” which then can be used in data collection and given to the frontline management team of the area initiating the call for further learning.

Data Collection Just as with other forms of response teams, data should be gathered from each RRT call. Data can then be analyzed to identify opportunities for improvement. The RRT data collection tool can include the following: • Situation or reason for the call – nursing, physician, or staff activation of MET, i.e. who initiated the call • Assessment of the patient (patient clinical findings, vital sign assessment), acuity status of patient when MET arrives, i.e. is the patient in shock

19  RRT: Nurse-Led RRSs

217

Table 19.5  Rapid response evaluation Pt. Sticker

Rapid Response Team Evaluation

Team Member Feedback 1. Communication with the patient’s Attending

Smooth

Difficult

Smooth

OK

Yes

No

Yes

No

Very good

OK

was: 2. Shifting my work assignment to accommodate

Chaotic

my role on the team was: 3. When I arrived in the patient care area it was

clear who was requesting assistance. 4. When I arrived in the patient care area, it was

clear which patient needed my clinical assessment. 6. I would rate this experience on the Rapid

Bad

Response Team as… 7. I have the following suggestions for improving

the Rapid Response Team:

_________________________________________ _________________________________________ _________________________________________

• Background of the patient (medical history and events leading up to the deterioration) • Recommendations and interventions (treatments delivered) • Outcomes (for example, transferred to a higher level of care, cardiac or respiratory arrest, and survival, observations from a follow-up visit, etc.) In addition to using these data to track the utilization and efficacy of the RRT, the information can also provide validation for the bedside nurse that the RRT call was beneficial, thereby giving her confidence to respond to her instincts and ensure that this behavior will be repeated. By reinforcing the behavior of attempting to rescue the patient, it will encourage continued placement of calls and may provide confidence to make the call even earlier in the patient’s deterioration. Review of the data collected from the nurse-led RRT described above can also help to reveal patterns or repetitions of problems that can be further evaluated and corrected in quality improvement initiatives such as:

218

K.D. Duncan and C. Levine

Table 19.6  Staff Member feedback Pt Sticker

Rapid Response Staff Evaluation **Name of Person Initiating RRT: __________________________________ (for follow up if needed)** Staff Member Feedback 1. The team responded promptly (within 10 minutes).

Yes

No

2. The team was respectful of my need for help.

Very much

Somewhat

Not

3. The team provided education and/or helpful clinical hints to help me in my practice.

Very much

Somewhat

Not

4. The patient’s outcome was impacted by this process.

Positively

Not sure

Negatively

5. My overall satisfaction with the rapid response team was…

Very satisfied

Somewhat satisfied

Not satisfied

6. What else might improve this service?

______________________________________________ ______________________________________________

1. Failure to communicate. For example, a failure in the nurse-to-physician or nurse-to-nurse communication, communication between departments, or failure to reach the provider may need to be improved. 2. Failure to recognize. These failures may include a failure to note a change in heart rate, blood pressure, respiratory status, behavior change, or change in level of consciousness and subsequently result in a failure to act. 3. Failure to plan. These failures may include a failure to place the patient in the appropriate level of care, failure to diagnose, or failure to assess or reassess the patient. 4. Failure to escalate. Many staff nurses are too timid to escalate patient care for a number of reasons including social relationships, not wanting to be wrong, or intimidation by staff of a higher level. 5. Failure to treat. This form of error may include incorrect interventions, undertreatment, or incorrect patient placement.

19  RRT: Nurse-Led RRSs

219

Once such trends are identified, leadership can work to develop mechanisms to address these system-wide failures and improve processes. For example, failure to recognize a crisis might lead to the development of mnemonic tools such as pocket cards or posters, electronic alerts, and the development of educational opportunities to foster better knowledge and performance. The use of data gleaned from RRT events and the institutional response to shortcomings can make the hospital environment safer for all patients. In Atlanta, the Rapid Response Team was rolled out simultaneously with education to all staff on recognition of shock, implementation of mock shock codes, and simulations with the PediaSims mannequins including calling the RRT as a part of simulation for these patient scenarios. Atlanta Children’s has a database that notes the entire assessment form so that each element can be queried – for example, how many 2-year-olds were found in shock with no IV access.

Efficacy As stated earlier, each institution will need to develop its crisis response system in relation to both its needs and resources. We have related anecdotal evidence of the efficacy of nurse-led RRTs. The literature also relates research determining RRT efficacy, although results are inconclusive; Chan et al8 related the impact of an RRT in a single 404-bed institution over 3 years, and although they did not find a reduction in the primary end-point of hospital-wide code-rates, they did demonstrate a reduction in non-ICU code rates.8 Bader et al. in 2000 reported upon the impact of a nurse-led RRT in a single institution, and demonstrated that the rate of floor codes decreased to 17/year from 36/year after RRT implementation, and that the mortality rate for floor-code patients decreased from 61% to 26%. Unanticipated transfers from the ward units to the ICU also decreased significantly.9 In another study of nurse-led RRTs, Bertaut et al. reported that 1 year after implementation, the incidence of code-blue events outside the ICU was reduced by 9%, and overall mortality by 0.12%. Although these results are somewhat mixed, they mirror the mixed results in the literature of the efficacy of physician-led METs.10,11 It is not clear whether these mixed results indicate that the efficacy of crisis intervention teams, or rather the difficulty in studying such a complex issue.

Summary Nurse-lead RRTs may be the ideal mechanisms for crisis intervention, particularly when physician resources are scarce. RRTs frequently utilize a variety of tools and mechanisms to increase effectiveness and ensure patient safety, such as communication pathways, treatment protocols, specialized training in crisis-response skills, physician chain of command documentation, and post-event debriefing of involved staff to improve patient care. Nurses can indeed bring a unique perspective to the

220

K.D. Duncan and C. Levine

leadership role in a Rapid Response Team process. Experience, instinct, determination, and a spirit of collaboration with the nurse at the bedside are attributes that can enhance and sustain the RRT process, improve patient outcomes and, over time, improve the patient safety culture of the facility.

References 1. Schein RM, Hazday N, Pena M, Ruben BH, Sprung CL. Clinical antecedents to in-hospital cardiopulmonary arrest. Chest. 1990;98(6):1388–1392. 2. Franklin C, Mathew J. Developing strategies to prevent in-hospital cardiac arrest: analyzing responses of physicians and nurses in the hours before the event. Crit Care Med. 1994;22(2):244–247. 3. Buist M, Bernard S, Nguyen TV, Moore G, Anderson J. Association between clinically abnormal observations and subsequent in-hospital mortality: a prospective study. Resuscitation. 2004;62(2):137–141. 4. Devita MA, Bellomo R, Hillman K, et al. Findings of the first consensus conference on medical emergency teams. Crit Care Med. 2006;34:2463–2478. 5. Institute for Healthcare Improvement (IHI). How to guide for rapid response systems. (2010) http://www.ihi.org 6. Society of Critical Care Medicine Guidelines Committee. Guidelines for granting privileges for the performance of procedures in critically ill patients. Crit Care Med. 1993;19:275–278. 7. DeVita MA, Smith GB. Rapid response systems: is it the team or the system that is working? Crit Care Med. 2007;35:2218–2219. 8. Chan PS, Khalid A, Longmire LS, Berg RA, Kosiborod M, Spertus TA. Hospital-wide code rates and mortality rates before and after implementation of a rapid response team. JAMA. 2008;300:2506–2513. 9. Bader MK, Neal B, Johnson L, et al. Rescue me: saving the vulnerable non-ICU patient population. Jt Comm J Qual Patient Safe. 2009;35(4):199–205. 10. Winters BD, Pham JC, Hunt EA, et al. Rapid response systems: a systematic review. Crit Care Med. 2007;35:1238–1243. 11. Hillman K, Chen J, Creitkos M, et al. MERIT Investigators. Introduction of the medical emergency team (MET) system: a cluster-randomized controlled trial. Lancet. 2005;365:2091–2097.

Chapter 20

MET: Physician-Led RRSs Daryl Jones and Rinaldo Bellomo

Keywords  Physician-led • Team • Medical • Emergency • Team • MET

Introduction Patients admitted to modern hospitals are complex and have increasing co-morbidity.1 Several studies from multiple countries around the world reveal that patients suffer serious adverse events (SAEs) that prolong length of hospital stay, and may result in permanent disability and death.2–13 Other studies have shown that these events may be preventable and avoidable14,15 and are often preceded by objective signs of deterioration that manifests as derangement in commonly measured vital signs.16–19 Further studies suggest that the response of ward doctors and nurses to these deteriorations may be suboptimal and not commensurate with degree of instability.14,15,20,21 Every hospital must develop strategies to deal with this problem. One such strategy is the physician-led medical emergency team (MET).

Principles Underlying the Physician-Led MET The principles underlying the physician-led MET are similar to those of other non-physician-led rapid response teams (RRTs) (Table  20.1). At most, 17.7% of patients admitted to hospital develop an SAE.2 Accordingly, there needs to be a screening mechanism to detect and identify these deteriorations reliably and promptly. Several studies have shown that in up to 80% of SAEs there are derangements of vital signs in the period leading up to the event.16–19 Such derangements

D. Jones (*) Department of Intensive Care, Austin Hospital, Studley Road, Heidelberg, VIC 3084, Australia e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_20, © Springer Science+Business Media, LLC 2011

221

222

D. Jones and R. Bellomo

Table 20.1  Principles underlying the physician-led MET Serious adverse events (SAEs) are seen in up to 17% of admissions SAEs are preceded by derangements in commonly measured vital signs in up to 80% of cases Development of such derangements predict an increased risk of in-hospital death The MET calling criteria are usually simple and based on derangements in vital signs The response of ward doctors and nurses leading up to an SAE may be suboptimal Early intervention in the course of deterioration improves outcome The common causes of MET calls have effective treatments The expertise to manage deteriorating patients resides in the ICU and can be rapidly deployed

form the basis for the calling criteria for MET/RRTs (Fig. 20.1). The criteria are usually simple, and do not require the ward staff to interpret, calculate, or formulate an “at-risk” score. At least three studies have shown that if a patient develops these criteria they are more likely to die than patients who do not develop them.22–24 Thus, the calling criteria are not only simple but are predictive of a subsequent SAE. Other studies have shown that the evolution of the deterioration is often slow, occurring over hours, and thus there is time to intervene.16,19,25 Examination of the actions of ward doctors and nurses during this period of clinical decline suggests that the response is frequently suboptimal, as they do not have sufficient skills and experience to identify and treat such deteriorations.14,15,20,21 In contrast, intensive care staff are highly trained in advanced resuscitation of acutely unwell patients. Such expertise can be deployed from the ICU when patients develop one or more RRS activation criteria. One of the most important tenets underlying the RRS concept is that early intervention in the course of deterioration improves outcome. This phenomenon has been seen in the early management of severe trauma,26 the resuscitation of patients presenting to the emergency department with sepsis,27 and in association with early administration of thrombolytic therapy for acute myocardial infarction28 and selected cases of ischemic stroke.29

What is a Physician-Led MET? The MET is one form of the response component of an RRS (see Chap. 1). The major difference between the MET and other RRTs is that the team leader of the MET is a physician (M = medical). In Australia and New Zealand, the doctor is usually a registrar (fellow) who is training to be a specialist in Intensive Care Medicine.30 Such trainees undertake rotations in internal medicine, anesthesia, and at least 2 years training in a University affiliated teaching hospital Intensive Care Unit. Accordingly, the MET team leader is ideally credentialed to supervise or perform all aspects of advanced resuscitation, including endotracheal intubation and insertion of invasive vascular access.

20  MET: Physician-Led RRSs

223

Fig. 20.1  Criteria for medical emergency team activation at the Austin Hospital

Ideally, the team should be available 24 h a day, seven days a week, as a number of studies have shown that many MET calls occur out of hours.31 In some cases, the MET supersedes the cardiac arrest team and the MET is called for all medical emergencies.31,32 In other cases, it runs in parallel with the cardiac arrest team and the MET is summoned to review all emergencies other than arrests.33,34 The team should review the deteriorating patient promptly (less than 5 min). According to the statement of the first international conference on METs, the

224

D. Jones and R. Bellomo

Table 20.2  Structure and roles of personnel comprising the MET Staff member Role/responsibility Intensive care fellow Thorough understanding of interplay between clinical medicine, mechanism of disease and therapies for reversal of acute physiological derangement (advanced resuscitation techniques) Skills in airway management and advanced cardiac life support Facilitation of advanced treatment directives Documentation of issues surrounding MET for ongoing audit and quality control Intensive care nurse Advanced knowledge in the application of therapies required in advanced resuscitation Provision of ongoing information and advice to ward nurses for patients remaining on the ward following MET call Liaising with intensive care unit regarding potential for patient admission Medical fellow Skills in diagnosis and management of underlying etiology of medical condition Follow up and ongoing management of patients remaining on ward following MET call Ward nurse Knowledge of patients nursing issues since admissions and leading up to MET call  Respiratory care practitioner (USA) Assistance with respiratory-related therapies

MET should have all of the following core competencies: (1) ability to prescribe therapy; (2) advanced airway management skills; (3) capability to establish central vascular lines; and (4) ability to begin an ICU-level of care at the bedside.35 The key members of the MET typically include the ICU fellow, the internal medicine fellow, and an ICU nurse. Additional members include the ward nurse, and in the USA, a respiratory care practitioner (Table 20.2). The major role of the MET is to act as a triage system in the early phases of clinical deterioration. As stated by England and Bion, the principle of the MET is to “take critical care expertise to the patient before, rather than after, multiple organ failure or cardiac arrest occurs”.36

What Does the Physician-Led MET Do? The MET provides prompt patient review, institutes acute resuscitation and formulates a management plan. In many cases, subsequent management is delegated to ward staff who then follow up the patient and liaise with ICU staff as required. The patient may remain on the ward for full active therapy, or have a limitation of medical therapy or “Do Not Resuscitation” order instituted.37

20  MET: Physician-Led RRSs

225

At risk

Unwell

patient

patient

• Staff member worried about the patient

• No palpable pulse

• HR<40 or >130 beats/min

• No detectable blood pressure

• Systolic blood pressure <90 mmHg

• Unresponsive; and

• Respiratory rate < 80 or > 30 breaths/min

• Not breathing

• Pulse oximetry saturation < 90% • Acute change in conscious state • Urinary output < 50mL in 4 hours

Basic life support commenced

Respond blue call Parent unit notified of MET call & outcome

MET call made Advanced life support commenced

Treated and remain on ward

Patient made Do Not Resuscitate (DNR)

Unplanned ICU admission

Patient dies

Fig. 20.2  Flow diagram showing course of deteriorating patient to subsequent MET criteria or cardiac arrest. Shown are the parallel functioning MET and cardiac arrest team, as well as the potential outcomes from both types of call. MET medical emergency team, HR heart rate

Alternatively, if the acuity exceeds the care that can be provided on the ward, the patient may be urgently transferred to the ICU (Fig.  20.2). In this regard, the MET is providing a prompt second medical opinion for the acutely ill. It also provides a mechanism for ­expediting transfer of the critically ill patient to a higher level of care. We have previously proposed that there should be a minimum standard for the management of an MET call38 including: (1) determining the cause of the deterioration; (2) documenting the events surrounding the MET call; (3) prescribing a management plan and ensuring that appropriate medical follow up occurs; (4) initiating a medical referral in cases where a surgical patient receives an MET call for a medical reason; (5) communicating the occurrence of the MET with the appropriate parent unit doctors; (6) liaising with the ICU consultant on call if the patient fulfills predefined criteria or remains unwell despite the initial resuscitation. An approach to the management of an MET has also been proposed (Table 20.3). This A ® G approach can be adapted to the management of commonly encountered “MET syndromes”.38

226

D. Jones and R. Bellomo

Table 20.3  An “A → G” approach to managing an MET call Ask and Assess   Ask the staff how you can help them   Ask about the reason for the MET call   Assess for the etiology of the deterioration Begin basic investigations and resuscitation therapy Call for help/call consultant if needed Discuss, Decide, and Document   Discuss MET with parent unit/consultant   Discuss advanced care planning if appropriate   Decide where the patient needs to be managed   Document the MET and subsequent frequency of observations Explain: the cause of the MET, the investigations required, and subsequent management plan Follow up: which doctor to follow-up the patient? What are the criteria for doctor re-notification? Graciously thank the staff at the MET MET Medical emergency team, ICU Intensive Care Unit

Like all RRTs, the MET should be part of an RRS. It should be reliably triggered by objective criteria, and linked to governance and quality arms and overseen by an administrative structure.35

Why Do Patients Need MET Calls? There are few studies that assess the cause and treatment of MET reviews.38 A study of 400 MET calls at our institution revealed that the trigger for calling were hypoxia (41%), hypotension (28%), altered conscious state (23%), tachycardia (19%), increased respiratory rate (14%) and oliguria (8%). Infection, pulmonary edema, and arrhythmias were responsible for 53% of all MET calls. At least two other studies have assessed the abnormalities leading to the activation of an MET service. In the original description of the MET, Lee et al31 analyzed the cause of 522 MET calls, 148 of which were cardiac arrests and 62% of which occurred in the emergency department. The most common causes of MET calls in this study were acute respiratory failure, status epilepticus, coma, and severe drug overdose. Kenward et al39 analyzed 136 MET calls over a 12-month period and found that altered conscious state, hypoxia, tachypnea, hypotension and tachycardia were the most common precipitants. There are also very few studies assessing the treatments initiated by the MET.33,34,39,40 Common interventions include administration of fluid boluses, diuretics, and oxygen. In most institutions, such interventions require physician prescription. In a small but significant number of cases, the MET is required to administer vasoactive medications, or perform advanced airway skills (including bag and mask

20  MET: Physician-Led RRSs

227

ventilation, and endotracheal intubation), or insert invasive vascular access en route to the ICU or operating room. The later interventions require the presence of a physician with critical care expertise.

What Are the Advantages and Disadvantages of Physician-Led METs? The rationale behind the RRS concept and the method of activation of the system are simple. The system promotes prompt review of a deteriorating ward patient by a team of experienced and senior staff, where traditionally they would have been seen by a junior member of the medical staff in isolation. The calling criteria are objective, simple, and stress the importance of addressing deranged vital signs. There is no gradation of response, or hierarchy of communication that may create delays in intervention. The calling criteria are meaningful for the ward nurse, who is the typical caller of the MET.34 The MET is a hospital-endorsed mechanism of rapid patient review that may empower nurses to seek help for their deteriorating patients.41 In a questionnaire of nursing staff at our hospital, nurses stated that the MET assisted them in managing sick ward patients, reduced their workload when managing such patients, and actually taught them how to better manage acutely ill patients.41 The MET may assist in establishing end-of-life care planning for patients in whom a switch to comfort care is most appropriate.37,42 Because the MET has a limited number of members, only a relatively small number of personnel need to be trained each year. In addition, physician involvement in the team allows prescription of medications and the conduct of complex and invasive procedures.35 These interventions may be delayed in the case of a nurse-led RRT. The major advantage of the physician-led MET relates to improved patient outcome. Of all the comparative studies of the RRT that demonstrate improved patient outcome, only one involved a nurse-led RRT.43 In contrast, multiple studies in both adults33,34,44–51 and children52,53 demonstrate that a physician-led MET can result in reductions in cardiac arrests, unplanned ICU admissions, in-hospital deaths, and complications following major surgery. There have been a number of stated disadvantages of the physician-led MET. For example, it has been argued that the MET is resource-intensive and deskills ward doctors and nurses.54–55 However, it is arguably more labor-intensive to train all ward doctors in the management of acutely ill ward patients. In addition, nurses at our institution state that the MET improves their skills in managing unwell ward patients.41 The MET review process may create conflict and the MET staff may make errors because of lack of familiarity with the patient.42 Involvement of the parent unit doctors in the assessment and management plan is strongly encouraged to limit this problem. In at least one institution, the composition and method of MET activation was modified because ward staff perceived it as an excessive ramp up for a minor problem.32

228

D. Jones and R. Bellomo

The MET may remove senior ICU staff from their primary role, impacting on the outcomes of those patients already in the ICU.42 It may also distract hospital administrations and clinicians from addressing other mechanisms of improving patient outcome such as staffing levels and training, and hospital-wide policies for end-of-life care planning.56 Finally, despite the simplicity of the approach, even in hospitals with a mature MET, there is frequently delayed activation of the MET, which has been associated with increased patient mortality (see Chap. 16). Developing alternative strategies for providing prompt and competent medical review of deteriorating ward patients appears daunting. This would require repeated education of multiple junior medical staff on multiple occasions per year. Even with such an approach, it is unlikely that ward staff would attain the competency of an ICU fellow. In addition, there would need to be a mechanism for reviewing deteriorating surgical patients in the event that the surgical staff were in the operating room and unable to attend. Finally, a prompt and reliable mechanism of referral to the ICU would need to be ensured. In conclusion, the physician-led MET appears to be the most tested and most reliably effective approach to the efferent arm of an RRS. We contend that it should be the standard and preferred approach until other, less intensive approaches can be demonstrated to be the equivalent in effectiveness and safety.

References 1. Zajac JD. The public hospital of the future. Med J Aust. 2003;179:250–252. 2. Andrews LB, Stocking C, Krizek T, et al. An alternative strategy for studying adverse events in medical care. Lancet. 1997;349:309–313. 3. Baker GR, Norton PG, Flintoft V, et al. The Canadian Adverse Events Study: the incidence of adverse events among hospital patients in Canada. CMAJ. 2004;170:1678–1686. 4. Bellomo R, Goldsmith D, Russell S, Uchino S. Postoperative serious adverse events in a teaching hospital: a prospective study. Med J Aust. 2002;176:216–218. 5. Brennan TA, Leape LL. Adverse events, negligence in hospitalized patients: results from the Harvard Medical Practice Study. Perspect Healthc Risk Manage. 1991;11:2–8. 6. Brennan TA, Leape LL, Laird NM, et al. Incidence of adverse events and negligence in hospitalized patients. Results of the Harvard Medical Practice Study I. N Engl J Med. 1991;324:370–376. 7. Davis P, Lay-Yee R, Briant R, Ali W, Scott A, Schug S. Adverse events in New Zealand public hospitals I: occurrence and impact. N Z Med J. 2002;115:U271. 8. Davis P, Lay-Yee R, Briant R, Ali W, Scott A, Schug S. Adverse events in New Zealand public hospitals II: preventability and clinical context. N Z Med J. 2003;116:U624. 9. Leape LL, Brennan TA, Laird N, et al. The nature of adverse events in hospitalized patients. Results of the Harvard Medical Practice Study II. N Engl J Med. 1991;324:377–384. 10. Schimmel E. The hazards of hospitalization. Ann Intern Med. 1964;60:100–110. 11. Thomas EJ, Studdert DM, Burstin HR, et  al. Incidence and types of adverse events and negligent care in Utah and Colorado. Med Care. 2000;38:261–271. 12. Vincent C, Neale G, Woloshynowych M. Adverse events in British hospitals: preliminary retrospective record review. BMJ. 2001;322:517–519. 13. Wilson RM, Runciman WB, Gibberd RW, Harrison BT, Newby L, Hamilton JD. The quality in Australian Health Care Study. Med J Aust. 1995;163:458–471.

20  MET: Physician-Led RRSs

229

14. Bedell SE, Deitz DC, Leeman D, Delbanco TL. Incidence and characteristics of preventable iatrogenic cardiac arrests. JAMA. 1991;265:2815–2820. 15. Hodgetts TJ, Kenward G, Vlackonikolis I, et al. Incidence, location and reasons for avoidable in-hospital cardiac arrest in a district general hospital. Resuscitation. 2002;54:115–123. 16. Buist MD, Jarmolowski E, Burton PR, Bernard SA, Waxman BP, Anderson J. Recognising clinical instability in hospital patients before cardiac arrest or unplanned admission to intensive care. A pilot study in a tertiary-care hospital. Med J Aust. 1999;171:22–25. 17. Hodgetts TJ, Kenward G, Vlachonikolis IG, Payne S, Castle N. The identification of risk factors for cardiac arrest and formulation of activation criteria to alert a medical emergency team. Resuscitation. 2002;54:125–131. 18. Nurmi J, Harjola VP, Nolan J, Castren M. Observations and warning signs prior to cardiac arrest. Should a medical emergency team intervene earlier? Acta Anaesthesiol Scand. 2005;49:702–706. 19. Schein RM, Hazday N, Pena M, Ruben BH, Sprung CL. Clinical antecedents to in-hospital cardiopulmonary arrest. Chest. 1990;98:1388–1392. 20. Hayward RA, Hofer TP. Estimating hospital deaths due to medical errors: preventability is in the eye of the reviewer. JAMA. 2001;286:415–420. 21. McQuillan P, Pilkington S, Allan A, et  al. Confidential inquiry into quality of care before admission to intensive care. BMJ. 1998;316:1853–1858. 22. Bell MB, Konrad D, Granath F, Ekbom A, Martling CR. Prevalence and sensitivity of METcriteria in a Scandinavian University Hospital. Resuscitation. 2006;70:66–73. 23. Buist M, Bernard S, Nguyen TV, Moore G, Anderson J. Association between clinically abnormal observations and subsequent in-hospital mortality: a prospective study. Resuscitation. 2004;62:137–141. 24. Goldhill DR, McNarry AF. Physiological abnormalities in early warning scores are related to mortality in adult in-patients. Br J Anaesth. 2004;92:882–884. 25. Franklin C, Mathew J. Developing strategies to prevent in-hospital cardiac arrest: analyzing responses of physicians and nurses in the hours before the event. Crit Care Med. 1994;22:244–247. 26. Blow O, Magliore L, Claridge JA, Butler K, Young JS. The golden hour and the silver day: detection and correction of occult hypoperfusion within 24 hours improves outcome from major trauma. J Trauma. 1999;47:964–969. 27. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med. 2001;345:1368–1377. 28. Fresco C, Carinci F, Maggioni AP, et al. Very early assessment of risk for in-hospital death among 11,483 patients with acute myocardial infarction. GISSI investigators. Am Heart J. 1999;138:1058–1064. 29. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N Engl J Med. 1995;333:1581–1587. 30. Jacques T, Harrison GA, McLaws ML. Attitudes towards and evaluation of medical emergency teams: a survey of trainees in intensive care medicine. Anaesth Intensive Care. 2008;36:90–95. 31. Lee A, Bishop G, Hillman KM, Daffurn K. The medical emergency team. Anaesth Intensive Care. 1995;23:183–186. 32. Jones DA, Mitra B, Barbetti J, Choate K, Leong T, Bellomo R. Increasing the use of an existing medical emergency team in a teaching hospital. Anaesth Intensive Care. 2006;34:731–735. 33. Bellomo R, Goldsmith D, Uchino S, et  al. Prospective controlled trial of effect of medical emergency team on postoperative morbidity and mortality rates. Crit Care Med. 2004;32: 916–921. 34. Bellomo R, Goldsmith D, Uchino S, et al. A prospective before-and-after trial of a medical emergency team. Med J Aust. 2003;179:283–287. 35. Devita MA, Bellomo R, Hillman K, et al. Findings of the first consensus conference on medical emergency teams. Crit Care Med. 2006;34:2463–2478.

230

D. Jones and R. Bellomo

36. England K, Bion JF. Introduction of medical emergency teams in Australia and New Zealand: a multicentre study. Crit Care. 2008;12:151. 37. Jones DA, McIntyre T, Baldwin I, Mercer I, Kattula A, Bellomo R. The medical emergency team and end-of-life care: a pilot study. Crit Care Resusc. 2007;9:151–156. 38. Jones D, Duke G, Green J, et al. Medical emergency team syndromes and an approach to their management. Crit Care. 2006;10:R30. 39. Kenward G, Castle N, Hodgetts T, Shaikh L. Evaluation of a medical emergency team one year after implementation. Resuscitation. 2004;61:257–263. 40. Cretikos M, Hillman K. The medical emergency team: does it really make a difference? Intern Med J. 2003;33:511–514. 41. Jones D, Baldwin I, McIntyre T, et al. Nurses’ attitudes to a medical emergency team service in a teaching hospital. Qual Saf Health Care. 2006;15:427–432. 42. Joyce C, McArthur C. Rapid response systems: have we MET the need? Crit Care Resusc. 2007;9:127–128. 43. Priestley G, Watson W, Rashidian A, et  al. Introducing critical care outreach: a ward-­ randomised trial of phased introduction in a general hospital. Intensive Care Med. 2004;30:1398–1404. 44. Bristow PJ, Hillman KM, Chey T, et al. Rates of in-hospital arrests, deaths and intensive care admissions: the effect of a medical emergency team. Med J Aust. 2000;173:236–240. 45. Buist M, Harrison J, Abaloz E, Van Dyke S. Six-year audit of cardiac arrests and medical emergency team calls in an Australian outer metropolitan teaching hospital. BMJ. 2007;335:1210–1212. 46. Buist MD, Moore GE, Bernard SA, Waxman BP, Anderson JN, Nguyen TV. Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. BMJ. 2002;324:387–390. 47. DeVita MA, Braithwaite RS, Mahidhara R, Stuart S, Foraida M, Simmons RL. Use of medical emergency team responses to reduce hospital cardiopulmonary arrests. Qual Saf Health Care. 2004;13:251–254. 48. Jones D, Bellomo R, Bates S, et al. Long term effect of a medical emergency team on cardiac arrests in a teaching hospital. Crit Care. 2005;9:R808–R815. 49. Jones D, Egi M, Bellomo R, Goldsmith D. Effect of the medical emergency team on long-term mortality following major surgery. Crit Care. 2007;11:R12. 50. Jones D, George C, Hart GK, Bellomo R, Martin J. Introduction of medical emergency teams in Australia and New Zealand: a multi-centre study. Crit Care. 2008;12:R46. 51. Jones D, Opdam H, Egi M, et al. Long-term effect of a medical emergency team on mortality in a teaching hospital. Resuscitation. 2007;74:235–241. 52. Sharek PJ, Parast LM, Leong K, et al. Effect of a rapid response team on hospital-wide mortality and code rates outside the ICU in a children’s hospital. JAMA. 2007;298:2267–2274. 53. Tibballs J, Kinney S, Duke T, Oakley E, Hennessy M. Reduction of paediatric in-patient cardiac arrest and death with a medical emergency team: preliminary results. Arch Dis Child. 2005;90:1148–1152. 54. Kerridge RK, Saul WP. The medical emergency team, evidence-based medicine and ethics. Med J Aust. 2003;179:313–315. 55. Chan PS, Khalid A, Longmore LS, Berg RA, Kosiborod M, Spertus JA. Hospital-wide code rates and mortality before and after implementation of a rapid response team. JAMA. 2008;300:2506–2513. 56. Winters BD, Pham J, Pronovost PJ. Rapid response teams – walk, don’t run. JAMA. 2006;296:1645–1647.

Chapter 21

Pediatric RRSs James Tibballs and Richard J. Brilli

Keywords  Pediatric • Rapid • Response • Systems

Introduction The incidence of cardiac arrest among hospitalized children is small but the outcome is poor, despite expert resuscitation. For example, in two large series of outcomes from many hospitals, only 24% of 544 children1 and 27% of 880 children2 experiencing cardiac arrest survived to hospital discharge. In the latter study, 34% of the survivors had severe neurological dysfunction. In smaller, single hospital series, survival at 1 year was 34%,3 19%4 and 15%.5 Cardiac arrest and life-threatening critical illness in children, as in adults, is often preceded by warning signs and symptoms.6 Although cardiac arrest in children may be sudden and without warning, it is often due to progressive hypoxemia or hypotension or both resulting from diverse illnesses. If the severity of a condition is recognized, it may be possible to intervene and prevent cardiopulmonary arrest and death. Around the world, various strategies have been developed to recognize critical illness and to mobilize early assistance with the aim of preventing cardiopulmonary arrest. These interventions are systems approaches to reducing adverse events in hospitals. The principles described for adult patient systems elsewhere in this volume apply equally to systems for children. This chapter describes the characteristics of pediatric rapid response systems, recognition of critical illness and the outcomes of established systems. Although various systems solutions have been adopted, all have the common themes of early recognition of acute illness and a triggered therapeutic response with follow up.

J. Tibballs (*) Intensive Care Unit, Royal Children’s Hospital, Flemington Road, Parkville, 3052, Australia e-mail: [email protected]

M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_21, © Springer Science+Business Media, LLC 2011

231

232

J. Tibballs and R.J. Brilli

Development and Operation of Pediatric Rapid Response Systems In response to the realization that some deaths in hospitals are preventable, the Institute for Healthcare Improvement recommended the creation of rapid response systems.7 Different names and compositions have been imparted to these around the world, but they all strive to prevent cardiopulmonary arrest. Typically, medical emergency teams (METs) are composed of doctors and nurses, rapid response teams (RRTs) are composed of either doctors and nurses or nurses alone, while critical care outreach teams (CCOTs) and patient-at-risk teams (PARTs) are usually composed of nurses alone but with rapid access to physician assistance. In this chapter, “MET” and “RRT” are used interchangeably. Some institutions utilize liaison8 or “rover”9 nurses whose tasks are to help colleagues identify children on wards with critical illness and encourage activation of their rapid response system. Setting up a system requires the consideration of numerous factors. From a pediatric practice perspective, the most prominent considerations are the ­immediacy of response (operation), composition of the team, and choice of system activation triggers or calling criteria.

Operational Team Responses: One-Tier Vs. Two-Tier Two different strategies have been adopted regarding the urgency of responding to a request for expert assistance. Some pediatric institutions consider a request for assistance as needing the same urgent response as that for a “code” – that is, for a respiratory or cardiopulmonary arrest. These are “single-tiered” systems.9–13 These teams are multidisciplinary, and include physicians, nurses, respiratory therapists, pharmacy personnel, security, social service, and sometimes surgeons. Other institutions have adopted a “two-tiered” system in which there is a dual response according to the urgency and severity of the patient’s condition.14,15 The first tier is a small focused team that responds to consults and requests for advice and is not required to attend immediately but must do so within a specified period, e.g., within 15  min. The second tier is a larger multidisciplinary team similar to the one-tier system and functions as a “code blue” team for cardiopulmonary arrest and must attend immediately. Improved outcomes have been observed with both one-tiered and two-tiered systems.14 There are advantages and disadvantages to both one- and two-tiered systems. Advantages of the one-tiered system include quick provision of definitive care and provision of all services. Disadvantages of the one-tiered system include: (1) requirement for highly skilled personnel, even for consultative function; (2) intimidating barrier for staff to call the team for a consultation or evaluation; and (3) high cost compared to two-tiered system. In contrast, the first tier of a two-tiered system may be less costly and less intimidating for clinical staff to activate when

21  Pediatric RRSs

233

consultation and advice are sought. However, in the two-tiered system, it is also possible that the initial smaller, more focused team will be under-skilled to handle a genuine life-threatening emergency. Each institution must balance the advantages and disadvantages of the two systems because the personnel who respond and the time commitment of the responders may be quite different.

Recognition of Children with Critical Illness If children at risk for cardiopulmonary arrest can be identified, more resources and effort can be devoted to their care in order to prevent respiratory, or cardiopulmonary arrest. Little research has been done on groups of patients at high risk of unexpected cardiac arrest. Kinney et al16 observed that children in need of urgent assistance from an established MET were infants less than 1 year of age, children with chronic or complex illnesses, or those in the immediate post-operative period. These children sustained critical events, including cardiac or respiratory arrest and required interventions that included the need for reversal of analgesia or sedation, resuscitation with fluid boluses or management of acute hyponatremia. Children sustaining a critical event had a significantly higher incidence of in-hospital mortality. In the absence of secure knowledge of children at risk of cardiopulmonary arrest, institutions have devised various schema to alert their medical and nursing staff that a patient is deteriorating and may be at risk for cardiopulmonary arrest. Two general approaches (actiation triggers or calling criteria and early warning scores) have been developed, but they are not mutually exclusive.

Activation Triggers or Calling Criteria One approach is to specify triggers or calling criteria to activate the rapid response system. In contrast to adult rapid response systems, pediatric systems are faced with the problem of adjusting the trigger or calling criteria according to different ages and sizes of patients. Some institutions have chosen quite specific age-related physiological variables including heart rate, blood pressure and respiratory rate (Table 21.1), while others have chosen more open-ended disturbances of cardiovascular and respiratory function (Tables  21.2–21.4). Both types of calling criteria include the option for a clinician to activate the team because they are “worried” even if the other calling criteria are not met. Some hospitals have allowed the “parent” to activate the team as well.13,14 The presence and recognition by staff or parent of any one criterion is the impetus to activate the rapid response system. However, to date, none of the triggers or calling criteria have been evaluated to determine their sensitivity and specificity in preventing cardiac arrest. Brilli et al14 attempted to identify reliable activation triggers from over 1,000 combinations of pre-arrest variables derived from previous

234

J. Tibballs and R.J. Brilli

Table 21.1  Criteria for activation of medical emergency team Royal Children’s Hospital, Melbourne Any ONE or more of: 1. Nurse, doctor or parent worried about clinical state 2. Airway threat 3. Hypoxemia: SpO2 <90% in any amount of oxygen SpO2 <60% in any amount of oxygen (cyanotic heart disease) 4. Severe respiratory distress, apnea or cyanosis 5. Tachypnea: Age Term-3 months 4–12 months 1–4 years 5–12 years 12 years+

Respiratory rate/min >60 >50 >40 >30

6. Tachycardia or bradycardia: Age Term-3 months 4–12 months 1–4 years 5–12 years 12 years+

Bradycardia (beats/min) <100 <100 <90 <80 <60

Age Term-3 months 4–12 months 1–4 years 5–12 years 12 years+

BP (systolic mmHg) <50 <60 <70 <80 <90

Tachycardia (beats/min) >180 >180 >160 >140 >130

7. Hypotension:

8. Acute change in neurological status or convulsion 9. Cardiac or respiratory arrest

Table 21.2  Medical emergency team activation criteria Cincinnati Children’s Hospital Increased work of breathing and any of the following   Worsening retractions   Saturations <90% despite supplemental oxygen   Cyanosis or gray skin color Agitation or decreased level of consciousness Heart rate ³ 180 in the absence of fever or agitation PEWS (pediatric early warning score) score ³ 7 Staff concern or worry about the patient Parental concern about the child

21  Pediatric RRSs

235 Table 21.3  Rapid response team activation criteria Lucile Packard Children’s Hospital, Stanford 1. Any staff member worried about a patient 2. Acute change in respiratory rate 3. Acute change in oxygen saturation 4. Acute change in heart rate 5. Acute change in blood pressure 6. Acute change in level of consciousness

Table 21.4  Medical emergency team (RRT) calling triggers Johns Hopkins Hospital, Baltimore 1. Respiratory distress/compromise 2. Abnormal or worsening respiratory symptoms 3. Decrease in saturations despite first-line interventions 4. Seizures with apnea 5. Progressive lethargy 6. Circulatory compromise/acute shock syndrome 7. Supraventricular tachycardia/other dysrhythmias 8. Acute change in neurologic/mental status 9. Respiratory arrest 10. Cardiac arrest 11. Worried staff 12. Worried family member

codes but were unable to do so with sufficient sensitivity and specificity. Instead, they chose criteria based on chart analysis and expert opinion. It is possible that the search for more reliable activation criteria will not impact significantly on patient outcomes. Nonetheless, studies with detailed pediatric early warning scores (PEWS) offer valuable insights into activation and calling criteria. In the absence of persuasive data regarding the sensitivity and specificity of the calling criteria, surrogate markers may be used. A surrogate marker of the sensitivity of calling criteria may be the reduction in proportion of preventable cardiac arrests and other adverse events. For example, Tibballs and Kinney13 found a 55% reduction in preventable cardiac arrest and Sharek et al,11 a 72% reduction for all cardiac arrests; however, no system has reduced preventable deaths to zero, regardless of how defined. A surrogate marker of specificity may be the disposition of patients following team activation (call-outs). A high rate of non-admission to the intensive care unit suggests over-utilization of the RRS and low specificity of the activation criteria. The optimal activation of the RRS, as might be indicated by the ratio of non-admission and admission to a critical care unit, is unknown. However, there is remarkable agreement between two large international units13,14 that suggest that their activation criteria are reasonably specific. At Cincinnati Children’s,14 52% of 27 RRS activations over 12 months did not result in ICU admission14 and 50% of 263 calls over 3  years (R. Brilli, unpublished data). Tibballs and Kinney13 found 53% of 809 calls over 4 years did not result in ICU admission. One interpretation

236

J. Tibballs and R.J. Brilli

of these data is that the RRT is being over-utilized; however, another interpretation is that in order to maintain a high degree of suspicion and awareness by the general care clinical staff, a 50% ICU admission rate may be optimal. Moreover, in the majority of call-outs, the RRT renders treatment beyond just advice. In the report of Sharek et al,12 advice only was given in 6% of RRT activations and in Tibballs and Kinney13 it was 16%.

Early Warning Scores Development of PEWS or criteria (tools) for detecting clinical deterioration may help improve the system of identifying the deteriorating child. The scores are compilations of points attributed to various physiological abnormalities. Four PEWS and one tool have been evaluated in terms of predicting serious adverse events outside the intensive care unit environment. The first PEWS was developed by Monaghan17 at Royal Alexandra Children’s Hospital in Brighton (UK). The score is based on the behavior of the child and on cardiovascular (skin color, capillary refill and heart rate) and respiratory parameters (chest wall retractions and relationship to normal respiratory rate). Additional ­scoring points are added according to treatment such as the need for added oxygen and nebulizers. According to the summed score, the nurse would inform the nurse in charge, or increase the frequency of observations, or call for a medical review or call out the full medical (treatment) team. The system has not been fully evaluated, but a small retrospective study of 30 patients appeared to achieve appropriate management decisions. Tucker et  al18 at Cincinnati Children’s Hospital (USA) evaluated Monaghan’s PEWS (with a minor adoption, Table  21.5) among 2,979 children admitted to a single medical unit over a 12-month period. The outcome measure was the need to transfer a patient to the PICU as a substitute for cardiac arrest. Fifty-one patients needed admission to the PICU and there was a significant association between this outcome and the magnitude of the PEWS. The area under a receiver-operating curve was 0.89 (95% CI 0.84–0.94, p < 0.001). Another early warning system was developed at Cardiff in Wales (UK) and is known as the Cardiff and Vale Paediatric Early Warning System (C&VPEWS).19,20 The eight trigger criteria, taking 15–30  s to calculate are airway threat; oxygen required to keep saturation >90%; respiratory rate; respiratory observation; heart rate; blood pressure; level of consciousness; and, nurse or doctor worried about the child’s clinical state. Each abnormal criterion scores one point. Edwards et  al20 evaluated PEWS for predicting adverse outcomes including respiratory arrest, cardiac arrest, admission to a high-dependency area or PICU, and death. Of 1,000 children studied, 16 had adverse outcomes but none died, and there were no cardiac or respiratory arrests. The area under a receiver operator curve for the system was 0.86 (95% CI 0.82–0.91). Interestingly, the reliability of any one trigger criterion alone was almost as good as the optimum number of triggers, which was two; the

Within normal parameters, no retractions

>10 above normal parameters, using accessory muscles OR 30+% FiO2 or 3+ l/min

Score 2 extra for half-hourly nebulizers or persistent vomiting following surgery Required minimum actions for scores 0–13 by bedside nurse: 0–2: no intervention 3: senior RN to assess patient 4: notify pediatric resident 5: senior RN and pediatric resident to assess patient 6: senior RN, pediatric resident, and senior resident to assess patient ³7: activate MET Bedside nurse may contact seniors or activate MET regardless of PEWS

Respiratory

Table 21.5  A pediatric in-patient early warning scoring system Components 0 1 Behavior Playing/ Sleeping appropriate Cardiovascular Pink or capillary Pale or capillary refill 3 s refill 1–2 s

>20 above normal parameters Retractions OR 40+%FiO2 or 6+ l/min

Gray or capillary 4 s OR tachycardia of 20 above normal rate

2 Irritable

3 Lethargic/confused OR reduced response to pain Gray and mottled or capillary refill 5 s or above OR tachycardia of 30 above normal rate or bradycardia Five below normal parameters with retractions Grunting OR 50% FiO2 or 8+ l/min

Score

21  Pediatric RRSs 237

238

J. Tibballs and R.J. Brilli

sensitivity of any one trigger criterion was 89% with a specificity 64%, while a score of 2 had a sensitivity of 70% and a specificity of 90% (figures rounded), indicating that any two of the triggers should prompt an RRT response. Nonetheless, with these derivations, 30% of children in need of assistance would not be attended and 10% would activate a system unnecessarily. A more complicated PEWS was developed at the Hospital for Sick Children at Toronto, Canada,21 to predict actual or impending cardiac arrest. The score, to a maximum of 26 points, is the summation of points for physiological status, treatment, and known medical conditions. It includes age-related heart rate, respiratory rate, blood pressure, general items including pulse, saturation, capillary refill, coma score, oxygen therapy, need for fluid bolus and temperature and treatment and condition items including abnormal airway, home oxygen, previous PICU admission, central venous line, transplant recipient, cerebral palsy, gastrostomy tube, number of medications and involvement of more than three medical specialities. The area under the receiver operator curve using up to 16 items scored was 0.90 with maximum reliability (summation of sensitivity and specificity) at a score of 5, with a sensitivity of 78% and specificity of 95%. This implies that 22% of children needing urgent assistance would not have triggered a response and 5% would have triggered a response unnecessarily. Although this study centered on what should be prevented – cardiac arrest – it may be too complicated for routine ward use, in addition to the issue of not identifying all children in need of assistance. Lastly, a PEW Tool developed and evaluated at Bristol Royal Hospital for Children22 used need for enhanced care, respiratory arrest, cardiac arrest or death as outcomes, while the triggers were any one of a combination of symptoms, signs and treatment, including airway obstruction, breathing, circulation, disability and laboratory values. Unfortunately, only patients who triggered the tool were studied, thus limiting evaluation of the tool’s positive predictive value.23,24 When the Bristol tool was applied to unplanned admissions to Liverpool Children’s Hospital PICU, it identified most but not all children at risk,25 but like the previous study, it was not applied to children not admitted to PICU, thus preventing comprehensive evaluation. Early warning scores thus have the potential to predict adverse events but thus far have not been evaluated for the prevention of cardiac arrest. Conclusions drawn from studies to date suggest that simple observations rather than complex scoring is sufficient to detect and trigger an expert response.

Outcomes of Some Pediatric Rapid Response Systems Although rapid response systems have been the subject of discussion and debate in adult hospitals for many years, the uptake of these systems by pediatric hospitals or combined adult-pediatric hospitals has been relatively slow. In North America, by 2005, 24% of 181 hospitals with more than 50 acute pediatric beds had established activation criteria in place for “code” teams but only 3% had activation criteria for a rapid response system.26 Similarly in the UK, only 22% of 144 hospitals caring for children had an early warning system.19

21  Pediatric RRSs

239

Thus far only a few pediatric institutions have reported their experience with rapid response systems to prevent cardiac and respiratory arrest.9–15 The main outcomes are summarized in Table 21.6. All have been before-and-after studies, that is, evaluating the incidence of cardiac arrest and death before and after the introduction of a rapid response system. Several have observed gratifying improvements in patient outcomes but a randomized trial is highly desirable.27 A rapid response system was started in 2002 at Royal Children’s Hospital, Melbourne, and preliminary results were published after 12 months of operation.9 It was commenced after analysis of general care unit adverse events revealed that in many cases warning symptoms and signs had not been appreciated or had resulted in ineffective resuscitative treatments. Some deaths may have been preventable. The system operated under a single-tier system – that is, every call-out was regarded as a need to treat a cardiac or respiratory arrest. It was composed of doctors from the PICU, emergency department and the admitting medical unit and a nurse from the PICU. After 12 months of operation, although there were decreases in the total number of cardiac arrests on wards, the decreases were not significant. However, there was complete elimination of so-called “preventable” cardiac arrests and death, that is, cardiac arrests occurring in children whose physiological parameters contravened the calling criteria. After another 3 years of operation,13 although there was a non-significant decrease in unexpected cardiac arrest, the decreases in general care unit preventable cardiac arrest (55%) and death (87%) were significant. Moreover, the total hospital death rate decreased significantly by 34% or 34 deaths eliminated per year. It was estimated that one life was saved for every 72 RRT call-outs. During this interval, deaths in the PICU also decreased (31%), so the reduction in general care unit deaths did not occur because of the transfer of patients to PICU. Another pediatric one-tiered rapid response system that has achieved notable success was commenced in 2005 at the 264-bed Lucille Packard Children’s Hospital in Stanford, California.12 The system operated on a single-tier basis and used as activation triggers any acute change in respiratory rate, oxygen saturation, heart rate, blood pressure or consciousness state or concern of a staff member. Its team, composed of a PICU doctor and nurse, respiratory therapist and nursing supervisor, attended 90 activations annually. It achieved significant reductions in-hospital mortality (18%) and in respiratory and cardiac arrest (72%). Of patients in need of treatment by the team, only 29% did not require admission to a level of higher care. A two-tiered system, operated by Cincinnati Children’s Hospital and commencing in 2005, achieved a significant decrease (59%) in code rate (respiratory + cardiopulmonary arrest) over a period of 8 months in the ward environment.14 This reduction was principally due to reduction of respiratory arrest. This system utilized activation criteria consisting of increased work of breathing combined with worsening retractions, oxygen saturation <90% or cyanosis, or agitation or decreased consciousness, or staff or parental concern about the child. From 27 MET calls, 14 (52%) were transferred to PICU. In total, these studies and others10,12,15 demonstrate that efforts to reduce preventable cardiac arrest and death in pediatric institutions are effective, but they also show that they have not entirely eliminated predictable and therefore possibly preventable

56 Lucile Packard Hospital, Stanford Zenker et al 22 Children’s 200815 Hospitals and Clinics, Minneapolis Hunt et al 12 Johns Hopkins 200912 Hospital, Baltimore Tibballs and 41 Royal Kinney13,a Children’s Hospital, Melbourne a  Continuation of initial study (Tibballs et al9)

Sharek et al 200711 19

12

12

48

 3

–

 2

 3

Table 21.6  Main outcomes from pediatric rapid response systems PostMET/ PreRRT ImpleMET/RRT comparison mentation operation (months) (months) (months) Publication Institution Tibballs et al Royal 41  3 12 20059 Children’s Hospital, Melbourne Mistry et al  6  5 Duke 200610 Children’s Hospital Brilli et al 15  4  8 Cincinnati 200714 Children’s Hospital

202

88

150

90

41

>24

Calls (per annum) 184

One-tiered Doctors, nurses, RT, pharmacist One-tiered Doctors and nurses

One-tiered Doctors, nurses, RT Two-tiered Nurses, RT

One-tiered Doctors and nurses Two-tiered Doctors and nurses

Response system One- tiered Doctors and nurses

Significant reductions in all hospital deaths (34%), preventable cardiac arrest (55%) and death (87%). Thirty-four hospital deaths prevented annually

Significant reduction of respiratory arrests (73%), no change in cardiac arrest

Significant decreases in non-PICU cardiac arrest (58%) with death (48%) and all in-patient non-PICU hospital death (12%) Significant decreases in cardiopulmonary arrest rates per 1,000 non-ICU admissions; nonsignificant decreases in preventable cardiac arrest (43%) and death (55%) Significant reductions in all cardiac arrests (72%) and all death (18%). Twenty-one hospital deaths prevented annually Non-significant decreases in respiratory and cardiac arrest (36%)

Main findings Non-significant decreases in all cardiac arrest and death but elimination of preventable cardiac arrest and death

240 J. Tibballs and R.J. Brilli

21  Pediatric RRSs

241

cardiac arrest and death. In some cases,12–14 cardiac arrest has occurred without warning, and these cases may reasonably be classified as unpreventable. In other cases however, cardiac arrest occurred despite the presence of premonitory symptoms and signs. Obvious explanations are that their symptoms and signs were not included in the activation criteria, or, if they were, they were not recognized in a timely fashion, were not acted upon, or intervention was unsuccessful.

Barriers to Implementation and Use of Rapid Response Systems The adoption of a rapid response system is a cultural change that must overcome many institutional barriers. The key feature is empowerment of any staff, including junior nurses and doctors, and parents, to summon help without deferring to senior colleagues or to medical staff. This traditional medical hierarchy is a formidable barrier to change. Junior personnel are not accustomed to questioning senior physician staff or even nurses questioning physicians. A key step in overcoming this barrier is to challenge all staff to “put the patient first.” What does the patient need from the hospital team? Not hierarchy, but rather dedication to keeping the child safe – in this case, the recognition that unexpected and preventable death can be reduced by the use of an RRS. The change requires input and acceptance from all groups of personnel directly involved in patient care. A multidisciplinary group of personnel (physicians, nurses, respiratory therapist, physiotherapists, managers) should convene to: (1) define the role of a rapid response system, its composition, the type of response (one-tier vs. two-tier); (2) activation criteria; (3) educate staff; (4) develop data collection methods; (5) evaluate, debrief and provide feedback to front-line clinical staff regarding the effects of the RRS. Like any institutional change, it will not be successful without the full support of senior hospital management whose influence is particularly important in breaking down communication barriers between disciplines and departments. It is essential that all personnel recognize that patients are the responsibility of the institution and not only the responsibility of a particular ward, department or physician or surgical team. Simply put, although patients are the responsibility of defined entities, they do not “belong” to them and patients should have the benefit of whatever other assistance the institution can offer. Summoning such other assistance may be perceived as “treading on another’s turf” or disempowering the attending primary treating team, but the system is not intended to supplant the traditional role of the resident/fellow/attending doctor. It is intended to forestall disastrous outcomes by bringing to the bedside experienced personnel to deal with situations ordinarily beyond the capability of regular wards. Further improvements in outcomes may relate to improved education of front-line nurses and doctors to recognize critical illness and to overcome activation barriers to a rapid response system. Team composition, active surveillance of patients, factors related to willingness to activate the system, and determination of preventability of adverse events are all likely factors that determine the success of rapid response

242

J. Tibballs and R.J. Brilli

systems. An institutional culture wherein junior staff members are comfortable calling for advice and help without fear of recrimination may also contribute to the success of a RRS. Rapid and constant feedback to staff are also important components of RRT success. Indeed, one may argue that irrespective of timely recognition of critical illness, the biggest challenge to RRS implementation and effectiveness is the reluctance to activate the system. There are no published studies regarding obstacles to activation of pediatric systems, but barriers that contributed to avoidable adult in-hospital cardiac arrest included reluctance by junior doctors to seek assistance from senior doctors.28 Although nurses readily accept rapid response services,29,30 the most important barrier is adherence to the traditional hierarchical referral to the parent medical service rather than fear of criticism.31 Less experienced nurses are reluctant to call MET,30 as are junior doctors.32,33 In the early stages of introduction of a MET service, nurses required reassurance that MET calls were warranted if the calling criteria are breached and that they would not be reproached by their supervisors or by senior physicians.32 Full acceptance of a service may take several years,34 with repeated education and periodic assessment of site-specific obstacles to utilization.35 It is likely that the same problems exist in pediatric institutions.

References 1. Young KD, Seidel JS. Pediatric cardiopulmonary resuscitation: a collective review. Ann Emerg Med. 1999;33:195–205. 2. Nadkarni VM, Larkin GL, Peberdy MA, et al. First documented rhythm and clinical outcomes from in-hospital cardiac arrest among children and infants. JAMA. 2006;295:50–57. 3. Tibballs J, Kinney S. A prospective study of outcome of in-patient paediatric cardiopulmonary arrest. Resuscitation. 2006;71:310–318. 4. Olkkola SP, KT VV, et al. Utstein style reporting of in-hospital paediatric cardiopulmonary resuscitation. Resuscitation. 2000;45:17–25. 5. Reis AG, Nadkarni V, Perondi MB, et al. A prospective investigation into the epidemiology of in-hospital pediatric cardiopulmonary resuscitation using the international Utstein reporting style. Pediatrics. 2002;109:200–209. 6. McQuillan P, Pilkington S, Allan A, et  al. Confidential enquiry into quality of care before admission to intensive care. BMJ. 1998;316:1853–1858. 7. Institute for Healthcare Improvement. Five million lives campaign. ; Accessed 29.06.09. 8. Caffin CL, Linton S, Pellegrini J. Introduction of a liaison nurse role in a tertiary paediatric ICU. Intensive Crit Care Nurs. 2007;23:226–233. 9. Tibballs J, Kinney S, Duke T, et al. Reduction of pediatric in-patient cardiac arrest and death with a medical emergency team: preliminary results. Arch Dis Child. 2005;90:1148–1152. 10. Mistry KP, Turi J, Hueckel R, Mericle JM, Meliones JN. Pediatric rapid response teams in the academic medical center. Clin Ped Emerg Med. 2006;7:241–247. 11. Sharek PJ, Parast LM, Leong K, et al. Effect of a rapid response team on hospital-wide mortality and code rates outside the ICU in a children’s hospital. JAMA. 2007;298:2267–2274. 12. Hunt EA, Zimmer KP, Rinke ML, et al. Transition from a traditional code team to a medical emergency team and categorization of cardiopulmonary arrests in a children’s center. Arch Pediatr Adolesc Med. 2008;162:117–122.

21  Pediatric RRSs

243

13. Tibballs J, Kinney S. Reduction of hospital mortality and preventable cardiac arrest and death on introduction of a pediatric medical emergency team. Pediatr Crit Care Med. 2009;10:306–312. 14. Brilli RJ, Gibson R, Luria JW, et al. Implementation of a medical emergency team in a large pediatric teaching hospital prevents respiratory and cardiopulmonary arrests outside the intensive care unit. Pediatr Crit Care. 2007;8:236–246. 15. Zenker P, Schlesinger A, Hauck M, et al. Implementation and impact of a rapid response team in a children’s hospital. Joint Comm J Qual Patient Saf. 2007;33:418–425. 16. Kinney S, Tibballs J, Johnston L, Duke T. The clinical profile of hospitalized children provided with urgent assistance from a medical emergency team. Pediatrics. 2008;121:e1577–e1584. 17. Monaghan A. Detecting and managing deterioration in children. Paediatr Nurs. 2005;17:32–35. 18. Tucker KM, Brewer TL, Baker RB, Demeritt B, Vossmeyer MT. Prospective evaluation of a pediatric in-patient early warning scoring system. JSPN. 2009;14:79–85. 19. Duncan HP. Survey of early identification systems to identify inpatient children at risk of physiological deterioration. Arch Dis Child. 2007;92:828. 20. Edwards ED, Powell CV, Mason BW, Oliver A. Prospective cohort study to test the predictability of the Cardiff and Vale Paediatric Early Warning System (C&VPEWS). Arch Dis Child. 2008; doi:10.1136/adc.2008.142026. 21. Duncan H, Hutchinson J, Parshuram CS. The pediatric early warning system score: a severity of illness score to predict urgent medical need in hospitalised children. J Crit Care. 2006; 21:271–279. 22. Haines C, Perrott M, Weir P. Promoting care for acutely ill children – development and evaluation of a paediatric early warning tool. Intensive Crit Care Nurs. 2006;22:73–81. 23. Tibballs J, Kinney S. Evaluation of a paediatric early warning tool – claims unsubstantiated. Intensive Crit Care Nurs. 2006;22:315–316. 24. Haines C. Evaluation of a paediatric early warning tool – claims unsubstantiated. Intensive Crit Care Nurs. 2006;22(6):317. 25. Tume L. The deterioration of children in ward areas in a specialist children’s hospital. Nurs Crit Care. 2007;12(1):12–19. 26. VandenBerg SD, Hutchinson JS, Parshuram CS, et al. A cross-sectional survey of levels of care and response mechanisms for evolving critical illness in hospitalized children. Pediatrics. 2007;119(4):e940–e946. 27. Bonafide CP, Priestley MA, Nadkarni VM, Berg RA. Have we MET the answer for preventing in-hospital deaths or is it still elusive? Pediatr Crit Care Med. 2009;10:403–404. 28. Hodgetts T, Kenwood G, Vlackonikolis I, et al. Incidence, location and reasons for avoidable in-hospital cardiac arrest in a district general hospital. Resuscitation. 2002;54:115–123. 29. Galhotra S, Scholle CC, Dew MA, et al. Medical emergency teams: a strategy for improving patient care and nursing work environments. J Adv Nurs. 2006;55(2):180–187. 30. Salamonson Y, van Heere B, Everett B, et al. Voices from the floor: nurses’ perceptions of the medical emergency team. Intensive Crit Care Nurs. 2006;22(3):138–143. 31. Jones D, Baldwin I, McIntyre T, et al. Nurse’s attitudes to a medical emergency team service in a teaching hospital. Qual Saf Health Care. 2006;15:427–432. 32. Daffurn K, Lee A, Hillman KM, et al. Do nurses know when to summon emergency assistance? Intensive Crit Care Nurs. 1994;10:115–120. 33. Buist MD, Jarmolowski E, Burton PR, et al. Recognising clinical instability in hospital patients before cardiac arrest or unplanned admission to intensive care. Med J Aust. 1999;171:22–25. 34. Jones D, Bates S, Warrillow S, et al. Effect of an education programme on the utilization of a medical emergency team in a teaching hospital. Intern Med J. 2006;36:231–236. 35. Jones DA, Mitra B, Barbetti J, et al. Increasing the use of an existing medical emergency team in a teaching hospital. Anaesth Intens Care. 2006;34:731–735.

Chapter 22

Sepsis Response Team Emanuel P. Rivers, David Amponsah, and Victor Coba

Keywords  Sepsis • Rapid • Response • System • Early • Goal-directed • Therapy

Introduction The transition from sepsis to severe life-threatening disease frequently develops well before admission to an intensive care unit (ICU), often in the pre-hospital setting, the emergency department (ED), general medical-surgical floors, operating­ room or the outpatient clinic setting. One would hope that as soon as possible after the sepsis syndrome occurs, treatment with resuscitation fluids, restoration of adequate oxygen delivery to the tissues, and antimicrobial therapy would begin. However, optimal care may be delayed for many reasons, including lack of recognition, ED overcrowding and long wait times for ICU beds. Delays to medical emergency teams (MET) and rapid response team involvement may further hinder the availability of expert care to prevent morbidity and mortality. For years we have recognized that delay in care negatively impacts outcome for trauma, myocardial infarction, and stroke.1–3 There is now robust evidence that treatment delay can also negatively impact outcome in sepsis.4–7 This chapter will review the evidence for early intervention in sepsis, models of delivery, and potential obstacles. While a specific team response is not required, a Rapid Response System with a trained efferent limb may provide one mechanism for providing a rapid, coordinated team response to patients presenting to the ­hospital with signs of sepsis.

E.P. Rivers (*) Wayne State University, Henry Ford Hospital, Emergency Medicine and Surgery, Detroit, MI, USA e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_22, © Springer Science+Business Media, LLC 2011

245

246

E.P. Rivers et al.

Early Goal-Directed Therapy In 2001, Rivers et al showed that Early Goal-Directed Therapy (EGDT) provided in the ED at the most proximal stages of severe sepsis and septic shock produced significant outcome benefit (Table  22.1).5 The interventions stipulated by EGDT were commonly used ICU therapies. What was novel in this report was earlier application of these therapies in the ED within 1  h of presentation. EGDT is an algorithmic approach to resuscitation designed to correct existing (and prevent the recurrence of) hemodynamic and oxygen delivery instability of severe sepsis/septic shock within the first 6 h of hospital care (Fig. 22.1). It may be especially helpful in high-risk patients. It is a “bundle-of-care” strategy based on early recognition and rapid resolution of inadequate systemic oxygen delivery and subsequent global tissue hypoxia that occurs in severe sepsis. There are a number of “triggers” to embark on EGDT; however, we favor in particular the use of a lactate greater than 4 mM/L because it has been shown to be associated with high mortality in both the pre-hospital and ED setting.8–11 EGDT targets achieving normal oxygen delivery by optimizing preload using central venous pressure monitoring, afterload using mean arterial pressure, and oxygen delivery using hemoglobin and inotropic agents guided by central venous oxygen saturation (SCVO2). The goal of these interventions is to correct high lactate levels and ameliorate global tissue hypoxia. In the Rivers et al study, patients presenting to the ED with severe sepsis and septic shock were randomized to standard therapy vs. an EGDT protocol. Mean ED length of stay was 6.3 h vs. 8.0 h, in the stangard and EGDT groups, respectively. The standard therapy group received volume resuscitation and vasoactive therapy guided by central venous pressure andt arterial pressure monitoring. The EGDT group received the same volume resuscitation and vasoactive therapy, but resuscitation in this group was guided by SCVO2 monitoring (Edwards Lifesciences, Irvine, CA) to assess global tissue hypoxia, which might occur despite normalized central venous pressure, blood pressure, and urine output. Additional fluid resuscitation, blood transfusions, ventilatory support, and inotropic therapy were used to reach a goal SCVO2 of at least 70%. All patients received the same standard of care after admission to the ICU, with no Table 22.1  Entry criteria for EGDT5 Clinical suspicion of infection Plus Two of four systemic inflammatory response syndrome (SIRS) criteria   Temperature ³ 38°C or <36°C   Heart rate > 90 beats∙min−1   Respiratory rate > 20 breaths·min−1 or PaCO2 < 32 mmHg   White blood cell count > 12,000 ml−1 or < 4,000 ml−1 or >10% immature forms Plus either Hypotension (systolic blood pressure < 90 mmHg after a 30 ml·kg−1 bolus) or Arterial lactate ³ 4 mmol/Liter−1

22  Sepsis Response Team

247

Suspected infection within 2 hours and Source Control within 6 hours

The High Risk Patient: Blood pressure < 90 mmHg after 20-40 cc/kg volume challenge or lactic acid > 4 mmole/liter

Antibiotics within 1 hour and Source Control

CVP

<8 mmHg

Crystalloid

>8-12 mmHg

Decrease Oxygen Consumption

MAP

<65 or >90 mmHg

Vasoactive Agent(s)

>65-90 mmHg

ScvO2

<70%

>70% Packed red blood cells to Hct >30%

>70% Ionotrope (s)

No

Fig. 22.1  The EGDT protocol

Goals Achieved

<70%

248

E.P. Rivers et al.

further involvement by the ED or the investigators. The ICU clinicians who assumed care were blinded to the randomization order. A significant absolute hospital mortality reduction of 16% was observed in the EGDT group. The EGDT group also required significantly less vasopressor therapy, mechanical ventilation, pulmonary artery catheterization and experienced a greater modulation of systemic inflammation after admission to the ICU.12 Of the patients who survived, the EGDT group had significantly shorter hospital and ICU lengths of stay. Numerous similar studies performed in the ED, general hospital floors and ICU have replicated these outcome findings since the original trial in adults.5,13–33 Children34 also have been shown benefit from EGDT. The “bundle” was found to be cost-effective across a broad range of assumptions.35–37 Furthermore, the rate of achieving normalcy is important: reaching the endpoint of ScvO2 within 6  h has been shown to be associated with increased survival over those patients who do not reach this endpoint in the first 6 h38 for patients with severe sepsis or septic shock, or those within 48  h of ­sustaining acute lung injury.39,40

Implementing EGDT Implementing EGDT requires several components: (1) early recognition of high-risk patients; (2) mobilization of resources for the required interventions; (3) performance of the early resuscitation interventions41; (4) a continuous quality improvement (CQI) program; and (5) a continuing education program. These components mirror the four limbs of the Rapid Response System, even if a specific response team is not utilized to achieve the goals. As with any therapy, a collaborative team approach for planning and intervention is required, with integration of medical, nursing, and support staff from both the ED and ICU. Effective EGDT teams are based on reliable mobilization of resources – both personnel and equipment – to perform the required tasks. In addition, an appropriately sized pool of health care providers (both physicians and nurses) who can effectively deliver EGDT is a key to successful implementation. Most importantly, a local “champion” is needed to be responsible for EGDT implementation and to communicate effectively with both the ED and ICU personnel. Although all successful EGDT programs share strong collaboration between the ED and ICU, there are basically three models of implementation: (1) an ED-based model, (2) a mobile ICU or MET, team-based model, and (3) a completely ICUbased model. (All these models fit the criteria for an RRS). In all models, it is the ED’s responsibility to identify patients with sepsis that require EGDT (afferent limb). The efferent limb of the three models may be different, however. In the ED-based model, the ED staff would then perform all the necessary steps to deliver EGDT before admission to the ICU. This is the model followed at Henry Ford Hospital in Detroit, MI, where the original EGDT study was carried out. This model has been shown to be feasible in other studies.

22  Sepsis Response Team

249

For some hospitals, an ICU-based “mobile sepsis team” model may be a better fit. In this model, an ICU-based physician or a surrogate would be notified by the ED and would go there to begin delivery of EGDT, prior to transfer to the intensive care unit. EGDT would be continued from the ED to the ICU, and transfer to the ICU would occur as soon as possible. In this regard, the “sepsis team” is essentially a specialized MET, organized to provide EGDT. Hospitals that have already successfully implemented a MET may be able to expand its role to the provision of EGDT. This model requires significant cooperation between the ED and intensive care unit, along with 24-h availability of healthcare providers capable of obtaining thoracic central venous access. It would likely work best in hospitals with a significant ICU bed wait-time in the face of the pressures for a rapid ED turnaround. Lastly, an ICU-based model would entail ED notification of ICU staff followed by transfer as rapidly as possible to the ICU, where EGDT would be instituted (only after arrival in the ICU). In effect, this model is the method that most hospitals follow for critically ill patients (i.e., ED recognition and initial stabilization of critically ill patients, followed by ICU admission for definitive care). The primary limitation of this model is that ICU beds may not be readily available,6 leading to delays in initiating potentially life-saving therapy and defeating the goals of the treatment regime.

Barriers to Implementation of EGDT As with any new therapy or process, implementation of EGDT would face the familiar barriers of inertia and lack of resources.42 However, EGDT also faces several unique challenges. Successful delivery of EGDT begins with the recognition that it is a hospital-wide problem. This requires open communication between the ED, general practice floors, operating room, recovery room and the ICU. This involves physician, nurses and allied healthcare professionals. Potential problems include miscommunication around such issues as billing and transfer of care between providers. Importantly, each department’s autonomy must be respected to avoid resentment and non-cooperation. Both teams of health care providers (ED and ICU) need to develop a sense of joint “ownership” of patients presenting with sepsis, similar to how patients with ST-segment elevation myocardial infarction (STEMI) are managed today (early ED recognition, work-up according to protocol and care in the ED, early cardiologist notification, with choice of revascularization technique determined by local resources).43 EGDT also requires the routine implementation of technologies that may be seen as “ICU-based” by many healthcare providers, such as measurements of central venous pressure, blood pressure (noninvasive and invasive), as well as continuous monitoring of SCVO2; up to 50% will be mechanically ventilated as well. The successful use of these technologies outside the ICU may require new levels of expertise among health care providers outside of their usual domain.

250

E.P. Rivers et al.

Summary Since 2001, EGDT has been repeatedly shown to significantly reduce mortality in patients presenting to the ED with sepsis or septic shock. However, widespread implementation of EGDT requires a collaborative effort between multiple disciplines (physicians, nurses, allied health personnel) across multiple specialties (emergency medicine, hospitalists and critical care medicine). This effort is best led by a local “champion” who can scientifically and diplomatically communicate with all the stakeholders. Models of EGDT delivery can be ED-based, team-based, or ICU-based. While all three approaches share elements of a successful Rapid Response System, the mobile team-based approach is similar to an MET and could be an extension of an already-existing MET service. It may also be the most amenable to a fluid delivery of care from the ED to the intensive care unit. Other effective solutions may be dictated by local resources.

References 1. Nathens AB, Jurkovich GJ, Maier RV, et al. Relationship between trauma center volume and outcomes. JAMA. 2001;285:1164–1171. 2. Anderson HV, Willerson JT. Thrombolysis in acute myocardial infarction. N Engl J Med. 1993;329:703–709. 3. Marler JR, Brott T, Broderick J, Kothari R, Odonoghue M, Barson W. Tissue plasminogen activator for acute ischemic stroke. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. N Engl J Med. 1995;333:1581–1587. 4. Han YY, Carcillo JA, Dragotta MA, et al. Early reversal of pediatric-neonatal septic shock by community physicians is associated with improved outcome. Pediatrics. 2003;112:793–799. 5. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med. 2001;345:1368–1377. 6. Chalfin DB, Trzeciak S, Likourezos A, Baumann BM, Dellinger RP. Impact of delayed transfer of critically ill patients from the emergency department to the intensive care unit. Crit Care Med. 2007;35:1477–1483. 7. Lundberg JS, Perl TM, Wiblin T, et al. Septic shock: an analysis of outcomes for patients with onset on hospital wards versus intensive care units. Crit Care Med. 1998;26:1020–1024. 8. Pearse RM. Extending the role of lactate measurement into the pre-hospital environment. Crit Care. 2009;13:115. 9. Mikkelsen ME, Miltiades AN, Gaieski DF, et al. Serum lactate is associated with mortality in severe sepsis independent of organ failure and shock. Crit Care Med. 2009;37:1670–1677. 10. Trzeciak S, Dellinger RP, Chansky ME, et  al. Serum lactate as a predictor of mortality in patients with infection. Intensive Care Med. 2007;33:970–977. 11. Shapiro NI, Howell MD, Talmor D, et al. Serum lactate as a predictor of mortality in emergency department patients with infection. Ann Emerg Med. 2005;45:524–528. 12. Rivers EP, Kruse JA, Jacobsen G, et al. The influence of early hemodynamic optimization on biomarker patterns of severe sepsis and septic shock. Crit Care Med. 2007;35:2016–2024. 13. Gao F, Melody T, Daniels DF, Giles S, Fox S. The impact of compliance with 6-hour and 24-hour sepsis bundles on hospital mortality in patients with severe sepsis: a prospective observational study. Crit Care. 2005;9:R764–R770. 14. Kortgen A, Niederprum P, Bauer M. Implementation of an evidence-based “standard operating procedure” and outcome in septic shock. Crit Care Med. 2006;34:943–949.

22  Sepsis Response Team

251

15. Sebat F, Johnson D, Musthafa AA, et al. A multidisciplinary community hospital program for early and rapid resuscitation of shock in non-trauma patients. Chest. 2005;127:1729–1743. 16. Shapiro NI, Howell MD, Talmor D, et  al. Implementation and outcomes of the Multiple Urgent Sepsis Therapies (MUST) protocol. Crit Care Med. 2006;34:1025–1032. 17. Trzeciak S, Dellinger RP, Abate NL, et al. Translating research to clinical practice: a 1-year experience with implementing early goal-directed therapy for septic shock in the emergency department. Chest. 2006;129:225–232. 18. Lin SM, Huang CD, Lin HC, Liu CY, Wang CH, Kuo HP. A modified goal-directed protocol improves clinical outcomes in intensive care unit patients with septic shock: a randomized controlled trial. Shock. 2006;26:551–557. 19. Micek ST, Roubinian N, Heuring T, et al. Before-after study of a standardized hospital order set for the management of septic shock. Crit Care Med. 2006;34:2707–2713. 20. Nguyen HB, Corbett SW, Steele R, et al. Implementation of a bundle of quality indicators for the early management of severe sepsis and septic shock is associated with decreased mortality. Crit Care Med. 2007;35:1105–1112. 21. Sebat F, Musthafa AA, Johnson D, et  al. Effect of a rapid response system for patients in shock on time to treatment and mortality during 5 years. Crit Care Med. 2007;35: 2568–2575. 22. Chen ZQ, Jin YH, Chen H, Fu WJ, Yang H, Wang RT. Early goal-directed therapy lowers the incidence, severity and mortality of multiple organ dysfunction syndrome. Nan Fang Yi Ke Da Xue Xue Bao. 2007;27:1892–1895. 23. He ZY, Gao Y, Wang XR, Hang YN. Clinical evaluation of execution of early goal directed therapy in septic shock. Zhongguo Wei Zhong Bing Ji Jiu Yi Xue. 2007;19:14–16. 24. Qu HP, Qin S, Min D, Tang YQ. The effects of earlier resuscitation on following therapeutic response in sepsis with hypoperfusion. Zhonghua Wai Ke Za Zhi. 2006;44:1193–1196. 25. Castro R, Regueira T, Aguirre ML, et al. An evidence-based resuscitation algorithm applied from the emergency room to the ICU improves survival of severe septic shock. Minerva Anestesiol. 2008;74:223–231. 26. Akinnusi ME, Alsawalha L, Pineda LA, El Solh AA. Outcome of septic shock in the elderly following the implementation of the sepsis bundle: a propensity-adjusted analysis. Chest. 2007;132:494. 27. Jones AE, Focht A, Horton JM, Kline JA. Prospective external validation of the clinical effectiveness of an emergency department-based early goal-directed therapy protocol for severe sepsis and septic shock. Chest. 2007;132:425–432. 28. Ferrer R, Artigas A, Levy MM, et al. Improvement in process of care and outcome after a multicenter severe sepsis educational program in Spain. JAMA. 2008;299:2294–2303. 29. Zubrow MT, Sweeney TA, Fulda GJ, et al. Improving care of the sepsis patient. Jt Comm J Qual Patient Saf. 2008;34:187–191. 30. Zambon M, Ceola M, Almeida-de-Castro R, Gullo A, Vincent JL. Implementation of the Surviving Sepsis Campaign guidelines for severe sepsis and septic shock: we could go faster. J Crit Care. 2008;23:455–460. 31. Thiel SW, Asghar MF, Micek ST, Reichley RM, Doherty JA, Kollef MH. Hospital-wide impact of a standardized order set for the management of bacteremic severe sepsis. Crit Care Med. 2009;37:819–824. 32. Focht A, Jones AE, Lowe TJ. Early goal-directed therapy: improving mortality and morbidity of sepsis in the emergency department. Jt Comm J Qual Patient Saf. 2009;35:186–191. 33. Moore LJ, Jones SL, Kreiner LA, et al. Validation of a screening tool for the early identification of sepsis. J Trauma. 2009;66:1539-1546. Discussion 1546–1547. 34. de Oliveira CF, de Oliveira DS, Gottschald AF, et al. ACCM/PALS haemodynamic support guidelines for paediatric septic shock: an outcomes comparison with and without monitoring central venous oxygen saturation. Intensive Care Med. 2008;34:1065. 35. Shorr AF, Micek ST, Jackson WL Jr, Kollef MH. Economic implications of an evidence-based sepsis protocol: can we improve outcomes and lower costs? Crit Care Med. 2007; 35:1257–1262.

252

E.P. Rivers et al.

36. Huang DT, Clermont G, Dremsizov TT, Angus DC. Implementation of early goal-directed therapy for severe sepsis and septic shock: a decision analysis. Crit Care Med. 2007;35: 2090–2100. 37. Talmor D, Greenberg D, Howell MD, Lisbon A, Novack V, Shapiro N. The costs and costeffectiveness of an integrated sepsis treatment protocol. Crit Care Med. 2008;36:1168–1174. 38. Pope JV, Jones AE, Gaieski DF, Arnold RC, Trzeciak S, Shapiro NI. Multicenter study of central venous oxygen saturation (ScvO(2)) as a predictor of mortality in patients with sepsis. Ann Emerg Med. 2009;55(1):40–46. 39. Varpula M, Tallgren M, Saukkonen K, Voipio-Pulkki LM, Pettila V. Hemodynamic variables related to outcome in septic shock. Intensive Care Med. 2005;31:1066–1071. 40. Grissom CK, Morris AH, Lanken PN, et al. Association of physical examination with pulmonary artery catheter parameters in acute lung injury. Crit Care Med. 2009;37:2720–2726. 41. Dellinger RP, Levy MM, Carlet JM, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008. Crit Care Med. 2008;36: 296–327. 42. Berwick DM. Disseminating innovations in health care. JAMA. 2003;289:1969–1975. 43. De Luca G, Suryapranata H, Ottervanger JP, Antman EM. Time delay to treatment and mortality in primary angioplasty for acute myocardial infarction: every minute of delay counts. Circulation. 2004;109:1223–1225.

Chapter 23

Other Efferent Limb Teams: (BAT, DAT, M, H, and Trauma) Daniel Shearn and Michael DeVita

Keywords  Specialized • Response • Critical • Needs • Efferent • Limb Most of the world’s literature on Rapid Response Systems (RRS), Rapid Response Teams (RRT), and Medical Emergency Teams (MET) focuses on in-hospital crises for patients with sudden physiological deterioration. At UPMC Presbyterian Hospital, we have found that certain low-frequency, high-risk events require specialized expertise not available on our “routine” MET. As a result, we created a series of teams, all part of the Rapid Response System, which can address a variety of these critical events. The purpose of this chapter is to describe how and why we developed our Medical Emergency Teams and rapid response system to promote patient safety. We will describe each team response, staffing, purpose, and outcomes. The RRS in its “native” form is a system that includes a team response that will rescue patients from critical or emergent situations. The goal was to bring the Critical Care Unit and resources to the bedside of the critical patient in need. At UPMC, we found that after the Rapid Response System had become part of our culture, there were crisis events that did not fit neatly into our existing system either because the team did not have the skills, or the crises were not medical in nature and not amenable to a clinical intervention team. We felt that these events could benefit from an organized response to a clear triggering event. So as crises occurred and repeated, we undertook to identify an RRS approach to their management. In this chapter, we will list the various teams that were developed at UPMC Presbyterian Hospital. We will assume that the reader has reviewed other chapters in this text to understand the components of the RRS, and how to develop a competent system. So we will begin by describing other teams that we developed, such as Stroke Team, Trauma Team, Blood Administration Team (BAT), Chest Pain Team, Condition L (for Lost patient) Team, Difficult Airway Team (DAT), Pediatric Response Team, and Condition M. While this list is not inclusive of all teams developed in the US, it does

D. Shearn (*) UPMC Presbyterian Hospital, Pittsburgh, PA 15213, USA e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, 253 DOI 10.1007/978-0-387-92853-1_23, © Springer Science+Business Media, LLC 2011

254

D. Shearn and M. DeVita

compromise a broader scope than most RRSs that have been described. One additional example of a specialized team is the “secondary victim” team described in detail in Chap. 28. We acknowledge that there may be a generalizability issue for some of these teams, especially in smaller hospitals. Our point in this chapter is to describe how the RRS model can be adapted to other crises, and show examples of this RRS structure.

Basic Condition Response At UPMC-Presbyterian Hospital, our MET (or Condition C) response has been in place for many years. The team consists of two physicians, a respiratory therapist (RT), two intensive care registered nurses (RN), a floor RN, and a nurse’s aide or personnel certified in cardiopulmonary resuscitation (CPR). Table 23.1 designates the roles and responsibilities of the team members. Note that sometimes roles are overlapping in nature and that consideration to manpower sometimes dictates personnel performing dual roles.

Stroke Team Why. The UPMC Presbyterian Hospital organized a Stroke Team to improve the hospital’s ability to manage efficiently and effectively patients brought to the Emergency Department. The team was very effective in speeding the process of Table 23.1  Team members and their responsibilities Role Personnel Airway Manager MD Airway Assistant

RT

Bedside Assistant

Floor RN

Crash Cart Mgr

ICU RN

Treatment Leader

MD or ICU RN until MD arrives

Circulation

CPR-certified personnel

Procedure MD

MD

Data Manager (ICU RN)

ICU RN

Responsibility Assess, assist ventilation, intubate Assist airway manager, oxygen and suction setup, suction as needed Check pulse, obtain vital signs, assess patient IVs, deliver meds Deploy equipment, prepare meds, run defibrillator Assess team responsibilities, data, direct treatment, set priorities, triage patient. Check pulse, place defib pads, perform chest compressions Perform procedures, IVs, chest tubes, ABGs Role tags, AMPLE, lab results, chart, record interventions

23  Other Efferent Limb Teams: (BAT, DAT, M, H, and Trauma)

255

intake, physical evaluation, diagnostic testing, and therapeutic intervention. Subsequent to the success of this ED-based team, we noted that patients already in the hospital who had stroke-like symptoms did not have the same efficiency of care. As a result, we sought to improve our management capabilities, and considered an in-patient Stroke Team. The Stroke Team was instituted to efficiently treat patients that have exhibited symptoms of a stroke. The plan was to get the stroke therapy experts to the bedside of a patient with emergent neurological symptoms suggestive of an acute neurovascular event, and enable prompt therapeutic interventions like tPA, angioplasty, or craniotomy. With the goal being early treatment, whether medicinal or surgical intervention, having the personnel that knew how to get the job done was essential. Who. This team consisted of a physician trained and expert in the treatment of strokes. The team is activated only by the physician in charge of the basic MET that was caring for the patient. This helped to ensure that calls would be activated that were legitimate, thus decreasing unnecessary calls. It also promotes use of the MET team and makes decision making regarding calls easier for lesser-skilled individuals: the “caller” has only to decide if there is a crisis, and not diagnose the crisis. How. If a patient has neurologic symptoms, an MET response is triggered, bringing critical care resources. Should the treating physician in the team response perceive an acute neurovascular event, then the Stroke Team is activated. In our model, the MET remains to provide basic critical care support (circulation, respiration, etc.), and the stroke team leader performs an assessment and determines neurovascular care. The Data. Data of outcomes is important in learning if the team is working. It is also helpful to continue to gain financial support for the team. Suggested Stroke Team outcome data can consist of monitoring time to treatment, number of calls activated or mortality of patients treated, and time lapse from onset of symptoms to initiation of treatment (e.g., tPA or surgery), if any.

Trauma Team Why. Rarely, patients have an acute traumatic event despite already being hospitalized. Examples include falls, suicide attempts, and assault. Should a severe traumatic event occur, often the resources brought by the MET are not sufficient to manage the crisis. Therefore, we created the ability to call to an in-patient’s bedside the responders usually delegated to respond to the emergency department for an out-of-hospital trauma. Like the Stroke Team initiation, only the MET team leader can trigger the In-Hospital Trauma team call. Who. We do not require the full Trauma Team to respond, because the MET is already on site and has significant critical care resources. Thus, only a scaled-down Trauma Team (usually the trauma resident and attending) responds, and assists the MET response. How. When a patient has a significant trauma, any staff member may trigger the rapid response system and our MET responds as usual. If the MET senior physician

256

D. Shearn and M. DeVita

feels it appropriate, then a “Trauma Team” call is initiated as well. The MET is responsible for circulation and respiration, while the Trauma Team has responsibility for diagnosis, decision making regarding needed diagnostic or therapeutic interventions, and performing procedures. In our experience, this team is usually called for sudden massive post-operative bleeding events. The Data. Data outcomes are measured by the amount of non-ED Trauma Team calls and the outcomes of individual patients. Patient safety events related to the Trauma Team activation are reviewed.

Blood Administration Team Why. The blood administration team (BAT) was developed to ensure that in the setting of acute large volume blood product resuscitation, blood would be made available and brought to the bedside, it would be given according to policy, and given rapidly. The recognition of a need for this type of service came about in review of cases of multiple unit transfusions wherein delays in obtaining blood were identified as well as failure to follow protocol for blood administration due to over-tasking of MET responders. Who. The BAT consists of a two ICU nurses and a nurse’s aide. The former’s responsibility is to administer the blood products, and the latter’s to obtain the blood from the blood bank expeditiously. How. The BAT is activated by the primary MET, or by any ICU nurse caring for an ICU patient with a sudden large blood replacement need. The team is required in any case where two units of blood are to be administered immediately, with the intention to promote the ability to bring additional resources to the bedside of a patient who has sudden needs. A BAT response is initiated by contacting the operator of the institution who notifies the team through overhead announcements and via the paging system. The Data. Data outcomes for this measure are the number of BAT calls activated and number of patient safety events that have occurred since the initiation of the team. One might also survey the time from bleeding event to infusion of blood products.

Chest Pain Team Why. The Chest Pain Team was developed to treat acute myocardial infarctions in a timely manner. Whether treatment requires catheterization lab intervention or medicinal therapy alone. We noted that our emergency department had built an efficient and effective system for ensuring that all patients with symptoms of acute coronary syndrome receive diagnostics and intervention in under 90 minutes. In contrast, hospitalized patients had delays from symptom onset to treatment. The leadership of our Rapid Response System decided to mimic the ED’s capabilities and created an identical response for the hospitalized patient. This drove down the time from onset of symptoms to intervention.

23  Other Efferent Limb Teams: (BAT, DAT, M, H, and Trauma)

257

Who. At our organization, critical care resources are immediately available via a MET call, so the Chest Pain Team consists only of a cardiologist to determine the correct mode of treatment and with the ability to enable emergent cardiac catheterization. The capabilities of the physician on the chest pain team included the ability to diagnose an acute coronary syndrome, order and perform an intervention, and mandate use of the catheterization laboratory as needed. How. The MET and specifically treatment team leader will determine the need to call for the Chest Pain Team. This team is called for by overhead paging in the institution. The level of cardiologist will be determined by the availability of the physician on call, but will always have the set of skills described above. As with our other supplementary teams, the MET remains with the patient to ensure circulatory and respiratory support. The Data. Suggested data collection to point out outcomes are time of initial condition call to cath lab or tPA, number of Chest Pain Team calls, and mortality data related to Chest Pain Team calls.

Condition L (Lost Patient) Why. This team was created after a patient left her room and wandered away. In spite of a large response taking hours, and which extended to the neighboring community, the patient was not found for some time. The patient was later found unresponsive and attempts at resuscitation were unsuccessful. The hospital performed a root cause analysis to understand how such a large-scale search had been unable to find the patient in a more timely manner. Indeed, the place where she was found had been searched three times, and she had not been there. What was recognized was that the patient could be wandering in a way as to be unintentionally or intentionally avoiding searchers. The new scheme followed national benchmarks for search and rescue teams; we created an organized response to simultaneously search all areas in the hospital and environs. Who. Condition L is an organized notification of all staff. It is an “all-hands-ondeck” approach to notifying staff of a lost patient and signals the initial search for lost patients. In essence, all staff have a search responsibility and the entire facility can be searched within a few minutes of the call. After an initial search, a huddle is completed to determine next steps in the search or clearing of the Condition L. The Condition L can be called for by any staff member in the institution. This RRT will alert each staff member of a lost patient. The call will alert security to become involved and look at exits in the institution, housekeeping to search the staircases in the building, nursing staff to search each room and the staircase above and below their floor, and the administrator on duty will coordinate the response and go to the floor by 30 min from the call. In addition, an overhead call is announced describing the lost patient enabling even visitors to help in the search. How. Condition L may be activated by any employee in the institution. The activation occurs through the hospital operator. The operator initiates the Condition L over the hospital paging system. In addition, the operator calls Condition L on

258

D. Shearn and M. DeVita

the overhead speaker system and provides the name of the patient, age of the patient and a general description of the missing patient. The Data. Suggested data collection for Condition L are the number of Condition L calls per month, time to find the patient, whether the appropriate assessments were completed prior to Condition L, and whether key process points were followed (examples: was the patient assessed as a wanderer; did the post-Condition L huddle occur on the floor the patient was missing from; did the patient have a wanderer identification gown on). Our data show a remarkably short search time is now needed. In our first mock event, it took under 10 minutes to locate a “patient” trying to avoid discovery. Mock events since that time have found the wanderer in about 3 minutes. In real “lost” patient events, the patient is typically found in the same time frame. We advocate this capability and system for all hospitals to prevent unintended harm.

Difficult Airway Team Why. The Difficult Airway Team’s (DAT) primary responsibility is to maintain oxygenation and ventilation and secure an artificial airway when the primary airway manager is unable to do so. While primary critical care resources including a critical care medicine physician are part of the basic condition response, on occasion the airway may be beyond that person’s skills. Thus there is always a need for availability and a planned response by experienced personnel. The hope is that a difficult airway not become a “failed” airway. The needs for other personnel to respond vary. Some reasons are a large patient, abnormal anatomy, or the need for surgical airway intervention. Who. The DAT is activated by the Treatment Team Leader at the advice of the Airway Manager. The DAT will bring an anesthesiologist, an anesthesia technician, and a “difficult airway cart,” which includes an intubating bronchoscope as well as other specialized airway equipment including a tracheostomy surgery tray. How. When an airway is unsuccessfully instrumented more than once, a difficult airway is considered to exist. In some cases, for example, a person with a massively bleeding airway, the DAT may be called prior to any attempt by the MET responders. The MET leader will call for the DAT. This is done by calling the operator and asking for the DAT to be paged overhead. The Data. Suggested data outcomes are the amount of times the DAT is activated, the success of obtaining an airway, and the need for surgical intervention at the bedside. DAT should be reviewed by an independent patient safety review panel.

Pediatric Response Team Why. Pediatric patients require a response that is capable of diagnosing and treating sudden critical events in a child. In an adult acute care facility, the expertise for caring for pediatric patients may be confined to a few staff members. Thus there is a need to define a specific trained pediatric response and develop the capability to

23  Other Efferent Limb Teams: (BAT, DAT, M, H, and Trauma)

259

maintain competency of individuals as well as efficiency of the team. The pediatric response team is activated when a pediatric patient (under 14 or under 40 kg.) meets our usual condition criteria. The goal of the team is to treat the patient acutely, rescue the patient from immediate danger and escort him promptly to trained staff in the ER who have pediatric critical care skills. The intention is to rapidly treat or triage the patients to a pediatric facility. We note that most of our pediatric MET calls are for children who were visiting relatives and had an event, or children who had an event in the vicinity of the hospital and just brought their child into the front door of the hospital. Who. Pediatric Team members consist of a (pediatric if possible) critical care physician, two critical care RNs, a respiratory therapist and security and escort personnel. The physician is trained in the general airway management but the team is aware of the Broselow pediatric resuscitation system. The team receives specific training on how to treat pediatric patients. A pediatric crash cart has been created and is brought to the scene. Nurse responders are limited to concentrate expertise and facilitate training. The staff are aware of pediatric medication dosing requirements and mnemonic devices like the “Broselow tape.” How. Pediatric condition response occurs when staff recognize a patient who meets pediatric criteria is in distress. The staff alerts the operator to activate the Pediatric Response Team. This is one of the few MET teams that does not first activate the basic condition response team. The Data. Suggested data outcomes to be measured are number of pediatric response team calls and resultant outcomes of the calls. A review of all pediatric response team calls should take place.

Condition M Why. Condition M is a crisis intervention that may be called for any patient or visitor who is experiencing a crisis that could pose a potential threat to themselves, patients, staff, or visitors. It is a “behavioral code”. While most of the events occur with an unruly individual who is angry and has lost control, it also may include individuals under the influence of psychotropic medications or patients who are having an acute psychiatric episode. The Condition M will provide resources trained in dealing with the situation, including a psychiatry liaison nurse or physician, a social worker, and security. In most cases, security is never needed, but it is reassuring to have them on site. Who. The Condition M response team consists of a behavioral resource nurse, psychiatric resident, other psychiatric responders, security staff, and an administrator on call along with in-unit resource personnel. The purpose of providing these personnel is to bring to the bedside the experts who have the knowledge to deal with the problem and also the knowledge of how to get the appropriate resources for the patient. How. Condition M response is called for by the staff on the unit who feels that an individual is in crisis. This crisis may be evidenced by a patient’s shouting, hitting objects

260

D. Shearn and M. DeVita

or other people, or throwing objects. Other criteria may be pacing, verbal threats, reddening face, increased heart rate or blood pressure associated with threatening behavior. The Data. Suggested data outcomes to consider are amount of Condition M calls and the outcomes of those calls. Analysis of each Condition M is required to ensure that all appropriate personnel responded and that all points of the policy were followed.

Summary As stated before, the low-frequency, high-risk nature of many events has led to the creation of many MET and RRT’s at UPMC Presbyterian. Below is Table  23.2 that gives the name of the team, reason developed, or purpose and who the responders are. Table 23.2  Teams and their responders RRT or MET name Reason developed or purpose Stroke Team (MET) To efficiently treat patients that have exhibited symptoms of a stroke Trauma Team (MET) The ability to call to an in-patient’s bedside the responders usually delegated to respond to the emergency department for an out of hospital trauma Blood Administration To ensure in the setting of acute Team (BAT) (MET) large volume blood product resuscitation, blood would be made available and brought to the bedside; it would be given according to policy, and given rapidly Chest Pain To treat acute myocardial Team (MET) infarctions in a timely manner Condition L (Lost) To find a patient who is Team (RRT) unaccounted for or lost

Difficult Airway Team (DAT) (MET) Pediatric Response Team (MET) Condition M – Behavioral Crisis Team (MET)

Responders Physician trained to treat strokes plus the basic condition response team Physician and trauma team specialist plus the basic condition response team. Security and escort personnel 2 ICU nurses and a nurse’s aide plus the basic condition response team

Cardiac MD plus the basic condition response team Unit staff, security, escorts, housekeeping staff, patient’s attending physician, Administrator on duty Anesthesia personnel plus To maintain oxygenation and basic condition response ventilation, and secure an airway team when primary team is unable Physician, 2 ICU RNs, The need for a specific trained respiratory therapist, pediatric response escorts, security was needed Behavioral resource nurse, A crisis intervention that may be Psychiatric C & L called for any patient or visitor Resident, other psychiatric who is experiencing a crisis responder, security staff, an that could pose a potential administrator on call, unit threat to themselves, patients, resource personnel staff or visitors

23  Other Efferent Limb Teams: (BAT, DAT, M, H, and Trauma)

261

Our RRS has used the MET to enhance patient safety within our institution. The RRS has also added other innovative teams to manage other crises that require immediate delivery of specialized skill sets or equipment.

Chapter 24

Other Efferent Limb Teams: Crisis Response for Obstetric Patients Gabriella G. Gosman, Hyagriv N. Simhan, Karen Stein, Patricia Dalby, and Marie Baldisseri

Keywords  Crisis • Teams • Obstetric • Patients Inpatient obstetric care involves periodic maternal and or fetal crisis situations. During events such as fetal bradycardia and maternal hemorrhage, patient care needs greatly exceed the resources allocated to routine care. Many of these crises require rapid, coordinated intervention of a multidisciplinary team to optimize outcome. Increasingly, hospitals have incorporated obstetric teams into their rapid response system to address these recurring, but unpredictable maternal and/or fetal events. The American College of Obstetricians and Gynecologists (ACOG) Committee Opinion “Medical Emergency Preparedness” emphasized the importance of crisis response teams for clinical emergencies relevant to obstetric patients.1 Many institutions established obstetric-specific crisis teams as local quality improvement initiatives. Thus, few reports on such teams appear in the published literature. This chapter describes the implementation, training and maintenance of an obstetric-specific crisis team at Magee-Womens Hospital of the University of Pittsburgh Medical Center (UPMC) from 2005 to 2009. This also includes descriptions of alternate obstetric team approaches chosen by other institutions. This information was solicited through query of several medical associations (Council of Women’s and Infants’ Specialty Hospitals, the Society for Obstetric Anesthesia and Perinatology, and the Society for Simulation in Healthcare), the Institute for Healthcare Improvement, and via personal communication by two of the authors (GG and PD).

G.G. Gosman (*) Department of Obstetrics, Gynecology and Reproductive Sciences, University of Pittsburgh School of Medicine, Magee-Womens Hospital, Room 2314, 300 Halket Street, Pittsburgh, PA 15213, USA e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concept and Implementation, DOI 10.1007/978-0-387-92853-1_24, © Springer Science + Business Media, LLC 2011

263

264

G.G. Gosman et al.

Background and Justification Magee-Womens Hospital added an obstetric-specific team response to its rapid response system for several reasons. 1. The single call system was the most rapid way to bring multidisciplinary providers­to the patient who needed care urgently. One call assembled the necessary personnel and expertise to provide optimal evaluation and intervention. After calling, the bedside provider could focus on immediate crisis care of his or her patient, rather than making multiple sequential phone calls. 2. The multidisciplinary response facilitated interdisciplinary communication, as team responders arrive and receive a briefing about the patient at nearly the same time. Poor communication was the number one root cause identified in the JCAHO Sentinel Event Alert “Preventing Infant Death and Injury During Delivery”.2 3. The UPMC health system had improved patient outcomes by introducing a crisis team response for inpatient pre-arrest medical emergencies (Condition C). After the institution of Condition C, inpatient cardiac arrests and deaths decreased.3 4. Incorporation of obstetric crises into the health system’s rapid response system held the promise to enhance data collection and quality efforts in obstetric care.4 5. The team response provided a designated team member to do real-time documentation of patient status and interventions during critical obstetrical events. This promised to improve documentation problems caused by multiple providers retrospectively recording their recall of critical events. 6. Crisis response teams provided a valuable resource for staff satisfaction and peace of mind. For a staff member who recognizes a patient in crisis, help is only a single call away.5

Design and Introduction Magee-Womens Hospital of UPMC is a full-service, academic urban hospital and a tertiary referral center. It is the largest maternity hospital in Western Pennsylvania, performing over 9,300 deliveries in 2008. At the time Magee considered adding an obstetric team, the UPMC health system already had a well-established rapid response system. The crisis team (Condition C) had been activated occasionally for obstetric patients. Event review of these calls suggested that the team composition was not ideal for obstetric patients. Key personnel not included on the Condition C team include obstetricians, obstetric nurses, a newborn resuscitation team, and anesthesia providers. The obstetric-specific team response was called “Condition O.” Table 24.1 compares personnel who respond to Condition O and Condition C. Some institutions have opted to use their medical crisis team to respond to obstetric patient events.

24  Other Efferent Limb Teams: Crisis Response for Obstetric Patients

265

Table 24.1  Medical crisis vs. obstetric crisis team responders at Magee-Womens Hospital Obstetric crisis Medical crisis Responder (condition O) (condition C) In-house obstetrician X Ob/Gyn resident (4th year) X Anesthesiology MD and/or nurse anesthetist X Critical care medicine MD X X Patient’s nurse X X Administrative clinician (nurse) X X L&D charge nurse X L&D clinician or manager (nurse) X Newborn resuscitation team X Respiratory X X Emergency medicine MD X Resident on service X Emergency department nurse X Telemetry unit nurse X Safety/security officer X Ob/Gyn, obstetrics and gynecology; MD, physician; L&D, labor and delivery

It took approximately 1 year to design and fully implement the team. The p­ rocess began in early 2005. Stakeholders in the patient care areas of obstetrics, maternal-fetal medicine, obstetric nursing, critical care, neonatology, and anesthesiology convened to determine the appropriate team composition and size for the characteristics of our institution. Key considerations in team composition included provision of adequate nursing manpower, involvement of anesthesia providers, involvement of newborn resuscitation providers, and capability for real-time documentation. Activation criteria for Condition O include clinical events such as shoulder dystocia, seizure, or hemorrhage. Additionally, staff members are encouraged to call Condition O for acute situations in which the physician or nurse believes immediate evaluation/intervention is needed to avoid fetal/maternal harm. Activation criteria at other institutions with obstetric teams include “staff concern” with or without a list of clinical events such as those listed above. Additional clinical event triggers reported by other institutions include fetal distress, prolonged fetal bradycardia, cord prolapse, absent fetal heart tones, vaginal bleeding, uterine rupture, emergent delivery, maternal respiratory distress, and maternal cardiac arrest. Table  24.1 shows the obstetric response team members at Magee-Womens Hospital. Table 24.2 shows the roles and duties that these responders fulfill during a crisis. Magee opted for a large team with full critical care capabilities. Nursing contributes the highest number of responders. This is because adequate nursing manpower is critical for implementing crisis interventions. Anesthesiology responds because they provide essential evaluation and intervention skills regarding anesthesia, analgesia, airway, and hemodynamic status. A critical care physician responds because about one-quarter of crises at Magee are maternal. Further, these individuals have extensive crisis team leadership experience. The neonatal resuscitation team was originally called only for gestations ³24 weeks. However, a revised

266

G.G. Gosman et al.

Nursing team members select one of these roles as appropriate

Table  24.2  Obstetric crisis medical emergency team personnel and duties at Magee-Womens Hospital Team member Duties Obtain briefing from appropriate person, assess Treatment leader (obstetrician team organization/composition, assess data, or critical care MD direct treatment, set priorities, collaborate with or anesthesiologist anesthesia team on patient plan, triage patient as indicated) Bedside nurse (usually Stay by patient, attach monitoring, deliver briefing to patient’s nurse) responders, report IV size/location, adjust IV rate, draw up and administer medications Runner (L&D clinician or Obtain medications and equipment, deliver to other personnel) appropriate person Call for/dismiss personnel and family, call for/facilitate Nurse responder equipment acquisition, call for/facilitate patient (usually L&D charge transfer, get results nurse) Obtain record sheet, document (team leader, situation, Documenter vital signs and clinical data, treatments), brief (usually administrative personnel who come later clinician) Procedure MD Examine patient, inform team of maternal/fetal (usually ob/gyn resident) assessment, perform procedures Anesthesia team Obtain briefing from obstetric team, assess analgesia, assess airway, perform anesthesia procedures, communicate anesthesia plan to team, collaborate with treatment leader on maternal issues Newborn resuscitation team Obtain briefing from obstetric team, assess newborn, resuscitate newborn MD, physician; IV, intravenous; L&D, labor and delivery; ob/gyn, obstetrics and gynecology

policy included them for every call. This change was ­motivated by the desire to eliminate an extra call, the need for timely arrival of the ­neonatal team, and gestational age uncertainty. Responders who are not needed for the crisis at hand are dismissed (“ramp down” strategy). This team composition suits Magee-Womens Hospital as a high volume, tertiary referral center with all of the described providers in the hospital around the clock. Many other institutions with a high volume of obstetrics have obstetric-specific teams with personnel similar to the Condition O team. The following considerations help institutions that care for smaller volumes of inpatient obstetric patients to design an appropriate obstetric team. Skills needed by a full team include (1) team leadership for data analysis and treatment decisions (this skillset may overlap with that of the team member serving in 2, 3 and 4 below); (2) adult medical crisis expertise; (3) delivery/obstetric management expertise; (4) provision of anesthesia for emergency operative procedures; (5) performance of nursing tasks; (6) coordination to acquire additional resources if needed; (7) acquisition of additional resources if needed; (8) newborn resuscitation; and (9) real-time documentation. Table 24.3 lists potential personnel who could participate. A single individual may play multiple roles, depending on the size of the team. As an alternate strategy, some institutions have opted for a “ramp up” approach with a small team (1–2 individuals). This smaller team responds, rapidly assesses, and has a process to ­summon additional personnel with the skills listed in Table 24.3 if indicated.

24  Other Efferent Limb Teams: Crisis Response for Obstetric Patients

267

Table 24.3  Core skills and potential personnel to include on an obstetric team Skill Possible provider type Team leadership Obstetrician Critical care physician Emergency physician Hospitalist physician Adult medical crisis expertise Anesthesiologist (airway, breathing, circulation) Critical care physician Emergency physician Hospitalist physician Delivery and other obstetric Anesthesiologist management Obstetrician Midwife Emergency physician Family medicine physician Anesthesia for emergency procedures Anesthesiologist (or preparation for this) Certified registered nurse anesthetist Nursing task performance Nurses from multiple units of the hospital (e.g. obstetric care, emergency department, medical-surgical unit, intensive care unit) Nurse administrator Coordination of additional resources (e.g. personnel, equipment, patient transfer, crowd control) Acquisition of additional resources Nurse (e.g. medications, equipment) Medical assistant Newborn resuscitation Providers trained in newborn resuscitation (e.g., obstetrician, neonatal intensive care unit team, pediatric team, nurses) Real-time documentation Nurse Nurse administrator Physician

Staff Education Prior to initiating Condition O, the institution held education sessions for potential callers and potential responders. Sessions occurred during staff meetings for these caregivers. Written material about Condition O appeared in institutional newsletters and as postings on patient care units. Condition O was added to the list of emergency preparedness resources on the back of employee badges. Both callers and responders expressed initial resistance during these sessions and during the first 6 months of Condition O use. During the first few activations of the team, some physician responders criticized the caller (usually an obstetric nurse). Fear of this negative physician reaction resulted in few Condition O calls for about 6 months after implementation. To overcome this, further educational efforts included casebased discussions of teamwork during critical obstetric events, review of videotaped simulated crises with Condition O vs. usual sequential calling pattern, and an institutional multidisciplinary patient safety day. Facilitators of these sessions focused on two major points: (1) Responders must not criticize the caller for summoning the team to help a patient; and (2) Condition O greatly streamlines the crisis response process.

268

G.G. Gosman et al.

Response Team Training Initial Condition O case review revealed opportunities for improvement in team organization, leadership, and crisis communication. In November 2005, a multidisciplinary team from Magee-Womens Hospital initiated simulation-based team training at the Peter M. Winter Institute for Simulation, Education and Research. The Obstetric Crisis Team Training course trains both potential Condition O callers and responders, including obstetricians, anesthesiologists, and obstetric nurses. The course format involves online and in-person didactic presentations combined with filmed simulated emergencies and debriefing. The course curriculum was designed to meet the needs identified during the initial Condition O experience regarding team performance and specific crises encountered. Course participants learn, practice, and debrief with a focus on the following key elements: crisis communication, team organization and leadership, and appropriate emergency care. A streamlined version of the course is also utilized to facilitate learning during obstetric crisis drills done in situ on Magee’s patient care units. Just like in the simulation center, these drills are filmed for video-based debriefing of team performance. Many other institutions with obstetric-specific teams use simulation-based crisis team training at simulation centers and/or in situ. Structured simulation-based team training for obstetric emergencies improves patient outcomes, including hypoxicischemic encephalopathy, 5-min APGAR scores <7, and shoulder dystocia.6,7

Data Collection, Review, and Process Improvement The hospital developed a peer review process of the obstetric crisis team calls, which was similar to the review of cardiac arrests and medical crisis team calls. Obstetric crisis event records and patient medical record are reviewed by a designated physician to detect opportunities to improve patient management, team function, or systems issues that may have resulted in or influenced the crisis. Details of each team response are entered into the hospital’s code response database. Multiple improvements have resulted from issues identified in obstetric crisis team events and crisis team drills. Additional major patient safety interventions that have resulted from this event review include interdisciplinary rounds every 4 hours on the labor and delivery unit and an increased number of attending obstetricians supervising care on the unit 24/7. Institutions initiating an obstetric-specific crisis response might consider collecting data on the following: (1) event rate; (2) event location; (3) reason for the call; (4) services provided by the team; (5) time (and duration) of initial call, team arrival, and end of resuscitation; (6) maternal and fetal clinical characteristics; (7) maternal and fetal/neonatal clinical outcomes; and (8) patient events for which the team should have been called but was not. Staff perception of the patient safety environment in obstetric care services may also give valuable feedback about the obstetric crisis team.

24  Other Efferent Limb Teams: Crisis Response for Obstetric Patients

269

Usage of Condition O at Magee-Womens Hospital and Discussion Providers began using Condition O regularly, beginning about 6 months after its introduction. Figure  24.1 shows crisis team call rates for obstetric patients since 2005. Since 2006, the rate of Conditions (O and C) called on obstetric patients has remained stable with 81–87 calls per 10,000 obstetric discharges. The exception to this is the calendar year 2008, in which crisis team calls peaked at 117 per 10,000 obstetric discharges. Table 24.4 shows the reasons for Condition O calls. Providers called Condition O primarily for threats to fetal wellbeing, 65%. Use of the team for this indication suggests that the single call mechanism allows the caregivers (especially nursing) to focus on patient care interventions instead of summoning each member of the team individually. Emergencies that pose a direct risk to both mother and fetus made up 7% of calls. Purely maternal crises constituted 27% of events. Obstetric staff called Condition C purely for maternal indications. Table 24.5 describes the reasons for these calls. The majority of Condition O calls come for patients on labor and delivery and the obstetric triage unit (69%). These units are adjacent to each other and staffed with multiple obstetricians, anesthesia personnel, and multi-tiered nursing providers. During the initial 6 months of Condition O, many nurses and physicians highlighted the ready availability of personnel on the unit. For this reason, they did not perceive that an obstetric crisis response team was needed. However, routine staffing on the unit is targeted to routine patient needs. Condition O use patterns demonstrate that providers’ attitudes have

Fig. 24.1  Crisis team calls on obstetric patients, rate per 10,000 obstetric discharges

270

G.G. Gosman et al.

Table 24.4  Indications for obstetric crisis team activation, June 2005 – June 2009 Indication Number of events Fetal Non-reassuring fetal heart rate (prolonged deceleration,   73 heart rate lost, etc.) Shoulder dystocia   24 Cord prolapse   19 Imminent or precipitous delivery term   19 Imminent or precipitous delivery pre-term   17 Imminent delivery of fetus with malpresentation   16 (footling breech, face, etc.) Difficult delivery during caesarean section    6 Pre-term labor    3 Rupture of membranes off of the obstetric unit    2 Contraction/abdominal/back pain off the    2 obstetric unit Head entrapment during breech delivery    1 Maternal Postpartum hemorrhage   32 Seizure   19 Syncope or lightheadedness   14 Patient unresponsive    4 Postpartum hypoglycemia    2 Respiratory distress    2 Chest pain/pressure    1 Possible intravenous injection of epidural    1 medications Upper gastrointestinal bleeding    1 Both Abruption   10 Antepartum vaginal bleeding    6 Hemorrhaging placenta praevia    3 Home delivery preterm    1 Total 278

Table 24.5  Indications for medical crisis team activation for obstetric patients, June 2005 – June 2009 Indication Number of events Syncope or lightheadedness 27 Seizure 21 Postpartum hemorrhage  5 Respiratory distress  4 Hemorrhage due to miscarriage  3 Patient unresponsive  3 Change in mental status  2 Trauma or fall  2 Anaphylaxis  1 Hypotension  1 Chest pain  1 Total 70

24  Other Efferent Limb Teams: Crisis Response for Obstetric Patients

271

Table 24.6  Obstetric crisis team activation locations, June 2005 – June 2009 Location Number of events Labor and delivery 146 Triage unit   46 Antepartum unit   27 Emergency department   27 Postpartum unit   23 Ultrasound department    5 Parking lot    1 Cafeteria    1 Staff meeting room    1 Ob clinic    1 Total 278

shifted to recognize the periodic need for a rapid bolus of care resources. Table  24.6 shows the patient location for Condition O and Condition C calls on obstetric patients. Condition O patient events divert patient care resources from lower to higher acuity patients. This has raised the concern that the crisis response abandons lower acuity patients. However, Condition Os have thus far not resulted in staffing shortages for other obstetric patients at Magee. The median duration of Condition O responses is less than 10 min. The majority of team members disband after a brief event review. Remaining members document the event for the medical record (physician team leader – several minutes) and for code response data collection (nurse administrator 5–30 min). Providers initially expressed concern that the Condition O response would frighten patients and families. However, unsolicited patient comments at our institution indicate that patients and family are not frightened by the sudden influx of personnel and flurry of activity during Condition O. This feedback suggests that patients and their families perceive the crisis team response as evidence of highquality emergency care. Providers have used Condition O at a steady rate at our institution. An obstetricspecific crisis response team may be particularly important for institutions with lower volume and/or lower acuity obstetric units. Such units have fewer staff ­present around the clock and fewer crises. For such institutions, a designated team may facilitate efficient training of team members and an improved response to obstetric crises. Part of the justification for the introduction of Condition O was to apply the health system’s successful experience with the rapid response system to a different clinical setting, widely recognized to be high risk and in need of periodic emergent multidisciplinary expert care. Condition O accomplished the goal of providing a reliable and effective resource that improves the patient care process. It is a resource that hospital staff members have used with consistent frequency since 2006. This provides evidence of both staff satisfaction with the process and a safety culture change. Condition O has not had a detectable impact on the perinatal quality and safety data collected and reported by our institution. These parameters include The Joint

272

G.G. Gosman et al.

Commission on Obstetrical Pregnancy and Related Conditions Core Measures, the Agency for Healthcare Research and Quality Perinatal Patient Safety Indicators, the National Perinatal Information Service/Quality Analytic Services Obstetric Quality Indicators, and the Adverse Outcome Index. Specifically, these include event rates of vaginal birth after caesarean section, third or fourth degree laceration, obstetric trauma, postpartum re-admissions, wound complications, anesthesia complications, neonatal mortality, birth trauma injury to neonates, and birth trauma linked to maternal shoulder dystocia. The Adverse Outcome Index includes maternal deaths, intrapartum neonatal death, uterine rupture, unplanned maternal intensive care admission, birth trauma, return to the operating room or labor and delivery, neonatal intensive care unit admission ³37 weeks and ³2,500  g, maternal blood transfusion, and third or fourth-degree laceration. Many of these measures are not directly related to the crisis team activities. Some of the potentially relevant outcomes such as maternal and neonatal death are such rare events that crisis team impact is difficult to assess. The institution also evaluated malpractice claims data for possible impact of Condition O on malpractice activity. However, not enough time has elapsed since 2005 to assess the impact on professional liability activity, given the long statute of limitations on obstetric cases. The following patient outcome measures may help assess the impact of an obstetric crisis team. For events with emergency concerns about fetal wellbeing that result in caesarean section (prolonged bradycardia, cord prolapse, etc.) collect data on time from decision for caesarean section to incision (“decision to incision”), Apgar scores, cord blood gases, and unanticipated neonatal intensive care admission. For patients with obstetric hemorrhage, collect data on the number and type of blood products administered, estimated blood loss, and unanticipated hysterectomy. For shoulder dystocia events, collect data on the time from delivery of the newborn’s head to delivery of the body (“head to body interval”), Apgar scores, cord blood gases, and neonatal fracture and nerve injuries. Patient safety initiatives are commonly criticized because of the inability to detect and demonstrate a causal link between process change and outcome. Obstetric quality and patient safety indicators are in need of evaluation and revision to boost their ability to discriminate between high- and low-quality care.8 This revision process is currently under way. Future measures may improve the ability to evaluate the impact of patient safety initiatives such as Condition O. However, in the modern patient safety environment, clinical areas often have ongoing multiple safety and quality improvement projects at any given time. Thus it can be difficult to show a causal link between a specific initiative and a specific patient outcome measure.

Summary An obstetric-specific crisis team allows institutions to optimize the care response for patients with emergent maternal and/or fetal needs. The team response provides a key resource to reassure staff, physicians, and patients that prompt crisis care is only

24  Other Efferent Limb Teams: Crisis Response for Obstetric Patients

273

a single call away. Data on the obstetric-specific crisis response at ­Magee-Womens Hospital show that team activation is common, improves the care process, and has promise to improve outcomes directly and/or through event analysis and subsequent process improvement.

References 1. ACOG. ACOG Committee opinion: Medical emergency preparedness. Obstet Gynecol. 2006;108(6):1597–99. 2. JCAHO. Preventing infant death and injury during delivery. JCAHO Sentinel Event Alert. 2004;30. 3. DeVita MA, Braithwaite RS, Mahidhara R, Stuart S, Foraida M, Simmons RL. Use of medical emergency team responses to reduce hospital cardiopulmonary arrests. Qual Saf Health Care. 2004;13(4):251–4. 4. Braithwaite RS, DeVita MA, Mahidhara R, Simmons RL, Stuart S, Foraida M. Use of medical emergency team (MET) responses to detect medical errors. Qual Saf Health Care. 2004;13(4):255–9. 5. Galhotra S, Scholle CC, Dew MA, Mininni NC, Clermont G, DeVita MA. Medical emergency teams: A strategy for improving patient care and nursing work environments. J Adv Nurs. 2006;55(2):180–7. 6. Draycott T, Sibanda T, Owen L, et al. Does training in obstetric emergencies improve neonatal outcome? BJOG. 2006;113(2):177–82. 7. Draycott TJ, Crofts JF, Ash JP, et al. Improving neonatal outcome through practical shoulder dystocia training. Obstet Gynecol. 2008;112(1):14–20. 8. Grobman WA. Patient safety in obstetrics and gynecology: The call to arms. Obstet Gynecol. 2006;108(5):1058–9.

Chapter 25

Personnel Resources for Responding Teams Andrew W. Murray, Michael A. DeVita, and John J. Shaefer III

Keywords  Personnel • Resources • Crisis • Response

Introduction Care for patients who are in a medical crisis has historically entailed management by their physicians and nursing staff in a way that was somewhat haphazard and reactive. The system’s success was based largely on whether the established hierarchy was followed in an appropriate and timely manner, allowing the patient to receive the required attention rapidly enough to prevent further deterioration. Staff learned how to respond to crises on the job: they were handed pagers and told to respond, but rarely received specific instruction regarding who else should respond, who had responsibility for what, and what was expected of them during a crisis event. It is no wonder that crisis team responses are often described as chaotic. There is a problem with identifying the need for a crisis team response, which is deeply imbedded in medical care. A strict hierarchy has made the patient’s individual attending physician the “captain of the ship.” This is a great strategy for coordinating care in routine situations, but can become an obstruction to rapidly responding to changes in a patient’s status. In some clinical settings, such as the operating room or intensive care unit (ICU), this hierarchical system further adds to the potential confusion when multiple “captains” are responsible for different aspects of a single patient’s care. The problems with this hierarchical style of management stem partly from the commonly held belief that the hierarchy must be “ascended” instead of being “transcended.” “For example, in a “medical-education care model,” a nurse assistant would first call the nurse, who would then call the intern, who would call the resident, who would then call the attending physician. Each step up the call ladder occurs only when the individual perceives that he or A.W. Murray (*) Department of Anesthesiology, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213, USA e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concept and Implementation, DOI 10.1007/978-0-387-92853-1_25, © Springer Science + Business Media, LLC 2011

275

276

A.W. Murray et al.

she is incapable of managing the situation. A similar problem occurs in a non-educational ­healthcare setting: the nurse assistant notifies the nurse who calls the primary physician, who then calls the consultant. For a patient with sudden onset of respiratory distress, one can imagine that it could take 30 min or more just to contact the person best able to manage the problem. By then, the patient’s condition could have significantly deteriorated, to the point where the patient could be in extremis. These hierarchical considerations are not limited to the administrative chain of command, but are found in many settings and cultures and reflect a social hierarchy that in a crisis setting can interfere with optimal patient care. For example, a nurse responsible for a patient experiencing a crisis may feel reluctant to offer suggestions and even important information to a perceived higher authority, because they feel that it is not their “place” in the social or professional order. The Rapid Response System (either using a Medical Emergency Team (MET) or a Rapid Response Team (RRT) as the responders) is an improved method of providing care for these patients, because it brings the necessary individuals to the bedside within minutes and establishes a “flat” hierarchy that prioritizes the patient’s immediate interests above any social or professional considerations. A number of studies have shown that the implementation of a rapid-response system (RRS) resulted in an inverse relationship wherein an increased number of team activations leads to a decreased incidence of cardiac arrest and unexpected death.1–4 Unfortunately, devising trigger and response strategies is less than half the battle, and change needs to occur in the minds of the caregivers to implement plans to best benefit the patient. While this paradigm shift starts with individuals, to be effective it must reach the institutional level. Implementation is difficult primarily because it requires a true institutional commitment of leadership and resources, as is discussed in other chapters in this book.

Shortcomings of the Current System It has been assumed that once physicians and nurses have completed their training, they are adequately equipped to deal with any crisis that may arise. The method of training has been based on the time-honored method of apprenticeship, which relies on individuals learning by observing more experienced individuals, assuming that the longer you are in the situation, the more expert you become. Some people do this very well and naturally, but others not at all. Inherently, this is a very time-inefficient process. The question is whether staff can be trained to a point where there is a greater pool of people able to deal effectively with crises.5 This is analogous to the concept of herd immunity – in this case, enough people are adequately trained in a system such that for any individual patient crisis, enough responders with adequate training show up so that the crisis team effectively and efficiently provides care. Most crisis responders have basic life support and/or advanced cardiac life-support training, and while this is important, crisis care is rarely delivered well by any one individual, i.e., an optimally functioning, well-coordinated team delivers optimal care. Knowledge of the causes of respiratory distress and what treatments are ­indicated may facilitate care

25  Personnel Resources for Responding Teams

277

management of patients. Advanced education and ­training in a profession, specialty, or subspecialty are also beneficial. However, an additional form of training is necessary: teamwork training. The current paradigm of delivering crisis care to a patient at risk of morbidity or mortality is analogous to asking a group of mechanics who do not know each other to show up when paged to a NASCAR pit stop, spontaneously organize into a team, get the work done as quickly as a trained team, and neither hurt themselves nor the driver through their actions or lack thereof. No matter how capable each of the individuals, he cannot be expected to be as effective as a trained team because he does not have a role-specific action plan that coordinates with others, and they have not practiced as a team. This type of ad hoc response methodology would be risky for both the driver and members of the crisis team. We believe the analogy holds true for medical crisis response. Untrained crisis response individuals may make myriad mistakes that are representative of latent errors in the system and that are only discovered when the ­individual is being stressed. The authors have observed that when first exposed to the crisis, inexperienced anesthesiology residents respond in the following ways: 1 . Compulsion to do something 2. Loss of routine 3. Fixation on a task 4. Loss of effective communication In a crisis, caregivers may feel compelled to do something to help the patient, even if they are not certain as to the best course of action. For example, it is drilled into all advance cardiac life support (ACLS) learners to perform A-B-C (airway-breathing-circulation). This, of course, is important from the patient’s perspective, but it does not consider team aspects of crisis response. As a result, all responders may attempt to perform the same interventions, or by chance neglect some necessary aspects of care. While this “need to act” is noble, it may not be done with forethought as to whether it is the most important task or what the ultimate outcome might be for the patient. The loss of the routine has the effect of putting blinders on the provider, as they no longer look at data streams and thus are unable to respond to them. When caregivers place themselves in a situation where they are physically engaged in an action, they can often become fixated on that task and stop paying attention to the situation as a whole. Communication breaks down to the point where no direction is given to any other responders to the crisis about what has happened to the patient, what has been done to rectify the situation, or what still needs to be addressed. Planning and practice are needed to overcome both intrapersonal and interpersonal barriers to effective crisis management.

How Organization Can Help in Crisis Response Organizing the response to a crisis is an attempt to ensure that a team can respond to an unexpected event and, once there, cooperate to perform a series of mindful

278

A.W. Murray et al.

responses that benefit the patient. Group training creates a coordinated response that allows people to act intelligently within the bounds of the team and adapt to the differences in each situation. This concept of adaptability, or flexibility, is a highly useful team attribute considering the potential myriad and complexity of medical crises. From an educational design perspective, this concept allows one to focus on role-oriented goals and objectives for the individual as they apply to the collective team goals and objectives. To maintain integration among roles and introduce enhanced collective responsibility for the patient, the goal is not to simply perform designated tasks, but also to constantly monitor the situation and note how an individual response has to change to coordinate with that of others and to benefit the patient. From a learner’s perspective, this collective approach allows them to focus on their own specified set of treatment, monitoring, and communication goals. Through a focused understanding of how their actions both influence and are influenced by other team members’ goals, the learner can become more efficient and flexible in both achieving individual goals and contributing to other team members’ performances, all to the benefit of the patient. The institutional or system advantage of training individuals in functioning as a team is that while the individual composition of a crisis team can be expected to change, both the individual and the team-training objectives are uniformly developed. As long as the responders know what their interactions should be within the group and also understand that they are to retain their own monitoring of the situation, then any combination of trained individuals can gather to solve the problem.6 This collective pool of capabilities transcends skills to include social or professional behaviors and expectations centered on the primary goal of the patient’s welfare, as a counterpoint to the typical, hierarchal professional and social barriers that can decrease a team’s flexibility. As a result of this perspective, we now teach O-ABC, where O equals “organize.” The few seconds it takes to organize responders according to roles enables more coordinated responses and better patient care. The process of training responders also provides an opportunity to uncover and correct latent errors. The individuals who arrive at the scene of a crisis are usually either overwhelmed into not doing anything, or they try to correct everything that they see as a problem. Both of these responses are undesirable, the former because the patient receives no care, and the latter because the care is haphazard, at great risk for things getting missed, inadequately managed, or an additional risk (i.e., a therapy that includes a risk of complication). Individuals who feel they have to do everything run the risk of providing inadequate, if not detrimental, care, although this is not their intent. If they are not made aware of the pitfalls of the situation, they cannot avoid them. In the training environment, this individual performance trait can be identified, corrected, and re-evaluated. One author’s observations of anesthesiology trainees in the training environment has been that they tend to become fixated on one task or piece of information while the situation requires that many tasks be completed in a prioritized manner. This can be referred to as a fixation error,5 where the operator becomes so engrossed in the task at hand that all other input is actively or passively ignored or rejected. Organizing crisis responders empowers them to focus more effectively on the tasks

25  Personnel Resources for Responding Teams

279

assigned. With fewer tasks to accomplish (due to better division of labor that occurs with better design and rehearsal), they become positioned to either help with another team member’s tasks or to perform their next required task. This directly improves the efficiency of both the individuals’ and the team’s performances, with the added potential benefit of improving the outcome for the patient. When the position of leader is predetermined or accepted by all members of the team, it allows that person to observe the situation and make clear-minded, informed decisions regarding care. This role should be regarded more as a coordinating position than a hierarchical one. In our hospital-crisis team training, each person already knows what his or her responsibility is and therefore self-assigns. Because team leaders do not have to organize personnel resources or perform a specific therapy, they can efficiently focus on the coordination of the collection of data, data interpretation, treatment prioritization and intervention, reassessment, etc. If the leader is instead focused on organizing the team, less time and thought is spent on the leader’s primary responsibility: orchestrating the treatment.

Rethinking the Thinking The thought processes involved in the management of a crisis need to be quick, reliable, and able to adapt easily to an ever-changing situation. Effective performance requires that the person be able to process simple routine tasks while simultaneously performing and using all sensory-motor input to adjust the way that they are performing. Cognitively, this implies the efficient combination of pattern recognition and reflexive response. Until now, we have expected this development to occur through experience in an apprenticeship training process. However, the author’s experience with anesthesiology residency training is that simple routines have often not been established through experience and they become allconsuming.5 The greatest challenge lies in the position of the leader. This individual needs to be able to step back from the situation and not do anything unless absolutely necessary. This will allow for him to be able to survey the situation and ensure that tasks assigned are being done, that appropriate interventions are taking place, data is being collected and relayed appropriately, treatment decisions are being made, and therapy is delivered as ordered. This can be especially difficult for the recently graduated provider who has never had the experience of being in a coordinating role. He functions on two different levels. The first level is that of: what is happening, what are we doing about it, are our actions effective, are we causing more problems, and are we on the right track. This focus is patient-centered and consists of doing appropriate assessments, collecting relevant data, analyzing findings to make a list of potential diagnoses, and determining the correct additional diagnostics and treatments. We term this role the “treatment leader.” The second level is that of team performance oversight. They need to develop supervisory control of the amount of attention that the team members devote to any given task. The better

280

A.W. Murray et al.

defined the roles, as well as the tasks that are attached to each role, the more likely that team members will complete their tasks. This role definition will allow the leader to decide what inputs are worth paying close attention to and create a priority list for the next point of care that needs to be addressed (rather than focusing on who needs to be doing what).5 Resource management is also included in this level of cognition, in which the leader uses all resources – human and otherwise – most effectively. This may occasionally entail changing the roles of the team members to ensure that the efficiency is maintained. An example would be calling additional help or requesting the departure of a member.

Structure One of the keys to caring for these patients better is being able to treat the patient early in the progression of the crisis. This entails equipping the nursing staff and junior physicians with the tools and training that they need to feel comfortable asking for help at the best possible time, i.e., earlier rather than later. However, once the communication has been improved, it has to be met with a structure that creates a meaningful response. Structure is necessary from various aspects. The obvious initial goal is deciding on the structure of the human resources to be applied to the medical emergency situation. A second aspect is the structure of team activation, from the initial call to their arrival. This needs to be reliable, constant, and unambiguous. The next aspect is the structure that needs to exist so that the patient is cared for in the most efficient manner, ensuring the best outcome from the emergent situation. This is probably the structure point that is best managed by training, so that the responders can function as a team. The final structure point should be that of the physical resources, such as intravenous fluid and supplies, medications, defibrillators, and monitoring and laboratory testing.

Human Resources Whereas it would be very comforting to have a dedicated team that has no responsibility other than responding to crises, this is not feasible in a system that is already laden with costly care. Cross-training is a must, and to this end the Medical Emergency Teams that have been created have consisted of people with the necessary expertise being called together at a moment’s notice from different ends of the hospital to focus their collective care and attention on a single individual. The team needs to be large enough to be able to provide all the technical and knowledge resources that are required, but not so large that it becomes cumbersome. The two required constants are a team leader and a record-keeper.5

25  Personnel Resources for Responding Teams

281

The treatment leader must remove him/herself from the fray to understand the larger picture of the unfolding crisis. This person’s performance can be crucial in the crisis’s being handled appropriately and efficiently. The person needs to be able to develop situational awareness not only to monitor what is happening or evaluate what has occurred, but, most importantly, also to extrapolate the results of their current direction. The leader must be able to avoid the trap of getting fixated on the minutiae of the problem, but at the same time be able to ensure that the little things do get done as requested. The remaining team members must assume identified roles and complete the essential tasks for each. The roles need to be designed to ensure application of monitors, airway management, intravenous access, and patient examination. The details of these roles and responsibilities are described below in terms of skills required to assume the task, and a separate chapter in this book focuses on team training in detail (see Chap. 26). A study done in New Zealand demonstrated that the providers were unclear of the roles and responsibilities of various team members, especially crossing professional boundaries indicating the continuing need for this to be addressed in the training phase.7 The team should consist of people who are familiar with the crisis situations and have the required knowledge. These would invariably, but not exclusively, be those who are frequently involved in the care of acutely ill patients. The authors’ institution, the University of Pittsburgh Medical Center (UPMC), employs the use of intensivist physicians (fellow and attending), an intensive care nurse, and the medicine residents on call for that day. Specialists in airway management and surgical management are also available, although they do not respond to every MET activation. An additional member is an administrator to facilitate the rapid transit of the patient to an ICU location to continue care of the patient.

Activation of the RRS A system is of no use if it is not activated, so the next step is to active the RRS when a crisis develops. Herein lies a great challenge, in that the standard approach is one of individual management of the patient’s care by the most junior member of the staff. This mindset needs to change, so that the MET is activated instead of “business as usual.” A study performed in the UK examined the results of the establishment of a dedicated code team. This study showed a more rapid response to the crisis, and the survival-to-discharge was 10%, which falls in the range of other studies that have investigated cardiac arrests.8 The amount of time that the MET takes to respond also affects the outcome, as illustrated in a study in Italy that showed that in cases where the crisis response took longer than six minutes, none of the patients survived.9 Where the absolute response time has not been established in such a way that a benchmark exists, certainly the concept of “more time is worse” must be accepted.

282

A.W. Murray et al.

The Ad Hoc Team The individuals will respond to a call and instantly become part of an ad hoc team (Table 25.1). The members who respond to a call bring with them experience and skill sets that can be very necessary and specialized. The varying skill sets and the nature of an ad hoc team can be both a strength and a potentially significant weakness. The strength is that enough people respond to be able to divide the workload and function as a collective mind (think in terms of the adages, “many hands make light work” and “two heads are better than one”). The responders need to enter the situation in such a way that they understand that their primary goal is to ensure that the patient’s care is the best possible, regardless of who is seen to be right or wrong in the management of the situation (one would want to avoid the adages of “too many cooks spoil the broth” or “too many chiefs and not enough Indians”). The individuals need to perceive themselves as useful to the structure that has been created. They need to feel that their opinions will be regarded thoughtfully and applied if appropriate. The ad hoc structure also presents the greatest challenge. Ideally, the people who arrive upon activation of the MET previously would have practiced working together. More often than not, however, they may have worked together only occasionally, and in some instances may have not even met. There needs to be a way of training and structuring so that the responders, who may have vast amounts of technical skills but very limited opportunities of working with the particular individuals in that situation, will be able to form a cohesive team in an extremely short time. Each institution must create its own structure in response to the perceived needs that they experience on a day-to-day basis.

Table 25.1  Roles and goals for crisis team in the operating room Responder roles Skills/expertise Team leader Experience in managing acutely ill patients, preferably current attending anesthesiologist/surgeon Advanced Cardiac Life Support training Technician Blood sampling (arterial blood gas, coagulants, thromboelastography) Troubleshooting equipment Supply equipment needs Scribe Keep accurate record of interventions and results Airway management If needed in airway emergency Medication delivery Knowledge of allergies Knowledge of ongoing medicinal therapy Knowledge of pharmacology Advanced Cardiac Life Support training Circulatory support Provide cardiopulmonary resuscitation support if needed Surgeon Surgical intervention to correct or prevent problems

25  Personnel Resources for Responding Teams

283

Changing the Existing Culture Much of the difficulty in altering the way things are done exists in changing the minds of the people involved. The longer things are done a certain way, the stronger the culture becomes. As stated previously, the historical culture is that a knowledgeable, skilled physician should be able to deal with any crisis, and this is to be achieved with the knowledge that has been obtained during the short years of training in medical school. All too often these years have been jammed full of other competing, yet important, concepts. The management of crises has only been addressed by providing the didactic knowledge. A study published in 2008 showed that ICU trainees felt that the training and time spent on medical emergency teams was beneficial for their training and the care of patients on the wards, but sounded the alarm about the potential for decreased care for patients in the ICU. An additional concern was that there was insufficient supervision during MET activations.10 Accordingly, there needs to be a modification at the training level if one expects the system change to be durable. Medical trainees should be introduced to the concept of a team approach and team training at a very junior level. This will allow for the development of respect for one’s limitations, as well as for the contributions of other responders. The training should stress that the concept of being “all-capable” is false: the well-trained team should perform better than the sum of its individuals. In a training setting, the student can be nurtured in an environment that is non-threatening to both the learner and the “patient,” rather than having to learn from poor performance in the real world. This also opens the possibility of constructive, retrospective review of the management of crises in a way that nurtures all team members. Unfortunately, human nature is often concerned with being the one with the correct answer in a given situation. While this is a noble goal, it can often be destructive in a team setting. MET- system training puts much focus on the non-medical, nontechnical aspects of team management of a medical crisis.11 The emphasis is on the concept of being a member of a team, a component of the system that is taking care of the patient. Communication is a key concept that needs to be taught. Communication needs to be clear and organized. Speaking should be limited to what is absolutely necessary in order to minimize confusion and missed orders. Such communication should function as a closed loop, so that the individual who asks the question or gives the order is absolutely sure that he was heard, understood, and that the request is being carried out or has already been completed. Resource management is also a concept that needs practice and training. The person functioning as the leader needs to be trained so that all their resources, human and otherwise, are managed effectively and brought to bear against the problem. Some of the pitfalls that face the leader are the risk of getting too involved in the micromanagement of patient care, forgoing the oversight function that they are responsible for, and becoming fixated on a certain aspect of the patient’s problem while the patient’s overall situation deteriorates. The focus must be on data acquisition, analysis, and treatment delivery.

284

A.W. Murray et al.

An important component of resource management is recognizing the crisis and reliably alerting the team. The UPMC had a system in place for responding to crises for a decade, but it saw limited use while physicians were still being paged urgently and sequentially to the bedside to care for their patients. The institution then started reviewing the incidences of sequential stat pages and reminding the nursing units to use the team response, and the individuals responsible for the delay in team activation were given feedback on their actions. However, the two most effective policies that were implemented were: (1) the establishment of calling criteria, and (2) the dissemination of these criteria to the nursing units and other caregiver groups. The result was a significant positive change in MET use, with a corresponding drop in the number of sequential stat pages.12 Because hospital resources differ, one would expect that their MET compositions would be different (Table  25.2), yet  all will have to fill the needs of the patients in crisis. In our courses for training teams, we have identified a number of different roles that must be filled: airway manager, airway assistant, circulatory support person, someone to deliver medications, someone to deploy medications and equipment from the crash cart, a person who makes therapeutic decisions after analyzing the data, and finally, someone to collect data and record the events. The skills and knowledge needed for each position are different. For example, the airway manager will be responsible for assessing the airway, positioning the head, choosing whether ventilation needs to be assisted, and if so, determine the ­methodology for that. The airway manager must know how to mask ventilate the patient, suction the Table 25.2  Roles and goals of MET responders, Pittsburgh methodology Responder roles Skills/expertise Treatment leader Management of acutely ill patients in ER, OR, or ICU Airway manager Mask ventilation Endotracheal intubation Neurological assessment Administer medications (sedative, neuromuscular blocking agents) Airway assistance Mask ventilation Bedside assessment Attain reliable intravenous access Delivery of medications Knowledge of allergies Obtain vital signs Equipment manager (crash cart) Deploy medications and equipment Run defibrillator Circulatory support Rapid assessment of the patient’s physical examination Mechanical circulatory support Place defibrillator pads Data manager Keep accurate recording of the events of the crisis Obtain key data from chart, caregivers Deliver data to treatment leader Check pulse, adequacy of chest compressions Procedure physician Perform procedures

25  Personnel Resources for Responding Teams

285

airway, and intubate the trachea if necessary. In addition, because of the location ­relative to the patient, the airway manager should be able to assess the patient’s level of consciousness, check pupils, and feel carotid pulses. Finally, the airway manager must have the capability to deliver medications for sedation, anesthesia, or neuromuscular blockade to facilitate intubation when necessary. Because of the complexity of the skills needed to maintain oxygenation and ventilation, the airway manager’s role can be filled by only a few types of professionals. Usually anesthesia, critical care, and emergency medicine physicians, as well as nurse anesthetists have this capability, and so they are usually responsible for this role. Respiratory care personnel have the training and experience to non-invasively manage the airway, and, at some hospitals, they also have the skills to perform endotracheal intubation using standardized physician order sets for sedation. Every team must have a person from one of these professions to manage the airway, oxygenation, and ventilation during a crisis response. The airway manager cannot function alone, because it is difficult to both manage an airway and simultaneously set up the necessary equipment for this. An airway assistant therefore is assigned to deploy and assemble a bag-mask unit for use, connect the oxygen source to the airway devices, set up suction, assist intubation, prepare the endotracheal tube and laryngoscope prior to insertion, test carbon dioxide in exhaled air, and connect the patient to a ventilator if necessary. In addition, because so many METs calls are due to respiratory crises, the team must have the knowledge and skills to deliver respiratory treatments and adjust oxygen levels. Usually a respiratory therapist, respiratory technician, or nurse can tend to these duties. Most hospitals have a crash cart that contains the medications and equipment needed for a crisis response. The medications must be located swiftly and mixed expertly, and then delivered via the appropriate route. The crash cart manager will be better able to manage the defibrillators if all of the institution’s defibrillators are the same. The crash cart manager should be trained and should practice frequently. Several types of professionals have the skills to run the crash cart, including nurses, pharmacists, respiratory therapists, and perhaps physicians. Regardless of the profession, the crash cart manager must be able to operate the defibrillator efficiently and effectively, and be knowledgeable about and trained on the crash cart equipment and medications. Creating appropriate admixtures are usually in the professional skills and experience of ICU nursing staff and pharmacy staff, so they may be preferred to take on this responsibility. Most teams have a person responsible for recording what happens during the crisis, usually focusing on the vital signs and treatments delivered. We had trained (and depended on) the crash cart manager to also keep the records. However, when the same person both manages the crash cart and the record keeping, there tends to be errors. We have therefore delegated two people to manage these responsibilities separately. We have also attempted to make the recorder a more active position, since one of the more common mistakes in a crisis response team is a failure to collect data and deliver it to the person who needs that information to make decisions. In our training, we have changed the “recorder” into a “data manager” who is active, collects a database that includes recent laboratory results, electrocardiogram and chest x-ray results, and physical findings by the team.

286

A.W. Murray et al.

A person is needed at the patient’s bedside to deliver the medications and to obtain vital signs. Most responders should be able to take vital signs and, if this person delivers medications under the supervision of another nurse (for example, when an ICU nurse is the equipment/medication manager), then medication delivery is within the scope of this person’s capabilities. We prefer that the patient’s bedside nurse assume this role because she or he usually is aware of the prior state of the patient and can tell the team if findings change. The bedside nurse also is usually aware of allergies and recently administered medication, and this knowledge, if relayed to the treatment leader, can prevent error or lead to a diagnosis of the etiology of the crisis. At the UPMC, there is sufficient personnel to have one person assume each of the eight roles. Other institutions may not have as many responders, and so one person may need to assume more than one role. In designing a crisis team response, the directors should attend to these roles, and make sure that responders with adequate skills and knowledge are assigned to them. Redundancy needs to be built in when establishing the teams; people should be adequately cross-trained to be able to fill in for other members until they arrive. A survey performed in Germany and published in 2009, although the response rate was low, emphasized that the training methods for METs are inconsistent and would likely benefit from standardization with possibly enhanced impact on patient outcomes.13 In addition, this study showed the make up of the MET teams that, in the author’s opinion, seemed very lightly staffed, possibly leading to the responders being overwhelmed by the circumstances.

Operating Room Crisis Teams Crisis teams may be needed for purposes other than a crisis on a hospital ward or lobby area. Much has been written about anesthesia (or operating room [OR] crisis teams).5 Personnel considerations for METs need to reflect the organizational and hierarchal realities of these environments in order to create effective training programs. OR crisis teams respond to anesthetic emergencies, such as a difficult airway with failure to ventilate, malignant hyperthermia, OR fires, massive hemorrhage, and obstetrical as well as surgical emergencies. In contrast to the floor crisis response teams described previously, OR personnel that could be used in an MET are often already present on-site or immediately available (with other duties) when the crisis emerges (Table 25.1). A surgical team is constructed, organized, and focused on a collective task, such as a surgical procedure. A typical one includes an anesthesiologist and/or anesthetist, a surgeon, a scrub nurse, and a scout nurse. The hierarchal structure is quite variable, with ranges of perceptions including a surgeon, anesthesiologist, or a nurse anesthetist who may all believe themselves to be “captain of the ship,” or some combination thereof. This often changes with the introduction of a crisis, where the assumed team leader during a regular surgical procedure – usually the surgeon – switches to the case anesthesiologist for ­management of the crisis.

25  Personnel Resources for Responding Teams

287

Additional resources are dictated by the type or severity of a crisis and may include equipment such as a defibrillator, a bronchoscope or perhaps a “difficult-airway cart,” central line placement kits, blood products, blood administration devices including rapid infusers, and emergency medications that require preparation, like dantrolene. The number and composition of additional personnel required, as well as specialized equipment, depends on the type and severity of the specific crisis. For purposes of discussion, an OR crisis can be viewed as any adverse event that requires additional personnel or equipment beyond that normally anticipated for the specific surgical procedure. OR crises can either present as a sudden, obvious event that requires significant immediate resources, or can arise as a slowly escalating continuum that requires a matching escalation of resources to manage appropriately. Similarly, an OR crisis usually occurs in the setting of an ongoing surgical procedure that requires its own resources at the same time. Crisis management and personnel decisions need to take this into account, as they draw resources from the same pool – whether from personnel involved in the case during which a crisis occurs, or from resources earmarked for other cases, either ongoing at the time or about to be started (Fig. 25.1). Administrative control and communication channels of these resources are sitespecific and often varied. Compared to a generic MET response system and training program, design and training for OR crisis management must include significant

Do I have all the right equipment? Do I need to call anybody else

What do I do if the airway manager does not arrive in time???

Team Leader

Airway Management

Airway management

Airway assistance

Fig. 25.1  Resource management is the key to successful treatment of a patient with an airway crisis

288

A.W. Murray et al.

leadership training for the anesthesia team to be able to rapidly recruit and organize resources. Institutionally, for the operating room environment, the burden of this can either be left to the individuals involved in a crisis, such as the anesthesiologist (currently the most predominant method), or addressed at the system level such as with the full MET concept level. For example, a hospital’s OR system could systematically address (1) the composition of an MET for the OR; (2) the method to activate this team and address effective cross-coverage for the responding members’ responsibilities, so that they can focus on the crisis; and (3) an administrative process for capturing the information about the root causes of the OR crisis, to benefit from the lessons learned. If systematically addressed, then the emphasis for training is less centered around how to de novo recruit and deploy an MET, but rather is focused on getting the optimal performance out of an MET designed for the OR. At our facility, we train around the existing operational, administrative, and hierarchical operating room environment, where there is no highly organized MET response for a crisis. Therefore, the crisis training program focuses on training the anesthesia care team, including (1) generic patient safety principles and concepts; (2) how to rapidly tap into existing resources to organize a scalable crisis team in the setting of an ongoing surgical procedure; and (3) expanding capabilities of effectively treating uncommon and common types of crisis through simulated experiential and reflective learning methods. This is an interdisciplinary approach including all members of the anesthesia care team. Over the last 10 years we have trained approximately 1,000 anesthesiology residents, student nurse anesthetists, and anesthesia technicians. The objective validity and reliability of the training methods and outcomes still have to be proven, yet the subjective survey and incidental case reports suggest a valuable outcome.14

Summary Crisis teams are an effective response system to emergency situations. For crisis teams to be effective and efficient, they must be well designed. The best design requires bringing together the personnel who will participate in the response and determine what is needed for that institution’s crisis team. We have delineated the roles needed for a ward crisis and an OR crisis team (Table 25.2). How each institution fills these roles depends on many local factors. The planning team needs to ensure that enough people with knowledge and skills to accomplish the task respond, and that appropriate equipment and resources are made available. Furthermore, there must be a strategy for bringing those resources to bear; simply making sure they are available is not sufficient. In addition, a plan must exist for both assembling the equipment and personnel and deploying them in a crisis. Finally, by their nature crises require more than one person to respond effectively; cooperation and coordination are imperative. Like any group effort, knowledge of what to do is insufficient, so group practice is needed to maximize potential. We believe that full-scale simulation is the best tool to enable the responders to perform

25  Personnel Resources for Responding Teams

289

the skills needed to save lives. The difference between the knowledge of what to do and the skill of actually doing it in a group setting is huge. Developing this group skill can save lives.

References 1. Chen J, Bellomo R, Flabouris A, et al. The relationship between early emergency team calls and serious adverse events. Crit Care Med. 2009;37(1):148–153. 2. Buist MD, Moore GE, Bernard SA, et al. Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. BMJ. 2002;324(7334):387–390. 3. Jones D, Bellomo R, Bates S, et al. Long-term effect of a medical emergency team on cardiac arrests in a teaching hospital. Crit Care. 2005;9(6):R808–R815. 4. Dacey MJ, Mirza ER, Wilcox V. The effect of a rapid response team on major clinical outcome measures in a community hospital. Crit Care Med. 2007;35(9):2076–2082. 5. Gaba DM, Fish KJ, Howard SK. Crisis Management in Anesthesiology. New York: Churchill Livingston; 1994. 6. Weick KE, Roberts K. Collective mind in organizations: heedful interrelating on flight decks. Adm Sci Q. 1993;38:325–350. 7. Weller JM, Janssen AL, Merry AF, Robinson B. Interdisciplinary team interactions: a qualitative study of perceptions of team functions in simulated anaesthesia crises. Med Educ. 2008;42(4):382–388. 8. Lee KH. Survival after cardiopulmonary resuscitation in the general wards – the result of a dedicated “Code” team. Ann Acad Med. 1998;2:323–325. 9. Sandroni C, Ferro G, Santangelo S, et al. In-hospital cardiac arrest: survival depends mainly on the effectiveness of the emergency response. Resuscitation. 2004;62(3):291–297. 10. Jacques T, Harrison GA, McLaws ML. Attitudes towards and evaluation of medical emergency teams: a survey of trainees in intensive care medicine. Anaesth Intensive Care. 2008;36(1):90–95. 11. Lightall GK, Barr J, Howard SK, et al. Use of a fully simulated intensive care unit environment for critical event management training for internal medicine residents. Crit Care Med. 2003;31:2437–2443. 12. Foraida MI, DeVita MA, Braithwaite RS, Stuart SA, Brooks MM, Simmons RL. Improving the utilization of medical crisis teams (Condition C) at an urban tertiary care hospital. J Crit Care. 2003;18:87–94. 13. Siebig S, Kues S, Klebl F. Cardiac arrest: composition or resuscitation teams and training techniques: results of a hospital survey in German-speaking countries. Dtsch Arztebl Int. 2009;106(5):65–70. 14. Byrne AJ, Greaves JD. Assessment instruments used during anesthetic simulation: review of published studies. Br J Anaesth. 2001;86:445–450.

Chapter 26

Equipment, Medications, and Supplies for an RRS Edgar Delgado, Wendeline J. Grbach, Joanne Kowiatek, and Michael DeVita

Keywords  Equipment • Medications • Supplies • Medical • Emergency • Team • Response

Introduction Once a patient has a crisis event, it is essential to deliver safe, accurate, and timely emergency care. Patient survival depends on the efficiency of the response.1 This response must include not only the appropriate personnel, but also the supplies, equipment, and medications that are needed. We have noted that most people who have been involved in a cardiac arrest (or Medical Emergency Team) situation characterize the crisis response using words like “disaster,” or “chaos;” at best, it may be termed “ineffective.” This does not have to be the case, as an organized response can be planned and achieved. This chapter will describe a methodology for providing an organized equipment and medication supply response.

Institutional Oversight of Equipment To obtain a quality emergency response team, education of the personnel deemed “responders” is critical. All staff members in contact with direct patient care are required to have basic life-support education, and critical care staff is also educated in advanced life support. However, depending upon the acuity of the patient’s condition and observations, one may not have experience in a crisis until it actually happens, and some medical staff may not feel adequately prepared to participate in a crisis event. To help staff understand the importance of correct procedures, everyone E. Delgado (*) Respiratory Care, UPMC Presbyterian Shadyside, Pittsburgh, PA, USA e-mail: [email protected]

M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_26, © Springer Science+Business Media, LLC 2011

291

292

E. Delgado et al.

should be educated during hospital and/or unit-based orientations regarding their specific role in a crisis. Annual competency evaluations are another method of training for crisis event management; to reinforce and test the knowledge base, we use “mock code” scenarios. The mock code can also determine system and personnel deficiencies, and promote processes to decrease or eliminate them. By using programmable, computer-based, full-scale human simulation, events can be repeated and data obtained regarding specific personnel tasks. For example, one may determine the range of how long it takes for a crash cart to get to the scene of a crisis, or how long it takes personnel to defibrillate a patient in ventricular fibrillation. This data can determine equipment needs and personnel education. By retesting, frontline staff can gain the comfort level and knowledge necessary to improve patient outcomes in any crisis situation, and hospital leaders gain confidence in the adequacy of their crisis response program. The Medical Emergency Response Improvement Team (MERIT) committee at the University of Pittsburgh Medical Center Presbyterian Hospital is the administrative structure charged with ensuring that personnel and equipment are prepared for crisis events. The committee noted that all nursing units and clinic settings had various types of equipment stock for crisis response, including intubation equipment, individually designed and stocked crash carts, and a variety of personnel designated by individual units to respond to crisis calls. This resulted in problems related to equipment mismatch and malfunction. First, we noted variation of the crash carts throughout the hospital, with each designed according to specific unit needs. Nursing staffs were responsible for restocking the carts after the event. At times, because of normal patient care activities, this task was delayed, perhaps posing a safety risk in that if a subsequent emergency occurred, the cart was not prepared. The organization’s pharmacy and central supply were enlisted to improve medication delivery and restocking. Fully-stocked crash carts were kept ready in the central supply area and exchanged for used carts, dramatically shortening the time it took a unit to have a restocked cart. When all the carts are prepared in a central area, it is possible to employ a systematic method for stocking: we use a sectional medical tray system that can quickly be exchanged, thus limiting the time it takes to restock the cart. The medication and equipment contents of the cart were determined by critical care nurses and physicians who respond to the events. Pharmacy’s role was to ensure sufficient supply of medications for each event (e.g., how many vials of each medication) and create a convenient location for each item. Central supply had a similar role for all other equipment. Finally, we went through a similar process for our pediatric crash cart. The contents of the crash carts and the layout for each drawer are in Tables 26.1–26.7. It was felt that access to intubating equipment on a crash cart enabled those who were not skilled to have the opportunity to intubate patients. To improve patient safety, the intubation equipment was removed from all non-intensive care units and placed in a portable orange airway bag. This bag was stocked in central supply to

26  Equipment, Medications, and Supplies for an RRS Table 26.1  Emergency airway bag contents Intubation kit Surgi-lube 6£ Tongue blade Gum bougie #3 Mac blade #4 Mac blade Laryngoscope handle with batteries #3 Miller blade Adult Magill forceps Green oral airway 80 mm Yellow oral airway 90 mm Red oral airway 100 mm Intubation stylet Yankauer suction #9 Endotracheal tube #8 Endotracheal tube #7 Endotracheal tube #7 Endotracheal tube

293

Other accessories CO2 detector Mask with filter set-up Mask (child) Mask (small adult) Mask (large) Exam gloves Suction catheter kit Salem sump 16 fr. 1£ tape Nasal airway 28 Fr. Nasal airway 30 Fr. Nasal airway 32 Fr. Syringe 30 cc Luer lock Syringe 10 cc Luer lock Biohazard specimen bag Splash mask with shield Endotracheal tube holder 20 gauge 1.5£ needles Jet ventilation kit

ensure reliable, functioning equipment. After every intubation, the old bag is exchanged for a new one. The bag was kept only by the on-call anesthesia and critical care personnel. This forced novice physicians to intubate patients under the supervision of physicians with airway management experience and training. (This is described in more detail below.)

Personnel Response Efficiency in personnel response during a crisis event is also desirable. There is evidence (Gaba, DeVita, Raemer) that teamwork during crisis events may be deficient, and consequently errors may be made. Personnel response and teamwork therefore should be improved. Two models for teamwork improvement have been advanced: the first depends on the training for the crisis team leaders, and the second emphasizes a flat hierarchy wherein each team member has a specified role and responsibilities. The personnel designated to respond, the rationale for choosing those responders, and team training are discussed in Chaps. 25 and 32 (Murray et al. and Fiedor et al.).

8075

8801

100078

8046

E051100

8254 8030 8040 8041

Resuscitator bag “AMBU” Airway 90 mm yellow

1

1

Each

Each

Swabs, Betadine 5 Each/box Tape, adhesive 1-in. 1 Roll/box Tape, adhesive 3-in. 1 Roll/box Tape, Transpore 1 Roll/box 2-in. Tape, Micropore 1 Roll 1-in. 1 Roll/box Tape, Micropore 2-in. Needle box w/lid 1 Each Please secure lid onto top of needle box NO CHG

No CHG NO CHG NO CHG NO CHG

8740

122851

122877

8806 6000

8251

103557

109730 8911 8913 6222

Table 26.2  Standard crash “code” cart supply contents DATE: U.P.M.C. HEALTH SYSTEM P.U.H.COST CENTER: CART NUMBER STRANDARD CRASH “CODE” CART ISSSUE# SUPPLY CONTENTS Expiration Item Description Qty. Issue unit Filled date Item Top of cart 8740 Electrodes 2 Each/box No CHG 8962 8838 Gloves, non-sterile 1 Box E056600 exam 9289 Wipes, alcohol 10 Each/box No CHG 100235

Tourniquet I.V D5 w/H2O 100 ml Needle, spinal 18ga × 3-1/2 Needle, filter 19ga × 1–1/2 Electrodes

Benzoin solution

Chloraprep swabs

Description Drawer 3 Interlink Luer Lock Vac holder (no charge) Vac needle (no charge) Blood gas kit Lever lock Threaded lock NACL 0.9% 30 ml

NURSING UNIT:

Each

3

2

4

1

2 2

1

3

Each/box

Each

Each

Each Each

Each

Each

Each each Each Each

Each Each

5 1

5 5 5 3

Issue unit

Qty.

No CGE

NO CGE

No CGE

No CGE

No CGE

No CGE

Filled

Expiration date

294 E. Delgado et al.

Airway 100 mm red

1

Each

1 Ultrasonic gel 1 Vacutainer. Breen too Vacu, purple top no charge Vacutainer, red top Vacutainer, red top serum Vacutainer, grey top Syringe, 10 cc slip-tip Syringe, 30 cc luer-lock Syringe, 10 cc luer-lock Cath. IV 14ga protective Cath. IV 16ga protective Cath. IV 18ga protective Gauze, pad 4 × 4 (2/pack) Blunt cannula Needle safety 21 GA × 1 1/2 Needle safety 25 GA × 5/8 3 cc syringe (lock)

19923 109772

122854

The outside of the cart

116539

3 CC blunt cannula 5 10 CC blunt 5 cannula 10 CC Syringe 8 flushers

B37375 B37385

each

each each

each

No CGE

No CGE No CGE

No CGE

122841

122875

122823 122865

9038

8507

Drawer 3

5 CC blunt cannula 5

8506

Any items in these two drawers

B37370

8505

Drawers 1 and 2 are for pharmacy use. please do not place

122847

9780

109760

THE bag must then be hung in plain sight on

Resuscitator bag must be assembled and placed in the draw string bag provided 70075 the red and yellow airways must be placed into the bag also 109763 109751

8076

Each Each

2 2

4

4

4 4

6

Each

Each

Each Each

Pack/box

Each

Each

5

2

Each

Each

1

5

Each

Each Each

2 2 2

Each

Each

Each

2

2

1

No CGE

No CGE

No CGE No CGE

No CGE

No CGE

No CGE

No CGE No CGE

No CGE

No CGE

26  Equipment, Medications, and Supplies for an RRS 295

296

E. Delgado et al.

Table 26.3  UPMC syringe drawer contents list, expiration dates, and billing Quantity Expiration # to Medication How supplied stocked date bill Albuterol inhalation 5 mg/ml 20 ml vial 1 Atropine syringe 1 mg/10 ml 3 Topex spray 54 gms 1 (Benzocaine 20%) Dextrose 50% syringe 50 ml 2 Dopamine premix 800 mg/250 ml 1 Epinephrine syringe 1 mg/10 ml (1:10,000) 8 Lidocaine jelly 2% 5 ml 1 Lidocaine premix 2 g/250 ml 1 Lidocaine syringe 100 mg/5 ml 4 Esmolol 2.5 g/250 ml or 10 mg/ 1 premixed bag ml 250 ml Magnesium sulfate 80 mg/ml 50 ml or 1 premixed bag 4 g/50 ml 2.25% 0.5 ml 1 Racemic epinephrine inhalation Sodium bicarbonate 50 mEq/50 ml 6 syringe 500 ml 2 Hextend (place in Drawer #5 on cart)

Mnemonic ALBT20L ATRP1S

D5050S DPMN800I EPNP10S LDCN8I LDCN25S

MGS80 RCPN225 SDBC50S HXTD500I

Table 26.4  UPMC vial tray contents, crash cart medication list, expiration dates, and billing Quantity Expiration Medication How supplied stocked date # to bill Mnemonic Adenosine 3 mg/1 ml 2 ml 5 ADNS6I Amiodarone 150 mg/vial 3 ml 3 AMDR3 Aspirin, chewable 81 mg 2 Calcium chloride 1 g/10 ml 2 CACH10I Diphenhydramine 50 mg/1 ml 1 ml 2 DPHNI Dobutamine 12.5 mg/ml 20 ml 2 DBT250 Epinephrine 1 mg/1 ml 30 ml 2 EPNPI (1:1,000) Syringe labels 10 N/A N/A N/A Flumazenil 0.1 mg/ml 10 ml 1 FLMZ1I Furosemide 10 mg/1 ml 10 ml 2 FRSM10S Heparin 1,000 units/ml 2 HPRN10I 10 ml Lidocaine 2% 20 ml 1 LD20I Methylprednisolone 125 mg 2 MTHL125I Metoprolol 1 mg/ml 5 ml 4 MTPR1I Midazolam 1 mg/ml 2 ml 5 MDZL2I Naloxone with 10 cc 0.4 mg/ml 1 ml 4 NLXN4I sodium chloride (continued)

26  Equipment, Medications, and Supplies for an RRS

297

Table 26.4  (continued) Medication

How supplied

Nitroglycerin

0.4 mg tablets SL #25 Norepinephrine 1 mg/ml 4 ml Phenobarbital 130 mg/ml 1 ml Phenylephrine 10 mg/ml 1 ml Phenytoin 50 mg/1 ml 5 ml Procainamide 100 mg/ml 10 ml Vasopressin 2000A0units/ml 1 ml Vecuronium 10 mg Verapamil 2.5 mg/ml 2 ml Bacteriostatic water 30 ml 0.9% sodium chloride 10 ml

Quantity stocked

Expiration date

1

# to bill

Mnemonic NTRG4

4 4 4 4 3 3

NRPNI PHNB130I PHNY10I PHNY250 PRCN10I

2 4 1 5

VCRN10I VRPM2I WTR30I SDCL10

Table 26.5  Cart replacement process Medication cart exchange process Emergency cart is opened by nurse caring for patient • At completion of crisis care, nurse calls pharmacy department to exchange the cart • Pharmacy technician brings new cart and a black plastic lock marked “do not use” (this denotes the cart as “used” and prevents unintentional redeployment) to central supply • Central supply department technician brings new cart and black seal to unit. New cart is left on unit and black seal is placed on used cart • Used cart is taken to pharmacy and two medication drawers/trays are removed • Used cart is taken to central supply for cleaning and to replenish supplies and equipment. A form is placed on top of the cart stating the first expiration date of the supplies • Newly-stocked cart is taken to pharmacy • Pharmacy replaces the two medication drawers/trays • An inventory-tracking sticker is applied. The sticker notes the date the cart was filled and checked, the supervising pharmacist’s signature, and date and name of the first medication to expire • The cart is secured with red plastic lock (denotes new cart that is ready to go)

Table 26.6  Syringe drawer layout Drawer 1 layout: 1 Dopamine 800 mg/250 ml premixed bag will be placed on the top of the Bicarb Atropine 1 mg/10 ml syringe 3 Epinephrine 1:10,000 1 mg/10 ml Syringe 8

Dextrose 50% 50 ml syringe 2 Lidocaine 2 g/250 ml premix bag 1 Lidocaine jelly 2% 5 ml 1

Lidocaine 100 mg/5 ml syringe 4

Albuterol inhalation 5 mg/ml 20 ml vial 1

Topex SPRAY 57 gm 1

Sodium bicarbonate 50 meq/50 ml syringe 6 Magnesium sulfate premix 80 mg/ml 50 ml 1 Esmolol premix 1 2.5 gm/250 ml Racemic epinephrine inhalation 2.25% 0.5 ml 1

1 Furosemide 10 mg/ml 10 ml

1 Furosemide 10 mg/ml 10 ml

1 Adenosine 3 mg/ml 2 ml 2

1 Flumazenil 0.1 mg/ml 10 ml

1 Adenosine 3 mg/ml 2 ml 3

1 Amiodarone 150 mg 3 ml 2

Vasopressin 20 units/ml 1 ml 3 Naloxone 0.4 mg/1 ml with 10 cc NSS

Table 26.7  Vial drawer layout Drawer 2 vial layout Procainamide Procainamide 100 mg/ml 100 mg/ml 10 ml 10 ml 1 1 Naloxone Naloxone 0.4 mg/1 ml 0.4 mg/1 ml with 10 cc with 10 cc NSS NSS

1 Amiodarone 150 mg 3 ml 1

Vecuronium 1 mg/ml 10 ml 1 Nitroglycerin 0.4 mg SL #25 Tab bottle 1 Heparin 1,000 units/ml 10 ml 1 Aspirin 81 mg Chewable 2

1 Heparin 1,000 units/ml 10 ml

Vecuronium 1 mg/ml 10 ml 1 Norepinephrine 1 mg/ml 4 ml

1 Lidocaine 20 mg/ml 2% 20 ml 1 Calcium Chloride 1 g/10 ml 1

Verapamil 2.5 mg/ml 2 ml 2 Norepinephrine 1 mg/ml 4 ml

1 Calcium Chloride 1 g/10 ml 1

1 Methylpred 125 mg

Verapamil 2.5 mg/ml 2 ml 2 Norepinephrine 1 mg/ml 4 ml

298 E. Delgado et al.

0.9% Sodium Chloride 10 ml 1 Phenobarbital 130 mg/ml 1 ml 4 Metoprolol 1 mg/ml 5 ml 2 Dobutamine 12.5 mg/ml 20 ml

1

1 Diphenhydramine 50 mg/ml 1 ml

2

Bacterio-static Water 30 ml 1 Norepinephrine 1 mg/ml 4 ml 1 Methylpred 125 mg

1

0.9% Sodium Chloride 10 ml 1 Phenylephrine 10 mg/1 ml 1 ml 4 Metoprolol 1 mg/ml 5 ml 2 Dobutamine 12.5 mg/ml 20 ml

0.9% Sodium Chloride 10 ml 1 Phenytoin 50 mg/ml 5 ml 2 Midazolam 1 mg/ml 2 ml 3 Epinephrine 1:1,000 (1 mg/ml) 30 ml Injection 1

0.9% Sodium Chloride 10 ml 1 Phenytoin 50 mg/ml 5 ml 2 Midazolam 1 mg/ml 2 ml 2 Epinephrine 1:1,000 (1 mg/ml) 30 ml Injection 1 10

0.9% Sodium Chloride 10 ml 1 Procainamide 100 mg/ml 10 ml 1 Naloxone 0.4 mg/1 ml with 10 cc NSS 1 Syringe Labels

26  Equipment, Medications, and Supplies for an RRS 299

300

E. Delgado et al.

Nursing Responder Equipment At the UPMC, equipment is provided from several sources. Intensive care nurse responders are designated to respond to geographic zones, resulting in a very rapid response time. Because the equipment cart may arrive after the ICU nurse, if the crisis is located outside a patient care area, the nurses have created an equipment bag that contains items necessary for immediate intervention, such as a blood pressure cuff, stethoscope, gloves, and equipment for intravenous catheter insertion and intravenous fluid resuscitation. However, the packs do not contain medications. The nurses are responsible for restocking the bag afterward.

Airway Equipment Success during a crisis response requires adequate ventilation and oxygenation as a first priority, and establishing a secure airway as part of that goal. Therefore, personnel, equipment, and medications must all be brought to the scene within about a minute of the onset of the event. Over the years, when attempting to establish an airway, clinicians have been frustrated by unfamiliarity with the equipment, lack of available equipment, or lack of process standardization. This presents a significant challenge in an institution where medical students, residents, or fellows are required to learn and perform airway management. To facilitate and improve this skill set, a system was set up which uses a portable emergency airway bag (identified house-wide as the “orange bag” mentioned previously), which contains the necessary equipment to allow a clinician to always quickly ensure adequate oxygenation and ventilation. The bag’s contents are standardized, so that any member of the response team will know where to find any needed item. The airway bag is divided into two compartments: a “quick intubation kit” and “other accessories or adjuncts.” The division improves organization and facilitates rapid intubation, since the equipment for most intubations is placed in one location. The bag contents are in Table 26.1. They include an intubation kit with laryngoscopes and blades, and a variety of endotracheal tube sizes; a mask with a bacterial filter for mouth-to-mask ventilation; gloves; nasal and oral airways; a CO2 detector; syringes; tape; an endotracheal tube fixation device, Magill forceps; suction equipment; and a hand-jet insufflator in case of an emergency cricothyrotomy. A secured side compartment contains medications to facilitate intubation, including etomidate, rocuronium, succhinylcholine, propofol, benzocaine topical metered dose spray, oxymetazoline nasal spray, and lidocaine jelly are included. The critical care medicine fellow must ensure the integrity of the bag at each exchange between fellows. Once the crisis event is over, the orange bag is returned to the central supply department and exchanged for a new, fully-stocked one. The used bag is then sent to the pharmacy for medication restocking and medication integrity assurance by means of a numbered, red plastic

26  Equipment, Medications, and Supplies for an RRS

301

seal, and then returned to central supply for equipment restocking. Then the bag’s separate, non-medication contents are sealed with a white plastic seal to assure the integrity of the contents.

Emergency Cart Standardization To organize crash carts, it is essential to thoroughly review all emergency carts in hospital departments and patient units, including the operating rooms and post-anesthesia care units. Emergency carts may differ in both organization and contents (disposable supplies, durable equipment, medications, and documentation). Following a review of the differences among the hospital’s emergency carts – and of individual units’ special needs – a central committee (often the hospital’s medical emergency response committee) can negotiate and develop a standardized emergency cart. Standardization includes medications, supplies, equipment, and layout for all general patient units, intensive care units, hospital departments, emergency department, post-anesthesia care units, operating room, and hospital-based outpatient clinics. On the outside of the cart, crisis algorithms and dosing charts can be securely attached for use by the emergency response team. This strategy facilitates the actions of the emergency team and reduces the potential for error. For example, if one unit stocks dopamine at 800 mg per 250 mL, and another stocks it at 400 mg per 250 mL, it is probable that a dosing mistake will occur due to look-alike intravenous bags being mixed up in the pharmacy. Another potential mistake is a misreading error. If magnesium is stocked on one cart, but morphine is placed in the same spot on another cart, it is more likely there will be a morphine-for-magnesium mix-up. In addition, the committee must decide whether certain medications that are needed in only a few sites are worth putting on the standardized cart. The cost goes up as additional medications are included, and the probability of a medication being used goes down as medications are excluded. Stocking crash cart contents are thus a collaborative and ongoing process. We continuously review and revise, and implement changes once a year. This facilitates and organizes the process for change and education.

Selecting an Emergency Cart The emergency cart we chose as a model was an anesthesia cart, with a workstation on the top (Fig. 26.1). We chose it because it was able to accommodate an oxygen tank, needle box, and backboard, such as our committee required. The emergency cart had five drawers, two for medications and three for supplies and equipment. The crash cart needs to be durable, mobile, and secure; it should have sufficient capacity for the equipment and medications, and accommodate a workspace. A number of suppliers make carts that meet these specifications.

302

E. Delgado et al.

Fig. 26.1  Adult crash cart

The two medication drawers hold trays (medication cassettes) that can be pre-stocked and sealed in clear plastic by the pharmacy for easy replacement. The medications contained within the cassettes are arranged alphabetically by generic drug name so that they can be found and easily viewed through the plastic (Table 26.2). Because the cart has a locking mechanism that can be secured with numbered, plastic, break-away seals, reviewers can determine whether the cart is “ready for use.” This facilitates central supply restocking and storage, as well as the patient unit auditors (daily crash cart checks).

Need for Specialty Carts Pediatric crash carts present difficult logistic problems because a wide array of equipment and medications is required to meet the needs of the large range of ages and sizes of the patients. While our institution is not a pediatric facility, we have

26  Equipment, Medications, and Supplies for an RRS

303

Fig. 26.2  Pediatric crash cart

nevertheless prepared for the pediatric crisis (Fig. 26.2). Although we rarely have a pediatric in-patient, children visitors are common, and they may have a medical crisis while visiting. The most common events are seizures, syncope, and asthma exacerbation. Obviously, for any child less than 40 kg, the medications and equipment used to care for adults are inappropriate. Per The Joint Commission’s Medication Management standards, it is important that emergency medications be available in unit-dose, age-specific, and ready-to-administer forms.2 There are two techniques for pediatric crash carts that we will briefly discuss. The first is a cart that is organized according to equipment type: for example, all airway equipment is stored together in a single drawer, and all medications in another. The second is the so-called Broselow cart. In this cart, each drawer is color-coded with Broselow tape, which delineates medication dosages based on the child’s weight and size. Each drawer contains the equipment and medications for a certain size child. At our institution, we use the former style of cart, but it is a different color than the standard adult cart to avoid confusion. These carts, like the adult carts, are standardized.

304

E. Delgado et al.

House-Wide Crash Cart This specialty crash cart is used by Respiratory Therapy, which is specially equipped to handle codes in off-site, non-patient care areas (i.e., lobbies, cafeteria, etc.) that do not stock a crash cart. It holds the two standard medication trays.

Medication Selection Every hospital must decide which emergency medications and supplies will be readily available in patient care areas. We referenced the medications and supply requirements from the Advanced Cardiac Life Support (ACLS) algorithms and our own clinical experience. Because 90% of the events we respond to are crises and not cardiopulmonary arrests, our cart is modified to accommodate our needs. Although our goal was to provide the necessary medications and equipment, another goal was to limit medication choices to one drug from each class where possible. The three reasons for this are (1) fewer medications reduces opportunity for error, (2) practice standardization also reduces error, and (3) fewer medications means lower costs. When medications are commercially available in a premixed dosage form (i.e., magnesium sulfate, esmolol, dopamine), we chose that formulation to reduce the chance for errors in admixing. A final goal was to create an emergency cart that requires minimum maintenance (except for the required daily checks by nursing staff to ensure that the cart is intact, up-to-date, and ready to go). The cart is organized to improve efficiency and limit errors. Medications on the emergency cart are arranged in alphabetical order by the generic (not trade) drug name. The individual drug vials are placed in the vial trays and are clearly labeled with the generic name, the drug concentration, and the stock quantity. The top medication drawer holds boxes of needle-less emergency medication syringes, premixed medication intravenous solutions, and odd-sized medications (i.e., topical gels or sprays that do not fit into the second drawer cassette). Table 26.3 displays the plan for the first drawer. The second drawer holds all of the medication vials and ampules in vial display racks as shown in the plan in Table 26.4. When medications are stocked in the cart, they must have a minimum of at least 6 months until their expiration dates. This reduces medication waste due to outdated items and reduces the frequency that carts need to be exchanged due to expiration (for example, at UPMC we have over 100 crash carts). We also created a pediatric crash cart following the principles of the adult crash cart with the exception that airway management equipment is included in the cart. We made this decision for two reasons – first, because pediatric events occur so infrequently, we wanted to streamline our process by only requiring one equipment source, and second, we have never had the experience of a novice attempting an intubation on a child. The contents and layout of the pediatric crash cart are in Tables 26.8–26.9, and Fig. 26.3.

Table 26.8  UPMC pediatric crash cart syringe medication list Qty. Expiration Medication How supplied stocked date Adenosine 3 mg/ml 2 ml syringe 2 Albumin 5% 250 ml bottle 2 Broselow pediatric Emergency tape 1 N/A Dobutamine 250 mg/250 ml , 1 mg/ml 1 premix bag Dopamine 400 mg/250 cc premixed 1 bag Epinephrine0.1 mg/ 1:10,000 10 ml syringe 6 ml Famotidine premix 20 mg/50 ml Lidocaine 2 gm/500 ml bag, 4 mg/ 1 ml Lidocaine 1% 10 mg/ml 5 ml syringe 1 Lidocaine 2% 20 mg/ml 5 ml syringe 1 2 Magnesium sulfate 0.325 meq 40 mg/ml 100 ml premixed bag Sodium bicarbonate 1 meq/ml 10 ml syringe 2 8.4% Table 26.9  UPMC pediatric crash cart vial medication list Qty. Medication How supplied stocked Albuterol inhalation 5 mg/ml 20 ml vial 1 Amiodarone 50 mg/ml 3 ml amp 2 Atropine 0.4 mg/ml 20 ml vial 1 Atropine 0.4 mg/ml 1 ml amp 2 Calcium chloride 100 mg/ml 10 ml vial 3 Calcium gluconate 100 mg/ml 10 ml vial 1 Dextrose 50% 0.5gm/ml 50 ml vial 2 Dipenhydramine 50 mg/ml 1 ml amp 1 Epinephrine 1 mg/ml 1:1000 30 ml vial 2 Furosemide 10 mg/ml 10 ml vial 1 Heparin 1000 units/ml 10 ml 2 Lidocaine 1% 10 mg/ml 5 ml vial 3 Mannitol 25% 0.25 g/ml 50 ml vial 2 Methylprednisolone 500 mg vial 1 Midazolam 1 mg/ml 2 ml 5 Naloxone & 10 cc NSS 0.4 mg/ml 1 ml amp 4 Norepinephrine 1 mg/ml 4 ml amp 2 Phenobarbital 130 mg/ml 1 ml vial 2 Phenytoin 50 mg/ml 2 ml vial 3 Physostigmine 1 mg/ml 2 ml amp 3 Racepinephrine inhalation 2.25% 0.5 ml 3 Sodium bicarbonate 1 meq/ml 50 ml vial 5 Sodium chloride 0.9% 10 ml vial 4 Sterile water for injection 10 ml vial 3 Vecuronium 1 mg/ml 10 ml vial 2

# to bill

305 MNEMONIC

ALBMN5250

DPMN400 EPNP1S

LDCN1S LDCN2SYR

Expiration # to MNEMONIC date bill ALBT20L AMDR3 ATRP20I ATPN4I CACH10I CAGL10I D5050V DPHNI EPNPI FRSM10S HPRN10I LDCN5I MNNT50I MTHY500I MDZL2I NLXN4I NRPNI PHNB130I PHNY2I PHYT1I RCPN225 NABC50I SDSL10 WINJ20I VCRN10I

Place in-patient billing bin for processing. Revised 1/12/06 JMP

306

E. Delgado et al.

Lidocaine 2% 5% Albumin syringe 250 ml 20mg/ml 5ml 2 1

Adenosine 3mg/ml 2ml Syringe 2 Epinephrine 1:10,000 0.1 mg/ml syringe 6

Lidocaine 1% Syringe 10mg/ml 5ml 1

Magnesium Sulfate 0.325 meq /40mg /ml 100ml bag 2

Dopamine 400mg/ 250cc Premix bag 1 Famotidine 20mg/50 ml Premix Bag 1

Lidocaine 2gm 500ml Bag 1 Sodium bicarbonate 1 meq/ml 10ml syringe 2

Dobutamine 250mg/250ml 1

Broselow Pediatric Emergency Tape 1

Fig. 26.3  Pediatric crash cart layout

The medications in the emergency cart are reviewed annually by the MERIT committee members and are based on the American Heart Association’s Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care.3 Changes to the medications and other emergency cart stock are permitted only once a year, due to the workload involved in updating each of the emergency carts in the hospital. The MERIT members, in particular a pharmacist and a critical care medicine physician, review requests to add or remove medications on the emergency cart. Medication requests that have been particularly troublesome include controlled substances (such as morphine), lorazepam (because of its activity against seizures), and insulin. Controlled substances are difficult to add to the emergency cart because the federal requirements for double-locking, daily audits, and concern for diversion create too much work. Medications that require refrigeration, such as fosphenytoin, lorazepam and insulin, and have a reduced stability when not refrigerated (i.e., 30 days) should not be stocked in the crash cart, so consider placing them in the automated medstation on “override” for quick access. Again, the additional workload of maintaining and tracking short expiration medications on emergency carts is prohibitive. In addition, some medications (such as insulin) have a high propensity for error, and yet may not be required in virtually any crisis situation. While insulin may often be helpful (as in hyperkalemia), there may be other medicines that can be used with fewer risks and logistic difficulties. Some medications in the cart require additional warning labels or information to ensure safety in preparation and administration. Examples are warning labels on phenytoin to note that the drug must be mixed in 0.9% sodium chloride solution, and specific dilution and administration instructions for the use of naloxone injection to reverse opioid overdose or appropriate use of flumazenil to reverse benzodiazepine overdose.

26  Equipment, Medications, and Supplies for an RRS

307

Pharmacy Emergency Cart Exchange Process Regulatory agencies like The Joint Commission2 have outlined standards for the managing emergency medications. These include restocking, maintaining appropriate inventory, ensuring that emergency medications and their associated supplies are readily accessible, and verifying that the carts themselves are secure in their locations within the hospital.2 After opening and using medications and equipment from the emergency cart, the supplies must be replaced as soon as possible in order to be prepared for the next event. We have an exchange process to meet performance standards developed with all of the involved disciplines and departments – nursing, respiratory therapy, central supply, pharmacy, risk management/patient safety, and critical care medicine physicians. At UPMC we utilize the transport tracking system run by the central supply department to assist in timely tracking and exchanging of crash carts after they have been used. The system tracks the number of requests for carts, and time from dispatch of pharmacy technician to completion of the job.

Restocking Medications in the Emergency Cart As previously discussed, medications placed on the cart must have at minimum a 6-month expiration. The outside of the cart contains the name and expiration dating of the first medication to expire in the cart. During the monthly pharmacy inspections of the patient units and hospital departments, the pharmacy checks for outdated carts and to ensure that nursing staff is performing the daily required emergency cart checks and documentation. Inside the emergency cart there are billing forms with the medication trays so that patients can be charged for medications used, and pre-printed labels that denote specific emergency response team member assignments during the code. Table 26.5 shows our cart replacement process. We offer it as an example; many different processes are possible, and the one chosen at a specific hospital depends on that institution’s resources. The pharmacy keeps a sufficient supply of backup emergency carts and back-up medication trays on hand for immediate exchange with units that have used their carts. In addition, the pharmacy maintains complete and sealed medication back-up trays, so that the exchange process within the stocking area can be performed quickly on all shifts.

Additional Methods for Supplying Emergency Medications In addition to the crash cart process, other methods have been developed to provide emergency medication before or during an MET response. We have created transport emergency boxes and orange airway management bags (Fig. 26.4). Both contain a small assortment of medications, and the orange bag also contains intubation

308

E. Delgado et al.

Fig. 26.4  Emergency bag layout

equipment and supplies. Transport boxes are for units that care for patients on monitors and must be transported for a test or similar transport. The box contains the three most often used medications for emergencies: atropine injection, epinephrine injection, and lidocaine injection. The emergency airway bag contains a sedating agent for intubation (etomidate), local anesthetic (benzocaine aerosol and lidocaine jelly), and a neuromuscular blocking agent (succinylcholine). Other orange bag contents are described earlier in this chapter.

Obstacles to Implementation The potential barriers to implementation of the emergency medications, equipment, and supply exchange systems include cost, ability to standardize contents (resolving the variation), dynamic administrative backing and leadership, education and training needs, knowledge deficits, time involved, and the staff needed to maintain the processes. To break through these barriers, the focus must be on a common goal – patient safety. Our approach to standardizing the emergency carts was to define a core group that shared the need to simplify the cart-restocking procedure, and improve the reliability of the equipment. With administrative help, other groups that

26  Equipment, Medications, and Supplies for an RRS

309

were also relatively easy to convince were included. Consensus regarding content was achieved, and the hospital administration provided the funds to purchase the carts and their content. After staff education, the carts were deployed only on those units. After observing the improved reliability and decreased effort by unit nursing staff, other units requested inclusion in the system. The few “stragglers” were then easily brought on board as their own system now fell outside the hospital standard. There is a cost associated with crash cart standardization. In many hospitals, cart purchases are part of a nursing unit’s budget, and the individual unit must prioritize purchasing a new crash cart above other expense items. Most units see use of the emergency carts as exceptions to the rule, and many may therefore consider it a low priority. The administration was essential in creating funds for the purchases and creating the imperative for units to make the appropriate purchase. While there were costs, there were also savings. First, standardizing the carts reduced the number of medications on carts, and provided a mechanism for systematically reviewing and choosing medications and supplies with global costs in mind. Second, it reduced nursing work for restocking and checking the medications. Third, less waste of outdated medications occurred. Fourth, because all the carts are the same, staff became very familiar with their contents so there were fewer episodes of two crash carts being opened to deal with a single crisis event. Crisis events are low-frequency, high-stress occurrences. Therefore training is important, and performance is unreliable unless practiced frequently. Training nurses, respiratory therapists, and physicians to perform well during crisis situations is an essential component to crisis response preparedness.4,5 Standardizing the medication and equipment response dramatically facilitates training. However, if the hospital’s educators are not involved in the process, there are two potential problems. First, they may not train people according to the hospital’s process design, and second, the design team may make decisions that seem sensible from one perspective, but may create huge training issues. Overcoming this obstacle is relatively simple if the education staff is included.

Supply Standardization in the Emergency Carts The normal supply stock for the cart needs to be uniform for the same reasons as outlined above. To achieve this goal, central supply staff was enlisted to develop consistency in all crash carts. The methodology was similar: review of the current carts’ contents, determine relevance and necessity, and consider which items could reasonably be provided from floor stock. The emergency cart drawers 3, 4, and 5 are standardized, which helps prevent restocking errors, limits the time crisis response staff needs to find items, and decreases the probability of error (either through misuse, or a mistaken impression that the equipment is not present). Tables 26.6 and 26.7 show the equipment drawer configuration and contents.

310

E. Delgado et al.

Emergency cart standardization in all areas of the institution is an important patient safety methodology. In our opinion, the most important piece of equipment to standardize was at first overlooked by our MERIT members. All crash carts in the institution are mandated to hold a defibrillator (with monitoring, pacing, defibrillation and synchronous cardioversion capabilities); however, they were not initially standardized. If cables, pads, paddles, and defibrillators are not standardized, mismatches result and the equipment cannot be used. Further, if the responding staff is not familiar with a particular model, an inability to use the equipment could result. Both problems might appear to be “equipment failure” although the equipment might be in perfect working order. At one point, a survey of all defibrillators in this institution yielded up to eight different varieties. Figure 26.5 shows several models, including the required assortment of pads, paddles, and cables that they use, and it is possible to imagine the confusion that might result from an assortment of equipment in an emergency. While stopgap measures like writing “No Pacing” or colored stickers to prompt recognition of compatible equipment might be enticing, they are unlikely to prevent error. Instead, standardized defibrillators provide an important patient safety measure. Standardization facilitates education. Unfamiliarity may contribute to hesitation on the part of less-experienced staff to perform defibrillation without an expert clinician. To address this, we chose to purchase “hands-free” and “auto-analyze” defibrillators. Because pads and not paddles are used, the staff is more willing to

Fig. 26.5  Various models of defibrillators

26  Equipment, Medications, and Supplies for an RRS

311

perform a “quick look” maneuver with the defibrillator. In addition, the hands-free pads are safer to use for staff and patients because there is less chance of electrical “arcing” or short circuit (particularly if the patient or the person delivering the shock is wet). The auto-analyze function tells the staff whether a shock is indicated; when a shock is recommended, the staff needs only to assess consciousness, and, if absent, defibrillate. In high-traffic areas such as lobbies, corridors, and cafeterias, neither personnel nor equipment are readily available. We have placed emergency airway equipment (like a bag-mask device) and automatic external defibrillators to promote early intervention. Automatic external defibrillators are easy to use by lay personnel as visitors in an institution or the site, and who may possibly find themselves to be the first responder in a crisis.

Summary To mount an effective emergency response, medication and equipment resources must be available, reliable, and organized in a way to make them easily usable. Staff must be trained adequately so that they know what their resources are and how to manage them. Standardizing the equipment and medications contributes to a safe system by improving a number of logistic issues, including staff training, performance, error reduction, equipment maintenance and replacement after a crisis, and finally the institution’s ability to revise medication and equipment resources for crises. We believe that improving efficiency and reliability can reduce delays and errors, and contribute to the primary goal of improving patient outcomes following a crisis event. Acknowledgments  We acknowledge the contributions of Jan Phillipps and Richard Snyder in developing and implementing the processes described in this chapter.

References 1. Abella BS, Alvarado JP, Myklebust H, et al. Quality of cardiopulmonary resuscitation during in-hospital cardiac arrest. JAMA. 2005;293:305–310. 2. Joint Commission on Accreditation of Healthcare Organizations (TJC). Comprehensive Accreditation Manual for Hospitals: The Official Handbook (CAMH). Oakbrook Terrace, IL: Joint Commission Resources; 2009. 3. Cummings RO, Hazinski F. Guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Currents. AHA 2000; Fall. 4. Lighthall GK, Barr J, Howard SK, et al. Use of a fully simulated intensive care unit environment for critical event management training for internal medicine residents. Crit Care Med. 2003;31:2437-2443. 5. Tham KY, Evans RJ, Rubython EJ, et al. Management of ventricular fibrillation by doctors in cardiac arrest teams. BMJ. 1994;309:1408–1409.

Chapter 27

The Administrative Limb Daryl Jones and Rinaldo Bellomo

Keywords  Administrative • Limb • RRS In this chapter, we discuss the role of the Administrative Limb of the RRS. Specifically, we outline why an administrative limb is needed, what its goals should be, its various components, and finally, the importance of senior hospital medical and nursing administration in ensuring successful RRS implementation and maintenance. The need to link audit activities of the RRT to research projects, quality improvement and clinical governance is also emphasized. The structure, members and roles of the administrative arm at our hospital are used as an example.

Why Is an Administrative Arm Needed? The RRS is an example of a complex intervention as it “developed from a number of components that may act both independently and inter-dependently”.1 All rapid response systems are composed of an afferent limb (including activations criteria and the trigger mechanism) as well as an efferent limb team of responders.2 To be successful in identifying and treating a deteriorating ward patient, there must be a timely progression of a complex chain of events. This includes measuring the abnormal vital signs, making the decision to activate the system, actually activating the system, timely review by the RRT, accurate patient assessment and initiation of appropriate therapy, and finally, appropriate patient disposition and follow-up after the RRT review has occurred (Fig. 27.1). The RRS is a resource-intensive system involving staff that may regularly rotate or turn over. It has been suggested that the institutionalization of system change

D. Jones (*) Department of Intensive Care, Austin Hospital, Studley Road, Heidelberg, VIC 3084, Australia e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concept and Implementation, DOI 10.1007/978-0-387-92853-1_27, © Springer Science + Business Media, LLC 2011

313

314

D. Jones and R. Bellomo Observations not taken

At risk patient

Spontaneous improvement

Not seen by doctor MET not activated

Unwell patient Observations taken

Seen by doctor

Deterioration

Improvement

Treated & remain on ward

MET activated

Conception of MET model

Education of nurses & doctors

Cardiac arrest Patient made NFR

Adequate staff & equipment

Implementation

Unplanned ICU admission

Fig. 27.1  Chain of events for identification, review and treatment of acutely deteriorating ward patient. MET medical emergency team, ICU intensive care unit, NFR not for resuscitation

may fail because of the turnover of key employees.3 It is essential that all RRSs have an administrative arm to ensure that the RRS is successfully implemented, efficiently run, and effective in achieving its stated goals. As stated by Berwick, “the effectiveness of these systems is sensitive to an array of influences: leadership, changing environments, details of implementation, organizational history, and much more”.4 The RRS cannot exist in isolation; it must be part of an overarching strategic plan to make the hospital safer.

What Should the Aims and Objectives of the Administrative Arm Be? The initial aim of the administrative arm is to ensure that the RRS is effectively implemented.2 This may initially involve a planning phase and assessment of the incidence of site-specific serious adverse events in the population the RRS will be targeted to treat. Such baseline information was obtained at our hospital prior to the implementation of the Medical Emergency Team (MET).5,6 All existing and new staff must receive education and induction into the RRS philosophy and the cultural shift that is required for reviewing acutely deteriorating ward patients.7 Calling criteria need to be developed and these should be displayed in convenient places such as on walls of hospital wards and the badges of hospital staff to assist in decision support.

27  The Administrative Limb

315

Table  27.1  Mission statement for administration of Rapid Response System at the Austin Hospital To make the wards of the Austin Hospital the safest in the world To ensure that there is prompt and reliable identification of acutely deteriorating patients on the ward To ensure that there is maximal MET activation for patients most likely to benefit from review and intervention To optimize the outcome of patients who are subject to MET review To ensure that patient comfort and dignity are maintained in instances where the condition causing MET activation is irreversible, or where aggressive and invasive treatment options are deemed inappropriate To support and educate ward medical and nursing staff in the management of unwell ward patients To link audit activity of the MET with quality improvement and clinical governance activities to develop and implement strategies to improve patient safety To episodically review the activities of the RRS to identify areas of need and develop potential research or quality improvement activities MET medical emergency team

A mechanism for identifying and auditing the characteristics and outcomes of patients subject to RRT review is essential. Of crucial importance is the need to adequately resource the RRS from the outset to ensure that it is adequately staffed with competent and senior clinicians and nurses and adequately equipped with all necessary medications and devices to initiate intensive care-level of care at the bedside. The RRT must be resourced adequately to provide cover 24 h a day, seven days a week.2,8 Infrastructure to document RRT reviews and subsequently assess and audit their epidemiology is also an important component for enabling process improvement. The switchboard operator should independently maintain a log of emergency calls to allow cross-referencing during the audit process. Once the RRS is established, the most important first role of the administrative arm is to ensure that the afferent and efferent limbs of the RRS are running efficiently, and are achieving the desired outcome measures – namely, reduction in the incidence of cardiac arrests, unplanned ICU admissions and unexpected deaths. We have recently developed a mission statement to guide the ongoing administration of the RRS at our hospital (Table 27.1).

What Are the Components of the Administrative Limb? At our institution there are three major components of the RRS administrative arm. These include the Intensive Care MET administrative group, the clinical governance unit, and the senior hospital administration group (Fig. 27.2).

316

D. Jones and R. Bellomo

Event

Ward NUM

Nursing Education

MET review group

MET call Issue identified

Medical supervisor

Senior Nursing link

Governance

Root cause analysis

Medical Education

Executive committees Senior Medical link SYSTEM CHANGE Fig. 27.2  Complex interactions between the three major components of the Austin RRS administrative arm. Shown are the Intensive Care MET administrative group (light grey), the clinical ­governance unit (medium grey) and the senior hospital administration group (dark grey)

The Intensive Care MET Administrative Group This is the central component of the RRS administrative arm at our hospital. It has a number of key members and its major roles are in education, audit and research, as well as liaising with the other arms of the administrative limb (Table  27.2, Fig. 27.2). The group ensures that all new hospital staff are educated about the background to the MET, the nature of the calling criteria, and the fact that the MET is a hospitalendorsed policy. During this phase it is emphasized that no member of the hospital staff should be criticized for initiating an MET call. At the completion of each MET review, the MET fellow completes an electronic record of the events surrounding the review. A dedicated database coordinator ensures that all records are completed and compiles a monthly audit of all emergency reviews. This is then presented to members of the MET review group who meet regularly to discuss issues around the MET and problems with its running. The MET review group subsequently liaises with other components of the administrative arm and with other key members of hospital staff (Fig. 27.2).

27  The Administrative Limb

317

Table 27.2  Members of the intensive care MET administrative group and their roles Group member Role Medical supervisor Educates new hospital medical staff about MET Educates ICU fellows about managing an MET call Reviews performance of MET fellows Provides mentorship to MET fellows Assists with complex MET calls as needed Together with MET fellows, identifies system problems uncovered by MET that require “political”/“administrative” intervention Hospital nurse educator Gives regular feedback to ward about MET activity Gives regular lectures to ward staff on MET Educates all new nurses entering hospital to MET concept Educates all new ICU nurses and gives feedback to ICU nursing staff ICU nurse educator Trains ICU nurses to be part of MET Does MET simulation training sessions Teaches advanced resuscitation skills Develops nursing research projects in relation to MET Conducts nursing research MET database coordinator Ensures all MET calls data are entered in computerized database Ensures completeness against ICU NUM shift report Generates monthly report Uses monthly report of all emergency activity (code blue calls and MET calls) to identify areas of concern Charge nurse group link ICU NUM meets with all NUMs from hospital wards Issues of mutual concern discussed Adverse events reported Unsatisfactory MET responses identified Problems reported to MET review group MET research fellow Identifies all aspects of MET system that require research Designs research project with MET medical supervisor Obtains ethics approval Conducts research work with MET administrative team Develops manuscripts with MET supervisor Meets to review issues related to MET MET review group (Medical supervisor, ICU NUM, Develops strategic plan to resolve them hospital nurse educator, Develops plans for future developments ICU nurse educator, ICU Identifies hospital-wide issues that require involvement of clinician) governance Communicates with ward NUMs through open forum ICU intensive care unit, NUM nurse unit manager, MET medical emergency team

Feedback is repeatedly given to all hospital staff about the activities of the MET. The ICU nurse unit manager (NUM) liaises with NUMs from other wards and units via regular meetings (charge nurse group link). Medical staff are regularly updated about MET-related issues in grand rounds, outcome review committees, and weekly morbidity and mortality meetings.

318

D. Jones and R. Bellomo

Coordinating the Efferent Limb Both medical and nursing members of the MET undergo regular training, ­particularly in the case of new ICU fellows, who may be relatively unfamiliar with the MET concept. An approach to managing an MET review is presented, based on a previous consensus statement produced by staff from our hospital and a nearby metropolitan teaching hospital.9 Simulation training of MET management is ­undertaken by members of the MET, which is coordinated by the ICU nurse educator. A consultant intensive care physician is available at all times to provide clinical medical backup for the MET fellow. The MET medical supervisor assesses the performance of MET fellows and provides mentorship to them. All of these ­processes are essential to ensure that the members of the MET have credibility in the eyes of hospital ward staff by being competent, collegial, compassionate, and capable of effective communication.

Monitoring of Outcomes The activity of the MET is regularly reviewed and presented to the MET review group. All cardiac arrest occurring in adult surgical and medical ward patients are reviewed and issues surrounding these events are discussed. Some of the findings of this group are canvassed for the basis of ongoing research projects.

Directing Future Research The medical supervisor is also responsible for coordinating MET-related research activities. For the past 4 years, our hospital has had a part-time MET research fellow. In addition, research projects have been undertaken by local10,11 and international12 research staff. The role of the MET research fellow is to conceive new research projects in conjunction with the medical supervisor. These are then developed into formal protocols that are presented to various members of the MET review group or other members of ICU staff. This ensures balance of opinion and direction, and optimizes stakeholder buy-in to ensure the ongoing success of the MET. All projects are submitted to the hospital research and ethics committee, and approved prior to commencement. Frequently, the data are collected by a member of the MET review group, a junior researcher undertaking a research degree, or a member of the ICU nursing staff. All research findings are presented for peer review publication and many are presented to clinical governance and in regular medical and surgical meetings (see above). The collected data also allow hospital administration to benchmark the performance of our hospital with other hospitals in our region, and to assess changes in performance within our hospital temporally, or in response to quality improvement initiatives.

27  The Administrative Limb

319

Linking with the Clinical Governance Unit The MET review group and medical supervisor frequently liaise with the hospital risk management and clinical governance unit. Significantly, the MET itself is not a mechanism for clinical governance. Instead, the MET identifies issues relating to patient safety that may potentially be repeated unless quality improvement measures are put into place to prevent recurrence. Peer review publications from our hospital involving the clinical governance department include the use of conscious sedation on the ward,13 the effect of the MET on long-term in-hospital deaths,14 and the effect of an education program on the ongoing utilization of the MET at our hospital.7 The clinical governance unit co-ordinates the process of hospital-wide system change, reviews all hospital deaths, develops root-cause analysis for problem areas identified by the MET, takes these findings to executive committees, and liaises with the hospital board and chief executive officer (Fig. 27.2).

The Role of Hospital Administration The MET and broader RRS cannot be effective without endorsement, support and leadership from senior medical and nursing staff.2 The senior medical staff link to the RRS in our hospital is the Chief Medical Officer. His or her role is to communicate with clinical governance and the MET medical supervisor and to mediate with a variety of executive committees (Fig. 27.2). The chief medical officer also helps resolve issues of MET and hospital ward medical manpower, tackles issues of medical politics and addresses disciplinary issues involving the behavior or clinical performance of individual clinicians. The senior nursing staff link to the RRS is the hospital’s director of nursing, whose role is to communicate with the NUMs of hospital wards and the intensive care unit (Fig. 27.2). She or he also assists in developing changes in nursing allocation and assists in developing new roles and ensuring there is adequate nursing manpower. Senior nurse administrative endorsement and support for the RRS is essential, as nurses are responsible for the majority of RRT call initiations in most of the published studies. Hospital administration also needs to advocate with the Departments of Health to obtain adequate funding to ensure the RRS is adequately staffed and resourced. In large hospitals within state-funded hospital systems, funding is often governmental and capped. Obtaining adequate resources for the RRS often involves reallocation of resources within the hospital (which may be difficult and political), and presenting the case for additional funding to governmental funding bodies. In conclusion, the success of an RRS heavily depends on the creation of a suitable administrative structure, which is essential to the integration of the RRS into the overall safety and governance strategy of the hospital and allows auditing and quality improvement. Without such an administrative structure, the performance of an RRS is likely to be suboptimal and its long-term viability poor.

320

D. Jones and R. Bellomo

References 1. Delaney A, Angus DC, Bellomo R, et al. Bench-to-bedside review: the evaluation of complex interventions in critical care. Crit Care. 2008;12:210. 2. DeVita MA, Bellomo R, Hillman K, et al. Findings of the first consensus conference on medical emergency teams. Crit Care Med. 2006;34:2463–2478. 3. Advances in Patient Safety: From Research to Implementation. 4. Berwick DM. The science of improvement. JAMA. 2008;299:1182–1184. 5. Bellomo R, Goldsmith D, Russell S, Uchino S. Postoperative serious adverse events in a teaching hospital: a prospective study. Med J Aust. 2002;176:216–218. 6. Bellomo R, Goldsmith D, Uchino S, et al. A prospective before-and-after trial of a medical emergency team. Med J Aust. 2003;179:283–287. 7. Jones D, Bates S, Warrillow S, et al. Effect of an education programme on the utilization of a medical emergency team in a teaching hospital. Intern Med J. 2006;36:231–236. 8. Lee A, Bishop G, Hillman KM, Daffurn K. The medical emergency team. Anaesth Intensive Care. 1995;23:183–186. 9. Jones D, Duke G, Green J, et al. Medical emergency team syndromes and an approach to their management. Crit Care. 2006;10:R30. 10. Downey AW, Quach JL, Haase M, Haase-Fielitz A, Jones D, Bellomo R. Characteristics and outcomes of patients receiving a medical emergency team review for acute change in conscious state or arrhythmias. Crit Care Med. 2008;36:477–481. 11. Quach JL, Downey AW, Haase M, Haase-Fielitz A, Jones D, Bellomo R. Characteristics and outcomes of patients receiving a medical emergency team review for respiratory distress or hypotension. J Crit Care. 2008;23:325–331. 12. Calzavacca P, Licari E, Tee A, et al. A prospective study of factors influencing the outcome of patients after a Medical Emergency Team review. Intensive Care Med. 2008;34:2112–2116. 13. Warrillow S, Bellomo R, Jones D. Conscious sedation on a general ward: the MET and clinical governance. Jt Comm J Qual Patient Saf. 2007;33:112–117, 161. 14. Jones D, Opdam H, Egi M, et al. Long-term effect of a Medical Emergency Team on mortality in a teaching hospital. Resuscitation. 2007;74:235–241.

Chapter 28

The Second Victim Susan D. Scott, Laura E. Hirschinger, Myra McCoig, Karen Cox, Kristin Hahn-Cover, and Leslie W. Hall

Keywords   Second • Victims • Designing • Emotional • First • Aid • Rapid • Response • Team The longer we dwell on our misfortunes, the greater is their power to harm us. Voltaire

Gary Donnell MD, is a second-year resident in Medicine caring for Mr. Pauley, a 64-year-old with a history of diabetes and chronic renal insufficiency. Mr. Pauley’s nurse is Katie, a new graduate. Mr. Pauley was admitted for sudden-onset left side-weakness. A CT scan of his brain on admission showed no evidence of hemorrhage/mass. Katie paged Dr. Donnell 4 h later, when she found the patient to have a distinct change in his speech patterns from her initial admission assessment. Busy with several ER patients, Dr. Donnell did not evaluate Mr. Pauley, but ordered a STAT repeat head CT scan based on Katie’s findings. Radiology assured Dr. Donnell that Mr. Pauley’s CT would be ­completed within the hour. Dr. Donnell instructed Katie to perform hourly neurologic exams, and to page him with any changes. Shortly before Katie was next due to perform a neurologic exam, Mrs. Pauley noticed her husband could not carry on a conversation. She found Katie in the hallway and asked for her help. Katie hurried to assess Mr. Pauley, and found him unresponsive. She immediately activated the rapid response team to help assess and stabilize her patient. A text page notification was also sent to Dr. Donnell informing him of the team’s activation. Upon Dr. Donnell’s and the rapid

S.D. Scott (*) University of Missouri Health System, Office of Clinical Effectiveness, MU Sinclair School of Nursing, Doctoral Student, Columbia, MO, USA e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concept and Implementation, DOI 10.1007/978-0-387-92853-1_28, © Springer Science + Business Media, LLC 2011

321

322

S.D. Scott et al.

response team’s arrival, Mr. Pauley was unarousable. His heart rate was 138, blood pressure 164/92, and he was afebrile. Dr. Donnell arranged immediate transfer to the Neurological Intensive Care Unit for progression of presumed stroke. The morning after the transfer, intensivist Erin Boyd, MD, paged Dr. Donnell with an update. Mr. Pauley’s blood glucose had been 20 when he arrived at the ICU. He was treated with a D50W bolus and a D5-1/2 NS drip. Repeat head CT showed no evidence of bleeding or ischemia. Mr. Pauley’s blood glucose has now normalized, but he remains unresponsive. The team is concerned about brain injury due to profound hypoglycemia.

Identifying Emotional Vulnerability and Recognizing Second Victims Dr. Donnell couldn’t believe it. How could he have missed a hypoglycemic event? What must Dr. Boyd and the ICU team think of him? He would have to tell his attending and intern. How would he ever regain his credibility? Would he ever be trusted again? Questioning one’s abilities to meet the intense demands required of healthcare professionals in the aftermath of an unexpected clinical event can be the beginning of a long emotional road of silent suffering. Dr. Donnell encountered an inner turmoil that he had never experienced before in his training or career. He had a sickening realization that he was responsible for not recognizing sooner the hypoglycemic event that may have caused serious harm to his patient. Dr. Donnell could only think about this case, repeating each action and thought, step by step, over and over. One perplexing question remained on his mind: If he missed this simple problem, was he good enough to be a doctor? Unexpected clinical events occur every day in hospitals; some involve mistakes, while others result from complications of the patient’s condition. We are just now learning more about the vast emotional suffering experienced by many healthcare workers, irrespective of level of experience, gender or professional type. The intense work of healthcare often exposes clinicians to emotionally laden situations. Resilient members of the healthcare professions can typically review case events and make sense of what has unfolded under their watch. Occasionally, specific patient experiences trouble even the most confident clinician, resulting in a vulnerable period in which the clinician faces haunting re-enactments of the event. The personal impact of emotionally devastating clinical events has been described in healthcare literature as the “second victim” experience.1 Second victims often experience a variety of physical and/or psychosocial symptoms in the aftermath of the event.2 Many clinicians are perplexed about whom they can safely turn to for support and guidance. As a result, it is common for them to suffer in silence. Clinicians, like Dr. Donnell, often question their clinical decision making

28  The Second Victim

323

and ultimately their professional abilities. These events can result in heavy personal toll and potential career-altering changes. In the 1980s, first-person stories began to appear in the literature, chronicling intense feelings of incompetence, inadequacy, or guilt following a medical error.3–6 Additional renditions of the traumatic impact of unexpected clinical events continued in the 1990s when unique needs of second victims were described and hypothesized support strategies for clinicians were offered.7–13 As patient safety and risk management professionals dealing with patient safety events, we witnessed emotional suffering first-hand, yet the literature lacked a clear roadmap of actions we could or should take when providing emotional support interventions during postevent investigations. Since 2006, the multidisciplinary team charged with overseeing safety event reporting and risk management at University of Missouri Health System, has been systematically studying this phenomenon in order to improve patient safety while also mitigating the impact of emotional trauma on clinicians. A steering team, consisting of interprofessional clinicians and individuals who provide crisis support as part of their job, was convened. Our team defined second victims as “healthcare providers who are involved in an unanticipated adverse patient event, in a medical error and/or a patient-related injury, and become victimized in the sense that the provider is traumatized by the event. Frequently, these individuals feel personally responsible for the patient’s outcome. Many feel as though they have failed the patient, second-guessing their clinical skills and knowledge base.”14 Using this definition, interviews were conducted with 31 clinicians known to have been personally affected by events in the workplace, to inquire about the experience and what institutional supports might have’ have mitigated the suffering. Most accounts clearly described specific career-jolting experiences, in meticulous detail, revealing the damaging impact of emotionally charged incidents. Unexpectedly, these narratives, once analyzed in the aggregate, revealed a largely predictable trajectory towards emotional recovery. These post-event stages include (1) chaos and accident response; (2) intrusive reflections; (3) restoring personal integrity; (4) enduring the inquisition; (5) obtaining emotional first aid; and (6) moving on (Fig.  28.1).14 Institutional support strategies were also described that match with the stages of emotional recovery. Interventions to help the second victim cope were desired from peers and supervisors immediately after the event as well as periodic support by valued colleagues long after the event.

Immediate Support During the Crisis At the time of event identification and during the initial aftermath, enormous ­confusion is likely. “Impact realization” encompasses the first three stages of the ­recovery trajectory, where the second victim may experience a variety of emotional and physical reactions. These responses are normal human reactions to abnormal or unusual traumatic stressful situations or events.15

324

S.D. Scott et al. Stages 1-3 Impact Realization

Stage 4

Stage 5

Enduring Obtaining the Emotional Inquisition First Aid

Stage 6 Moving On

Chaos & Intrusive Restoring Accident Reflections Personal Response Integrity

Dropping Surviving Thriving Out

(Individual may experience one or more of these stages simultaneously)

(Individual migrates toward one of three paths)

Fig. 28.1  The second victim recovery trajectory

Upon event discovery, patient stabilization is the first priority. However, the second victim is frequently distracted from immediate patient care demands by becoming engrossed in self-reflection replaying the activities preceding the event. Some providers may be unwilling or unable to discuss the event, compounding their anxiety and ability to work through the case events.16 Once the patient’s condition stabilizes, clinicians begin to worry about how the case might impact their careers. Specifically, physicians described apprehension about medical-legal proceedings, while nurses expressed fear of losing their jobs and/or licenses. Over days, the second victim tends to self-isolate in an attempt to concentrate and focus on precisely what transpired (Fig.  28.2). Over time, clinicians move from these worries to an intense fear that they will no longer be viewed as trusted colleagues. Ultimately, second victims desperately want to know post-event details so they can make a positive difference in future outcomes. During these three stages of impact realization, peers serve an important role in emotional “first aid.” In a study of physicians who experienced the second victim phenomenon, four specific needs were identified (1) the need to talk to someone; (2) the need for validation of decisions made during patient’s care; (3) the need for professional reaffirmation of competence; and (4) the need for reassurance of self-worth.9 Not addressing the specific unique needs of the second victim could leave an enduring emotional scar on the clinician that can be imprinted in their memories forever.17,18 Support by peers that is timely and effective can provide much-needed reassurance. Some clinicians benefit from immediate relief of patient care duties to allow them time to collect their thoughts prior to resuming patient care. After this respite from clinical care, a “safe zone” can be offered to allow second victims to openly express their feelings and concerns regarding the event in a ­confidential and non-judgmental environment. Given an opportunity to unload their emotional response with a trusted peer is likely to result in less emotional suffering and improved feelings of self-worth.

28  The Second Victim

325

Fig.  28.2  Immediately following an adverse event, clinicians frequently reflect on the case to enhance their understanding of what has transpired

This interaction requires a peer who can facilitate a dialogue with someone in crisis, someone trained in active listening skills that maximally allow event reflection, and scripted responses to reassure the victim that he is not alone. Peer supporters should reinforce their confidence in the second victim. Astute supervisors can offset much emotional trauma. Supervisors should be aware of common high-risk situations likely to evoke a second victim response so that clinicians at the epicenter of such events can be deliberately identified, monitored and supported in a timely manner. Figure  28.3 identifies events from our interviews that are likely to evoke emotional turmoil. It is important to touch base with potential victims more than once, as there may be cases in which an emotional impact is delayed or cases in which the victim is unable to talk about the emotional trauma immediately after the event. Second victims articulated the need for specific key messages from their supervisor after a traumatic clinical event. For example, second victims desperately need to hear that their supervisor continues to have confidence in their clinical skills, that they were still trusted, and that the victim remains a valued member of their clinical team. Supervisors should also introduce the victim to formal investigative steps following safety events. It is important for the supervisor to be readily accessible during event investigations to explain steps in the process, make introductions with key safety and risk management personnel, and provide reassurance/support.

326

S.D. Scott et al. · Failure to identify patient deterioration and/or failure to rescue · Medical error or preventable harm to patient · Any patient which ‘connects’ a staff member to his/her own family (share mother’s name, age of children, same appearance, red curly hair of grandchild, etc.) · Cases involving children · First death under their watch · Unexpected patient demise · Multiple patients with same poor outcomes within a short period of time within one clinical area

Fig. 28.3  High-risk clinical events for evoking second victim response

Support Long After the investigation Well after the chaos of the event has settled and after the second victim has had time to personally process what happened, interviewees expressed the feeling that, given a choice, they would welcome conversations about their emotional response with a trusted peer or colleague. The ideal peer or colleague is someone familiar with the victim’s specific professional role and setting, so that insights could be shared concerning common reactions and responses to similar case or event types. Through peer collaboration and reflection, the second victim senses a colleague who cares about him as an individual and as a professional. It is important for this person to practice active listening skills that will allow the second victim to share responses and feelings about the clinical event. Peers, especially those who have experienced this phenomenon, can offer powerful healing words and special ­support based on their personal experience. If the colleague has experience as a second victim, sharing personal stories can be powerful. Even without personal experience, the colleague can still be supportive while anticipating common ­second victim needs, free from judgment and blame. These brief conversations may take place spontaneously (Fig. 28.4). Ultimately, ongoing reassurance that the clinician remains a respected and trusted member of the clinical team is essential. However, some conversations may extend far beyond a few minutes and the ­second victim may rapidly lose composure, requiring the arrangement of a followup, by way of a more structured meeting. Some second victims – we ­estimate 10% – may benefit from support and counseling by professional experts. In these cases, the colleague must have the skills to recognize when additional assistance is ­warranted and be prepared to propose a professional referral, with contact information about available resources. Healing for the second victim may take months or even years. We believe that professionals who are not supported after emotional trauma and who are allowed to suffer alone are at risk for premature career departure. On the other hand, future professional direction may be significantly influenced by support and guidance received within those institutions that prepare for and manage a formal second victim response team.

28  The Second Victim

327

Fig. 28.4  Professional colleagues and peers can offer powerful healing words for their distressed co-worker

Emotional First Aid When Entire Teams Are Suffering Occasionally, an entire healthcare team can be impacted by a clinical event, such as a rapid response team activation for a hospitalized co-worker who ultimately dies, or trauma team activation for a maternal trauma that results in the deaths of both mother and baby. Dramatic events such as these may bring about long-lasting impact on team function and morale. In these situations, a team debriefing may be an effective and efficient approach for beginning the healing process as team members reflect on components of the event. Specifically, the team debriefing should focus primarily on the emotional impact, not individual team members’ performances during the crisis. Also referred to as team defusings, team debriefings should be planned within 8–12 h of the event.19 All team members closely involved, including licensed and non-licensed personnel, students or volunteers, should be individually contacted and invited to a 45–60  min team debriefing. Participation should be on a voluntary basis and no team member should be forced to speak up. A meeting location should be selected for the debriefing, which promotes an environment free of interruptions and other clinical distractions (Fig.  28.5). During these emotionally charged discussions, a facilitator skilled in group dynamics and critical incident stress management (CISM) should coordinate and lead the team discussion. CISM has been used effectively for years in the ­emergency medical community and was developed to help prevent potentially ­disabling stress responses among emergency responders to traumatic community disasters such as the Columbine shootings, the Oklahoma City bombing, and the 9/11 World Trade Center attack.20

328

S.D. Scott et al.

Fig. 28.5  Team debriefings are opportunities for participants to discuss their unique perspectives regarding the event as well as their feelings/reactions to it

How to Formalize a Support Network Clinicians should not have to seek support on their own. Key principles should serve as a guide for the institution’s formalizing a peer support infrastructure, keeping in mind that affected staff should always be treated with respect, ­compassion and administrative support immediately following adverse clinical events. 21 Denham proposes basic rights of second victims that should be guaranteed­by one’s workplace and summarizes these rights using the acronym “TRUST” (Treatment that is just, Respect, Understanding and compassion, Supportive care and Transparency) and the opportunity to contribute to learning.22(p.107) To support a rapid response for second victims at all times, formal peer support teams should be designed to complement existing support infrastructures within the facility. Appropriate resources can be allocated to ensure peer availability and support for members of the clinical team as well as the necessary team framework of polices and prcocedures.23 Successful peer-to-peer support programs can be found in community emergency response systems and could serve as useful examples for designing a hospital-based support network. Brigham and Women’s Hospital pioneered a formalized support structure, the Medically Induced Trauma Support Services Team (MITSS), to provide support for patients/families as well as healthcare clinicians following an unexpected clinical event.24 The University of Missouri Health System’s support structure, the forYOU team, was designed in 2006 to increase awareness of the second victim phenomenon and to provide consistent “in-time” support for second victims.25

28  The Second Victim

329

Putting It All Together Based upon emotional injury and recovery information acquired from second ­victims, institutions should formalize a second victim support network across every worksite to ensure that every healthcare clinician, student or volunteer is monitored for second victim reactions. Institutions should design a structured response plan that results in ongoing surveillance for potential second victims and acts to mitigate emotional suffering immediately upon second victim identification. Using a peerto-peer support model with specially trained individuals embedded in high-risk clinical environments (such as Rapid Response Teams, ICU, ER, OR, helicopter/ ambulance services, etc.) provides opportunities for continual surveillance as well as the capability for immediate basic emotional first aid. Healthcare institutions more than likely have internal resources available that could provide the support structure needed by caregivers. Developing an infrastructure of support requires leadership and coordination by skilled personnel representing patient safety, risk management, Employee Assistance Programs (EAP), chaplaincy, social work, and other holistic or mental health clinicians. These experts normally work with individuals during emotionally devastating times and have the passion to support these types of programs. These individuals, once organized, can serve as mentors for colleagues and on-unit peers who are interested in providing emotional first aid to identified second victims. An understanding of the second victim phenomenon, together with peer and supervisory surveillance in the aftermath of high-risk clinical events, provides a unique opportunity to enable second victims to return to full, rewarding professional roles. To have someone call me out of the blue, just to offer support, was a wonderful thing. It was like a burden was lifted off me, knowing I didn’t have to get through it alone. Physician in response to peer support team activation I don’t think I’ve met a doctor over the age of 30 that hasn’t experienced at least 2 memorable adverse patient events. By the time folks get to my point in their careers they will probably have experienced at least 4–6 such events. Even many years after these have occurred, I find myself thinking about them at least 2–3 times per year, rehearsing once again if there is anything else I could have done to avoid the negative outcome. Physician with more than 25 years’ experience

References 1. Wolf ZR, Serembus JF, Smetzer J, Cohen H, Cohen M. Responses and concerns of healthcare providers to medication errors. Clin Nurse Spec. 2000;14(6):278–287. 2. Scott SD, Hirschinger LE, Cox KR. Sharing the load: rescuing the healer after trauma. RN. 2008;71(12):38–43. 3. Levinson W, Dunn PM. A piece of my mind. Coping with fallibility. JAMA. 1989;261(15):2252. 4. Hilfiker D. Facing our mistakes. N Engl J Med. 1984;310(2):118–122. 5. Hilfiker D. Healing the Wounds: A Physician Looks at His Work. New York, NY: Pantheon; 2009.

330

S.D. Scott et al.

6. Elizabeth R. The mistake I’ll never forget. Nursing. 1990;20(9):50–51. 7. Christensen JF, Levinson W, Dunn PM. The heart of darkness: the impact of perceived mistakes on physicians. J Gen Intern Med. 1992;7(4):424–431. 8. Engel KG, Rosenthal M, Sutcliffe KM. Residents’ responses to medical error: coping, learning, and change. Acad Med. 2006;81(1):86–93. 9. Newman MC. The emotional impact of mistakes on family physicians. Arch Fam Med. 1996;5(2):71–75. 10. West CP, Huschka MM, Novotny PJ, et al. Association of perceived medical errors with resident distress and empathy: a prospective longitudinal study. JAMA. 2006;296(9):1071–1078. 11. Wu AW, Folkman S, McPhee SJ, Lo B. Do house officers learn from their mistakes? JAMA. 1991;265(16):2089–2094. 12. Wu AW, Folkman S, McPhee SJ, Lo B. How house officers cope with their mistakes. West J Med. 1993;159(5):565–569. 13. Wolf ZR, Serembus JF, Smetzer J, Cohen H, Cohen M. Responses and concerns of healthcare providers to medication errors. Clin Nurse Spec. 2000;14(6):278–287. 14. Scott SD, Hirschinger LE, Cox KR, McCoig M, Brandt J, Hall LW. The natural history of recovery for the healthcare provider “second victim” after adverse patient events. Qual Saf Health Care. 2009;18(5):325–330. 15. Mitchell JT, Everly GS. Critical Incident Stress Debriefing: An Operations Manual for Cisd, Defusing and Other Group Crisis Intervention Services. 3rd ed. Ellicott City, MD: Chevron Publishing; 1997. 16. Delbanco T, Bell SK. Guilty, afraid, and alone – struggling with medical error. N Engl J Med. 2007;357(17):1682–1683. 17. Schwappach DL, Boluarte TA. The emotional impact of medical error involvement on physicians: a call for leadership and organisational accountability. Swiss Med Wkly. 2009;139(1–2):9–15. 18. Serembus JF, Wolf ZR, Youngblood N. Consequences of fatal medication errors for health care providers: a secondary analysis study. Medsurg Nurs. 2001;10(4):193–201. 19. Mitchell J. Critical Incident Stress Management(CISM): Group Crisis Intervention. 4th ed. Ellicott City, MD: International Critical Incident Stress Foundation; 2006. 20. Mitchell JT. Advanced Group Crisis Intervention: Strategies and Tactics for Complex Situations. Participant Manual. 3rd ed. Ellicott City, MD: International Critical Incident Stress Foundation; 2006. 21. Conway JB, Weingart SN. Leadership: assuring respect and compassion to clinicians involved in medical error. Swiss Med Wkly. 2009;139(1–2):3. 22. Denham C. Trust: the 5 rights of the second victim. J Patient Saf. 2007;3(2):107–119. 23. Robinson R, Murdoch P. Establishing and Maintaining Peer Support Programs in the Workplace. 3rd ed. Ellicott City, MD: Chevron Publishing; 2003. 24. Medically Induced Trauma Support Services (MITSS) Organization located in Boston, Massachusetts, USA. http://www.mitss.org/; Accessed updated to October 6, 2010. 25. forYOU team. Columbia, MO. www.muhealth.org/secondvictim; Accessed updated to October 6, 2010.

Part III

Monitoring of Efficacy and New Challenges

Chapter 29

RRSs in Teaching Hospitals Max Bell and David Konrad

Keywords  Rapid • Response • Teams • Teaching • Hospitals

Introduction Although institutions for caring for the sick are known to have been around much earlier in history, the first teaching hospital was reportedly the Academy of Gundishapur in the Persian Empire during the Sassanid Era. In addition to systemizing medical treatment and knowledge, the scholars of the Gundishapur Academy also transformed medical education; students were authorized to methodically practice on patients under the supervision of physicians as part of their education. And rather than apprenticing with just one physician, medical students were required to work in the hospital under the supervision of the whole medical faculty. This early teaching hospital is seen to have been the most important medical center of the ancient world (defined as Europe, the Mediterranean, and the Near East) during the sixth and seventh centuries.1 Some things have changed since then, but not everything. Most teaching hospitals pursue three related enterprises: • Teaching: Training nurses, medical students and resident physicians. • Research: Conducting both basic science and clinical investigation. • Patient care: Delivering healthcare services through a network that may include one or more hospitals, satellite clinics, and physician office practices. This is sought to result in a health system that encourages the highest standards of quality and provides access to the most up-to-date treatments. It it also meant to

M. Bell (*) Department of Anesthesia and Intensive Care, Karolinska University Hospital, Solna, Stockholm, Sweden e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concept and Implementation, DOI 10.1007/978-0-387-92853-1_29, © Springer Science+Business Media, LLC 2011

333

334

M. Bell and D. Konrad

foster an environment where the staff is always thinking about the patient’s ­condition, leading to more thorough patient care. From one point of view, a teaching hospital is a high-risk environment. Usually these hospitals are large, with many beds, and complex medical conditions in patients with serious co-morbidities are treated there. This, coupled with high numbers of trainees, can create what might be a malignant combination. There have indeed been studies reporting adverse events in teaching hospitals.2–4 To complicate matters, teaching hospitals have a high turnover in their ward staff; both junior doctors and nurses-in-training migrate to new positions as part of their education. The local know-how in a single ward is thus changed – and drained – on a regular basis. Junior doctors and nurses must also be considered by default to be inferior (as compared to those who have completed training) when it comes to medical judgement, practical skills or knowledge of treatment protocols. The system of medical management in teaching hospitals may then per se represent a milieu where the signs and symptoms of deterioration in patients-at-risk could be overlooked. Yet another system deficiency is how junior staff are trained to deal with the critically ill patient. Nobody would expect a junior doctor to handle a brain hemorrhage by her- or himself. A neurosurgeon would immediately be alerted if a patient undergoing a CT scan showed cerebral bleeding. The deteriorating patient in the general ward is expected to (at least initially) be treated by the physicians that happen to work in that environment: those without critical care training, or perhaps those who are junior trainees. Indeed, critical care physicians may not be included in the care team until much later after further untoward events occur. Another point of view is this: Teaching hospitals are the perfect place to implement new systems of care, such as the RRS. One could argue that the environment, by its very nature, is open to change. Teaching hospitals are expected to frequently and critically review the way patients are treated. Pharmacological innovations, new treatment techniques, and systemic changes in how to manage patients represent changes in practice. The doctors and nurses in teaching hospitals may be eager to learn new things and are used to the constant struggle of life-long learning. Another benefit of teaching hospitals is the fact that research, and research projects, are present in the day-to-day care. This is helpful, as the introduction of an RRS should always be preceded by the recording of baseline patient characteristics and outcome. Moreover, after the RRS is in place, it is recommended to continue the data collection to ensure that the RRS works and to make feedback possible. In reality, however, teaching hospitals – just like any other hospital organiza­ tion – often do not implement new systems of care without a “fight.” There will always be a reluctance to move from a familiar system to something new, especially if this new system invites a whole new look on the responsibilities of the physicians, the nurses and other ward personnel. An introduction of a rapid response system represents a massive system change. In teaching hospitals – and in all hospitals – winning the hearts and

29  RRSs in Teaching Hospitals

335

minds of key persons is paramount! The joke that eminence-based medicine is more important than evidence-based medicine works here, too. Naturally, it is equally important to convince the nursing staff that this change will help them to help their patients. An insight from the Karolinska University Hospital, recently reported by Jones and colleagues, is that even after the implementation of an RRS, there is a perpetual need for continuous educational efforts. It may, in fact, take years before a rapid response system is up and engrained into the culture of care at a hospital.5

Implementing RRSs in Teaching Hospitals Due to the complexity, heterogeneity, and size of teaching hospitals, the process of change within them can be both time and energy consuming, but at the same time very rewarding. The key issue is changing the culture of an entire institution and not merely adding another process. Wide acceptance of a new course of action is crucial and depends largely upon educational efforts and persistent reinforcement. There is a need to explain the expected effects of RRS implementation to a very diverse audience – physicians, nurses, department leaders and hospital management – all of whom will have different viewpoints and reasons to accept or reject the proposals. Arguments in favor of the new system need to be focused and titrated in order to make all stakeholders embrace the common goal, that of enhancing the hospital’s level of patient safety. Implementing an RRS is a major step in that direction.

The Afferent Limb In order to have an efficient RRT, there need to be clear and readily available triggers. The ward caregivers, including physicians, should be well informed of these and encouraged to use them. Well-displayed posters and pocket-size stickers with the calling criteria and page-procedure will facilitate RRT activation. In a teaching hospital, there are a wide variety of wards and a high degree of specialization. Nonetheless, the triggers used should be the same throughout the institution. A major deterioration in physiological state constitutes a medical emergency, regardless of cause and location. In teaching facilities, as mentioned above, there tends to be a high turnover of staff and change of working places. Also, in such settings, many students will rotate among wards. Having identical triggers, regardless of various hospital locations, will increase clarity, minimize confusion, and reduce the risk of delay in calling the response team. Continuous feedback to the wards will also encourage the use of the RRS. At our institution, we use contact persons to ease communication between wards and the RRT, creating a channel for reciprocal feedback.

336

M. Bell and D. Konrad

The Efferent Limb The responding arm of the RRS, the efferent limb, should be multi-disciplinary and have critical care skills. The most common setup is using an ICU-based team consisting of an ICU physician and/or an ICU nurse who will meet up with the nursing and medical staff of the ward at the patient’s location. This will ensure ample competence as well as staff familiar with the patient, thus facilitating assessment and treatment. Including the medical staff of the ward at the scene will also decrease the risk of the response team being considered “intruders” and will aid in planning the remainder of the patient’s stay, possible referrals, and sometimes a reassessment of DNR status. It is wise for the responders to behave in a professional and inclusive manner in order to facilitate further collaboration rather than create a power struggle. It is very important that whoever triggered the RRS be encouraged and thanked rather than criticized. We have found it fruitful to think of every MET or RRT call as an educational opportunity regarding management of the critically ill. The introduction of the RRS is likely to cause an increased workload for the ICU staff if the response team is indeed made up of critical care professionals. On the other hand, the excursions outside the ICU will also provide insight into working conditions and patient population at the general wards. The team may recognize factors that can lead to crisis situations, and help create collaborative ways to prevent future events. One form of response team, the Critical Care Outreach team, is an example of this collaborative work. ICU nurses may visit patients at high risk of having an event – before a crisis occurs – in an effort to bring additional resources to the bedsides of patients and prevent an event. In our experience, ICU nurses consider RRT work to be a privilege.

Hospital Culture and Management In trying to change hospital culture and to implement the RRS, the value of ­information and education cannot be over-emphasized. There will certainly be opposition, since the implementation of an RRS affects current social and medical practices, possibly even financial aspects, and as a result may be seen as an intrusion. In our experience, a productive approach is to run a pilot study or otherwise assess an institution’s baseline in order to highlight problems regarding patient safety. Patient safety – and changes in patient safety – can be measured in many ways, such as adverse events, cardiac arrest rate, length of stay, standardized mortality ratios, or other aspects that may benefit from RRS implementation. The focus of any assessment or intervention should always be on improving a faulty system, not person, moving away from a culture of individual blame. Correct and timely feedback will also encourage hospital governance in facilitating the RRS process. The potential for cost savings may also be relevant.

29  RRSs in Teaching Hospitals

337

Experiences with the RRS Most of the current literature regarding RRSs comes from single-center, before-and-after studies conducted at teaching hospitals. Many of these studies report beneficial effects on unexpected deaths, cardiac arrest rate, ICU length of stay and some even on reduced hospital mortality. A more thorough review of the evidence is found in Chap. 6. Clinical deterioration and abnormal vital signs are associated with increased risk of cardiac arrest and death.6–8 Also, there seems to be a correlation between meeting one or more RRS-trigger criteria and longerterm mortality. At the Karolinska, meeting such criteria even once during a hospital stay is associated with an increased risk of death that persists for up to 6 months9 (Fig.  29.1). When acting upon such triggers, several authors have reported significant reductions in cardiac arrest rates.10–12 Hospital mortality has been shown to decrease significantly,11 as well as morbidity and mortality for patients undergoing major surgery.13 Interestingly, a dose-response relationship appears to exist: the more frequent an RRS is activated, the greater the reduction in cardiac arrest rates.14,15 A notable effect of implementing the RRS is the increase in NFR orders and the role of the RRS in end-of-life care. This area is addressed in detail in Chap. 34. The great majority of these studies comes from centers in Australia or North America. The reproducibility of these results in other centers in other parts of the world is probable but not certain. For this reason, it is imperative to continue to pursue

Fig.  29.1  Kaplan–Meier graph of survival for patients meeting any MET criteria. MIG MET criteria fulfilled; not MIG no MET criteria fulfilled. Modified from Bell et  al.9 MIG mobile intensive care group is the Swedish acronym for MET

338

M. Bell and D. Konrad

e­ vidence regarding the efficacy of an RRS. There is a need to understand by which mechanisms RRS functions; in order to do so, attention should be paid to the RRS process itself. The RRS, once implemented, is generally well accepted and often with great enthusiasm. For the nursing staff at the wards, the RRS constitutes a safety net that they can use, either directly, or when not getting the appropriate attention from their medical staff. It is therefore imperative that all the healthcare providers can rely on that activation of the RRS will lead to proper and swift action. The RRS can offer support to junior medical staff when they are faced with difficult situations. In addition, the RRS may also enhance the education if the responders, take advantage of the situation, and teach as well as treat. In our experience, the ICU nurses are looked upon as experts by their peers at the ward, contributing to the work satisfaction of both groups. At our ICU, it is considered to be a privilege to be part of the response team; only those nurses with at least 1 year of ICU experience may be candidates for being part of the team. Their skills and competence set an example for the ward staff and may also facilitate recruitment of ICU nurses in the future. For ICU residents and fellows on the response team, the experience is a very good way of learning how to deal with patients with a serious deterioration in vital organ physiology, and learning how to prevent further deterioration and ICU admission. These experiences likely contribute to their training.

Summary Adverse events and ward patient crises still pose a great challenge in modern teaching hospitals. Perhaps unlike the ancient academic institutions, such as the Academy of Gundishapur in the sixth century, modern, inexperienced healthcare providers during their training may not be supervised or adequately backed up. Also, experienced staff in the wards may not be sufficiently trained in the subspecialty of managing critically ill patients. The responding team has a great potential to intervene to reduce potential harm to patients. In many single-center studies, early recognition and timely treatment have shown to decrease cardiac arrest rates, morbidity and mortality. Apart from providing state-of-the-art health care, the teaching hospitals have a significant educational task where the RRS may serve two purposes: facilitating the process of learning by a generous attitude in sharing knowledge of illness, and averting the possible consequences of inexperienced staff. The success of rapid response systems depends largely on hospital-wide acceptance, reliability and feedback. To achieve this, a continuous educational effort is of utmost importance.

References 1. Frye R. The Period from the Arab Invasion to the Saljuqs, The Cambridge History of Iran, vol. 4. Cambridge: Cambridge University Press; 1975.

29  RRSs in Teaching Hospitals

339

2. Brennan TA, Leape LL, Laird NM, et  al. Incidence of adverse events and negligence in hospitalized patients: results of the Harvard Medical Practice Study I. 1991. Qual Saf Health Care. 2004;13(2):145–51 [discussion 151–2]. 3. Wilson RM, Runciman WB, Gibberd RW, et al. The Quality in Australian Health Care Study. Med J Aust. 1995;163(9):458–71. 4. James K, Bellomo R, Poustie S, Story DA, McNicol PL. The epidemiology of major early adverse physiological events after surgery. Crit Care Resusc. 2000;2(2):108–13. 5. Jones D, Bates S, Warrillow S, et al. Effect of an education programme on the utilization of a medical emergency team in a teaching hospital. Intern Med J. 2006;36(4):231–6. 6. Schein RM, Hazday N, Pena M, Ruben BH, Sprung CL. Clinical antecedents to in-hospital cardiopulmonary arrest. Chest. 1990;98(6):1388–92. 7. Hillman KM, Bristow PJ, Chey T, et  al. Antecedents to hospital deaths. Intern Med J. 2001;31(6):343–8. 8. Kause J, Smith G, Prytherch D, et al. A comparison of antecedents to cardiac arrests, deaths and emergency intensive care admissions in Australia and New Zealand, and the United Kingdom – the ACADEMIA study. Resuscitation. 2004;62(3):275–82. 9. Bell MB, Konrad D, Granath F, Ekbom A, Martling CR. Prevalence and sensitivity of METcriteria in a Scandinavian University Hospital. Resuscitation. 2006;70(1):66–73. 10. Buist MD, Moore GE, Bernard SA, Waxman BP, Anderson JN, Nguyen TV. Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. BMJ. 2002;324(7334):387–90. 11. Bellomo R, Goldsmith D, Uchino S, et al. A prospective before-and-after trial of a medical emergency team. Med J Aust. 2003;179(6):283–7. 12. DeVita MA, Braithwaite RS, Mahidhara R, et al. Use of medical emergency team responses to reduce hospital cardiopulmonary arrests. Qual Saf Health Care. 2004;13(4):251–4. 13. Bellomo R, Goldsmith D, Uchino S, et al. Prospective controlled trial of effect of medical emergency team on postoperative morbidity and mortality rates. Crit Care Med. 2004;32(4):916–21. 14. Jones D, Bellomo R, Bates S, et al. Long-term effect of a medical emergency team on cardiac arrests in a teaching hospital. Crit Care. 2005;9(6):R808–15. 15. Chen J, Bellomo R, Flabouris A, et al. The relationship between early emergency team calls and serious adverse events. Crit Care Med. 2009;37(1):148–53.

Chapter 30

The Nurse’s View of RRS Donna Goldsmith and Nicolette C. Mininni

Keywords  Nurse • View • Rapid • Response • Systems

Introduction During the past two decades, healthcare cultures have evolved to recognize the nurse’s critical thinking skills and to promote independent functioning of nurses. Nurses, like all other healthcare professionals, focus on patient safety and deliver the right care at the right time. One example of this new cultural focus of patient safety is introduction of Rapid Response Systems (RRSs) that include Medical Emergency Teams (METs) and Rapid Response Teams (RRTs). Both teams rely on nurse calling for assistance at the first signs of clinical deterioration in hospitalized patients to prevent serious adverse events (such as cardiac arrests) from occurring. Without nurses, METs/RRTs probably would not be successful in achieving this goal. The architects of the RRS programs incorporate the fact that nursing intuition is no myth. Instead, it is recognized as a form of critical thinking, defined as the blending of knowledge, skills, and attitude. High-level critical thinking as is described as evidence of “unconscious competency” because the person may demonstrate competency in situation analysis without having to think consciously about each thought process must occur and in which order. Unconscious competency comes from experience and being empowered to use the knowledge and skills that the person has developed. The nurse relies on critical thinking to help achieve the best possible patient outcomes. Because nurses are in a position to make frequent patient observations, they are often the first to recognize subtle changes that indicate that a patient is improving or getting worse. If the nurse has the power to act on this information at the right time, immediate help for a patient headed for trouble can be available.1 RRSs promote intervention by needed personnel not only by

D. Goldsmith (*) Austin Health, Heidelberg, VIC, Australia e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concept and Implementation, DOI 10.1007/978-0-387-92853-1_30, © Springer Science+Business Media, LLC 2011

341

342

D. Goldsmith and N.C. Mininni

giving nurses the power to call whenever they are worried about the patient, but also providing expert critical care service when the call for assistance is made. To support (or sometimes validate) the nurses’ activation of the MET/RRT, specific criteria have been identified that are either single-parameter (one value outside a predetermined range) or multi-parameter scoring systems developed to work within hospital specific environments. Most sets of criteria for RRS activation include any time a nurse is concerned about a patient’s condition. “Nurse worried” as an activation criterion for the MET/RRT is both acceptable and desirable because it supports the ability of nurses to recognize impending instability even if other criteria are not met, and because it promotes patient safety.

The Nurse’s Point of View Several researchers have found that nurses who have more experience are more likely to activate an MET or RRT.2–4 Nurses with more experience tend to have higher expertise in recognizing subtle changes. Nurses with 10 or more years’ experience and new nurses (<1 year’s experience) were more likely to perceive an improvement in in-patient care when an MET or RRT was implemented than other nurses (P = 0.025).5 Other researchers have found that educational preparation was an even stronger predictor of independently calling the MET or RRT than was years of experience, with baccalaureate (BSN) prepared nurses almost five times more likely to call independently than were associate degree (AD) nurses.2 In a study funded by the Robert Wood Johnson Foundation Rapid Response Team Evaluation Project, a convenience sample of staff nurses reported activating RRTs for three reasons: (1) clearly identified changes in the patient’s vital signs or mental status; (2) the nurse had a “gut feeling” that something was wrong; or (3) primary contact physicians were either not available or not responsive to urgent pleas for assistance.6 Additional information obtained from interviews of nursing staff found that METs and RRTs provided invaluable reassurance and support. Staff nurses were able to describe how they felt before, during and after activating a RRT. Many nurses went from being concerned, anxious, frustrated, uneasy, and nervous, to feeling relieved, better, supported and confident that they had done the right thing by activating the RRT.6 There is a great deal of evidence supporting the notion that most nurses consider the RRS extremely valuable. Both published and non-published results of nursing surveys document how much nurses value the response teams, whether they are physician-led (MET) or nurse-led (RRT). Nurses recognize that the RRS improves patient outcomes. In a small (n − 248), simple study conducted at the University of Pittsburgh Medical Center in 2005, 93% of the responding nurses reported that medical emergency teams improved patient care, and 84% felt that they improved the nursing work environment. Nurses who had called an MET on more than one occasion were more likely to value their ability to call a team (P = 0.002).4 In ­addition, 84% of nurses felt that it improved their work environment, and 65% of

30  The Nurse’s View of RRS

343

nurses took the availability of an RRS into consideration when seeking future employment. Similar positive attitudes towards an RRS were found in a major teaching ­hospital in Australia. A simple survey of 351 ward nurses (51% of the nursing population at the hospital) was carried out 4  years after the introduction of the MET. The survey determined that 91% felt that the MET prevented cardiac arrests, and 97% felt that the team helped manage unwell patients. Only 2% of the surveyed nurses felt that they would restrict the number of MET calls for fear of being criticized.7 The key to the success of this team project is thought to be the 1-year preparatory education program that was undertaken prior to the commencement of the service to ensure that a major “no-blame” culture change was adopted by all medical, nursing and allied health staff. In contrast, surveys conducted on the general nursing staff from all 12 of the MERIT study intervention hospitals after both the 4-month implementation period and after the 6-month study period did not show that same unwavering support of the RRS. Importantly, these surveys also showed that there was a significant relationship between this lack of positive attitude as well as a general awareness and understanding of the existence of the MET and the utilization of the system.8 RRSs are most commonly multi-disciplinary in composition. Research currently underway at the University of Pennsylvania Medical Center supports the view that not only nurses but residents also believe that METs/RRTs improve patient safety, with nurses holding this belief more strongly than residents. Positive attitudes are more prevalent when residents and nurses have more involvement in the events and the decision to activate the team.

Steps to Ensuring a Successful Rapid Response System Without any debate, the vast majority of MET/RRT activation calls are made by nurses. In order to continue to move forward and support the culture of patient safety and ensure the ongoing utilization of MET/RRT programs, progress in changing hospital culture must continue to be made. Successful RRS programs exist in hospital environments that support the staff nurse to call, call early and believe that no call is a bad call. There are several keys to successful programs. Physicians, because they hold social power, can potentially help or hinder implementation of the system. The administrative component of the RRS, which oversees its function, must be prepared to support nursing staff decisions to trigger a response by the MET or RRT. The RRS needs to include a safe reporting structure to help overcome barriers to its use. Collaborative practice relationships need to be developed between the MET or RRT and the medical, surgical or ward nurses that activate the team. Communication is an essential component of a successful RRS. Staff nurses interact with physicians every day. During a crisis, communication patterns may change in a way that can render it ineffective. Communication and hierarchy may

344

D. Goldsmith and N.C. Mininni

be challenged and result in poor transmission of important facts to key personnel who are making decisions. The Five E’ successful RRS have been described as Education, Empowerment, Efficiency, Equipment and Evaluation.1 Even with education, experience, and “empowerment,” there are still delays in activation of MET/RRTs, which must be overcome. Staff nurses will usually need supportive education on when to activate the team and how to effectively communicate with the team even as it is already in place. Despite the fact that nurses value the ability to activate a RRT for their patients and that nurses with varying experience and education will activate a team, there remain documented delays in activation. Culture change, empowering staff and re-education are ongoing needs. Institutions that have implemented an MET or RRT have reported documented decreases in calls within the first year due to waning awareness.9 Maintaining and sustaining activation of the MET/RRT requires ongoing communication programs to maintain awareness. Areas other than the in-patient nursing units, i.e., radiology, physical therapy, and physician office areas, have special education needs. With the potential for a lower frequency of calls, sustaining the education and change can be a challenge. Without ensuring that the MET/RRT is hospital policy, that all must adopt a “not-an-option-to-not-call” practice, the system will not succeed. Nurses, nor any member of staff must never be reprimanded or criticized for making a call for help. The concept of “there is no such thing as a bad MET/RRT call” must be adopted. Similarly, the message that the MET/RRT is not attempting to take over the care or management of all hospitalized patients, that they are only there to help when a patient is at risk of becoming critically unwell, must also be adopted to prevent any “turf” wars.10

Summary In conclusion, while it appears that the key to the success of an RRS lies with ­ensuring that all staff, and in particular nurses, are aware of its existence, a further challenge for ensuring the success of an RRS is the ability to persuade bedside nurses to accept this responsibility and act on their new empowerment when patients show signs of deterioration. Nurses will only do this if they feel secure enough that they are not going to be reprimanded for activating the MET or RRT. If the nursing staff is supported, and as they gain experience with the RRS, they will become the key link in the success or failure of the intervention.

References 1. Scholle CC, Mininni NC. Best-practice interventions: how a rapid response team saves lives. Nursing. 2006;36(1):36–40. 2. Wynn JD, Engelk MK, Swnaso M. The front line of patient safety: staff nurses and rapid response team calls. Qual Manage Health Care. 2009;18(1):40–47.

30  The Nurse’s View of RRS

345

3. Galhotra S, Scholle CC, De MA, Mininni NC, Clermont G, DeVita MA. Medical emergency teams: a strategy for improving patient care and nursing work environments. J Adv Nurs. 2006;55(2):180–187. 4. Jones DA, Mitra B, Barbetti J, Choate K, Leong T, Bellomo R. Increasing the use of an existing medical emergency team in a teaching hospital. Anaesth Intensive Care. 2006;34(6):731–5. 5. Jones D, Baldwin I, McIntyre T, et al. Nurses’ attitudes to a medical emergency team service in a teaching hospital. Qual Saf Health Care. 2006;15:427–432. 6. Donaldson N, Shapiro S, Scott M, Foley M, Spetz J. Leading successful rapid response teams. J Nurs Adm. 2009;39(4):176–181. 7. Cretikos MA, Chen J, Hillman KM, et al. The effectiveness of implementation of the medical emergency team (MET) system and factors associated with use during the MERIT study. Crit Care Resusc. 2007;9(2):206–12. 8. Scott SA, Elliott S. Implementation of a rapid response team: a success story. Crit Care Nurse. 2009;29(3):66–76. 9. Jones D, Bellomo R. Introduction of a rapid response system: why we are glad we MET. Crit Care. 2006;10:121. 10. Salamonson Y, van Heere B, Everett B, Davidson P. Voices from the floor: nurse’s perceptions of the medical emergency team. Intensive Crit Care Nurs. 2006;22(3):138–143.

Chapter 31

Resident Training and RRSs Geoffrey K. Lighthall

Keywords  Resident • Training • Rapid • Response • Systems

Introduction An increased focus on patient safety on the part of regulatory and accreditation agencies has led to a host of reforms aimed at improving patient welfare. Rapid response systems (RRSs) have emerged at the same time as resident work hour restrictions have come into effect, public awareness of medical error has increased, and new models of residency program accreditation have been instituted.1,2 While well-intentioned, these reforms and improvements in healthcare have not emerged as a coherent and user-friendly package. For example, concerns over medication error have led to directives for computer entry of drug orders, but this is not always compatible with a desire to maximize time at the bedside in the face of work-hour limitations. Likewise, the implementation of protocols and pathways that have ­provided higher quality of care may pose a threat to the concept of applying and individualizing basic and clinical science at the bedside. And while resident workhour restrictions have been promulgated as a measure to improve patient safety, compliance with these new rules requires even greater reliance on patient handoffs – a class of events believed to be associated with risks of their own.3,4 While numerous other challenges abound for those involved in resident education, the question to be dealt with here is whether the implementation of medical emergency teams or rapid response systems – a classically patient-centered intervention – interferes with medical education, or whether there are ways in which medical education can be enhanced through the existence and operation of a medical emergency team.

G.K. Lighthall (*) Anesthesia and Critical Care, Department of Anesthesia, Stanford University School of Medicine, Palo Alto, CA 94305, USA e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concept and Implementation, DOI 10.1007/978-0-387-92853-1_31, © Springer Science+Business Media, LLC 2011

347

348

G.K. Lighthall

Origins of Rapid Response Systems: A Solution to a Real Problem Medical emergency teams and rapid response systems (METs/RRSs) were created in response to a number of studies examining the quality of care and clinical decision making in patients who experienced cardiopulmonary arrest, or who had unplanned admission to an intensive care unit. The studies were notable for demonstrating great heterogeneity in quality of care, and, in particular, the widespread finding of care that was inadequate. Studies evaluating patterns of ward care prior to ICU admission show a general lack of time urgency in evaluating and treating patients with abnormal vital signs and other forms of deterioration.5–7 Patients initially admitted to hospital wards (as opposed to ICU) had up to a fourfold increased risk of mortality, suggesting that the nature of the care was a more significant determinant of ultimate clinical trajectory than the admitting diagnosis.6 In a study done by McQuillan et al., patients considered to have “suboptimal” care, had twice the ICU mortality rate of the other groups.8 Areas considered problematic were delays in admission, and improper attention to oxygen therapy, airway, breathing, circulation, and monitoring. Reasons underlying suboptimal care were “failure or organization, lack of knowledge, failure to appreciate clinical urgency, lack of experience, supervision, and failure to seek advice.” Our own experience in examining dynamic decision making of house staff in a fully-simulated ICU revealed similar classes of deficiencies, including non-adherence to established protocols.9 Studies of antecedents to cardiac arrest demonstrated that between 75 and 85% of the affected patients had some form of deterioration in the hours prior to the arrest.5,10 Nearly a third of such abnormalities persisted for more than 24 h prior to arrest, with a population mean of 6.5 h.5 In one series, the vast majority (76%) of the disease processes eventually progressing to cardiac arrest were not considered to be intrinsically rapidly fatal.10 In another series, over half of the arrests presented ample warning of decompensation; the majority had uncorrected hypotension, and half of these had systolic blood pressures less than 80 mmHg for more than 24  h.11 Other patients in this series had severe, but correctable abnormalities such as hypokalemia, hypoglycemia and hypoxemia. Problems with establishing proper care were found to exist at multiple levels: nurses were not calling physicians for patients with abnormal vital signs or changes in sensorum; physicians did not fully evaluate abnormalities when they were contacted; ICU consultants were not called in routinely; and even senior-level or consulting ICU caregivers did not obtain routine laboratory studies that would have defined the patient’s problem. In cases when laboratory studies were done, they were not always interpreted correctly, and when they were, therapy was not always initiated.7 All of the aforementioned studies were conducted in academic centers where junior team members are traditionally called to evaluate a patient and there is a graded engagement of more senior staff members. Loss of valuable time in

31  Resident Training and RRSs

349

patient evaluation and stabilization may have been further compounded by attending staff that lack knowledge of seriously ill patients and their problems, and who lacked the skills to direct an appropriate resuscitation.8,12 Further, teaching hospitals have also increased their reliance on cross-coverage schemes, which, too, have been associated with an increased incidence of potentially preventable adverse events.13 The collective experience suggests that the quality of care, more than the disease, may be responsible for morbidity and mortality in some patients. Inattention to, or unawareness of a developing serious problem causes the additional problems of hasty decision making at the time of cardiac arrest. It was with these concerns in mind that the creation of METs to mobilize key personnel earlier in the evolution of critical illness were inspired.

Concerns Over Implementing Medical Emergency Teams and Rapid Response Systems Acceptance and sustained success of rapid response systems has historically involved the cooperation and support of all major departments and services involved in patient care. As METs are typically an ICU outreach effort, a recurrent concern over their implementation is whether their activity will deprive trainees of valuable patient care experience. Below, we will describe how the RRS is, at its core, more about restructuring resource delivery than about radical change. Will an RRS interfere with resident training, or more specifically, deprive the junior-level trainee with practical problem-solving experience? Medical training is structured around the concept of graded increases in responsibility, concurrent decreases in the need for supervision, and a belief in learning from experience, including mistakes. Given this tradition, it is reasonable to consider whether the RRS approach to managing unstable patients may somehow undermine the fabric and outcome of medical education by depriving trainees of the ability to make medical decisions and to learn from such. We have serious doubts concerning the efficacy of the current system. If it were truly reliable, appropriate feedbacks would be in place and assure that delays in care would be minimized over time and that outcomes would be improved. This is clearly not the case. Instead, the status quo functions under the assumption that each trainee has an unwritten right to work through the management of severe medical problems on his or her own. Unfortunately, management of these illnesses is time-sensitive, and the process of unsupervised discovery and learning compromises patient outcome. More and more, educators are recognizing that having experts on hand to recognize and correct errors facilitates competence and learning to a greater extent than allowing unrecognized errors to play out.14,15 Training is being redesigned to meet such challenges.16 With advancing medical specialization, one frequently sees interdepartmental cooperation at the bedside, where the care-giving team is happy to acquiesce

350

G.K. Lighthall

s­ ignificant portions of a patient’s management to the cardiologist, nephrologist, ­surgeon or oncologist. Indeed, the latter specialties possess certain types of ­technology, protocols and detailed knowledge of pathophysiology that can be pivotal to the optimal management and survival of a patient. In the context of specific organbased derangements, it is rational to seek this type of support, and most do so ­willingly, without questioning the loss of educational benefit. In fact, most specialty services offer trainee rotations knowing that they carry great educational value that is maximized by interaction with experts in the field. The RRS is an analogous attempt on the part of the critical care community to educate others to the presence of critical illness outside of the ICU, and to bring the necessary resources to the patient at a time when he or she is likely to benefit.17 Additionally, there are patients with illnesses that are not likely to respond to intensive care; receiving the input of critical care professionals has allowed many to institute comfort measures with greater confidence.

Opportunities for Resident Involvement in METs/RRSs The subspecialty model for resident education and mentorship is certainly an appropriate starting point to discuss how to build a training mission into the RRS without compromising efficient patient care. There is plenty of flexibility in RRS design, with some, but not all containing junior trainees. There is no evidence suggesting the superiority of one design over another. If adequately staffed with attending or fellow-level staff, there should be plenty of room for resident involvement. Some teams include the ICU resident, and others have used a senior ward resident. A number of different means for incorporating trainees into RRS function have been developed, and include the following: Team model Intern + ward attending Senior float resident on MET ICU trainee (resident/fellow) on MET

Supervision Internal medicine (hospitalist) faculty ICU faculty, fellow ICU faculty

Additionally, the common denominator of nearly all rapid response systems in academic centers is a concurrent summons of the primary ward team.18–20 From a training perspective, having the primary team work with the RRS not only maintains their exposure to interesting cases and their proper management, but also respects the fact that the best solutions to patient problems involve sharing of information and establishing a consensus regarding appropriate goals of care. Additionally, cross-discipline training is also achieved when trainees from one discipline are called to care for patients admitted to other disciplines. Despite a great deal of resistance during the initial phase of MET operation, ­resident acceptance of its rationale and educational benefit has been surprisingly high (Fig. 31.1). Reported levels of cooperation between the primary teams and the MET have been equally encouraging.

31  Resident Training and RRSs

351

Fig. 31.1  Residents’ perceptions of the impact of rapid response systems. Mean responses ± SE are shown for a confidential Web-based survey of internal medicine residents. Residents were asked a series of questions including (1) whether they thought RRSs improved patient care; (2) whether the RRS had an adverse or positive impact on their education; and (3) to estimate the level of cooperation between the MET and the primary team. Survey questions were anchored to a 1–10 scale where 1 represented the lowest or most negative response, 5 was listed as neutral, and 10 was the most positive or highest level of agreement. Twenty surveys were completed, representing a response rate of 33%

When the first edition of this book was published, this chapter intended to “sell” the idea of the MET/RRS as a way of improving patient safety, and to facilitate RRS implementation by persuading training programs that their educational missions would not be compromised. At this writing, training programs are probably no longer in a position to successfully resist RRS implementation. Accreditation bodies, such as the Joint Commission and others, have changed the dialogue from whether an RRS should be implemented or not, to what it should look like. Departments that are interested in mining the educational potential of this type of system are encouraged to think creatively about how their trainees can be involved, and to take a leadership role in creating systems that fulfill such needs. Periodic surveys of trainees’ attitudes toward the RRS will help assure that the system continue to meet both educational and patient safety goals.21,22

A Win–Win Situation Many of the changes in healthcare delivery associated with medical emergency teams are likely to facilitate the development and assessment of competencies now required by the US Accreditation Council of Graduate Medical Education (ACGME) and the American Board of Medical Specialties. As part of the Outcomes Project, residency review committees of the ACGME have required programs to develop methods for residents to attain and document competency in six different areas2:

352

• • • • • •

G.K. Lighthall

Patient care Medical knowledge Practice-based learning and improvement Interpersonal and communication skills Professionalism, and System-based practice

Accordingly, residencies must define specific skills or knowledge that define each competency, and provide training experiences to attain those objectives. Viewed as a whole, the ACGME not only expects trainees to assimilate medical knowledge in specific areas, but also to develop a more dynamic style of practice. For application of medical knowledge, this means the trainee must be able to understand the natural history of a certain disease or finding, evaluate medical literature, critically evaluate the contexts of a given study with the context of the current patient, and from this, synthesize a treatment plan. Finally, the resident is also required to attain some understanding of healthcare systems in general, and how their economics and structure impact on the delivery of patient care.

What a Rapid Response System Can Teach Residents About Patient Safety Emergency ward care also provides a microcosm of healthcare in which other competencies can be mastered. “Practice-based learning” competency requires not only evaluation of medical evidence in the literature, but improvement in care based on the analysis of experience using a systematic methodology. As part of “systemsbased practice,” residents are required to “effectively call on system resources to provide care that is of optimal value.”2 The existence of an RRS in a given institution certainly reflects its commitment to self-examination and its willingness to rearrange resources in order to create more efficient care and better outcomes for patients – but what is really learned from this? Some direct learning objectives can and should be built into the operation of the RRS, but a trainee will also benefit from the culture of an institution that encourages: • Doing things the right way – placing a premium on patient safety over training. • Understanding the connection between events and errors, and how sources of errors are analyzed and how change comes from this. • Understanding how patient safety overlays the patient-system, doctor–patient, and doctor–system relationships and how institutional, cultural and individual change has come from this focus. • Multidisciplinary teams and teamwork. While the overall intent of an RRS is to provide immediate service to patients in need, an educational component can obviously be designed. First, the RRS should

31  Resident Training and RRSs

353

engage in research that monitors patient outcomes and assures that the calling ­criteria for the RRS is properly aligned to the target patient population. Data from cardiac arrests, ICU admissions, quality assurance projects, and disease-specific treatments can be analyzed to understand patient care characteristics, and gaps that need to be filled. Second, the RRS should have routine debriefings (immediately after the event being the best time) of recent efforts with an emphasis on self-­critique, team dynamics, and error analysis. Error analysis can provide important insight into whether improvements need to be made in technical training, teamwork, or organizational structure.23,24 The overall handling and discussion of errors and “near misses” is likely to increase the extent of error reporting if done in a proper manner. A system that is non-punitive and that rewards the admission of mistakes creates a healthy climate in which there is a greater likelihood of ­identifying and correcting “latent” sources of future errors.23 Residents exposed to this “culture of safety” may let this philosophy shape behavior and improve patient care in other settings. Overall, critical care outreach provides outstanding opportunities to understand the interaction between individual needs and disease processes, care delivery systems, available resources, care models, and patient outcomes. As a system that optimally includes data collection and analysis and evaluation of disease and event patterns, the RRS provides a wealth of opportunities to explore the socio-medical interactions stressed by the ACGME competencies – and probably provides opportunities that might not have existed prior to the RRS! It is important to note that the RRS is not a panacea. Avoidable cardiac arrests and late calls continue to exist in even the most seasoned efforts,25,26 so the need to understand and improve systems is a constant need. Trainees may also be interested in using elective time to participate in some of the data capture and analysis functions of the RRS to improve their efficacy.27 Valuable electives on patient safety or crisis management can be structured around research, education, data analysis, and team participation.

Summary The need to train and develop residents for independent practice may conflict with the needs of patients who require rapid stabilization. Finding a healthy balance between the two needs, and one where the patient receives the best possible care is a challenge to those in academic medicine. An increasing abundance of data suggests that the traditional models of resident training have in part failed to place the patient first. The contributors to this text believe that the implementation of Rapid Response Systems will offset some of these shortcomings in care and improve the public accountability of medical education. In this chapter, there is a description of how the RRS can synergize with resident training goals, and even enhance the acquisition of system-based care competencies for both trainees and established physicians. Likewise, the RRS can provide an educational structure from which the house staff can learn a great deal more about interdisciplinary teamwork, patient safety, and the responsiveness of healthcare to patient needs.

354

G.K. Lighthall

References 1. Corrigan JM, Kohn LT, Donaldson MS, eds. United States Institute of Medicine, To Err is Human – Building a Safer Health System. Washington, DC: National Academy Press; 2000. 2. Accreditation Council for Graduate Medical Education, Outcome Project. http://www.acgme. org/outcome/project/OPintrorev1 7-05.ppt; Accessed 2.10.09. 3. Cook RI, Render M, Woods DD. Gaps in the continuity of care and progress on patient safety. BMJ. 2000;320(7237):791–794. 4. Harrison M, Eardley W, McCarron B. Time to hand over our old way of working? Hosp Med. 2005;66(7):399–400. 5. Buist MD, Jarmolowski E, Burton PR, Bernard SA, Waxman BP, Anderson J. Recognising clinical instability in hospital patients before cardiac arrest or unplanned admission to intensive care. A pilot study in a tertiary-care hospital. Med J Aust. 1999;171(1):22–25. 6. Sax FL, Charlson ME. Medical patients at high risk for catastrophic deterioration. Crit Care Med. 1987;15(5):510–515. 7. Franklin C, Mathew J. Developing strategies to prevent in-hospital cardiac arrest: analyzing responses of physicians and nurses in the hours before the event. Crit Care Med. 1994;22(2):244–247. 8. McQuillan P, Pilkington S, Allan A, Taylor B, Short A, Morgan G. Confidential inquiry into quality of care before admission to intensive care. BMJ. 1998;316(7148):1853–1858. 9. Lighthall GK, Barr J, Howard SK, et al. Use of a fully simulated intensive care unit environment for critical event management training for internal medicine residents. Crit Care Med. 2003;31(10):2437–2443. 10. Schein RM, Hazday N, Pena M, Ruben BH, Sprung CL. Clinical antecedents to in-hospital cardiopulmonary arrest. Chest. 1990;98(6):1388–1392. 11. McGloin H, Adam SK, Singer M. Unexpected deaths and referrals to intensive care of patients on general wards. Are some cases potentially avoidable? J R Coll Physicians Lond. 1999;33(3):255–259. 12. Hillman KM, Bristow PJ, Chey T, et  al. Duration of life-threatening antecedents prior to intensive care admission. Intensive Care Med. 2002;28(11):1629–1634. 13. Petersen LA, Brennan TA, O’Neil AC, Cook EF, Lee TH. Does house staff discontinuity of care increase the risk for preventable adverse events? Ann Intern Med. 1994;121(11):866–872. 14. Wu AW, Wu AW, McPhee SJ, Lo B. Do house officers learn from their mistakes? JAMA. 1991;265(16):2089–2094. 15. Iqbal Y. While teams may reduce the rates of codes and mortality, house staff are still finding out where they fit in. ACP Observer; March, 2006. 16. Mcmahon GT, Katz JT, Thorndike ME, Levy BD, Loscalzo J. Evaluation of a redesign initiative in an internal medicine residency N Engl J Med. 2010;362:1304–1311. 17. Hillman K. Critical care without walls. Curr Opin Crit Care. 2002;8(6):594–599. 18. Bellomo R, Goldsmith D, Uchino S, et al. A prospective before-and-after trial of a medical emergency team. Med J Aust. 2003;179(6):283–287. 19. Bristow PJ, Hillman KM, Chey T, et al. Rates of in-hospital arrests, deaths and intensive care admissions: the effect of a medical emergency team. Med J Aust. 2000;173(5):236–240. 20. Goldhill DR, Worthington L, Mulcahy A, Tarling M, Sumner A. The patient-at-risk team: identifying and managing seriously ill ward patients. Anaesthesia. 1999;54(9):853–860. 21. The Joint Commission. National patient safety goals. http://www.jcrinc.com/common/PDFs/ fpdfs/pubs/pdfs/JCReqs/JCP-07-08-S1.pdf; 2009 Accessed 02.02.09. 22. Institute for Healthcare Improvement. Protecting 5 million lives from harm. http://www.ihi. org/IHI/Programs/Campaign; Accessed December, 2008. 23. Helmreich RL. On error management: lessons from aviation. BMJ. 2000;320(7237):781–785. 24. Braithwaite RS, DeVita MA, Mahidhara R, et  al. Use of medical emergency team (MET) responses to detect medical errors. Qual Saf Health Care. 2004;13(4):255–259.

31  Resident Training and RRSs

355

25. Galhotra S, DeVita MA, Simmons RL, Dew MA, Members of the Medical Emergency Response Improvement Team (MERIT) Committee. Mature rapid response system and potentially avoidable cardiopulmonary arrests in hospital. Qual Saf Health Care. 2007;16(4):260–265. 26. Downey AW, Quach JL, Haase M, Haase-Fielitz A, Jones D, Bellomo R. Characteristics and outcomes of patients receiving a medical emergency team review for acute change in conscious state or arrhythmias. Crit Care Med. 2008;36(2):477–481. 27. Devita MA, Bellomo R, Hillman K, et al. Findings of the first consensus conference on medical emergency teams. Crit Care Med. 2006;34(9):2463–2478.

Chapter 32

Optimizing RRSs Through Simulation Melinda Fiedor Hamilton, Elizabeth A. Hunt, and Michael A. DeVita

Keywords  Teaching • Organized • Crisis • Team • Functioning • Using • Human • Simulators

Introduction A recent report by the National Registry of Cardiopulmonary Resuscitation (NRCPR) of 207 hospitals within the U.S. revealed that the majority (86%) has an organized team to respond to in-hospital cardiac arrests.1 Despite the existence of these teams, there is mounting evidence that errors in the management of care for patients with in-hospital cardiac arrests and other medical crises may contribute to poor outcomes.2–8 Currently, no standards exist in terms of how such “code teams” are dispatched, how many members are on the team, or the team’s composition. There are even fewer reports regarding the make-up of Medical Emergency Teams (METs). Training to enhance the quality of care delivered by crisis teams in hospitals is essential. Although the composition of these two types of hospital teams varies from place to place, the principles of team training remain the same, and are reviewed in this chapter. The word “team” typically refers to a group of people that work together on a regular basis, in a coordinated and coherent fashion. Professional athletic teams provide an example of how a classic team functions: team members all have a common goal of winning their athletic events; they practice together regularly; individuals typically have one designated role or position that they play, and in which they become a true expert; team members often develop some type of shorthand to aid in communication; and the team typically functions best with a good team leader or captain.

M.F. Hamilton (*) Department of Critical Care Medicine, University of Pittsburgh Medical Center Director, Pediatric Simulation, Peter M. Winter Institute for Simulation, Education, and Research (WISER) and Children’s Hospital of Pittsburgh Simulation Center, Pittsburgh, PA, USA e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concept and Implementation, DOI 10.1007/978-0-387-92853-1_32, © Springer Science+Business Media, LLC 2011

357

358

M.F. Hamilton et al.

Unique Aspects of Hospital Crisis Teams Hospital crisis teams serve a critical purpose: to prevent death in suddenly critically ill patients. To succeed, they must function effectively and efficiently. Delays in action, miscommunication, and errors can increase the likelihood of a fatal outcome for the patient. Such a critical function should be the target of frequent and effective training programs, but, ironically, the training of crisis teams is particularly difficult because of the dynamic nature of the teams. In many ways, crisis teams represent the antithesis of the well-trained athletic team and must overcome major barriers to function effectively.

The Ad Hoc Nature of Crisis Teams Code teams are ad hoc teams – they consist of people who are brought together in a crisis, although they may never have worked together previously. Once the crisis is over, they return to their other activities and may not work together again. It seems impossible to train all of the possible combinations of a code team. A study by Pittman et al of a cardiac arrest team revealed that 67% of team leaders had had no communication with the team members prior to the cardiac arrest event;9 33% had “informal” communication prior, and only 7% had a debriefing session after the cardiac arrest.9 The ad hoc nature of crisis teams makes it difficult to practice communication, organization, group problem-solving, and the integrated functioning skills necessary for teamwork. The difference between a team that practices until they function like a well-oiled machine and the team whose members have literally never met becomes obvious when video recordings of crisis events are reviewed. Crisis team training programs must directly address the fact that the team that assembles for a crisis may not have worked together previously.

Simulation of Crises as Diagnostic Tool Sullivan and Guyatt published one of the first studies of the use of cardiac arrest simulations in the hospital setting to identify deficiencies in the crisis team response.4 They concluded: “Mock arrests are an extraordinarily powerful means of revealing suspected and unsuspected inadequacies in resuscitation procedure and equipment, and in motivating physicians and administrators to correct the deficiencies rapidly”.4 Subsequent work by other teams has similarly revealed that simulation can be a powerful diagnostic tool in revealing inadequacies in a hospital’s crisis response.5–8 The mock code work by Dongilli et al provides an example of applying crisis team training principles successfully as a diagnostic tool.6 Mock codes were run over several months in an adult hospital that is part of a 22-hospital health institution.6 The aim was to evaluate the first responders and the elapsed time until

32  Optimizing RRSs Through Simulation

359

a­ ppropriate resuscitation maneuvers. Measurements included elapsed time: time to call the operator about the crisis, time to send a voice page to the code team, time to code pagers’ tones, and time to first-responder arrival.6 Analysis of the results revealed that it took up to 4 min for the first responder to arrive at the crisis.6 In reviewing the hospitals’ operator notification procedure, it was discovered that the operators received the call, entered information into the paging system, triggered the belt paging system (which requires 3 min to deliver the page alert), and then paged the code team on the overhead speaker system.6 This process caused an unnecessary delay in the time to the first responder’s arrival to the crisis. The procedure was changed, and the operators were instructed to voice-page the code team immediately after receiving the call and to set off the pagers afterwards.6 After this protocol change, time to arrival of the first responder at the next mock code was improved to 1 min and 46 s, and has reliably remained around 90 s since.6 Hunt performed a series of 34 surprise, multidisciplinary mock codes over a 3-year period at three children’s hospitals.8 The mock codes consisted of scenarios of pediatric respiratory distress, respiratory arrest, or cardiopulmonary arrest.8 Evaluation revealed delays to the assessment of airway, breathing, circulation, administration of oxygen, initiation of chest compressions, and decision to defibrillate, as well as errors in leadership and communication.8 Hunt concluded that the study identified “targets for educational interventions to improve pediatric cardiopulmonary resuscitation and, ideally, outcomes”.8 Crisis team training should focus on organizational skills, not on medical and nursing assessment and treatment skills. Required elements in a crisis team training session include well-written, simulated resuscitation scenarios and skillful debriefing that focuses on organization. General principles to follow for crisis team training scenarios include using life-like situations, having a specific learning objective or objectives for each scenario (or, more accurately, each debriefing), and incorporating team quality improvement goals. In addition, real-life errors can be replicated to train teams to avoid similar mistakes. Finally, scenarios should be focused: they can be as short as 1 min if the learning objective is to see how long it takes a person to recognize a crisis, request an MET response, and have the first responder arrive. On the other hand, if the goal is to focus on diagnosis and triage to an appropriate ICU, the scenario may require as long as 20 min. In any case, one should not expect to use a simulator and suddenly be able to train teams. Team training requires an organized curriculum and effective teaching directed at achieving specific behaviors. This chapter will address some key elements in developing a crisis team training curriculum.

What to Teach The first step in developing a curriculum for crisis team training is to create clearly defined educational objectives for the session. Although these objectives will vary slightly at each institution, based on identified deficiencies, the training must

360

M.F. Hamilton et al.

openly acknowledge the unique nature of crisis teams and provide a road map to enable the ad hoc group to function well together. The following principles will invariably need to be covered: the team’s goals, designated role assignment, communication, and leadership.

Goals of Crisis Response Teams For crisis teams to function in an organized manner, they must devote time and effort to organization at the beginning of a crisis response. In our opinion, the major reason that crisis team responses tend to be disorganized is that individual members jump into medical and nursing actions prior to organizing the team. Teams are more likely to function well if the individuals have common goals, and they coordinate their efforts. Once organization has occurred, then medical and nursing interventions will proceed more efficiently. The first and most important goal of the team is to deliver effective and efficient basic life support throughout the entire episode. This includes assessment of airway, breathing, and circulation, and, if necessary, rapid initiation of bag-mask ventilation and chest compressions, rapid defibrillation, and frequent reassessment. If the patient is in a shockable rhythm, the specific goal should be for the first shock to be delivered within 3 min of patient collapse. The second goal is the effective and efficient delivery of appropriate advanced life support, including diagnosis of the underlying problem and delivery of definitive care. Finally, appropriate triage must occur. Merely making team members aware of their overall goals and of the time intervals by which a specific resuscitation maneuver should be performed likely will be associated with better performance.10 In addition to the team goals, each individual member should be aware of the goals for his specific role, as each specific job, if done appropriately, will help meet the overall goals.

Designated Roles: Assignment and Definition When a crisis occurs in the hospital, the worst-case scenario is to have no plan for dealing with it. This is very unlikely in the twenty-first century. However, it is surprisingly common for a hospital to make a concerted effort to determine who will carry code beepers and how they will be activated, and yet neglect to plan who will perform which resuscitation job, or even what the jobs are. Unfortunately, this often results in important jobs or resuscitation tasks being left undone – for example, a delay of the performance of chest compressions or placement of intravenous access. The key to designing a crisis team response is to determine the specific roles and corresponding responsibilities that are desired, and then designate them ahead of time, during training. There are two effective approaches to role assignment. The first

32  Optimizing RRSs Through Simulation

361

is to have very clear roles that need to be assumed; often it does not matter who assumes which role, only that each role is filled. The second approach is teach each person who carries the code beeper what his expected specific role will be when he arrives, so he knows what he should do. If that role is filled while another remains open, the responders must recognize that circumstance and fill the empty role. Failure to do this may result in failure to perform key tasks and lead to patient harm. Each institution must also determine how its team will be structured – that is, who will participate in the response. Organizing the team and choosing roles go hand in hand. One must be familiar with available personnel for a crisis team response and choose roles accordingly. For example, one institution may only have four available crisis team responders, while another may have ten. The latter team will be able to expand the number of available roles, i.e., having two crash cart managers instead of one. Once it is clear how many people (and from which disciplines) are available to the team, assignment of the responsibilities for each role is the next step. Explicitly designating roles and responsibilities will remove the ambiguity that contributes to the chaos often seen during a crisis situation. At the University of Pittsburgh, we have identified eight roles that are needed for every crisis team response. We initially used six members, but found repeated specific errors during training. We adjusted the team composition and added members until the training teams reliably succeeded in meeting pre-set task-completion goals. We also decided that it is important that all team members be trained to assume any one of several roles that they are capable of performing. This crosstraining improves the team’s flexibility and the understanding of the roles that other team members play, and failure to cross-train leads to errors of omission. For example, if a team has no anesthesiologists or intensivists, it is common that no one manages the airway; instead, they wait for the airway expert to arrive, even though team members often possess sufficient skills to manage the airway until an expert arrives. We acknowledge that some hospitals may not have eight responders ­available for a team response. In that case, we suggest that responders take on more than one role. For example, one person may take on both the airway manager and airway assistant roles. The use of roles to “bundle” responsibilities can help ease the allocation of tasks to team members. All members of the Medical Emergency Team – including residents, fellows, attending physicians, critical care nurses, respiratory therapists, and pharmacists – take part in crisis team training sessions at the WISER Simulation Center. These full-scale, human simulation training sessions have allowed us to analyze team function and have been the impetus for changes in team structure. At the beginning of the team training sessions, MET members who have usually never worked together during a medical crisis are required to complete an online didactic program. Upon arrival at the simulation center, the facilitator reviews key concepts of the didactic, and orients the group to the simulator. They then participate in a simulated crisis scenario. We currently have eight scenarios (including a “null” scenario in which the simulated patient is not in a crisis). We have observed that the first attempt by the team is invariably chaotic, and many important resuscitation tasks are either delayed or will not be performed at all. For example, after

362

M.F. Hamilton et al.

Fig. 32.1  Performance improvement of role-related tasks from the first through the third sessions of a human-simulator crisis team-training program

training over 500 Advanced Cardiac Life-Support trained individuals, only one team (usually 16–20 people participate in a crisis team training course) has successfully defibrillated ventricular fibrillation in under 3  min in their first simulator ­session; by the end of the training program, virtually all teams are successful. (See Fig. 32.1 for crisis task performance and Fig. 32.2 for simulated survival during the training sessions.) During the training session, crisis team members familiarize themselves with all goals and the roles that they are individually capable of performing; we ensure this by not allowing any person to play the same role twice during training. After debriefing in which participants determine whether they assumed all the roles of our response, and whether they completed all the tasks associated with each role, they then move on to more simulations. The team participates in four simulations, with debriefings to assess role assignment, task responsibility, and team interaction. Because participants take on different roles in each simulation, they develop an understanding of how the team begins to function more effectively and efficiently when each role is filled and the responsibilities clearly defined. This “roles and goals” approach teaches ad hoc teams to work well together, since no matter who arrives to the crisis response (or the order in which they arrive), they will be able to step into a role and perform the associated responsibilities. One of the key tasks during training is to teach the members important steps in effective management of a crisis team response (Table 32.1). The University of Pittsburgh has reclassified responsibilities at a crisis response to remove professional “tags” and promote the assumption of roles, regardless of which professionals arrive11 (Table 32.2 and Fig. 32.3). Training sessions are multidisciplinary, with nurses, respiratory therapists, and physicians all attending a single training session. The group becomes adept at filling roles according to their individual capabilities, and they begin to see how individuals can come together for

32  Optimizing RRSs Through Simulation

363

Fig. 32.2  Overall team task completion rate during the first, second, and third scenario session of a 3-h training program Table 32.1  Key lessons for crisis team training Organize the team Choose roles Identify and complete responsibilities Communicate data to relevant person Analyze data Diagnose patient Treat patient

the first time and function as a team. We have also discovered that mapping out positions for each crisis team member is vital; in this way, confusion is kept to a minimum and each team member can be in the proper position at the bedside to perform his responsibilities.11 Full-scale human simulation can be used to train pediatric crisis response teams as well. At the Johns Hopkins Children’s Center, we approach the team structure and function by training individuals on what their jobs will be upon arrival at every crisis they attend. We also avoid the term Pediatric Code Team, and have explicitly renamed our team the Pediatric Rapid Response Team. The reason is that the team is ideally called for pre-arrest medical crises, and prior to any cardiac arrest or “code” occurs. (The medical literature would call for our team being named a Medical Emergency Team rather than a Rapid Response

364

M.F. Hamilton et al.

Table 32.2  Roles and goals of crisis team members Roles Goals Airway manager (#1 in Fig. 32.3) Manages ventilation and oxygenation, intubates if necessary Airway assistant (#2) Provides equipment to airway manager, assists with bag-mask ventilation Bedside assistant (#3) Provides patient information including AMPLE*, medications delivery Equipment manager (#4) Draws up medications, supplies crash cart contents to appropriate team members Data manager/recorder (#8) Records vital signs, exam findings, test results, chart Circulation assistant (#6) Evaluates pulses, performs chest compressions Procedure MD (#7) Performs procedures such as central lines, chest tubes, pulse check Treatment leader (#5) Analyzes data, diagnoses, and directs patient treatment *Allergies, Medications, Past history, Last meal, and Event-what happened.

Performance improvement of role related tasks during a crisis team training program Task Completion % 100%

Completion %

80%

60%

40%

20%

0%

S1

S2

Session #

S3

Airway Manager

Airway Assistant

Bedside Assistant

Treatment Leader

Data Manager

Procedure MD

S4 Crash Cart Manager Criculation

Aide

Fig. 32.3  Graphic representation of team roles and goals. Numbers correspond to roles listed in Table 32.2

32  Optimizing RRSs Through Simulation

365

Team because it is physician-led. However, in our organization, we want to make it clear that this team is for children, and so we utilize the term “pediatric,” and as the team is for both medical and surgical patients, we have avoided the term “medical.” Thus the full name at our facility is the “Pediatric Rapid Response Team”.) The team has ten members, in addition to the floor team, and each member has a specific role. The roles are defined in a hospital protocol, described to individuals during orientation and are practiced during various team-training sessions. For example, pediatric residents are trained to understand that during their internship year, any time they respond to a medical crisis their sole job will be to perform compressions if needed. The second-year pediatric residents are trained such that their job will be to defibrillate the patient if needed and otherwise assist with placement of vascular access. In addition, there are monthly resuscitation training sessions for the residents rotating through the wards. They participate in a series of short mock codes and take turns practicing: (1) identifying themselves as the leader; (2) fulfilling the roles they will be expected to perform during actual events; (3) communicating during crises; and (4) performing important resuscitation tasks such as CPR, bag-mask ventilation, defibrillation, and placement of intraosseous needles. Developing a method to ensure that every important resuscitation task will be completed at every crisis depends on creating: (1) clear role definitions; (2) a clear method for determining who will fill each role; and (3) training exercises to allow team members to practice filling the roles and completing the tasks as a cohesive unit.

Communication Training to ensure effective communication is the next piece to organize. Closedloop feedback communication is an important method; this term refers to the ­ process whereby a team member will speak to another member and the second member will confirm hearing the message and confirm when he or she has completed the assigned task. For example, the team leader says, “Jimmy, can you check for a pulse.” Jimmy then states, “I will check the pulse” and completes his job by saying “I still do not have a pulse.” The leader should then confirm receiving this information, “Okay, there is no pulse. Jimmy, please continue CPR.” Jimmy confirms, “I am continuing CPR.” The closed-loop method of communication serves to lessen chaos and to ensure safe and appropriate management of patient care. Not only should communication occur in a closed-loop manner, it should be aimed at the specific team member who requires the information, such as the team leader or the crash cart manager; use of first names is important and effective. The team leader can then analyze data provided by the team members via closed-loop communication, assimilating the information and diagnosing the patient, and then again via closed-loop communication give instruction about appropriate patient treatment.

366

M.F. Hamilton et al.

Simulation of cardiopulmonary arrests and medical crises allows medical e­ mergency team members to practice these communication skills. A particularly effective method is to allow code team members to visualize the effect of communication errors that occurred in the safe environment of the simulation center. For example, during a mock code of a patient in profound septic shock, a pediatric resident orders a 20 cc/kg bolus of normal saline, and when the blood pressure remains low, repeats the order 2 more times; at the end of the mock code, the resident believes the patient has received 60 cc/kg of normal saline. Upon discussion with the nurses, it becomes clear that they were almost finished preparing the first bolus, but none of the fluid had actually been delivered. They had not even heard the second or third orders, and did not realize that the normal saline was a priority, believing that the antibiotics were more important. Within a medical emergency team, there are “mini-teams” that have more specific functions, and communication channels must be developed within the group as well as among groups. The three mini-teams are the breathing team, the circulation and patient assessment team, and the diagnosis and treatment team. The breathing team consists of an airway manager and an assistant. They obviously must work closely together, and must have one-to-one communication independent of other communication. The circulation team is responsible for determining the presence and quality of circulation, and delivering both circulatory assistance and medications. They need to cross-check findings and coordinate tasks that might interfere with each other, like doing chest compressions, placing a central venous catheter, and placing defibrillator pads. Finally the diagnosis and treatment team must assemble all the data, make treatment decisions, and implement them. Recognition of the presence and role of these mini-teams can help improve communication. The goal of the emphasis on organization and communication is to foster a collective consciousness, in which team members coordinate to collect, transmit, and act on data as a group rather than as individuals. Such a goal has been described in the military for aircraft carrier flight crews,12 but it is as yet not well described in the Medical Emergency Team literature.

Leadership Multiple studies demonstrate that poor leadership during cardiopulmonary resuscitation efforts is associated with poor team function.3,13 Cooper and Wakelam observed actual cardiopulmonary arrests and used the Leadership Behavior Description Questionnaire.3 This study revealed that leaders who participated in tasks in a “hands-on” manner were “less likely to build a structured team, the teams were less dynamic, and the tasks of resuscitation were performed less effectively”.3 Marsch et al studied simulated cardiopulmonary resuscitation efforts and observed that “absence of leadership behavior and absence of explicit task distribution were associated with poor team performance”.13

32  Optimizing RRSs Through Simulation

367

While good leadership can help ensure good team functioning, a system must be in place to make sure that the team functions well even without an effective leader. For this reason, team members should be taught to assume their roles without needing to be reminded to do it. However, simulation of medical crises and cardiopulmonary arrests will allow the medical emergency team members to practice their leadership skills and actually see the effect of poor leadership. More importantly, training can allow leaders to practice until they can competently head a team. However, at the University of Pittsburgh we avoid using the “team leader” designation for several reasons. First, if there is a team leader, other participants hesitate to assume responsibilities until they are assigned; this in turn can lead to delays. Second, if the team leader is assigning roles and tasks, then the leader is not attending to the patient (he or she is caring for the team instead). Third, if the team leader responds late, key treatments may be delayed while people await the leader’s arrival. Additionally, a hierarchical role structure assumed during codes may impede communication, especially when the team leader is perceived to be wrong. We choose a “flat” hierarchy: if everyone knows his role and self-assigns, then the treatment leader can attend to diagnostics and appropriate therapeutics. The terminology we prefer is “treatment leader” for the person who designates treatments, and “data manager” for the person who ensures that all roles are assumed and obtains the data from each team member. The flat hierarchy aids communication and the cross-checking of data and treatment decisions. No study has yet been performed to determine what kind of team training results in best performance.

Debriefing Debriefing is an essential component of medical emergency team training. Marteau et al state that there is a “well described tendency to invoke competence after success but not question it after failure.” 14 Their data demonstrated that resuscitation experience without feedback increased confidence, but not competence.14 The principles of debriefing include timeliness, objectivity, specificity, and balance. Timeliness means that debriefing is most effective for the learner if it occurs immediately after the scenario. This ensures that the experience is fresh in the minds of the trainees, allowing for the greatest learning from feedback and selfassessment. Debriefing should be objective and specific. It should focus on particulars – not “You did a good job,” but “You appropriately applied oxygen to the patient within 30 s.” Debriefing can be made very specific using simulation and video recordings. Video review of medical emergency team performance using an objective evaluation instrument allows exact and detailed debriefing. Finally, debriefing should be balanced. Both successes and errors should be discussed. The key to successful debriefing is to be positive, even if an error is the learning point. The error should be noted, but the focus should be on correct actions. Prior to team training, it should be reinforced that errors will occur during

368

M.F. Hamilton et al.

the simulation scenarios just as they occur in the real world. Simulation, however, allows for errors to happen in a safe and educational environment. In the simulated setting, errors can be powerful instructional aids and provide motivation. At the University of Pittsburgh, we believe that it is difficult, if not impossible, to objectively debrief team members after a real crisis, especially if there is a bad outcome. First, they are not objective, balanced, and specific because there are usually no video and audio recordings upon which to base an objective assessment. Second, because a live person is involved, team members may have significant emotional or psychological responses to the event and the outcome that impair their ability to be objective or to learn.

What to Measure Medical emergency team training is important for education, patient safety, and research. For each of these areas, training is most effective if its components can be measured. Measurement can identify deficiencies and demonstrate improvements, if they occur. The key is to measure the correct outcome. If the aim of medical emergency team training is education, scenarios should be written with tasks centered on a specific educational goal, such as correct bag-mask ventilation. The measurement will ultimately be a dichotomous value, i.e., “yes” or “no,” as to whether the skill was performed effectively. The results help to identify areas on which to focus during debriefing. Institutions may use this technique of successful skill completion to qualify individuals for privileges in a hospital. For example, anesthesiologists may have to successfully demonstrate difficult airway maneuvers in order to receive privileges in their institution. If the goal of a medical emergency team training is patient safety, measurement could be of patient outcome. However, since many of the outcomes we seek to avoid are rare, it may not be reasonable to look solely at changes in the rates of these outcomes. A second approach can be to observe the adherence to desired procedures. For example, if simulation of airway management during a crisis increases the proportion of times that the team remembers to apply cricoid pressure during bag-mask ventilation, it is probable that aspiration events will be decreased. Research using medical emergency team training can focus on combinations of the above. Measurement of successful tasks completed by a team member, or the entire team, can be compared before and after team training, as well as over time. Studies by Gaba et al show that both technical and behavioral performance can be assessed via evaluation of videotapes of simulated crisis events.15 Their results show that cognition and crisis management behaviors vary considerably.15 This has been seen before and demonstrates the ability of simulation to be used as a “needs assessment tool.” Blum et al describe the development of an anesthesia crisis resource management course.16 The course objectives were to understand and improve participant skills

32  Optimizing RRSs Through Simulation

369

in crisis resource management and learn debriefing skills.16 Course usefulness, debriefing skills, and crisis resource management principles were highly rated by participants.16 It is interesting to note that course participants were eligible for malpractice premium reductions.16

Summary Data from both real and simulated events reveals that errors in the management of in-hospital cardiac arrests and other medical crises may contribute to poor patient outcomes.2–8 Successful medical emergency team training may improve the effectiveness and efficiency of these teams, and ultimately improve patient outcomes. Medical emergency team training should include clear delineation of the goals of the team, designated role assignments of crisis team members, communication training, and leadership training. The training can be successfully achieved using simulation in combination with well-written scenarios, skillful debriefing, and specific measurements of deficiencies and achievements related to training.

References 1. Peberdy MA, Kaye W, Ornato JP, et al. Cardiopulmonary resuscitation of adults in the hospital: a report of 14, 720 cardiac arrests from the National Registry of Cardiopulmonary Resuscitation. Resuscitation. 2003;58:297–308. 2. Abella BS, Alvarado JP, Myklebust H, et al. Quality of cardiopulmonary resuscitation during in-hospital cardiac arrest. JAMA. 2005;293:301–310. 3. Cooper S, Wakelam A. Leadership of resuscitation teams: “Lighthouse Leadership.”. Resuscitation. 1999;42:27–45. 4. Sullivan MJ, Guyatt GH. Simulated cardiac arrests for monitoring quality of in-hospital resuscitation. Lancet. 1986;2:618–620. 5. Palmisano JM, Akingbola OA, Moler FW, Custer JR. Simulated pediatric cardiopulmonary resuscitation: initial events and response times of a hospital arrest team. Respir Care. 1994;39:725–729. 6. Dongilli T, DeVita M, Schaefer J, Grbach W, Fiedor M. The use of simulation training in a large mult-ihospital health system to increase patient safety. Presented at International Meeting for Medical Simulation. Albuquerque, NM; 2004. 7. Misko L, Molle E. Beyond the classroom – teaching staff to manage cardiac arrest situations. J Nurses Staff Dev. 2003;19:292–296. 8. Hunt EA. Simulation of pediatric cardiopulmonary arrests: a report of mock codes performed over a 40-month period focused on assessing delays in important resuscitation maneuvers and types of errors. Presented at International Meeting for Medical Simulation. Albuquerque, NM; 2004. 9. Pittman J, Turner B, Gabbott DA. Communication between members of the cardiac arrest team – a postal survey. Resuscitation. 2001;49:175–177. 10. Kinney KG, Boyd SY, Simpson DE. Guidelines for appropriate in-hospital emergency team time management: the Brooke Army Medical Center approach. Resuscitation. 2004;60:33–38. 11. Fiedor M, DeVita M. Human simulation and crisis team training. In: Dunn WF, ed. Simulators in Critical Care and Beyond. Des Plaines, IL: Society of Critical Care Medicine; 2004:91–95.

370

M.F. Hamilton et al.

12. Weick KE, Roberts KH. Collective mind in organizations: heedful interrelating on flight decks. Adm Sci Q. 1993;38:357. 13. Marsch SCU, Muller C, Marquardt K, Conrad G, Tschan F, Hunziker PR. Human factors affect the quality of cardiopulmonary resuscitation in simulated cardiac arrests. Resuscitation. 2004;60:51–56. 14. Marteau TM, Wynne G, Kaye W, Evans TR. Resuscitation: experience without feedback increases confidence but not skill. BMJ. 1990;300:849–850. 15. Gaba DM, Howard SK, Flanagan B, Smith BE, Fish KJ, Botney R. Assessment of clinical performance during simulated crises using both technical and behavioral ratings. Anesthesiology. 1998;89:8–18. 16. Blum RH, Raemer DB, Carroll JS, Sunder N, Felstein DM, Cooper JB. Crisis resource management training for an anaesthesia faculty: a new approach to continuing education. Med Educ. 2004;38:45–55.

Chapter 33

Evaluating Effectiveness of Complex System Interventions Jack Chen

Keywords  Evaluating • Complex • System • Interventions Following the landmark reports from the Institute of Medicine (IOM)1,2 and other studies, numerous intervention programs have been introduced to improve patients’ safety.2–6 The evaluation of the impact of such complex system interventions continues to be a challenge. There is little robust evidence about the effectiveness of intervention programs aimed at system improvement.7 Rigorous evaluation of the systems we use for delivering healthcare presents a different challenge from the ones used for evaluating simple interventions such as a new drug or new medical technology.

Characteristics of Complex System Interventions There is no single or standard way of defining a complex system intervention. A complex system intervention in health care often requires changes in the structure, culture and organizational behavior of an institute, as well as changes in individual practices, all aimed at improving the quality of care.8,9 Most of the quality improvement programs also share some of these characteristics. Implementing complex system intervention is quite demanding. It may require systematic, multifaceted change strategy including building coalition, academic detailing, 10 timely feedback, 11 involving opinion leaders, 12 extensive and targeted educational activities, changing organizational structure such as team building, streamlining processes, and providing appropriate staffing.

J. Chen MBBS, PhD., MBA (*) The Simpson Centre for Health Services Research, Liverpool Health Service, Locked Bag 7103, Liverpool BC, Sydney, NSW 1871, Australia e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concept and Implementation, DOI 10.1007/978-0-387-92853-1_33, © Springer Science+Business Media, LLC 2011

371

372

J. Chen

However, it should be noted that some seemingly simple interventions may also have distinctive system components. For example, there were some system elements necessary to deliver the two different fluids in the SAFE study.13 This trial involved more than simply delivering a single drug at one point in time. Apart from the fluid type, the fluid regimens were largely determined by the individual hospitals and by individual clinicians within their institutions. Thus the results depended partly on the protocols or systems used for fluid delivery protocols for each patient as well as the two different fluids being compared. Analysis of the effectiveness then requires assessment of both the medications and the “system” for the fluid delivery. Evaluating complex systems is sometimes known as health services research, defined as “a multidisciplinary field of enquiry, both basic and applied, that examines the use, cost, quality, accessibility, delivery, organization, funding and outcomes of health care services to increase the knowledge and understanding of the structure, processes, and effects of health services for individuals and populations.”14 An alternative and less formal definition of health services research states that it provides a framework for exploring the processes and dynamics underlying complex health systems.15 Thus, this type of research may generate new knowledge about the barriers and enablers to establishing and maintaining best practice within complex systems.

Defining the Components of Complex System Evaluation We must define the objectives of the evaluation clearly. Are we satisfied with simply determining clinical outcomes of the intervention, or are we also interested in gaining new knowledge about the processes and their impact within complex systems? Health promotion experts have argued the importance of assessing all three aspects.16,17 The conceptual drawback of a focus only on outcome is the lack of any knowledge about why the intervention did or did not produce the expected results. Even if the intervention finally delivers what it promises, we still do not have knowledge of to what degree the intervention could have been improved. Focusing on the final outcomes may also overlook important questions, such as how the intervention worked and on what sub-groups the intervention worked best. Evaluation of the Medical Emergency Team (MET) intervention in the MERIT trial is a good example of complex system research.18,19 The selected clinical outcomes were unplanned admissions to the ICU, unexpected deaths, and cardiac arrests. However, if the implementation of the MET into a hospital does not produce these expected outcomes, how will we know whether the failures were due to factors such as inadequate implementation, or sub-optimal calling criteria, or a relative short study period, or the lack of critical care skills of the MET, or the tardiness of the MET response, or whether the basic concept of early intervention in serious illness just does not work? To understand the

33  Evaluating Effectiveness of Complex System Interventions

373

critical linkages and processes is perhaps as important as assessing the clinical outcomes (Fig. 33.1). Other organizational characteristics may also contribute to outcomes. For example, the value placed on building capacity for patient safety may have contributed to the effectiveness of the MET intervention.20,21 The nature of the hospital culture may also be an important variable. To complicate matters even more, the implementation of the RRS in itself results in cultural change, such as the need to modify the traditional balance of power between medical and nursing staff and the need to reconfigure interdisciplinary relationships. All of these factors may impact upon both the processes and outcomes of a complex system intervention such as the RRS. Therefore, we need to consider how many factors other than the simple outcomes can be taken into account when we plan complex system evaluation.

Valid calling criteria: with good

Thorough and adequate implementation of the Rapid Response System

Frequent observations of the criteria with high documentation rate as required clinically

Readiness to assess patients’ status against ‘worried’ RRS criteria

Willing to make a RRS call should the documented vital sign over the thrash-hold value of the criteria

Willing to make a RRS call if the patients’ overall status causes concern and make one ‘worried’

The MET or RRT arrives in a timely fashion and has appropriate knowledge, skills and expertise

Clinical improvement of the selected outcomes

Fig. 33.1  The critical links leading to successful RRS outcomes

374

J. Chen

Choosing the Appropriate Research Methodology We must decide which paradigms we wish to use for evaluation. For example, do we use quantitative or qualitative methods, or perhaps both? If we decide in favor of the latter, what is the sequence, what will the scope be, and how might we go about integrating the two methods? Within an evidence-based approach to medicine, we tend to classify research evidence based on the design and the features of the evaluation. The single, large randomized controlled trial (RCT) – preferably double-blinded – is the “gold” standard of excellence. In randomized controlled, trials the emphasis is on the design and the scale of the study. This is very different to a research paradigm based on organizational development and organizational learning.22 The evaluation of the organizational aspects of systems could also adopt the participatory research and action research approach.23,24 This paradigm will have less emphasis on the rigor of the design, and more emphasis on the development of an iterative and cyclical program with the potential for supporting a learning culture (such as the Plan-DoStudy-Act [PDSA] cycle). Researchers who lead and design research exploring organizations and complex systems need to base their thinking within a systems perspective. They need to focus on factors such as the interactions among subsystems, and the unexpected benefits or harms of an intervention that may go beyond the original expectations.25,26 RCTs are useful for comparing a simple intervention such as a new drug or procedure. Experimental and quasi-experimental designs have the highest internal validity for the purposes of complex system evaluations. Complex interventions such as a new system to evaluate early intervention in serious illness usually require complex research methodologies. An intensive care unit is a complex organization, and patient care can be influenced by many factors, such as standards of nursing care and staff morale within the unit. Some of these factors may be as important, or even more important, than a simple intervention such as a new drug. It is often impossible and unethical to adopt a simple RCT design27 for the evaluation of a system intervention. A system intervention often requires a cluster-randomized controlled trial (cRCT) design.28 However, there can be challenges in using such research methodologies, including: 1. Limited choices in recruiting a control arm. There may be difficulty in finding suitable controls for a cRCT, as the number of units are often relatively small, each with a unique set of characteristics. 2. Contamination (the control group may also adopt the intervention). It is often very difficult to prevent the control units from learning and mimicking the intervention applied to the experimental groups. This is especially true when the intervention has good face validity and is intuitively appealing. During the MERIT study intervention phase,19 we found that half of the cardiac arrest team calls were not related to a cardiac arrest or death but rather were typical METtype calls of a patient in physiological crisis. This may have contributed to the failure to demonstrate a difference between the control and intervention groups,

33  Evaluating Effectiveness of Complex System Interventions

375

as well as explain the unexpected improvement seen in the control group. A similar finding also occurred in the ANZ Feeding Guideline Trial.10 The study found significant changes in the physicians’ behaviors in feeding the patients in the interventions groups but not in the hospital discharge mortality rate. However, the data also showed that within patients from control-arm ICUs, 37.3% of them (95% CI: 28.1–49.6%) had been fed within 24  h after being admitted to the ICU. This was exactly what the evidence-based feeding guideline intended to achieve in the intervention group (60.8%: 95% CI: 45.7–80%). Despite other significant improvement in the process measures, the above two 95 CIs were overlapped, signifying an insignificant difference. This may also in part explain why the study failed to achieve significant improvement in the patient outcome (i.e., hospital discharge mortality). 3. Incomplete implementation of intervention. For a complex system intervention that targets structural, cultural and behavioral change, not all the units within the experimental group will embrace the changes with the same willingness and commitment. Skepticism, ineptness, and a dislike of the unknown may easily derail an otherwise great initiative. Furthermore, individuals may also bring their own understanding and value judgments into the recommended changes. The skill sets and the knowledge of individuals may also decide the way that they will bring about the required changes. In the MERIT study, close to half of the adverse events in the intervention (MET) arm in which patients demonstrated MET activating criteria did not result in an MET call. In the ANZ Feeding Guideline Trial, although the proportion of “getting fed within 24 h of admission to an ICU” is higher in the intervention group (60.8% compared to 37.3% in the control group), it also indicated that close to 40% of patients within the intervention group were not treated according to the best available evidence. 4. The need for a large sample size. Due to a large variation among units (e.g., ICUs and hospitals) in terms of their size and organizational characteristics, a large number of units is often required in the study in order to achieve the ability to generalize the study’s results. These attributes impact significantly on the power of a cRCT. The greater the variation in these characteristics, the greater the ­number of the hospitals needed to achieve sufficient statistical power to test the primary and secondary outcomes.29 Also, the existence of the possible “conta­ mination” in the control group will effectively dilute the power of a study and ­confound the results. 5. The limited information needed to accurately estimate the power of the study. As a result of the relatively high costs of cluster randomized studies, one frequently does not have reliable information on the variations or intra-class correlation coefficients (ICCs)30,31 from previous research. The power issue could become even more complicated, as there are often no reported crude ICCs (the ICCs calculation based on the whole sample or the control and the intervention group separately without controlling for other covariates.28–32) Even if such information may be available from other studies, the applicability of such ICCs to the current research settings is often questionable. As ICCs reflect certain organizational characteristics, they could be quite different among different healthcare systems

376

J. Chen

and settings.33 Moreover, the information for the conditional ICCs (after adjusting for the key confounding factors such as hospital status – teaching, urban or rural, in the multivariate analysis model) is even scarcer. If one mixes all of these hospitals, one may run the risk of comparing “apples with oranges.” Using conditional model (conditional on the key confounding factors) could reduce this problem. It is important to note that a study that is underpowered based on the raw ICCs may have adequate power based on the conditional ICCs. 6. Time frames: Internalization of system change is a function of time, and thus may often require a lengthy timeframe in order to succeed. For example, it took more than 10 years to demonstrate the effectiveness of trauma systems.34,35 This requirement translates into significant research costs. For large-scale evaluation projects, such costs may become prohibitive, especially as the interventions themselves are already expensive and cumbersome. Therefore, the conventional gold standard, a simple RCT, may be impossible or impractical.

Sub-system Interactions After a Complex System Intervention When an intervention is applied to a unit or system, there may be unintended or unwanted results stemming from the changes introduced. While the RRS was originally designed to reduce the incidence of cardiac arrest, unexpected deaths and unplanned ICU admissions, the introduction of such a system may change some aspects of the care in hospitals. For example, several studies explored the relationship between the introduction of a MET system and the issuing of the Not-ForResuscitation (NFR) Order by the team.36–40 A study from the MERIT trial showed that the incidence of patients being issued an NFR when attended by an MET team was ten times as high as the rate when attended by a cardiac arrest team.41 This raises the possibility that an MET team may initiate the process of the end-of-life discussion and advanced-care planning. Thus, it may act as a de facto palliative care team and could potentially prevent unnecessary resuscitations for some patients who are highly unlikely to benefit from CPR. Another study from the MERIT data also showed that the introduction of an MET system was associated with significant improvement in the documentation of vital signs at the study period.42

Cost and Cost-Effectiveness Implementation of complex system intervention requires leadership and resources. The cost-effectiveness of system interventions should be taken into consideration. It has been widely recognized that well-considered and coordinated interventions are the prerequisite of a successful system change. However, it is also important to know whether the organization can afford to implement and sustain an intervention

33  Evaluating Effectiveness of Complex System Interventions

377

such as the MET or RRT. This raises the issue of whether we should incorporate a health economic perspective into evaluation projects. We also need to consider the marginal gain when implementing protracted and extensive system change.43,44 For example, the ANZ feeding guideline study applied 18 different implementation strategies in order to achieve a group of recommendations. It may be useful to know what strategy or what combination of these 18 strategies contributed mostly to the observed improvement. Furthermore, are there any strategies that are redundant and that may be left out without significant reduction of the impact?

Interpreting Study Results of Complex System Interventions We briefly discussed the possible difficulties in designing and conducting a rigorous study evaluating complex system interventions. There is also the challenge of interpreting the results of an evaluation. 1. Absence of evidence is not evidence of absence. We deliberated on the difficulties in conducting an ideal trial; in having enough power to prove the primary hypothesis; in avoiding contamination and other confounding factors; in analyzing complex relationships; and in dealing with multifaceted outcomes. It is more likely than not that a complex system intervention may produce negative results. The word “negative” is often used to incorrectly connote no impact rather than use the term “inconclusive.” In particular, it is incorrect to interpret those inclusive results (such as inadequate powered results) as a negative result.45 2. Under what types of conditions could the intervention work? There are often multiple conflicting results from single-center studies. It may be misleading to propose either positive or negative conclusions on a single-center trial. Similarly, in complex interventions, it might be more meaningful to discuss results in terms of why and how such interventions may or may not work in certain types of conditions instead of a simple blanket conclusion. For example, using the published data from the MERIT study,19 we might explore the hypothesis that the effectiveness of implementation may be related to the initial incidence of adverse events and that there may be a higher benefit among hospitals that have a higher incidence of the adverse event at baseline. This could mean that for those hospitals, which already have successful quality-improvement programs, the net benefit of an RRS may be less significant. The MERIT study has shown that the improvement in outcomes after introduction of an MET is largely dependent on the successful implementation of such a system. In other words, as the proportion of emergency team calls (defined as the calls not associated with a cardiac arrest or death) increases, the rate of cardiac arrests and unexpected deaths decreases. This inverse relationship provides support for the notion that early response to acutely ill ward patients by an emergency team is desirable.46 3. The burden of approval and a Bayesian approach. In view of the evidence-based approach in evaluating complex system intervention, one has to consider the

378

J. Chen

d­ ifficulties in assessing the evidence base. The challenge of conducting an adequately powered study in generating level-one evidence has not only limited the available studies, but has also made meaningful interpretation of results difficult. Planning and conducting a new trial may take several years and the availability of the results may take even longer. In the meanwhile, for frontline clinicians, decision makers and the campaigners for the quality improvement, the decision as to whether to implement an intuitively appealing intervention, such as intervening early in the case of serious illness, remains difficult. In the meantime, researchers and clinicians may turn to other forms of evidence such as those based on the observational studies, process-control techniques and organizational research evidence, qualitative research, or expert based consensus. Other than a deterministic general view of the “right” or “wrong” regarding any particular choice, a Bayesian approach may be employed.47 This involves summarizing the current knowledge, taking into consideration the individual organization’s own commitment and capacity in instituting the proposed changes, coupled with their own PDSA model. This type of approach may be a better guide in deciding whether a proposed complex system intervention should be adopted.

Summary • Evaluating complex system intervention poses unique challenges both to the internal and external validities of the study findings. • Appropriate designing and conducting a cRCT is demanding. So are the interpretations of the results from such a study. These difficulties may stem from the complexities involving inadequate power of a study, contamination, incomplete implementation of intervention, multiplicity of the results (both planned and unexpected, primary and secondary) and different types of the results (i.e., process, impact and outcomes). • Unplanned and unexpected results are common after a complex system intervention and the consequences of such results should also be explored and understood. Sometimes, the question asked should be “Under what types of conditions might such an intervention work?” instead of “Should such an intervention work anywhere, and all the time?” • A broader evidence base in assessing the results of a complex system intervention should be embraced and a Bayesian approach adopted in considering whether an intervention be applied to a particular local setting.

References 1. Donaldson MS, Kohn LT, Corrigan J. To Err is Human. Building a Safer Health System. Washington, DC: National Academies Press; 2000.

33  Evaluating Effectiveness of Complex System Interventions

379

2. Institute of Medicine. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academies Press; 2001. 3. Aspden P, Institute of Medicine (Committee on Data Standards for Patient Safety). Patient Safety: Achieving a New Standard for Care. Washington, DC: National Academies Press; 2004. 4. Byers JF, White SV. Patient Safety: Principles and Practice. New York: Springer; 2004. 5. Child AP, Institute of Medicine (Committee on the Work Environment for Nurses and Patient Safety). Keeping Patients Safe: Transforming the Work Environment of Nurses. Washington, DC: National Academies Press; 2004. 6. United States, Congress, House, Committee on Energy and Commerce. Patient Safety and Quality Improvement Act Report (To Accompany H.R. 663) (Including Cost Estimate of the Congressional Budget Office). Washington, DC: U.S. G.P.O; 2003. 7. John O. Safety Deficiencies in Healthcare – A Review of Research. Stockholm: Karolinska Institute Medical Management Centre; 2004. 8. Grimshaw J, Eccles M, Thomas R, et  al. Toward evidence-based quality improvement. Evidence (and its limitations) of the effectiveness of guideline dissemination and implementation strategies 1966–1998. J Gen Intern Med. 2006;21(suppl 2):S14-S20. 9. Grimshaw JM, Thomas RE, MacLennan G, et  al. Effectiveness and efficiency of guideline dissemination and implementation strategies. Health Technol Assess. 2004;8:iii-iv. 1–72. 10. Doig GS, Simpson F, Finfer S, et al. Effect of evidence-based feeding guidelines on mortality of critically ill adults: a cluster randomised controlled trial. J Am Med Assoc. 2008;300:2731–2741. 11. Thomas RE, Croal BL, Ramsay C, Eccles M, Grimshaw J. Effect of enhanced feedback and brief educational reminder messages on laboratory test requesting in primary care: A cluster randomised trial. Lancet. 2006;367:1990–1996. 12. Doumit G, Gattellari M, Grimshaw J, O’Brien MA. Local opinion leaders: Effects on professional practice and health care outcomes. Cochrane Database Syst Rev. 2007;CD000125. 13. Finfer S, Bellomo R, Boyce N, French J, Myburgh J, Norton R. A comparison of albumin and saline for fluid resuscitation in the intensive care unit. N Engl J Med. 2004;350:2247–2256. 14. Institute of Medicine National Academy of Sciences. Health Services Research: Workforce and Educational Issues. Washington, DC: National Academies Press; 1995. 15. Lomas J. Health services research: A domain where disciplines and decision makers meet. In: Sibbald WJ, Bion JF, eds. Evaluating Critical Care. Using Health Services Research to Improve Quality. Berlin: Springer; 2000:6–19. 16. Green LW, Kreuter MW. Health Program Planning: An Educational and Ecological Approach. 4th ed. New York: McGraw-Hill; 2005. 17. Windsor RA. Evaluation of Health Promotion, Health Education, and Disease Prevention Programs. 3rd ed. Boston: McGraw-Hill; 2004. 18. Hillman K, Chen J, Brown D. A clinical model for health services research – the medical emergency team. J Crit Care. 2003;18:195-199. 19. Hillman K, Chen J, Cretikos M, et al. Introduction of medical emergency team (MET) system – a cluster-randomised controlled trial. Lancet. 2005;365:2091–2097. 20. Basanta WE. Changing the culture of patient safety and medical errors: A symposium introduction and overview. J Leg Med. 2003;24:1–6. 21. Cohen MM, Eustis MA, Gribbins RE. Changing the culture of patient safety: Leadership’s role in health care quality improvement. Jt Comm J Qual Saf. 2003;29:329–335. 22. Argyris C. On Organizational Learning. 2nd ed. Malden, MA: Blackwell; 1999. 23. Howard JK, Eckhardt SA. Action Research: A Guide for Library Media Specialists. Worthington, OH: Linworth; 2005. 24. Johnson AP. A Short Guide to Action Research. Boston: Pearson/Allyn and Bacon; 2005. 25. Haslett T. Implications of Systems Thinking for Research and Practice in Management. Caulfield East, VIC: Faculty of Business and Economics, Monash University; 1998. 26. Jackson MC. Systems Thinking: Creative Holism for Managers. Chichester: Wiley; 2002. 27. Donner A, Klar N. Design and Analysis of Cluster Randomization Trials in Health Research. London: Arnold; 2000.

380

J. Chen

28. Campbell MK, Elbourne DR, Altman DG, Group C. CONSORT statement: Extension to cluster randomised trials. BMJ. 2004;328:702–708. 29. Kerry M, Bland JM. Statistical notes: sample size in cluster randomization. BMJ. 1998;316:549. 30. Kerry M, Bland JM. Analysis of a trial randomized in clusters. Br Med J. 1998;316:54. 31. Kish L, Frankel M. Inference from complex samples. J R Stat Soc. 1974;36:1–37. 32. Simpson JM, Klar N, Donner A. Accounting for cluster randomization: A review of primary prevention trials, 1990 through 1993. Am J Public Health. 1995;85:1378–1383. 33. Campbell MK, Fayers PM, Grimshaw JM. Determinants of the intracluster correlation coefficient in cluster randomized trials: the case of implementation research. Clin Trials. 2005;2:99–107. 34. Nathens AB, Jurkovich GJ, Cummings P, Rivara FP, Maier RV. The effect of organized systems of trauma care on motor vehicle crash mortality. JAMA. 2000;283:1990–1994. 35. Lecky F, Woodford M, Yates DW. Trends in trauma care in England and Wales 1989–97. UK Trauma Audit and Research Network. Lancet. 2000;355:1771–1775. 36. Young L, Donald M, Parr M, Hillman K. The medical emergency team system: a two-hospital comparison. Resuscitation. 2008;77:180–188. 37. Parr MJA, Hadfield JH, Flabouris A, Bishop G, Hillman K. The medical emergency team: 12-Month analysis of reasons for activation, immediate outcome and not-for-resuscitation orders. Resuscitation. 2001;50:39–44. 38. Kenward G, Castle N, Hodgetts T, Shaikh L. Evaluation of a medical emergency team one year after implementation. Resuscitation. 2004;61:257–263. 39. Jones DA, McIntyre T, Baldwin I, Mercer I, Kattula A, Bellomo R. The medical emergency team and end-of-life care: A pilot study. Crit Care Resusc. 2007;9:151–156. 40. Calzavacca P, Licari E, Tee A, et al. A prospective study of factors influencing the outcome of patients after a medical emergency team review. Intensive Care Med. 2008;34:2112–2116. 41. Chen J, Flabouris A, Bellomo R, Hillman K, Finfer S. The medical emergency team system and not-for-resuscitation orders: Results from the MERIT study. Resuscitation. 2008;79:391–397. 42. Chen J, Hillman K, Bellomo R, Flabouris A, Finfer S, Cretikos M. The impact of introducing medical emergency team system on the documentations of vital signs. Resuscitation. 2009;80:35-43. 43. Drummond MF, McGuire A. Economic Evaluation in Health Care Merging Theory with Practice. Oxford: Oxford University Press; 2001. 44. Folland S, Stano M, Goodman AC. The Economics of Health and Health Care. 4th ed. Upper Saddle River, NJ: Pearson Prentice Hall; 2004. 45. Altman DG, Bland JM. Absence of evidence is not evidence of absence. BMJ. 1995;311:485. 46. Chen J, Bellomo R, Flabouris A, Hillman K, Finfer S, MERIT Investigators for the Simpson Centre and ANICS Clinical Trials Group. The relationship between early emergency team calls and serious adverse events. Crit Care Med. 2009;37:148–153. 47. Spiegelhalter DJ, Abrams KR, Myles JP. Bayesian Approaches to Clinical Trials and HealthCare Evaluation. New York: Wiley; 2004.

Chapter 34

RRS Education for Ward Staff John R. Welch and Gary B. Smith

Keywords  Rapid response systems • Education • At-risk • Deterioration • Patients

Introduction Medical, nursing and other staff working on general wards form the most important component of hospital-wide patient safety systems. In relation to patient deterioration, ward staff roles include the regular monitoring, charting and interpretation of patients’ vital signs and other clinical variables; the identification of “at-risk” and deteriorating patients; the timely administration of simple, first-line treatments; recognizing the need for additional, often more experienced help; and activating the rapid response system (RRS). Without ward staff involvement, the rapid response team (RRT) and Medical Emergency Team (MET) cannot be effective. (In this chapter, when we refer to the response team, we will use RRT to mean RRT, MET, and Critical Care Outreach Team (CCOT) – in other words, any responder team that an institution utilizes). Ward staff also provides important clinical and resource support to RRT members on the team’s arrival. Finally, if a patient remains on the general ward following a visit by the RRT, the ward staff must assume ­responsibility for the patient’s continuing surveillance and care. Unlike specialist staff in emergency and critical care departments, ward staff is typically generalists and usually work with patients receiving routine, though often complex, care. For any single member of the ward staff, the experience of sudden patient deterioration is relatively uncommon. If deterioration occurs, the staff must adjust their working patterns rapidly in order to manage the crisis effectively and prevent life-threatening complications. This is a challenging role and requires that

G.B. Smith (*) Department of Critical Care, Queen Alexandra Hospital, Portsmouth, PO6 3LY, UK e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_34, © Springer Science+Business Media, LLC 2011

381

382

J.R. Welch and G.B. Smith

the staff is prepared appropriately, especially as ward settings are less well resourced than specialist areas. Ward staff have usually not enjoyed the same educational opportunities as specialist staff and numerous reports have identified that ward staff do not always possess the knowledge, psychomotor skills and attitudes necessary to recognize or respond to “at-risk” or deteriorating patients.1–4 Furthermore, ward staff may have difficulty in identifying their own learning needs.5 This chapter discusses the crucial contribution of education of ward staff in ensuring safety and improving outcomes of at-risk patients, as well as in developing the staff themselves.

The Challenge for Ward Staff Modern healthcare systems demand that many medical problems are now managed in community settings and many operative procedures are undertaken in day-­ surgery units, thus avoiding hospital admission. Large and increasing proportions of hospitalized patients are elderly, suffer from multiple co-morbidities, have complex medical interventions and are admitted as emergencies.6 All such patients will spend at least some of their hospital stay in a general ward. In the UK, patients who give considerable cause for concern are those designated as Level 1 by the UK Intensive Care Society (Table 34.1), as they are “...at risk of their condition deteriorating…” or have been “…recently relocated from higher levels of care…” with care needs that “…can be met on an acute ward with additional advice and support from the critical care team…”7 (n.b.: the term “critical care team” would in many circumstances also refer to an RRT). Studies of a group of British hospitals found that in 2001 12.2% of ward patients required Level 1 care (or more); re-audit in 2006 indicated that this had risen to 21.3%.8,9 Table 34.1  Examples of level 1 patients7 Requiring a minimum of 4-hourly observation Requiring continuous oxygen therapy Requiring boluses of intravenous fluid Epidural analgesia or patient-controlled analgesia in use Receiving parenteral nutrition Requiring bolus intravenous drugs through a Central Venous Catheter With tracheostomy in situ With chest drain in situ Requiring a minimum of 4-hourly GCS assessment, Receiving a continuous infusion of insulin At risk of aspiration pneumonia On established intermittent renal support Requiring physiotherapy to treat or prevent respiratory failure Requiring greater than twice daily peak expiratory flow measurements

34  RRS Education for Ward Staff

383

The Evidence for Improving Education of Ward Staff in Acute Care Early recognition and timely treatment of patients at risk of deterioration can ­significantly improve patient outcomes.10 Additionally, delays in intensive care unit (ICU) admission are associated with poor clinical outcomes.11,12 However, two separate confidential enquiries have demonstrated significant deficiencies in ward care for critically ill patients.1,2 In the first, 54% of emergency ICU admissions received sub-optimal care before their transfer to ICU and many referrals to ICU were late.1 In the other, the initial assessment and/or examination of 1,677 acute medical patients were deemed unacceptable in 13% of cases, and initial treatment was delayed and/or inappropriate in 26%.2 In both studies, underlying factors leading to substandard care included failures of organization; lack of supervision; lack of knowledge; failure to appreciate clinical urgency; and failure to seek advice. The management of the airway, breathing, circulation, monitoring and oxygen therapy were rated as poor in many cases. Small studies show that trainee doctors often possess poor knowledge of aspects of acute care, e.g., pulse oximetry, and oxygen therapy.13,14 Needs assessments also suggest significant gaps in nurses’ knowledge and competences.15–17 Some of the deficiencies in doctors’ acute care knowledge may exist because texts available to medical students and junior doctors have not provided sufficient information regarding the assessment of critically ill patients in the past.18 However, newly qualified doctors have expressed significant disquiet with the level of acute care training provided at medical school19,20 and there is concern that acute care training should have increased emphasis at the undergraduate and pre-registration level.21–25 While it is unlikely that an RRS can function effectively without sensitive and specific RRT activation criteria, a reliable RRT and education of the RRT members, there is evidence that education of ward staff may be a very, perhaps the most, important component of an RRS.26–30 In a single-center, prospective before-andafter trial of a medical emergency team from Australia, almost all of the observed reduction in the hospital cardiac arrest rate occurred before the introduction of the MET during the period when ward staff were being educated about, and prepared for its implementation.26 In another Australian hospital, with an established RRT (ICU Liaison Team), the introduction of a clinical marker tool designed to assist in the early identification of unstable patients in the general wards facilitated proactive admission of patients to the ICU and was associated with a reduction in the number of ward cardiac arrests.27 Similarly, in a US hospital where an established MET had functioned for some time, there was improved use of the MET and a significant reduction in cardiac arrests following the establishment of specific, objective criteria for activating the MET.28,29 Finally, a Portugese group has shown a reduction in the long-term effectiveness of an RRS, suggesting that the critical factor was probably “…the staff education, awareness, and responsiveness to physiologic instability of

384

J.R. Welch and G.B. Smith

the patients…”.30 The authors concluded that the ­effectiveness of an RRS program “…is dependent not only on the existence of an MET but mainly on the periodic and continued education and training of the entire hospital staff …”.30

What General Ward Staff Need to Know Although the exact competencies required by individual members of ward staff in relation to managing the acutely ill patient will vary, for most they will include: • Correct observation of patients, including vital signs measurement and recording; • Interpretation of observed signs; • Recognition of the signs of deterioration; • The use of a track-and-trigger system (e.g., early warning score or MET calling criteria); • Appreciating clinical urgency; • When and how to utilize simple interventions (airway opening, oxygen therapy, intravenous fluid administration, etc); • Successful team-work; • Organization; • Knowing how to seek help from other staff; • Knowing how to use a systematic approach to information delivery, e.g., RSVP 31 or SBAR32; and • End-of-life care. Detailed guidance by health policy developers regarding the specific acute care competencies required by ward staff is difficult to find. In 2000, the report of a national review of critical care services by the Department of Health (DoH) for England emphasized the need to “…share critical care skills with staff in wards and the community ensuring enhancement of training opportunities and skills practice …”; however, it provided little detail about how this might happen.33 Specifically, it identified that all ward staff should have competence-based training in high dependency care within 4  years of the report,33 but this ambitious aim has still to be achieved. In 2005, using a modified Delphi technique, a group of U.K. healthcare professionals established the Acute Care Undergraduate Teaching (ACUTE) Initiative, which identified 71 essential and 16 optional competencies that students should possess at the time of graduation. It was proposed that these competencies form the core set for undergraduate training in acute care and resuscitation.23 This provided a significant challenge as teaching in acute care is often poorly co-ordinated, even among university medical schools. In 2007, the National Institute of Health and Clinical Excellence (NICE) for England and Wales issued guidance on “Acutely ill patients in hospital: recognition of and response to acute illness,” recommending that staff working with acutely ill

34  RRS Education for Ward Staff

385

patients on general wards should be provided with education and training to include competencies in “...monitoring, measurement, interpretation and prompt response to the acutely ill patient appropriate to the level of care they are providing...” NICE also recommended that staff “...should be assessed to ensure they can demonstrate them...”.34 Subsequently, the DoH in England issued a framework of competencies targeted at hospital staff to complement the NICE guidance on managing acute illness.35 The DoH competencies are based on the principles that patients have needs related to their risk of deterioration, and that interventions appropriate to meet those needs must be readily available, as and when required.35 The competences are based on a “chain of response” that reflects “...escalating levels of intervention in the care of a patient who becomes acutely ill...”.35 The six identified grades of competencies are not assigned to specific staff groups, professions, grades or levels of seniority, but are aggregated, as follows, on the basis of the role of the staff required to possess them: • • • • • •

Non-clinical staff (this group may include the patient’s relatives) Recorder Recognizer Primary responder Secondary responder Tertiary responder (n.b.: this grade of response is not provided by ward staff).

The competences are cumulative and, as the patient severity increases, the competencies also advance in complexity, responsibility and clinical risk. The role of staff designated as “recorders” will generally involve the taking and recording of vital signs, observations and other data, and the use of a defined “track-and-trigger” system (usually an early warning score in the UK). “Recognizers” will monitor the patient’s condition, interpret data in the context of the individual patient, and adjust the frequency of observations and/or level of monitoring. “Primary responders” will be capable of interpreting data and starting first-line treatments, e.g., use of airway adjuncts, oxygen therapy, a fluid challenge. “Secondary responders” would be called if the patient fails to improve with first-line treatments and will be capable of formulating a diagnosis, refining and developing the treatment plan, and recognizing when referral to critical care is needed. “Tertiary responders” are staff with critical care competencies, as defined by the Competency Based Training programme in Intensive Care Medicine for Europe,36 or similar initiatives. Therefore, with the exception of tertiary responder roles and some “secondary responder” functions, ward staff make up the bulk of the components of the “Chain of Response.” The role of “recorder” can be performed by any member of staff, although it is often likely to be done by a nursing assistant. Most commonly, the “recognizer” or “primary responder” will be a nurse. However, in some situations, this role could be undertaken by a doctor and the “secondary responder” could be from a different professional background (e.g., nurse practitioner or consultant nurse), provided they have the certified skills, experience and training to respond appropriately.

386

J.R. Welch and G.B. Smith

Although clinical staff will perform most of the actions in the “chain of response,” some competencies might be undertaken by non-clinical staff (e.g., portering staff knowing which gas cylinder to obtain), or even by the patient’s visitors (e.g., calling for help37,38). The DoH competencies are conveniently organized under the following headings: (1) airway, breathing, ventilation and oxygenation; (2) circulation; (3) transport and mobility (4) acute neurological care; and (5) patient-centered care, teamwork and communication.35 Table 34.2 provides an example of the roles relating to high flow and controlled oxygen therapy for “recorders,” “recognizers,” “primary responders,” and “secondary responders.” Examples of some suggested primary-responder competencies to deal with breathing and circulation problems and issues of teamwork and communication are set out in Table 34.3 (from 35).

Challenges in Training Ward Staff in the Immediate Management of Acute Illness It is a major undertaking to get a critical mass of ward staff trained, and able to practice and retain essential knowledge and skills in the recognition and management of acute illness, especially if staff has received insufficient undergraduate or pre-registration education and training. This was highlighted in the failure of the multi-center MERIT study to demonstrate that MET systems improve outcomes.39 Despite 4 months of training, ward staff called the MET to only 30% of patients that were transferred to ICUs, even though these patients met agreed-upon referral criteria in the preceding period.39 There are also substantial costs and other challenges involved in developing and sustaining an effective, clinically credible education service, and in releasing both trainee staff and trainers while still maintaining the patient service. Organizations should analyze the education needs that follow from regular review of patient safety issues and service priorities, and develop a coordinated approach to providing accessible training that meets those needs – in as cost-effective a way as possible. This can be done only if there is dedicated senior level leadership of the process, as well as engagement with the whole range of clinical staff.35 Mechanisms for evaluating the effectiveness of any education and training – particularly with regards to clinical practice – are crucial. Baseline knowledge and clinical experience affect learning, but changes in the way that staff are trained have reduced “hands-on” time with patients. These points have practical applications in education and in deciding what is clinically and cost effective for different individuals. For example, the study of advanced life support training has shown that learning is better retained if delayed until trainees have 6 months’ clinical experience.40 Organizations should detail the roles and responsibilities of all involved in acute care, clarifying the skills that are essential for every member of the ward team, and identifying those staff that can be equipped to reliably deliver other, additional

Table 34.2  Examples of the roles relating to respiratory rate, oxygen saturation, and high flow and controlled oxygen therapy for “recorders”, “recognizers”, “primary responders”; and “secondary responders” (modified 35) Competency group Non-clinical staff “Recorder” “Recognizer” “Primary responder” “Secondary responder” Evaluates effectiveness Identifies inadequate Interprets trigger in Measures Respiratory rate Recognizes of treatment, refines respiratory effort context of patient respiratory rate. respiratory arrest treatment plan if and institutes and responds in Records result and calls 2222 necessary, formulates clinical management accordance with local and assigns a diagnosis and therapies escalation protocols. trigger score for recognizes when Adjusts frequency respiratory rate. referral to Critical Care of observations in Has knowledge of is indicated keeping with trigger what constitutes an abnormal value Has detailed knowledge Follows oxygen prescription. Prescribes oxygen Identifies and collects Identifies and uses High flow and and evaluates effectiveness of the use of controlled masks /nasal cannulae/ Understands the context medical gases if controlled and high flow oxygen when controlled oxygen is venturi adapters at designated oxygen therapy therapy. Evaluates appropriate oxygen flow required and applies high effectiveness of oxygen rates. Records oxygen flow oxygen effectively in therapy and revises emergencies concentration/flow treatment accordingly Identifies possible cause Formulates diagnosis, Interprets measurements Oxygen saturation Measures oxygen evaluates effectiveness of hypoxia, prescribes in context and intervenes saturation. Records of treatment, refines oxygen therapy and with basic measures in result and assigns treatment plan if institutes clinical accordance with local trigger score. necessary and recognizes management therapies escalation protocols Has knowledge when referral to Critical including oxygen and of limitations of Care is indicated airway support. Adjusts pulse oximetry and frequency of observations recognizes abnormal in keeping with trigger result

34  RRS Education for Ward Staff 387

388

J.R. Welch and G.B. Smith

Table 34.3  Example competencies for ward staff caring for the acutely ill35 Management of breathing Identifies inadequate respiratory effort/inadequate ventilation/cause of breathlessness, and institutes clinical management therapies Has knowledge of which additional diagnostic tests are appropriate (e.g., peak flow, spirometry), institutes them and formulates management plan Prescribes nebulizers including appropriate driving gas Identifies possible cause of hypoxia, prescribes oxygen therapy and institutes clinical management therapies – and evaluates effectiveness Undertakes arterial blood gas sampling and measurement, interprets blood gas values Has knowledge of indications for continuous positive airway pressure and non-invasive ventilation – and identifies risks Requests and interprets chest radiograph Uses basic airway adjuncts and suction Assists with urgent intubation Prepares equipment for/assists with chest drain insertion. Manages patient with drain Identifies tension pneumothorax as a possible cause of breathlessness. Has knowledge of the management of a tension pneumothorax Management of circulation Identifies abnormal heart rate and institutes clinical management therapies Has knowledge of common abnormalities and can interpret ECG in the context of the patient. Responds in accord with local protocols and institutes clinical therapies Has knowledge of causes of abnormal blood pressure, and which diagnostic investigations are appropriate. Institutes clinical management therapies In cases of abnormal fluid balance, identifies when clinical intervention is required and institutes diagnostic investigations and a management plan Inserts urinary catheter Identifies source of bleeding, clinical impact and initiates definitive management. Commences resuscitation Has knowledge of which blood tests are required in both elective and emergency situations. Can request test(s), performs venesection Inserts IV cannula in “difficult” cases Identifies need for, and initiates fluid challenge for resuscitation and institutes clinical management plan. Prescribes maintenance fluids – and drugs Has knowledge of indications for, and risks of blood products. Prescribes blood products Identifies low cardiac output, institutes diagnostic investigations and a management plan Identifies potential causes of collapse/unresponsiveness relevant to the individual patient In-hospital resuscitation Understands rationale for use of emergency drugs and can administer Team working and communication Recognizes leadership role within team and responsibility to refer to secondary responder. Provides information in a structured format that conveys clinical urgency Incorporates within the documentation a management plan and timescale for reassessment. Identifies when referral to the secondary responder will be indicated

skills when needed. Training can be tailored accordingly, so that individuals focus on what is really necessary for their particular roles and get more opportunities to practice what is important to them, rather than receiving training in skills they are unlikely to use. Thinking about the individual can also be applied to the selection

34  RRS Education for Ward Staff

389

of trainers for particular staff groups; it seems that learners prefer to be assessed by their peers, even though peers may be more critical of performance.41

Education Essential to the Implementation of an RRS The development of RRTs as the primary solution to preventing adverse patient outcomes has meant that education programs for ward staff have usually focused on the theory behind the RRS, and the criteria and process for activating the RRT.42,43 Such training is necessary for successful RRT utilization,44,45 but rarely provides staff with the knowledge of how to prevent deterioration or the actions to take between calling the RRT and its arrival. The latter is important, as no matter how fast an RRT responds, there will always be an opportunity for ward staff to act before their arrival at the patient’s bedside. Indeed, the commencement of interventions on the general wards before the RRT arrived appeared to be the most important factor leading to a reduction in in-hospital mortality in one study.22 Evidence suggests that the common causes of patient deterioration on general wards lie within the scope of the patient’s primary teams’ practice.46 Unfortunately, evidence from ward cardiac arrests suggests that ward staff often perform poorly in the moments between patient discovery and arrival of the cardiac arrest team.47

Current Initiatives in Acute Care Education Training in acute care should be structured around a systematic approach to assessment and management of Airway, Breathing, Circulation, Disability, issues revealed by Exposure of the patient (i.e., the ABCDE system), and methods of effective communication and team-work.33 These principles underpin many of the educational programmes in acute care that are now offered to undergraduate, postgraduate and post-registration staff working outside critical care areas.17,48–56 Many of these courses utilize techniques derived from long-standing, usually generic, resuscitation courses.57,58 For example, on a typical course, learners may be required to read a manual, attend a number of short lectures and workshops, view demonstrations of practical techniques, and practice some skills, usually using a variety of patient mannequins. Some courses incorporate e-learning51,53 and, more recently, there has been an emphasis on small-group interactive teaching and the use of scenarios akin to the real-life practice of learners using immersive human patient simulation.54,56 However, simulation-based courses are both labor-intensive and costly, making it difficult to train large numbers of staff. Consequently, hospitals can usually only provide this specialized type of training to individuals once a year at best, despite evidence that knowledge and skills deteriorate over time.59 A focused “little but often, in the clinical area” approach to training may be useful in maintaining some, mainly psychomotor, acute care skills. The use of “just

390

J.R. Welch and G.B. Smith

in time” and “just in place” training, in which skills training is provided in situ ­immediately before its likely use, is currently being investigated in ICU pediatric airway training 60 and cardiopulmonary resuscitation.61 Such techniques may find use in acute care. Studies from resuscitation training suggest that self-instruction may be an ­effective way of training staff in discrete psychomotor tasks.62 Self-instruction also has the attraction that it is likely to be cheaper, as it involves fewer or no instructors.63 However, acute care emergencies inevitably involve complex interactions of knowledge, psychomotor skills and attitudes, as well as demanding considerable inter-professional teamwork. Researchers studying interactive internet-based, online training by junior doctors have concluded that social interaction plays a part in reinforcing learning, which is difficult to achieve via a personal computer.64

Short Courses in Acute Care The development of specific, short, multi-professional courses that teach a preemptive approach to critical illness to general ward staff commenced in 1999 with the development of the Acute Life-threatening Events: Recognition and Treatment (ALERT) course in the UK.48 This 1-day course is designed to give ward staff greater confidence and ability in the recognition and management of adult patients who have impending or established critical illness. Using pre-course reading, informal and interactive seminars, practical demonstrations and role-playing, ALERT focuses on common problems encountered during patient deterioration e.g., “the blue and breathless patient,” “the hypotensive patient,” or “the patient who does not pass urine.” Because it trains staff from different disciplines together, it also focuses specifically on improving communication skills and multi-professional teamwork. Although developed in the UK, ALERT now runs successfully in several other countries and an e-learning version of the course is in development. The ALERT course has also formed part of a longer preparation program for newly-qualified doctors.65 The content of numerous other short courses, often uni-professional in nature, overlaps with that of ALERT 50,51,66 but, generally, their content is broader and often allocates only a small portion of course time to teaching about the prevention of critical illness.

The Role of the Response Team in Educating Ward Staff The responding team members have an important role in the education of ward staff, by teaching from a position of clinical expertise and providing confidence to the ward team that referring a patient will be a positive experience. In the UK, critical care outreach team members are expected to “share” critical care skills with

34  RRS Education for Ward Staff

391

ward staff; this is usually done via a combination of formal and informal teaching opportunities.67 There is great benefit to regular, inter-professional reviews of cases in individual departments by outreach team members and ward staff to discuss “What happened?” “What should have happened?” and “What could be done better in the future?” Findings of sub-optimal care will almost always require an educational response, tailored to addressing skill or knowledge deficits or system failures. Outreach staff often teach on acute care courses, such as ALERT or its derivatives, but also provide rapid feedback on emergencies at the bedside or work with ward staff to assess and manage more complex “at-risk” patients. The more reactive nature of the rapid response system is, perhaps, less well designed to permit such prolonged interactions between ward and RRT or MET staff. Finally, seconding ward staff to work on the RRT 68 or making members of a patient’s primary team lead the RRT 69 could provide additional opportunities for training for a limited number of staff.

Evidence for Benefit in Acute Care Educational Interventions It is difficult – if not impossible – to separate the effects of education from other developments aimed at improving patient outcomes in acute care.70 Perhaps as a result, there is a dearth of published evidence regarding the benefits of acute care training for ward staff, and those results that exist give a mixed message. The 1-day multi-professional acute care course, ALERT, has been shown to improve staff knowledge of acute care, attitudes towards managing the deteriorating patient and improved documentation in patient records following the ALERT principles of ABCDE assessment.71–74 ALERT has also been used as a component of a strategic approach to reducing hospital mortality.75 Furthermore, the authors of a recent 6-year observational study at a large university hospital that trained over 75% of its staff on Immediate Life Support courses demonstrated a significant decrease in cardiac arrest rates and a significant increase in survival from all emergency calls.76 On the other hand, despite considerable care in the development of a multiprofessional education intervention involving high fidelity simulation in Denmark,54 and the short and intense timeframe for education of staff, there was no improvement in nursing staff’s awareness of patients at risk on wards or in patient outcomes.77

Summary In summary, education of ward staff is a crucial and necessary part of any rapid response system. Evidence suggests that ward staff can influence the outcome for patients, not only by intervening early with simple procedures, but also by ­involving

392

J.R. Welch and G.B. Smith

the RRT when necessary. It is clear that a range of educational methods need to be used for ward staff, with the provision of flexible and easy access to learning materials, to suit different learning styles, and to meet different sorts of need. Skill acquisition must be seen to be valued by employers and staff, with an appropriate resource provided to learn and practice. Some theoretical knowledge can be gained by self-instruction, be it using written materials, video or on-line packages; but motor skills and integration of different components of the required skill-set need hands-on practice.

References 1. McQuillan PJ, Pilkington S, Allan A, et al. Confidential inquiry into quality of care before admission to intensive care. BMJ. 1998;316:1853–1858. 2. National Confidential Enquiry into Patient Outcomes and Death. An Acute Problem? London: National Confidential Enquiry into Patient Outcome and Death; 2005. 3. National Patient Safety Agency. Safer Care for the Acutely Ill Patient: Learning from Serious Incidents. London: NPSA; 2007. 4. National Patient Safety Agency. Recognizing and Responding Appropriately to Early Signs of Deterioration in Hospitalized Patients. London: NPSA; 2007. 5. Cox H, James J, Hunt J. The experiences of trained nurses caring for critically ill patients within a general ward setting. Intensive Crit Care Nurs. 2006;22:283–293. 6. Hospital Episode Statistics: Headline figures. Available from: http://www.hesonline.nhs.U.K./ Ease/servlet/ContentServer?siteID=1937&categoryID=193. Accessed 16.01.10. 7. Intensive Care Society. Levels of Critical Care for Adult Patients. London: Intensive Care Society; 2009. 8. Chellel A, Fraser J, Fender V, et al. Nursing observations on ward patients at risk of critical illness. Nurs Times. 2002;98:36–39. 9. Smith S, Fraser J, Plowright C, et  al. Nursing observations on ward patients – results of a five-year audit. Nurs Times. 2008;104:28–29. 10. Chen J, Bellomo R, Flabouris A, et al. The relationship between early emergency team calls and serious adverse events. Crit Care Med. 2009;37:148–153. 11. Chalfin DB, Trzeciak S, Likourezos A, et al. Impact of delayed transfer of critically ill patients from the emergency department to the intensive care unit. Crit Care Med. 2007;35:1477–1483. 12. Young MP, Gooder VJ, McBride K, et al. In-patient transfers to the intensive care unit: delays are associated with increased mortality and morbidity. J Gen Intern Med. 2003;18:77–83. 13. Smith GB, Poplett N. Knowledge of aspects of acute care in trainee doctors. Postgrad Med J. 2002;78:335–358. 14. Howell M. Pulse oximetry: an audit of nursing and medical staff understanding. Br J Nurs. 2002;11:191–197. 15. Wood I, Douglas J, Priest H. Education and training for acute care delivery: a needs analysis. Nurs Crit Care. 2004;9:159–166. 16. Derham C. Achieving comprehensive critical care. Nurs Crit Care. 2007;12(3):124–131. 17. O’Riordan B, Gray K, McArthur-Rouse F. Implementing a critical care course for ward nurses. Nurs Stand. 2003;17:41–44. 18. Cook CJ, Smith GB. Do textbooks of clinical examination contain information regarding the assessment of critically ill patients? Resuscitation. 2004;60:129–136. 19. Cooper N. Medical training did not teach me what I really needed to know. BMJ. 2003;327:190s.

34  RRS Education for Ward Staff

393

20. Goldacre MJ, Lambert T, Evans J, Turner G. Preregistration house officers’ views on whether their experience at medical school prepared them well for their jobs: National questionnaire survey. BMJ. 2003;326:1011–1012. 21. Harrison GA, Hillman KM, Fulde GW, Jacques TC. The need for undergraduate education in critical care. (Results of a questionnaire to year 6 medical undergraduates, University of New South Wales and recommendations on a curriculum in critical care.). Anaesth Intensive Care. 1999;27:53–58. 22. Smith CM, Perkins GD, Bullock I, Bion JF. Undergraduate training in the care of the acutely ill patient: a literature review. Intensive Care Med. 2007;33:901–907. 23. Perkins GD, Barrett H, Bullock I, et al. The Acute Care Undergraduate Teaching (ACUTE) Initiative: consensus development of core competencies in acute care for undergraduates in the United Kingdom. Intensive Care Med. 2005;31:1627–1633. 24. Higginson R, Lewis R, De D, Jones B. The need for critical care nurse education at preregistration level. Br J Nurs. 2004;13:1326–1328. 25. Shen J, Joynt GM, Critchley LA, Tan IK, Lee A. Survey of current status of intensive care teaching in English-speaking medical schools. Crit Care Med. 2003;31:293–298. 26. Bellomo R, Goldsmith D, Uchino S, et al. A prospective before-and-after trial of a medical emergency team. Med J Aust. 2003;179:283–287. 27. Green AL, Williams A. An evaluation of an early warning clinical marker referral tool. Intensive Crit Care Nurs. 2006;22:274–282. 28. DeVita MA, Braithwaite RS, Mahidhara R, et al. Use of medical emergency team responses to reduce hospital cardiopulmonary arrests. Qual Saf Health Care. 2004;13:251–254. 29. Foraida MI, DeVita MA, Braithwaite RS, et  al. Improving the utilization of medical crisis teams (Condition C) at an urban tertiary care hospital. J Crit Care. 2003;18:87–94. 30. Campello G, Granja C, Carvalho F, et al. Immediate and long-term impact of medical emergency teams on cardiac arrest prevalence and mortality: a plea for periodic basic life-support training programs. Crit Care Med. 2009;37:3054–3061. 31. Featherstone P, Chalmers T, Smith GB. RSVP: a system for communication of deterioration in hospital patients. Br J Nurs. 2008;17:860–864. 32. Thomas CM, Bertram E, Johnson D. The SBAR communication technique: teaching nursing students professional communication skills. Nurse Educ. 2009;34:176–180. 33. Department of Health. Comprehensive Critical Care: A Review of Adult Critical Care Services. London: Department of Health; 2000. 34. National Institute for Health and Clinical Excellence. Nice Clinical Guideline 50. Acutely Ill Patients in Hospital: Recognition of and Response to Acute Illness in Adults in Hospital. London: National Institute for Health and Clinical Excellence; 2007. 35. Department of Health. Competencies for Recognising and Responding to Acutely Ill Patients in Hospital. London: Department of Health; 2009. 36. Competency Based Training programme in Intensive Care Medicine for Europe and other world regions. Available from: http://www.cobatrice.org/en/index.asp. Accessed 16.01.10. 37. h t t p : / / w w w . i h i . o r g / I H I / T o p i c s / C r i t i c a l C a r e / I n t e n s i v e C a r e / T o o l s / ConditionHBrochureforPatientsandFamilies.htm. Accessed 16.01.10. 38. Ray EM, Smith R, Massie S, et al. Family Alert: implementing direct family activation of a pediatric rapid response team. Jt Comm J Qual Patient Saf. 2009;35:575–580. 39. Hillman K, Chen J, Cretikos M, et  al. Introduction of the medical emergency team (MET) system: a cluster-randomized controlled trial. Lancet. 2005;365:2091–2097. 40. Jensen ML, Lippert F, Hesselfeldt R, et al. The significance of clinical experience on learning outcome from resuscitation training – a randomized controlled study. Resuscitation. 2009;80:238–243. 41. Bucknall V, Sobic EM, Wood HL, et al. Peer assessment of resuscitation skills. Resuscitation. 2008;77:211–215. 42. Jones D, Baldwin I, McIntyre T, et al. Nurses’ attitudes to a medical emergency team service in a teaching hospital. Qual Saf Health Care. 2006;15:427–432.

394

J.R. Welch and G.B. Smith

43. Johnson AL. Creative education for rapid response teams. J Contin Educ Nurs. 2009;40:38–42. 44. Cretikos MA, Chen J, Hillman KM, et al. The effectiveness of implementation of the medical emergency team (MET) system and factors associated with use during the MERIT study. Crit Care Resusc. 2007;9:205–212. 45. Jones D, Bates S, Warrillow S, et al. Effect of an education program on the utilization of a medical emergency team in a teaching hospital. Intern Med J. 2006;36:231–236. 46. Prado R, Albert RK, Mehler PS, Chu ES. Rapid response: a quality improvement conundrum. J Hosp Med. 2009;4:255–257. 47. Einav S, Shleifer A, Kark JD, Landesberg G, Matot I. Performance of department staff in the window between discovery of collapse to cardiac arrest team arrival. Resuscitation. 2006;69:213–220. 48. Smith GB, Osgood VM, Crane S. ALERT™ – a multiprofessional training course in the care of the acutely ill adult patient. Resuscitation. 2002;52:281–286. 49. Ellison S, Sullivan C, Quaintance J, Arnold L, Godfrey P. Critical care recognition, management and communication skills during an emergency medicine clerkship. Med Teach. 2008;30:e228–e238. 50. Cave J, Wallace D, Baillie G, et  al. DR WHO: a workshop for house officer preparation. Postgrad Med J. 2007;83:4–7. 51. Gruber P, Gomersall C, Joynt G, Shields F, Chu M, Derrick J. Teaching acute care: a course for undergraduates. Resuscitation. 2007;74:142–149. 52. MacDowall J. The assessment and treatment of the acutely ill patient – the role of the patient simulator as a teaching tool in the undergraduate programme. Med Teach. 2006;28:326–329. 53. Shah IM, Walters MR, McKillop JH. Acute medicine teaching in an undergraduate medical curriculum: a blended learning approach. Emerg Med J. 2008;25:354–357. 54. Fuhrmann L, Østergaard D, Lippert A, Perner A. A multi-professional full-scale simulation course in the recognition and management of deteriorating hospital patients. Resuscitation. 2009;80:669–673. 55. Collins TJ, Price AM, Angrave PD. Pre-registration education: making a difference to critical care? Nurs Crit Care. 2006;11:52–57. 56. McGaughey J. Acute care teaching in the undergraduate nursing curriculum. Nurs Crit Care. 2009;14:11–16. 57. Fothergill J, Touquet R, Stacey S. Advanced life support courses. J Accid Emerg Med. 1995;12:111–114. 58. Gwinnutt CL, Driscoll PA. Advanced trauma life support. Eur J Anaesthesiol. 1996;13:95–101. 59. Hamilton R. Nurses’ knowledge and skill retention following cardiopulmonary resuscitation training: a review of the literature. J Adv Nurs. 2005;51:288–297. 60. Nishisaki A, Colborn S, Watson C, et al. Just-in-time simulation training improves ICU physician trainee airway resuscitation participation without compromising procedural success or safety. Circulation. 2008;118:S1453. 61. Niles D, Sutton RM, Donoghue A, et al. “Rolling Refreshers”: a novel approach to maintain CPR psychomotor skill competence. Resuscitation. 2009;80:909–912. 62. de Vries W, Schelvis M, Rustemeijer I, Bierens JJ. Self-training in the use of automated external defibrillators: the same results for less money. Resuscitation. 2008;76:76–82. 63. Bjørshol CA, Lindner TW, Søreide E, Moen L, Sunde K. Hospital employees improve basic life support skills and confidence with a personal resuscitation manikin and a 24-min video instruction. Resuscitation. 2009;80:898–902. 64. Jensen ML, Mondrup F, Lippert F, et al. Using e-learning for maintenance of ALS competence. Resuscitation. 2009;80:903–908. 65. Berridge EJ, Freeth D, Sharpe J, Roberts CM. Bridging the gap: supporting the transition from medical student to practising doctor – a two-week preparation programme after graduation. Med Teach. 2007;29:119–127. 66. Soar J, Perkins GD, Harris S, Nolan J. The immediate life support course. Resuscitation. 2003;57:21–26.

34  RRS Education for Ward Staff

395

67. McDonnell A, Esmonde L, Morgan R, et al. The provision of critical care outreach services in England: findings from a national survey. J Crit Care. 2007;22:212–218. 68. Plowright C, O’Riordan B, Scott G. The perception of ward-based nurses seconded into an Outreach Service. Nurs Crit Care. 2005;10:143–149. 69. Sarani B, Sonnas S, Bergey MR, et al. Resident and nurse perceptions of the impact of a medical emergency team on education and patient safety in an academic medical center. Crit Care Med. 2009;37:3091–3096. 70. Hutchings A, Durand MA, Grieve R, et al. Evaluation of modernization of adult critical care services in England: time series and cost effectiveness analysis. BMJ. 2009;339:1130. 71. Hutchinson S, Robson WP. Confidence levels of PRHOs in caring for acutely ill patients. Postgrad Med J. 2002;78:697. 72. Viner J. Implementing improvements in care of critically ill, ward-based patients. Prof Nurse. 2002;18:91-93. 73. Smith GB, Poplett N. Impact of attending a one-day course (ALERTTM) on trainee doctors’ knowledge of acute care. Resuscitation. 2004;61:117–122. 74. Featherstone P, Smith GB, Linnell M, Easton S, Osgood VM. Impact of a one-day interprofessional course (ALERT™) on attitudes and confidence in managing critically ill adult patients. Resuscitation. 2005;65:329–336. 75. Wright J, Dugdale B, Hammond I, et al. Learning from death: a hospital mortality reduction programme. J R Soc Med. 2006;99:303–308. 76. Spearpoint KG, Gruber PC, Brett SJ. Impact of the Immediate Life Support course on the incidence and outcome of in-hospital cardiac arrest calls: an observational study over 6 years. Resuscitation. 2009;80:638–643. 77. Fuhrmann L, Perner A, Klausen TW, Østergaard D, Lippert A. The effect of multi-professional education on the recognition and outcome of patients-at-risk on general wards. Resuscitation. 2009;80:1357–1360.

Chapter 35

Standardized Process and Outcome Assessment Tool Gabriella Jäderling and David Konrad

Keywords  Standardized • Process • Outcome • Assessment • Tool

Introduction The introduction and maintenance of a rapid response system (RRS) in a hospital, regardless of the hospital’s size, is an introduction of a complex change of system and as such poses many difficulties in the measurement of its possible success. Before embarking on a hospital-wide modification, opponent voices require hard evidence of the beneficial effects of such a system change. This raises the questions: how then do we best measure the effects of what we do and what tools do we have to assess our outcomes? Should we measure the process or the outcomes, or both? And what are the best outcomes to measure? System changes of a whole culture of a hospital may not lend themselves to being studied as easily as a novel drug or a single intervention. It is hardly ethical to perform a randomized controlled trial at the individual level to evaluate the effect of an intervention such as giving highly qualified care when an acute life-threatening condition is identified, nor does it lend itself to the experimental methodology of blinding. A cluster-randomized controlled trial is an appealing design and was the basis of the MERIT study, involving 23 hospitals in Australia.1 Although a huge and commendable effort, the study encountered numerous difficulties concerning statistical power, contamination, and inter- as well as intra-hospital variation. Many questions still remain to be answered and the traditional “gold standard” of evidence-based medicine – the blinded randomized controlled trial – may not provide the proper tools to answer these. It may be time to reconsider the conventional grading system of evidence2 when evaluating complex interventions and

G. Jäderling (*) Department of Anesthesia and Intensive Care, Karolinska University Hospital, 17176 Stockholm, Sweden e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_35, © Springer Science+Business Media, LLC 2011

397

398

G. Jäderling and D. Konrad

accept that one large study will not provide us with “the truth” but that multiple studies with different methodologies, both quantitative and qualitative, will be necessary to form the larger picture. Well-designed and performed “observational” studies from different types of hospitals and different parts of the world will be valuable and should be regarded as such.3 We need to assess all aspects of introducing a fundamental system change: has the process itself been successful? Are the calling criteria adequate? Are they being used? Are they being responded to in a timely fashion? These questions might be best assessed using qualitative studies and will be fundamental when trying to evaluate whether the intervention has any effect on the outcome. If we cannot confirm that the process has been successful, we cannot say whether or not the intervention is effective. To date, there have been a number of single-center studies with positive results, but it is not always easy to draw general conclusions as different outcomes have been used, inclusion criteria vary, and cultural and structural differences among hospitals may play an important role in the implementation and maintenance of the RRS.4 Did the investigators measure the same things? The use of historical controls in before-and-after studies is an important limitation as changes in case mix and seasonal variation must be taken into account. Generalization of single-center studies is problematic and the restraints must be understood. The commitment and enthusiasm of a local investigator may be key to the success of implementation in one setting that may not be reproducible elsewhere.3 The study population may differ greatly as well as the local resources, such as staffing of academic versus non-teaching hospitals. With these limitations in mind, there is no need to dismiss these studies. Commonly viewed, single center studies do not hold the same scientific worth as large RCTs and may not lend themselves to generalization as easily, although they do give important input as to whether implementation was successful and relevant goals for outcome have been reached. This gives important feedback to the system in question and may also point to possible areas of improvement as well as being able to inspire and promote further research.

Standardization of the RRS Process Initiating RRS When setting up an RRS in a hospital, a considerable amount of time must be put into preparations before the launch, and a thorough literature review should include methods of organizational change. Based on a hypothesis (early identification and intervention of the at-risk patient in the general ward improves outcome), the problem is identified. In our institution, this was done by performing a prevalence study to calculate how many patients fulfilling the criteria we would find at any given time.5 This also allowed us to validate the criteria chosen by restricting and ­extending them within the study.

35  Standardized Process and Outcome Assessment Tool

399

The introduction of a complex intervention is largely dependent upon good support among all involved: clinicians, nursing staff and management. It is helpful if a ­network is formed for continuous evaluation and feedback processes, since a ­multidisciplinary approach is necessary in the implementation as well as in the evaluation and the research. If a reliable baseline of the relevant outcomes is not available, a pilot study should be performed in order to assess these. This will also aid in determining possible barriers to implementing the protocol and offer insight into how to overcome them. Surveys of the opinions of clinicians and nursing staff can provide useful information for the present implementation and also outline practice variability and differences in attitudes among different centers.6 Trial design is a crucial step and should involve epidemiological and statistical expertise from the start. As with any other clinical research, the trial is to be designed to answer a clearly defined question.

Data Collection A great deal of information can be gathered from each call and later sorted and compiled for different purposes. Quantitative information based on the events ­surrounding the call is well suited to integration into a database. Standardization of the collection and reporting of RRS data is important for consistency and will facilitate comparisons across institutions. Guidelines for this have been proposed by an international consensus group in the Utstein style, with a comprehensive set of recommended core and supplemental data elements.7 The minimum data ­necessary for process and outcome evaluation includes information on the setting, patient demographics, events and outcome (Table  35.1). Data may need to be gathered from different sources, such as an RRS paper chart and patients’ electronic charts. A protocol should exist to fuse these bits of information reliably and accurately into a usable database. It should be limited to a few trained staff to ­collect and record data in order to ease standardization. This database can then be used in itself to describe the RRS process for internal evaluation and analysis of events or fused with further information, such as a hospital database for outcome assessment of mortality and cardiac arrest. A standardized and reliable documentation for Not-For-Resuscitation (NFR) orders is paramount but can be a challenging issue, which should be dealt with on a hospital-wide basis. This may differ greatly between hospitals and even between different departments within the same hospital and is possibly the most difficult factor to measure. A channel for qualitative information-gathering can be through regular meetings with contact persons in each ward and by setting up an easily accessible way to reach the RRS team – for instance, through an e-mail address on the hospital’s intranet. Further useful information to assess is changes in attitudes, decision ­making and acceptance of the system. This could all be explored through social studies and in-depth interviews and is a field well worth investigating.

400

G. Jäderling and D. Konrad

Table 35.1  RRS database case form – minimum set Hospital information   Bed allocation   Outcome statistics – annually Response team information   Composition   Structure   Coverage   Activation criteria Patient information   I.D. number   Age   Gender   Location/type of ward   Type of admission (acute/elective)   Admitted – date   Discharged – date   Previous RRS call   Previous ICU care RRS call information   Call date   Call time   Response time   Time spent on ward   Call reason (criteria activated) RRS outcome   Clinical findings   Interventions   Transfer to/left on ward Patient outcome   Discharge diagnosis (ICD-10)   Operated – date   NFR status – date   Mortality – death date

Evaluation The success of the system is dependent upon all parts of it being functional; thus, all the different parts need to be evaluated. Demographic data on the hospital as well as the setup of the team needs to be reported in order to describe the context in which the system is working.7 The afferent limb of any RRS is defined as the function that detects and identifies a medical emergency and consequently triggers an adequate response.8 It needs to be evaluated on a regular basis to see whether it is functioning properly. Have the educational efforts been adequate? Is the awareness widespread? Are the

35  Standardized Process and Outcome Assessment Tool

401

calling criteria being used enough? This is hard to measure accurately without considerable effort and we may see more of automated and continuous monitoring in the future.9 Delays in triggering need to be recognized as they may have an impact on the result; the underlying causes might best be assessed with qualitative studies. The efferent limb is the crisis-response component and can be of different constitutions, nurse-led (RRT) or physician-led (MET),8 but share the property of responding to a “red flag raised” and bringing critical care hands and experience to the acutely ill patient wherever the patient may be. Response times should be measured and reported as well as the competency of the team and the equipment immediately available. Cooperation and communication skills connecting the two limbs should be given consideration and perhaps assessed using video analysis of the system at work10 or other forms of team training.11,12 The time frame may be of importance and reflect a direct dose-response relationship.13 The effects of implementing system changes are not instantaneous. In comparison, one can look at the introduction of organized trauma systems that have taken at least 10 years before a reduction in mortality can be shown. 14 Any new system introduction needs time to mature and set properly.15

Outcome To set up standardized tools of assessment, we first need to decide what outcomes are of importance. The next step must be to have an adequate baseline of what we want to measure. Clinical outcomes commonly used are unexpected deaths, cardiac arrests, and unplanned ICU admission. Mortality can be easily assessed but the time to follow up should be taken into account. How long of a follow-up is adequate? ICU mortality or even hospital mortality may not be the most useful measures of the success of a treatment as they are influenced by local admission and discharge policies. The optimal follow-up time is yet to be decided. NFR status must be addressed and reported. Cardiac arrests are also commonly used due to their dichotomous nature, but the form of registration needs to be considered. Are there reliable accounts of the number of cardiac arrests in the hospital? Again, NFR status must clearly be stated as this could influence a change in the number of unexpected deaths/cardiac arrests. It is becoming apparent that part of the RRS activities include discussions of limitations in care and patients’ NFR status.16 An NFR order can be present before the call or decided during the call, in which case it is recorded with ease. However, a discussion initiated during the call could actually trigger a decision being made at any point thereafter. It can also be reconsidered and changed several times during a patient’s stay, depending on the clinical conditions. This heterogeneity is difficult to control, and thus hard to measure in any assessment of system function.

402

G. Jäderling and D. Konrad

Summary The process of implementing and upholding a functioning RRS is clearly a ­multi-dimensional web of events, and to make an assessment of it, the process needs to be defined down to its separate elements and viewed accordingly. Three phases can be discerned: pre-implementation, implementation and evaluation. During the period preceding the start lies the preparation and definition phase. It must be decided which type of RRS is suitable for this particular institution, and valid calling criteria chosen. Education and information for the entire hospital includes sessions with staff where questions and concerns can be addressed directly as well as reinforcement on the intranet and proper material handed out. The implementation is performed and data collection standardized, for which there are comprehensive recommendations to aid in the setting up of useful information-gathering.7 A good database format from the start is the key for future research. Standardization will provide us with a common nomenclature and enable easier comparisons between different centers around the world. Evaluation and re-evaluation is important for understanding the process and for being able to amend it if necessary. Questionnaires can be used to assess staff satisfaction. Feedback from clinicians and nursing staff is reviewed and answered as soon as possible. Continuous education is important considering staff turnover. The educational part of the RRS is probably the most crucial part of success, and as such needs to be a never-ending process.

References 1. Hillman K, Chen J, Cretikos M, et  al. Introduction of the medical emergency team (MET) system: A cluster-randomised controlled trial. Lancet. 2005;365(9477):2091–2097. 2. Atkins D, Eccles M, Flottorp S, et  al. Systems for grading the quality of evidence and the strength of recommendations I: Critical appraisal of existing approaches. The GRADE Working Group. BMC Health Serv Res. 2004;4(1):38. 3. Bagshaw SM, Bellomo R. The need to reform our assessment of evidence from clinical trials: a commentary. Philos Ethics Humanit Med. 2008;3:23. 4. Winters BD, Pham JC, Hunt EA, et  al. Rapid response systems: A systematic review. Crit Care Med. 2007;35(5):1238–1243. 5. Bell MB, Konrad D, Granath F, et  al. Prevalence and sensitivity of MET-criteria in a Scandinavian University Hospital. Resuscitation. 2006;70(1):66–73. 6. Delaney A, Angus DC, Bellomo R, et al. Bench-to-bedside review: The evaluation of complex interventions in critical care. Crit Care. 2008;12(2):210. 7. Peberdy MA, Cretikos M, Abella BS, et  al. Recommended guidelines for monitoring, ­reporting, and conducting research on medical emergency team, outreach, and rapid response systems: An Utstein-style scientific statement: A scientific statement from the International Liaison Committee on Resuscitation (American Heart Association, Australian Resuscitation Council, European Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Council of Southern Africa, and the New Zealand Resuscitation Council); the American Heart Association Emergency Cardiovascular Care Committee; the Council on Cardiopulmonary, Perioperative, and Critical Care; and the

35  Standardized Process and Outcome Assessment Tool

403

Interdisciplinary Working Group on Quality of Care and Outcomes Research. Circulation. 2007;116(21):2481–2500. 8. DeVita MA, Bellomo R, Hillman K, et al. Findings of the first consensus conference on medical emergency teams. Crit Care Med. 2006;34(9):2463–2478. 9. Hravnak M, Edwards L, Clontz A, et al. Defining the incidence of cardiorespiratory instability in patients in step-down units using an electronic integrated monitoring system. Arch Intern Med. 2008;168(12):1300–1308. 10. Hillman K, Chen J, May E. Complex intensive care unit interventions. Crit Care Med. 2009;37(1 Suppl):S102–106. 11. DeVita MA, Schaefer J, Lutz J, et al. Improving medical emergency team (MET) performance using a novel curriculum and a computerized human patient simulator. Qual Saf Health Care. 2005;14(5):326–331. 12. Wallin CJ, Meurling L, Hedman L, et al. Target-focused medical emergency team training using a human patient simulator: Effects on behavior and attitude. Med Educ. 2007;41(2):173–180. 13. Jones D, Bellomo R, Bates S, et al. Long-term effect of a medical emergency team on cardiac arrests in a teaching hospital. Crit Care. 2005;9(6):R808–R815. 14. Nathens AB, Jurkovich GJ, Cummings P, et al. The effect of organized systems of trauma care on motor vehicle crash mortality. J Am Med Assoc. 2000;283(15):1990–1994. 15. Jones D, Bates S, Warrillow S, et al. Effect of an education programme on the utilization of a medical emergency team in a teaching hospital. Intern Med J. 2006;36(4):231–236. 16. Chen J, Flabouris A, Bellomo R, et al. The Medical Emergency Team System and not-forresuscitation orders: results from the MERIT Study. Resuscitation. 2008;79(3):391–397.

Chapter 36

The Impact of RRSs on Choosing “Not-for-Resuscitation” Status Arthas Flabouris and Jack Chen

Keywords  Impact • Rapid • Response • Systems • Resuscitation • NFR

Background Rapid Response Teams (RRTs) evolved as a system-based approach for the recognition of, and response to, the acutely ill hospital in-patient. They evolved following revelations that hospital in-patients who suffer adverse events, such as cardiac arrest or unanticipated admission to an intensive care unit (ICU), had documented physiological abnormalities prior to these events,1–10 or had an inadequate or no response to these physiological abnormalities4 when a timely response may be more beneficial11. Similarly, cardiac-arrest teams were associated with a low rate of success,12 and when successful were likely to promote patient suffering and prolong the time until death.13 Formal hospital policies relating to orders of not-for-resuscitation (NFR) were first published in the mid 1970s.14,15 Interestingly, these policies were preceded by a period of less formal application of the withholding of cardiopulmonary resuscitation from patients by staff who would deem it to be of no benefit to certain patients.16 Such decision making was occurring in isolation to the patient and their wishes. At that same time, a strong ethical framework was evolving that was promoting the aspect of patient autonomy. Unilateral decision making was being less ­supported in favor of decisions that involved a multidisciplinary input, the wishes and values of the patients or, in the absence of a competent patient, their surrogates. Guidelines for the process of generating formal NFR orders were developed so as to increase competent patient participation in the decision making process. Formal NFR orders

A. Flabouris (*) Intensive Care Unit, Royal Adelaide Hospital, North Terrace, Adelaide, SA 5001, Australia e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_36, © Springer Science+Business Media, LLC 2011

405

406

A. Flabouris and J. Chen

increased patient participation in such decision making, reduced the likelihood of terminally ill patients receiving inappropriate resuscitation measures, and improved quality of life at the terminal stages and bereavement adjustment.17–19

Not-for-Resuscitation Decision Making However, subsequent studies have revealed highly variable patterns of NFR ­prescribing. Factors such as the patient’s age,20,21 gender,22 and diagnosis,23,24 as well as physician specialty25,26 and hospital characteristics,27 have been shown to contribute to this variability. The prescription of NFR orders often conflict with patients’ prior expressed wishes,25 and to documented patient advanced directives.28 Patients are typically admitted to a hospital under a senior clinician who is primarily responsible for the patient’s care while in the hospital. These consultant physicians provide the clinical leadership for a team of staff who manage the daily medical needs of the patient. In the not-too-distant past, senior clinicians would make unilateral patient treatment choices, including those relating to resuscitation 16, which would provide their team with the objectives of subsequent patient management. With the increased emphasis on patient autonomy, such decision making has subsequently been less supported. These changes have, in part, contributed to the variability and a change in the direction of the overall delegation of decision making, particularly for the incompetent patient. Physicians are also less confident in discussing resuscitation orders with patients and their surrogates 29, and not good at explaining end-of-life issues to patients.30 It is thus not unusual for the delegation of responsibility for decision making to be left to poorly informed relatives and patient carers who are ill-equipped and too emotionally attached to make appropriate decisions. Similarly, most of the day-to-day care of the patient is undertaken and supported, by junior medical staff and nursing staff who are often under-resourced, poorly trained for and disempowered to make significant decisions, such as initiating discussions relating to patient treatment choices.31 Thus patients, their relatives, caretakers and junior hospital medical and nursing staff look towards senior clinicians to lead such decision making. These circumstances not only leave the patient exposed to unnecessary aggressive invasive resuscitation measures at or near the time of death, but also other “avoidance” practices, such as initiation of invasive organ support and admission to an ICU. Once invasive organ support – “life support” – has been instituted, patients will appear calm and physiologically stable, death can effectively be delayed. In this setting, with an incompetent patient, decision making involves acute care physicians not previously involved in the patients’ continuum of care and surrogates who often don’t reflect the patients’ true wishes. Critical-care physicians have been shown to contribute more positively towards the process of end-of-life decision making. Such decisions occur more commonly

36  The Impact of RRSs on Choosing “Not-for-Resuscitation” Status

407

within the ICU environment 32; patients cared for by trained critical-care physicians are more likely to die with an NFR order, and have a shorter ICU length-of-stay and fewer interventions.33 Critical-care physicians are not infrequently involved in end-of-life decision making and care. They are also more readily available within the ICU and are thus more comfortable in dealing with the issue of withholding and withdrawal of life support than other specialty-based physicians, most of whom have less experience with end-of-life decision making. The current hospital in-patient population is older and more likely to have chronic disease and significant co-morbidities. Whereas in the past, death was relatively sudden, being in a younger group and usually hospitalized due to serious infection, trauma or obstetric-related problems, nowadays death is associated with the gradual deterioration from a chronic illness. This course may be predictable for certain conditions and demonstrate typical physiological disturbances as death approaches.34 It may also be precipitated by acute deterioration from acquiring an acute and unrelated illness (e.g., sepsis and septic shock due to a respiratory infection in a patient with severe congestive cardiac failure). Understanding the trajectory of an illness may allow preparation for and initiation of end-of-life decisions. Critical-care physicians are often accustomed to seeing a pattern of progressive organ dysfunction that is inevitably associated with a high mortality.

Rapid Response Teams and Not-for-Resuscitation Orders These circumstances described above are not dissimilar to the antecedents for RRTs. For example, the latter evolved from a failure of a hospital-wide systematic approach to potentially preventable cardiac arrests and unanticipated ICU admissions. The factors that contributed to such failures included increasing medical specialization, aging patient population with more co-morbidities, and the reliance on junior medical staff who are often ill-equipped and trained in the recognition and management of an acutely ill patient.35,36 A program based on identifying hospital in-patents for whom advanced-care planning could be initiated, training of staff to recognize, respond to and facilitate end-of-life discussions, documentation and supportive clinical care has been developed with the expressed intention of improving patient end-of-life treatment option documentation, reducing risk of unnecessary aggressive medical care, and increasing the likelihood of attaining the patients’ expressed wishes for end-of-life care.37 As intended RRTs are called to attend acutely ill patients, based on criteria that reflect acute physiological disturbance and thus expose the patient to increased risk of harm or even death. It is not improbable that some such patients may in fact be dying, as expected, but as yet have been failed to be recognized as dying and failed to have a management plan that excludes aggressive life-support measures. As RRTs typically consist of critical care-based staff, who are potentially better ­positioned to identify and manage dying patients and to withhold futile ­resuscitation

408

A. Flabouris and J. Chen

measures, it is theoretically plausible that RRTs may influence an NFR order. In fact, observational studies have shown that an NFR order is issued at the time of an RRT call for between 4 and 25% of all calls.38–40

Evidence for the Impact of Rapid Response Teams on Not-for-Resuscitation Orders The MERIT study 41 was a large-cluster randomized controlled trial of 23 Australian hospitals that tested the effect of introducing a Medical Emergency Team (MET) system. It utilized a standardized multi-faceted implementation strategy to put the MET into service within each of the 12 intervention hospitals. This implementation strategy was composed of traditional in-service education of existing clinical (medical and nursing) staff and any new staff. Data was collected related to the primary outcomes of the composite incidence rate of any cardiac arrest without NFR, unplanned ICU admission without NFR, and death without NFR. The secondary endpoints identified at that time were the incidence rates of any cardiac arrest without NFR, unplanned ICU admission without NFR, and death without NFR, respectively, over three study phases: 2-month baseline measurements; 4-month implementation of the MET system in the randomized MET hospitals; and 6-month collection of the data after the full implementation of the MET system. This study showed the rate of NFR orders issued at the time of an event was almost three times higher within MET hospitals compared with that of the nonMET hospitals. A subsequent analysis of the MERIT study data relating to NFR orders, with the objective of testing the hypothesis that there was a difference in the issuing of NFR orders by emergency teams when comparing MET to non-MET hospitals, was conducted.42 The principle findings of this study were that 90% of all deaths in both MET and control hospitals were preceded by NFR orders and among MET hospitals there was a significant increase in the proportion and rate of NFR orders issued for all the events combined, when compared to control hospitals. This difference was due predominately to a significantly higher number of NFR orders issued during emergency team attendance at emergency calls that were not related to cardiac arrest, death or unplanned ICU admission (referred to as an adverse event-free calls). Among MET hospitals, for almost one in every 12 MET attendances (or 8%), an NFR order was issued. This was in comparison to one NFR order being issued for every 33 similar emergency team attendances (3%) in the control hospitals (Fig. 36.1). Among the study hospitals, the proportion of patients who died and had NFR orders was approximately 90%. Thus, the MET-related issuing of NFR orders is a relatively small measurable component of overall hospital-wide issuing of NFR orders. MET teams issued NFR orders for approximately one patient out of every 1,449 admissions, or an additional 100 orders among the 12 MET hospitals over the

36  The Impact of RRSs on Choosing “Not-for-Resuscitation” Status

409

12 10

Control hospitals MET hospitals

p=0.048

8

% 6 4

p=0.005

2 0 Aggregated events

Unplanned ICU admissions

Cardiac arrests

Event

Adverse eventfree emergency team calls

Fig. 36.1  Proportion of each event for which an NFR order was documented, at the time of that event by the MET or the cardiac arrest team

6-month study period, compared to 1,793 NFR orders issued prior to an MET ­attendance (Table 36.1). On an individual patient level, this remains significant, as theoretically 100 patients were spared unnecessary cardiopulmonary resuscitation or other resuscitation measures during that period. In reality, this figure may be an under-estimation, as it is possible that emergency team attendance may have prompted a subsequent consideration of an NFR order being issued after an emergency team attendance. A future exploration of such processes would be important and would require consideration of the many logistic and educational factors that appear to influence the activities and impact of an RRT.43,44 The study also pointed towards a potentially as-yet unexplored and evolving consequence of RRTs upon the broader hospital practice. The allocation of NFR orders did not preclude a patient from having an emergency team attendance, particularly so for hospitals allocated an MET, which were three times more likely to attend to patients with a prior NFR order than a typical cardiac-arrest team. This would suggest that although there may be a proportion of patients with an NFR order for whom an RRT is called in error, there are also such patients for whom it is deemed appropriate that they receive critical care-type supports other than cardiopulmonary resuscitation. There is also a potential third group, that is patients with an NFR order that typically excludes cardiopulmonary resuscitation, and for whom other critical care interventions are also not appropriate, but are not stipulated to be so within the existing framework for NFR orders. This would indicate that there may be a need for specific orders in hospitals with an RRT relating to the appropriateness or not of the attendance of an RRT, and, if so, the desirable level of RRT intervention, e.g., a “Not-for-RRT” order or “modified RRT” order. Such a process would require either modification of existing NFR discussions and documentation ­processes, or development of a separate parallel process. These are important

Aggregated events 0.31 (0.26) 0.00–0.80 0.69 (0.51) 0.15–1.94 0.52 (0.14–0.90) Deaths 0.16 (0.11) 0.00–0.34 0.18 (0.11) 0.00–0.37 0.00 (−0.09–0.10) Unplanned ICU 0.06 (0.13) 0.00–0.40 0.07 (0.12) 0.00–0.37 0.02 (−0.06–0.09) admissions Cardiac arrests 0.16 (0.11) 0.00–0.37 0.15 (0.14) 0.00–0.46 0.01 (−0.10–0.11) 0.04 (0.09) 0.00–0.27 0.40 (0.42) 0.00–1.32 0.49 (0.20–0.78) Adverse event-free emergency team calls *The P values were derived using the weighted t-test to compare the rates between 11 control hospitals and 12 MET hospitals **P < 0.01

Table 36.1  Number of events per 1,000 admissions during which an NFR order was issued (aggregate and by each of our event types) Control hospitals MET hospitals Weighted mean Difference Outcome Mean (SD) Range Mean (SD) Range (95%CI)*

P

0.892 0.002**

0.009** 0.969 0.621

410 A. Flabouris and J. Chen

36  The Impact of RRSs on Choosing “Not-for-Resuscitation” Status

411

c­ onsiderations since RRTs are becoming increasingly prevalent and information related to them more public as it spreads beyond medical media and to the broader media. As discussed earlier in this chapter, there is a wide extent of variability in the issuing of NFR orders for which many factors have been implicated. In this study, the model examined, taking teaching hospital status, location (urban vs. metropolitan), size (number of beds) and MET allocation into consideration could only account for <50% of the variance in such orders. As only cases where NFR orders were issued at the time of MET attendance were examined, it is possible that an RRT attendance may have provided a trigger for subsequent review and reassessment of the patient’s prognosis soon after RRT attendance and thus precipitate an NFR order. In essence, RRTs appear to not only identify and respond promptly to acutely ill patients with the purpose of preventing progression to more serious event (e.g., cardiac arrest, ICU admission), but also identify and respond to patients with RRT calling criteria and maybe dying, with the purpose of preventing progression to another type of serious event (e.g., inappropriate resuscitation during the end-stage of life). With the former, acute resuscitation measures are indicated, while with the latter, patient-specific palliative care services and/or management are indicated. It is important to interpret these findings within the context of the broad crosssection of Australian hospitals in which the study was conducted. Within the Australian and New Zealand context, the proportion of hospital deaths with an NFR order is almost 90% 41,45 and this rate differs significantly from that of other countries – in comparison with 79% of hospital deaths in the U.S.46 and approximately 50–60% among a group of European countries, which varied from 73% in Switzerland to 16% in Italy. 47 In light of this high variability of NFR orders and the many factors already identified that contribute to this variability, it would be expected that the introduction of an RRT would have a differential impact upon each environment into which it is introduced. It also highlights the importance of the need for generic and uniform international guidelines for the issuing of NFR orders, and not for RRT orders. As such, since an intervention is not an insignificant component of RRTs, and RRTs are now internationally accepted, such guidelines could be incorporated as part of a standardized RRT implementation package.

Summary • There are similarities between the health system factors that contribute to variability and weaknesses in the application of NFR orders to hospital in-patients and those that form the basis of the evolution of RRT. • There is theoretical plausibility for RRTs to influence the issuing of NFR orders. • Observational studies have revealed that up to 24% of all RRT calls involve the issuing of NFR orders.

412

A. Flabouris and J. Chen

• More recently, a large cluster-randomized trial showed that with the ­introduction of an MET, and in comparison to conventional cardiac arrest teams, more NFR orders were issued at the time of an MET call. • MET-related issuing of NFR orders was a relatively small measurable ­component of overall hospital-wide issuing of NFR orders. However, this was in a hospital system that had a background of high proportion of deaths with NFR orders; the relative proportion of NFR orders issued at the time of an RRT call may differ within hospitals with a lower level of hospital deaths with an NFR order. • RRT may be associated with other hospital system changes that could potentially further influence the issuing of NFR orders. • RRT implementation and associated education could be potentially used as a vehicle for a standardized NFR implementation package. • There is the potential of patient benefit for a closer association between critical care based-RRT and palliative care services.

References 1. Schein RMH, Hazday N, Pena M, et al. Clinical antecedents to in-hospital cardiopulmonary arrest. Chest. 1990;98:1388–1392. 2. Franklin C, Mathew J. Developing strategies to prevent in-hospital cardiac arrest: analysing responses of physicians and nurses in the hours before the event. Crit Care Med. 1994;22:246–247. 3. Smith A, Wood J. Can some in-hospital cardio-respiratory arrest be prevented? Resuscitation. 1998;37:133–137. 4. McQuillan P, Pilkington S, Alan A, et  al. Confidential inquiry into quality of care before admission to intensive care. BMJ. 1998;316:1853–1858. 5. Goldhill DR, White SA, Sumner A. Physiological values and procedures in the 24 hours before ICU admission from the ward. Anaesthesia. 1999;54:529–534. 6. McGloin H, Adam SK, Singer M. Unexpected deaths and referrals to intensive care units of patients on general wards. Are some cases potentially avoidable? J R Coll Physicians Lond. 1999;33:255–259. 7. Hillman KM, Bristow PJ, Chey T, et  al. Antecedents to hospital deaths. Intern Med J. 2001;31:343–348. 8. Hillman KM, Bristow PJ, Chey T, et  al. Duration of life-threatening antecedents prior to intensive care admission. Intensive Care Med. 2002;28:1629–1634. 9. Hodgetts TJ, Kenward G, Vlackonikolis I, et al. Incidence, location and reasons for avoidable in-hospital cardiac arrest in a district general hospital. Resuscitation. 2002;54:115–123. 10. Kause J, Smith G, Prytherch D, Parr M, Flabouris A, and for the Intensive Care Society (UK) and Australian and New Zealand Intensive Care Society Clinical Trials Group ACADEMIA Study Investigators. A comparison of antecedents to cardiac arrests, deaths and emergency intensive care admissions in Australia and New Zealand, and the United Kingdom – the ACADEMIA study. Resuscitation. 2004;62:275–282. 11. Rivers E, Nguyen B, Havstad S, et  al., for the Early Goal-Directed Therapy Collaborative Group. Early Goal-Directed Therapy. N Engl J Med. 2001;345(19):1368–1377. 12. Nichol G, Stiell IG, Laupacis A, et  al. A cumulative meta-analysis of the effectiveness of defibrillator-capable emergency medical services for victims of out-of-hospital cardiac arrest. Ann Emerg Med. 1999;34:517–525. 13. Symmers WS. Not allowed to die. BMJ. 1968;1(5589):442.

36  The Impact of RRSs on Choosing “Not-for-Resuscitation” Status

413

14. Clinical Care Committee of the Massachusetts General Hospital. Optimum care for hopelessly ill patients: a report of the Clinical Care Committee of the Massachusetts General Hospital. N Engl J Med. 1976;295:362–364. 15. Rabkin MT, Gillerman G, Rice NR. Orders not to resuscitate. N Engl J Med. 1976;295:364-366. 16. Burns J, Edwards J, Johnson J, Cassem N. Do-not-resuscitate order after 25 years. Crit Care Med. 2003;31:1543–1550. 17. Fukaura A, Tazawa H, Nakajima H, et al. Do not resuscitate orders at a teaching hospital in Japan. N Engl J Med. 1995;333:805–808. 18. Stern SG, Orlowski JP. DNR or CPR – the choice is ours. Crit Care Med. 1992;20:1263–1272. 19. Wright AA, Baohui Z, Ray A, et  al. Associations between end-of-life discussions, patient mental health, medical care near death, and caregiver bereavement adjustment. JAMA. 2008;300(14):1665–1673. 20. Bedell SE, Pelle D, Maher PL, et al. Do-not-resuscitate orders for critically ill patients in the hospital. How are they used and what is their impact? JAMA. 1986;256:233–237. 21. Youngner SJ, Lewandowski W, McClish DK, et al. “Do not resuscitate” orders: incidence and implications in a medical-intensive care unit. JAMA. 1985;253:54–57. 22. Stolman CJ, Gregory JJ, Dunn D, et  al. Evaluation of the do-not-resuscitate orders at a community hospital. Arch Intern Med. 1989;149:1851–1856. 23. Schwartz DA, Reilly P. The choice not to be resuscitated. J Am Geriatr Soc. 1986;34:807–811. 24. Wachter RM, Luce JM, Hearst N, et  al. Decisions about resuscitation: inequities among patients with different diseases but similar prognoses. Ann Intern Med. 1989;111:525–532. 25. Hofmann JC, Wenger NS, Davis RB, et al. Patient preferences for communication with physicians about end-of-life decisions. SUPPORT Investigators. Study to understand prognoses and preference for outcomes and risks of treatment. Ann Intern Med. 1997;127:1–12. 26. Murphy BF. What has happened to clinical leadership in futile care discussions? Med J Aust. 2008;188:418–419. 27. Sidhu NS, Dunkley ME, Egan MJ. “Not-for-resuscitation” orders in Australian public hospitals: policies, standardised order forms and patient information leaflets. Med J Aust. 2007;186:72–75. 28. Morrell ED, Brown BP, Qi R, Drabiak K, Helft PR. The do-not-resuscitate order: associations with advance directives, physician specialty and documentation of discussion 15 years after Patient Self-Determination Act. J Med Ethics. 2008;34:642–647. 29. Sulmasy DP, Sood JR, Ury WA. Physicians’ confidence in discussing do-not-resuscitate orders with patients and surrogates. J Med Ethics. 2008;34:96–101. 30. Lynn J, Teno JM, Phillips RS, et al. Perceptions by family members of the dying experience of older and seriously ill patients. SUPPORT Investigators. Study to understand prognoses and preferences for outcomes and risks of treatments. Ann Intern Med. 1997;126:97–106. 31. Tulsky JA, Chesney MA, Lo B. See one, do one, teach one? House staff experience discussing do-not-resuscitate orders. Arch Intern Med. 1996;156:1285–1289. 32. Prendergast TJ, Luce JM. Increasing incidence of withholding and withdrawal of life support from the critically ill. Am J Respir Crit Care Med. 1997;155:15–20. 33. Kollef MH, Ward S. The influence of access to a private attending physician on the withdrawal of life-sustaining therapies in the intensive care unit. Crit Care Med. 1999;27:2125–2132. 34. Murray SA, Kendall M, Boyd K, Sheikh A. Illness trajectories and palliative care. BMJ. 2005;330:1007–1011. 35. Hillman K, Flabouris A, Parr M. A hospital-wide system for managing the seriously ill: a model of applied health systems research. In: Sibbald WJ, Bion JF, eds. Update in intensive care and emergency medicine – evaluating critical care, using health services research to improve outcome, vol 35. Berlin: Springer; 2000. 36. Lam S, Flabouris A. Medical trainees and public safety. In: DeVita MA, Hillman K, Bellomo R, eds. Medical emergency teams, implementation and outcome measurement. New York: Springer; 2006. ISBN: 0-387-27920-2.

414

A. Flabouris and J. Chen

37. Austin Health, Respecting Patient Choices. http://www.respectingpatientchoices.org.au/. Accessed 12.08. 38. Parr MJ, Hadfield JH, Flabouris A, Bishop G, Hillman K. The Medical Emergency Team: 12-month analysis of reasons for activation, immediate outcome and not-for-resuscitation orders. Resuscitation. 2001;50(1):39–44. 39. Kenward G, Castle N, Hodgetts T, et al. Evaluation of a medical emergency team one year after implementation. Resuscitation. 2004;61(3):257–263. 40. Buist MD, Moore GE, Bernard SA, Waxman BP, Anderson JN, Nguyen TV. Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. BMJ. 2002;324:387–390. 41. The MERIT Study Investigators. Introduction of medical emergency team (MET) system A cluster-randomised controlled trial. Lancet. 2005;365:2091–2097. 42. Chen J, Flabouris A, Bellomo R, Hillman K, Finfer S, and The MERIT Study Investigators. The Medical Emergency Team System and not-for-resuscitation orders: results from the MERIT Study. Resuscitation. 2008;3:391–7. 43. Cretikos MA, Chen J, Hillman KM, Bellomo R, Finfer SR, Flabouris A. The effectiveness of implementation of the medical emergency team (MET) system and factors associated with use during the MERIT study. Crit Care Resusc. 2007;9(2):206–212. 44. Buist M, Bellomo R. The medical emergency team or the medical education team. Crit Care Resusc. 2004;6:88–91. 45. Jones DA, McIntyre T, Baldwin I, Mercer I, Kattula A, Bellomo R. The medical emergency team and end-of-life care: a pilot study. Crit Care Resusc. 2007;9:151–156. 46. The Support Principal Investigators. A controlled trial to improve care for seriously ill hospitalized patients. JAMA. 1995;274:1591–1598. 47. van Delden JJ, Lofmark R, Deliens L, et  al. Do-not-resuscitate decisions in six European countries. Crit Care Med. 2006;34:1686–1690.

Chapter 37

The Costs and the Savings Dana Edelson and Rinaldo Bellomo

Keywords  Rapid • Response • Systems • Costs • Savings Limited data exist regarding the cost-efficacy of changes in hospital systems. Such calculations are complex and require accounting, not just for the costs of implementing the therapy and the patient outcome benefits, but also for the subsequent resource utilization associated with any patient outcome benefits. The collection of such data is extraordinarily burdensome, complicated and subject to bias.1,2 In addition, it requires a level of long-term follow-up that is often too difficult or itself too costly to accomplish. Yet, despite the above caveats, in an environment where return on investment and providing the best possible care at the lowest possible price is paramount, no discussion of a medical intervention is complete without analysis of its associated costs. Therefore, this chapter will attempt to shed some light on the cost implications of implementing a Rapid Response System.

The Cost of Adverse Events Hospitals can be dangerous places, and even if one is conservative on the number of preventable deaths or injuries, the cost to the hospital is still staggering. At the extreme, an adverse event would result in death. This might appear to be an event with low cost, but, in fact, the loss of income, taxation, family support, and time invested in retraining and re-education is very high for many deaths. Adverse events occur in all hospitals and the reasons for these may not always be evident. Nevertheless, the cost has been estimated to be in the millions of dollars. In 1999, the Committee on Quality of Health Care in America estimated that as a result of adverse events, between 44,000 and 98,000 patients die every year, and millions

R. Bellomo (*) Department of Intensive Care, Austin Hospital, Heidelberg, 3084, VIC, Australia e-mail: [email protected] M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1_37, © Springer Science+Business Media, LLC 2011

415

416

D. Edelson and R. Bellomo

more are injured.3 Adverse events may not always lead to death but will almost certainly lead to increased length of stay and increased hospital costs. These excess costs have been estimated to range between US $95 and $4,485 per patient.4–7 These costs, however, relate mainly to drug-related adverse events and errors. A study conducted in New Zealand found that the average direct cost for adverse events is NZ $10,264 (approximately US $5,455) per patient. Adverse events are estimated to cost the New Zealand health system NZ $870 million (approximately US $462 million) per year.8 In Australia, a study conducted in 1995 showed that between 10,000 and 14,000 Australians are likely to die of preventable deaths in acute-care hospitals each year. The Quality in Australian Health Care Study found that adverse events would likely account for 3.3 million bed days and equate to AUD $867 million (US $576 million) per year.9 In a 2006 report, the Institute of Medicine estimated that preventable medication errors cause 3.5 billion USA ­dollars in additional treatment costs, not including lost wages or productivity.10 Similar to the above observations, the National Health Service (NHS) in the UK stated that 10.8% of all admissions had an adverse event.11 And because these numbers are reported adverse events, they likely represent an underestimation of actual adverse events, resulting in even higher associated costs. These adverse events have also been recognized by the World Health Professional Alliance which has recorded adverse events in many countries and believes that poor quality healthcare is responsible for more than 30% of avoidable deaths in Italy.12

Evolution of the Rapid Response System Most hospitals employ a team that responds to an emergency that occurs within the hospital. Most of these teams are historically based on the cardiac arrest team concept; in other words, they respond when the patient has had a cardiac arrest. This concept has been accepted by hospitals since the late 1960s and represents a delayed reaction of the healthcare in patients who are already clinically “dead.” The problem with this approach is that it is reactive, that it responds too late, and that hospitals have changed substantially since it was introduced. To confirm these concerns, studies have shown that between 66% and 84% of patients who have a cardiac arrest in a hospital have signs of serious physiological deterioration up to 24 h prior to their arrest.13,14 Other studies have shown that the outcome from these in-hospital cardiac arrests remains extremely poor and, in fact, has changed little since the first implementation of cardiac arrest teams.15,16 As stated previously, hospitals have changed a great deal since the first introduction of cardiac arrest teams. Many patients who used to require intensive care (ICU) treatment 20 years ago are now ward patients, and many current ICU patients would never have survived 30–40 years ago. Thus, patients in our general wards are much sicker than they used to be and a new approach is necessary. These observations mean that the confines of the ICU must now move outside of its four walls and offer a service to all patients in the hospital in a proactive, trained and coordinated way.

37  The Costs and the Savings

417

It would make sense that if we could identify these “cardiac-arrest patients” early and treat them appropriately before they become too sick, we may then prevent their cardiac arrest and prevent them being transferred to an ICU for further expensive definitive treatment. These considerations are logical, clinically plausible and make obvious economic sense. No business model, no large manufacturer would tolerate regular failure, and only delayed intervention when the assembly line broke down, for example, two or three times a day before implementing preventive policies. It is simply foolish, from a cost-effectiveness point of view, to pursue such policies. The same should be true of hospitals. The Rapid Response System (RRS) with its efferent arm, either the Medical Emergency Team (MET) or Rapid Response Team (RRT), is one way hospitals are attempting to improve patient safety. The Medical Emergency Team (MET) is a different approach to current in-hospital patient emergency care. It is a more proactive system (or at least a system that reacts earlier), using pre-determined criteria to identify at-risk patients and activate a response that is time-appropriate and ensures that the team members are skilled to assess and treat the patients they see. The RRS comprises four processes: identification of an event and trigger of a response team (afferent arm); reaction of a trained response (efferent arm); clinical training of team members and hospital staff education (quality improvement arm); and use of outcome indicators to monitor the system (administrative arm).17

Costs Associated with a Rapid Response System When considering the costs, it is important to clarify the perspective of the payer since the accounting can vary when viewed from the standpoint of the individual patient, the hospital, the third-party payer, or society. The societal perspective is generally recognized as the preferred perspective since it represents the public interest and is most inclusive.18,19 Accordingly, a cost-effectiveness analysis requires determination of the net societal costs of an intervention divided by a standardized benefit.19 Generally, this benefit is defined in terms of life-years saved, or as is now preferred, quality-adjusted life-years (QALYs). Using this methodology, it has been shown that dialysis for end-stage renal patients costs approximately $51,000/life-year saved, while vaccinating children against measles, mumps and rubella saves more money than it costs.20 Unfortunately, critical care interventions are difficult to pursue in this way because long-term outcomes are rarely known, let  alone quality-adjusted ones.18 This is certainly the case with the RRS. The outcomes can only be crudely estimated given the current data, and the intervention is still too new to have long-term outcomes available. As a result, it makes more sense, at this point, to attempt to quantify the net costs/savings of the intervention. Additionally, the decision to introduce an RRS is frequently made at the level of the hospital, taking into account hospital-level costs and savings. Therefore, to the extent that there is a discrepancy between the societal and hospital perspective, we will attempt to highlight both, but will focus predominantly on the

418

D. Edelson and R. Bellomo

Table 37.1  Potential rapid response system costs and savings Potential costs Potential savings Coordinator Decreased ICU utilization Additional medical personnel Decreased overall length of stay Equipment (computer, defibrillator, etc.) Post-resuscitation care Disposables (medications, catheters, etc.) Liability costs for medical errors Incidentals (photocopying, stickers, etc.) Increased non-urgent surgeries Training (course materials & coverage) Salary earned by survivors Subsequent healthcare utilization by survivors Malpractice insurance premiums

costs and ­savings to the hospital. Table 37.1 summarizes the costs and savings that are ­discussed in greater detail below.

Efferent Arm Costs Potentially, the most costly aspect of the RRS is the response arm, also known as the MET or RRT, depending upon the responders involved. The MET is generally considered to be more expensive than the RRT because of the inclusion of a physician on the MET. However, the cost analysis of the two teams has not been performed. Nevertheless, each hospital may have different sources of personnel and therefore different costs. The MET may simply replace the existing cardiac arrest team or consist of an entirely different team. The team frequently includes at least one critical care nurse but can include a physician (ideally at least at the level of a critical care fellow), a respiratory therapist, a pharmacist and/or chaplain. The nurses and physician may have other duties or may be tasked solely to the response team. Obviously, the costs vary depending upon the associated work the team may be performing. The members of the team will likely depend on the size of the hospital, its staffing levels, and number of MET calls. There are several options for covering the team. The least expensive version involves utilizing existing resources by giving the role of responding to MET calls to individuals with other primary responsibilities who can more easily put those on hold when an MET is called. Ideally, this would be someone with a lighter clinical load or no patientcare responsibilities during dual role periods, such as a critical care nurse manager or educator. This requires that there be slack in the current system, which may not be the case, especially when resources are already tight. Also, since the responder will need to return to his/her primary responsibility upon completion of the MET call, s/he is unlikely to be able to assume any of the responsibilities associated with the remaining two arms of the RRS, the ones not directly associated with patient care. An alternative to this low-budget version is to create a position that provides coverage 24 h a day, 7 days a week. This would require 4.5 full-time equivalents (FTEs) in order to staff the calls. In the case of critical care nurses, 4.5 FTEs would likely cost over US $350,000/year in salary and benefits just for the nurse ­responders.21 This model has been employed by at least one prominent medical center in the United States.22 While this

37  The Costs and the Savings

419

model ensures that there will be time to address the remaining arms of the RRS, there would likely be significant time remaining after all RRS work has been done, given average MET workloads. The number of MET calls has been shown to range from 2.5/1,000 admission, in a new system, to >40/1,000 admissions, in a mature system.23–26 Depending on the size of the hospital, this correlates to an average of <1–3 or more calls per day. Given that the average time of the team at the bedside is approximately 30 min, it would be difficult to fully occupy such a position solely with MET responsibilities.27 The authors therefore favor a compromised approach, which involves creating one new position in the form of a coordinator (0.5–1.0 FTE). In most large hospitals that have an RRS, this is usually a nurse who has critical care experience. The coordinator’s position, depending on the size of the hospital, can be full- or parttime, but a hospital with more than 400 beds would likely require a full-time coordinator. Hospitals larger than 800 beds may require more than one staff member and this staff member may include another nurse or ICU fellow or clerical support. The coordinator, in addition to staffing MET calls, is responsible for ensuring that the system is implemented and monitored and thus that all four arms of the RRS are functioning. The average cost of the coordinator’s position is around US $50–60,000/year, plus an estimated US $20,000/year for benefits. Most hospitals that utilize cardiac arrest teams have advanced resuscitation equipment available on every ward. This may not be cost-effective as the upkeep of all this equipment is costly, and the checking of the equipment by ward staff may in fact be time-consuming. It is also important to note that many hospitals utilize agency staff and in fact this equipment may go unchecked for some time.28 The MET/RRT members take their own equipment and rely on ward staff to provide basic life support until the team arrives with all its resuscitation equipment. It is important to note that in general, all nursing and medical personnel in developed countries’ hospitals are accredited each year on basic life support. The RRS does not advocate the removal of advanced resuscitation equipment from critical care or high dependency areas of the hospital. The RRS equipment is comprised of trolley, bag and mask ventilation equipment, monitor, defibrillator and backpack containing circulation and respiratory packs, emergency drugs and other consumables. Most large hospitals would require two of these setups in case there is more than one MET call at the same time. The estimated cost of the MET equipment is approximately $60,000 with a further $20,000 on consumables per year.

Afferent Arm Costs Every RRS requires a way to “activate” the responders, also known as the afferent arm. The simplest and most widely-used afferent arm involves selection and dissemination of calling criteria that then trigger a manual activation of the team by the clinician who witnessed the presence of a trigger. The calling criteria are used to identify physiological changes in a patient’s condition, which if not acted upon immediately would ultimately lead to death or serious illness. These criteria help take the guesswork out of

420

D. Edelson and R. Bellomo

what constitutes a critically ill patient and allow any clinician to seek assistance ­whenever required. When these criteria are present, staff can then activate the MET in the same way they used to activate the cardiac arrest team, except that the patient is resuscitated early, long before suffering serious complications, cardiac arrest and death. Many different versions of calling criteria have been identified,29–31 and although the search for the most sensitive and specific signs of clinical deterioration continues, the coordinator can choose a system that best fits the patient population and staff at hand. Accordingly, a process to ensure that all hospital staff are educated and orientated to these criteria is important for ownership and use. The associated costs for this arm therefore involve primarily dissemination of the chosen triggers. It should be the coordinator’s responsibility to ensure that information flyers, calling criteria wall charts, reminders and ID badges are available and that time is allotted in orientation programs to ensure all personnel commencing employment are aware of the system and of how to use it. On the basis of previous experiences, the cost of wall charts, flyers, ID badges and other related items in a large hospital would equate to approximately US $10,000/year. There are a few experimental/theoretical afferent arms techniques that bear mention here. One of the main limitations of the RRS has been underutilization in the form of failing to activate the MET when trigger criteria are present.25,28 The two main barriers to calling the MET are (1) the ability on the part of the primary provider (generally the bedside nurse) to recognize the presence of a trigger, and (2) the activation energy associated with placing a call for help. The first can be partially overcome using physiology-based multiple parameter scoring systems in which all of an individual’s vital signs and clinical observations are input into a formula that calculates a score.29,32–35 The subjectivity of the decision to call is then replaced with an objective threshold score that requires a call. This calculation can be done manually at the bedside (which would require a few minutes per patient with every vital sign check – generally 3–6 times a day on the general wards), automatically using an electronic medical record with customized programming (generally a multi-million dollar investment on the part of the hospital), or continuously using a device that pulls data directly from telemetry monitors to calculate a score (which requires the cost of telemetry monitoring and the additional device for every patient monitored). Any of these systems can theoretically be tied to a hospital paging system, with the use of additional software and/or programming, to automatically trigger the call to the MET. However, the sensitivity and specificity of these scoring systems is at best 84 and 89%, respectively.29,32 Furthermore, there are no data, at the present time, to suggest that the addition of these costly features is superior to a simple set of activation criteria with manual calling.

Quality Improvement Arm Costs A hospital must establish a monitoring system to ensure that the skills and ­knowledge of each team member are sufficient to deliver high-quality care 24 h a day, 7 days a week. There is no point relying on partly-trained staff to respond to

37  The Costs and the Savings

421

these calls, as has been the case with many cardiac arrest teams. In most cases, there is little, if any resuscitation training (excluding Advance Cardiac Life Support – ACLS) in the undergraduate medical programs or outside the specialties of Intensive Care, Emergency Medicine and Anesthesiology. Some hospitals have some staff available during the day but not 24 h/day. This requires educating new staff, collecting and regularly analyzing data on all RRS activations, ensuring that there is a roster of appropriately trained staff 24/7, as well as organizing regular MET/RRT meetings, following up on links between MET/RRT call-identified systemic problems and clinical governance, and following up on any identified missed or unduly delayed calls. The coordinator position should include these responsibilities. Such a coordinator may also need to review notes and disseminate the latest information on advanced resuscitation and maintain the equipment used by the MET or RRT. The MET/RRT needs to be aware of the latest aspects of advanced resuscitation and should be trained in all in-hospital emergencies. Therefore in addition to the cost of the coordinator position, which has already been factored into the efferent arm costs, one needs to factor in training costs, which tend to come in two forms: the cost of the course itself, and coverage to pull nurses away from clinical service in order to train them. There are many training programs available in many countries that could be utilized to ensure that members of the staff are trained in the latest knowledge and skills of advanced resuscitation. These courses vary in cost, but an ACLS re-certification course costs approximately US $100/person and requires 8 h of clinical coverage, at a cost of approximately $300 for a critical-care nurse. Weekly 1-h case-review conferences do not generally require paid coverage. Quality is maximized and costs minimized by restricting the MET members to only as many as are needed to reliably cover the team at all hours. In a mediumsized hospital, this requires approximately 8–12 individuals. Therefore, allowing for 8 h of covered training each year for all responders would cost approximately $4,000/yr in a medium-sized hospital.

Administrative Arm Costs The final component is the feedback or outcome monitoring system. Using outcome indicators such as all deaths, all cardiac arrests, and unplanned ICU admissions, it is crucial to collect data on how many patients who have one of these outcomes had the calling criteria present prior to their event. This information needs to be fed back to the users of the system on a regular basis and can be facilitated by a weekly MET/ RRT meeting in which patients, treatment and outcomes can be discussed. Unlike mortality and morbidity meetings, the MET/RRT meetings need to be an open forum to all staff and can be used as an educational session. Information gathered can provide us with information on where these events most commonly occur in the hospital, what time they occur, and a systematic approach to solutions can be discussed.

422

D. Edelson and R. Bellomo

The costs for this arm are primarily contained within the already-accounted-for coordinator position. However, a support computer and printing costs for the year would add around US $10,000. In addition, part-time clerical support to assist the coordinator would cost approximately US $20,000. Therefore, the total cost for the start-up in a teaching hospital is around US $250,000, while the maintenance each year is approximately US $150,000. In the hospital of one of the authors, in 2008, there were approximately 1,500 MET calls. Given the recurrent expenditure of $150,000 as outlined before, once the system is set up, the cost of each call is $100. This is extraordinarily cheap (cheaper than the daily dose of some antibiotics), given the overall cost of other procedures and the potential benefit. Interestingly, hospitals all over the world spend millions of dollars on setting up fire safety programs and the maintenance of these early warning systems to provide patient safety, yet the numbers of deaths from in-hospital fires is extremely low.12 It should be noted that the forecasting of costs for additional specific medical staff have not been included. This would certainly depend on the numbers of calls. A large hospital might need to employ a dedicated ICU fellow to be on the team (at least for a significant part of the 24-h cycle) to assist with calls during the busiest times of the 24-h cycle. This depending on the healthcare system might translate into additional costs of US $60,000 to $90,000/year.

Societal Costs In considering societal costs, a proper accounting should factor in long-term costs as well as the more immediate ones described above. Therefore, any patient who survives a hospitalization because of the RRS who otherwise would not have, will likely incur future medical costs. For example, the cost of hemodialysis includes the cost of the equipment and medical personnel, patient earning potential lost during therapy and the additional healthcare expenditures that patients with end-stage renal disease, who previously would have died, now incur. This is not trivial; a failure to account for these costs has been shown to bias cost-effectiveness analyses.36 Unfortunately, there is insufficient data at present regarding long-term patient outcomes and therefore costs associated with those outcomes. When these data become available they should be included in all future analyses.

Potential Hidden Costs A common concern about the MET system is that it might “de-skill” ward staff in relation to advanced resuscitation. In other words, having resources take over ­management of suddenly critically ill patients might lead to a reduction in

37  The Costs and the Savings

423

knowledge and capability of clinicians on the ward. This begs the questions, “Do ward staff have those skills in the first place, and if so, how often do they get to ensure that their skills are maintained?” We argue that it makes more sense that ward staff be trained to identify potentially critically ill patients through clinical observation, and then alert appropriately trained personnel with the necessary skills and knowledge who practice these skills on a daily basis. Another concern is that the RRS may be turning many “unexpected deaths” into “expected deaths” simply by making more patients “Not For Resuscitation” (NFR) when the team arrives and evaluates the situation.37 This would decrease a hospital’s unexpected death rate in an artificial way. On the other hand, if the MET system facilitates the making of appropriate, compassionate, dignified and respectful end-of-life decisions, then we consider such changes desirable.

Savings Hospital Savings While the costs of setting up and maintaining an RRS are not negligible, there are significant associated savings that need to be taken into account. These savings are found primarily in decreased resource utilization associated with preventing medical complications as well as increased opportunity for revenue generation, such as elective surgery due to bed availability. Several studies have shown that with the implementation of an RRS, there is a reduction in unplanned ICU admissions, in-hospital cardiac arrests and deaths.38–40 In a recent study comparing the incidence of cardiac arrests in a large teaching hospital, the introduction of an MET system was associated with an 80% decrease in post-cardiac arrest ICU bed-days, and an 88% decrease in hospital bed-days (Figs. 37.1 and 37.2).41 This decrease represented a savings of approximately US $3,000,000 per year for cardiac-arrest patients alone. In the same institution, the introduction of an MET system also reduced overall length of stay for patients having major surgery (Fig. 37.3).42 Given the yearly throughput of such patients in that teaching hospital, this would translate into a savings of approximately US $9,000,000/year. In the same study, unplanned ICU admissions in patients having major surgery were reduced by 44.4%, representing a further approximate saving of AUD $1,000,000. Given these observations, it is difficult to argue that the introduction of an MET system is too costly. In addition, costs attributed directly to in-hospital cardiac arrest were recently studied in the UK and found to average £928.81 per resuscitation. However, for the patients who survived more than 24  h after cardiac arrest, the cost was £1,589.72.43 If these costs were similar in other countries allowing for

424

D. Edelson and R. Bellomo

c­ onservative currency exchange for the MET system to be cost-effective in a large hospital, it would need to prevent 40 cardiac arrests or 20 unplanned ICU admissions or a combination of both. In the study quoted above,41 it prevented 133 cardiac arrests per year and also a similar number of unplanned ICU admissions per year. In another hospital, a 128-bed, non-teaching, community hospital in the US, investigators estimated their annual RRS savings at $171,480.27 This was calculated based on the prevention of six cardiac arrests per year (at a cost of $1,000 per patient) and 24 fewer ICU transfers (mean length of ICU stay 3.5 days), with an increased cost of $1,160 per patient per day in the ICU, as compared to a medicalsurgical unit day. A sample template for calculating costs and savings in this manner can be found on the Institute for Healthcare Improvement Web site (Fig. 37.4).44 Change over a four-month period P<0.0001

MET system Control

ICU bed days

ICU bed days 0

50

100 DAYS

150

200

Fig. 37.1  Diagram illustrating the impact of an RRS on ICU bed-days after cardiac arrest

Change over a four-month period P<0.0001

MET system Control

Hospital bed days Hospital bed days 0

500

1000

1500

DAYS

Fig. 37.2  Diagram illustrating the impact of an RRS on hospital bed-days after cardiac arrest

37  The Costs and the Savings

425 Change over a four-month period P<0.009 MET system Control

Length of stay

Length of stay 0

5

10 15 DAYS

20

25

Fig. 37.3  Histogram illustrating the effect of an RRS on length-of-stay for patients having major surgery in a teaching hospital

Indicator

Key

Formula

A

Annual Number of Cases Benefiting from RRT

Input Data

B

Average Patient Days Saved via RRT in these cases

Input Data

C

Potential Bed Days of RRT-Associated Added Capacity

D

Overall Hospital ALOS

E

Potential Annual Additional Admissions Because of RRT

F

Average Hospital Net Revenue per Case

Amount

AxB Input Data C/D Input Data ExF

G

Potential Annual Incremental RRT-Associated Revenue

H

Annual Cost of Rapid Response Team

I

Profit (Loss) associated with RRT

J

One Time RRT Implementation Costs

Input Data

K

Annual Return on Investment of RRT

I,J

Input Data G-H

From: http//www.ihi.org/NR/rdonlyres/DEAD74FE-089B-4954-9A84-E40B538FD79B/3670/ RRTROIPresentation ppt#427,26,RRT Benefit Calculator Ward WJ. Accessed 2/26/09

Fig. 37.4  Template for calculating costs and savings associated with an RRS

“Societal” Savings While malpractice claims and payouts (and certainly insurance premiums) are felt by the hospital, the brunt of the costs is likely carried by society as a whole. Therefore, savings associated with decreased claims from fewer adverse events and poor outcomes seem better represented from the societal perspective. According to the US

426

D. Edelson and R. Bellomo

Department of Health and Human Services, there were more than 16,000 paid claims against healthcare providers in 2000, amounting to a total payout of approximately $4 billion dollars.45 Similarly, in 1999 there were £386 million of damages assessed for 3,254 claims in the UK.46 These numbers do not include expenses incurred in the process of investigation and defense of claims. In another US report, the total cost of medical malpractice torts were estimated at $28.7 billion in 2004.47 Similar to the cost analysis, the savings analysis should include long-term ­savings associated with patient outcomes. These would primarily be included as earning potential and work productivity saved for all patients who would otherwise have died in the hospital without the RRS. This will again require long-term outcome data to estimate costs in the future.

Summary Clearly, all of the cost and savings issues hinge, in the end, on whether or not the RRS “works.” While it must be acknowledged that the evidence available is the subject of controversy,48 it must also be clearly stated that these are now studies from multiple institutions that, if properly applied with adequate education, commitment and enthusiasm, it can work – and work dramatically.39,42,48–51 We are not aware of any institutions that, having introduced an RRS and having applied it to patient care for at least a year, have then gone back to an “RRS-free” environment. Furthermore, any underpowered multi-center studies, which failed to demonstrate an effect on direct comparison, have more recently clearly shown a strong link between pre-emptive activity and a decrease in adverse events, especially cardiac arrests.52 The available data suggest that the cost of implementing a properly resourced and trained RRS is relatively small. They suggest that the savings are likely to far exceed the costs. More importantly, they suggest that lives are saved, cardiac arrests prevented, and complications of surgery decreased. We believe that a well-organized, well-resourced, well-trained RRS is a highly cost-effective intervention for any modern hospital.

References 1. Raftery J. Costing in economic evaluation. BMJ (Clinical research ed). 2000;320(7249):1597. 2. Clement FM, Ghali WA, Donaldson C, Manns BJ. The impact of using different costing methods on the results of an economic evaluation of cardiac care: microcosting vs. grosscosting approaches. Health Economics. 2009;18(4):377–388. 3. Institute of Medicine. To Err Is Human: Building a Safer Health System: Committee on Quality of Health Care in America. Washington, DC: National Academy Press; 1999. 4. Bates DW, Spell N, Cullen DJ, et al. The costs of adverse drug events in hospitalized patients. Adverse Drug Events Prevention Study Group. JAMA. 1997;277(4):307–311. 5. Classen DC, Pestotnik SL, Evans RS, Lloyd JF, Burke JP. Adverse drug events in hospitalized patients – excess length of stay, extra costs, and attributable mortality. JAMA. 1997;277(4):301–306.

37  The Costs and the Savings

427

6. Schneider PJ, Gift MG, Lee YP, Rothermich EA, Sill BE. Cost of medication-related problems at a university hospital. Am J Health Syst Pharm. 1995;52(21):2415–2418. 7. Senst BL, Achusim LE, Genest RP, et  al. Practical approach to determining costs and frequency of adverse drug events in a health care network. Am J Health Syst Pharm. 2001;58(12):1126–1132. 8. Brown P, McArthur C, Newby L, Lay-Yee R, Davis P, Briant R. Cost of medical injury in New Zealand: a retrospective cohort study. J Health Serv Res Policy. 2002;7(Suppl 1):S29–S34. 9. Wilson RM, Runciman WB, Gibberd RW, Harrison BT, Newby L, Hamilton JD. The quality in Australian health care study. Med J Aust. 1995;163(9):458–471. 10. Committee on Identifying and Preventing Medication Errors PA, Julie Wolcott, J. Lyle Bootman, Linda R. Cronenwett, Editors. Preventing Medication Errors: Quality Chasm Series. National Academies Press. 2006 11. Shaw R, Drever F, Hughes H, Osborn S, Williams S. Adverse events and near-miss reporting in the NHS. Qual Saf Health Care. 2005;14(4):279–283. 12. Alliance WHP. World Health Professions Alliance Fact Sheet: Patient Safety. 2002 13. Franklin C, Mathew J. Developing strategies to prevent in-hospital cardiac arrest: analyzing responses of physicians and nurses in the hours before the event. Crit Care Med. 1994;22(2):244–247. 14. Schein RMH, Hazday N, Pena M, Ruben BH, Sprung CL. Clinical antecedents to in-hospital cardiopulmonary arrest. Chest. 1990;98(6):1388–1392. 15. Bedell SE, Delbanco TL, Cook EF, Epstein FH. Survival after cardiopulmonary resuscitation in the hospital. N Engl J Med. 1983;309(10):569–576. 16. Paradis N, Halperin H, Nowak R. Cardiac Arrest. The Science and Practice of Resuscitation Medicine. Baltimore: Williams and Wilkins; 1996:28–46 17. Hillman K, Brown D, Daffurn K. Implementing the Medical Emergency Team System into Your Hospital. Liverpool: South Western Sydney Area Health Service Publication; 1999 (ISBN:187590976). 18. Angus DC, Rubenfeld GD, Roberts MS, et al. Understanding costs and cost-effectiveness in critical care – Report from the Second American Thoracic Society Workshop on Outcomes Research. Am J Respir Crit Care Med. 2002;165(4):540–550. 19. Weinstein MC, Siegel JE, Gold MR, Kamlet MS, Russell LB. Recommendations of the panel on cost-effectiveness in health and medicine. JAMA. 1996;276(15):1253–1258. 20. Tengs TO, Adams ME, Pliskin JS, et al. Five-hundred life-saving interventions and their costeffectiveness. Risk Anal. 1995;15(3):369–390. 21. Salary Wizard: Staff Nurse-RN-CCU. Available at: http://swz.salary.com/salarywizard/layouthtmls/swzl_compresult_national_HC07000202.html. Accessed February 16, 2009. 22. Wachter R. Rapid Response Teams: Ready for Prime Time? The Hospitalist. Available at: http://www.the-hospitalist.org/blogs/wachters_world/archive/2007/11/27/rapidresponse-teams-ready-for-prime-time.aspx. Accessed February 16, 2009. 23. Chan PS, Khalid A, Longmore LS, Berg RA, Kosiborod M, Spertus JA. Hospital-wide code rates and mortality before and after implementation of a rapid response team. JAMA. 2008;300(21):2506–2513. 24. Jones D, Bellomo R. Introduction of a rapid response system: Why we are glad we MET. Crit Care. 2006;10(1):121. 25. Hillman K, Chen J, Cretikos M, et  al. Introduction of the medical emergency team (MET) system: a cluster-randomised controlled trial. Lancet. 2005;365(9477):2091–2097. 26. Kenward G, Castle N, Hodgetts T, Shaikh L. Evaluation of a medical emergency team one year after implementation. Resuscitation. 2004;61(3):257–263. 27. Thomas K, VanOyen Force M, Rasmussen D, Dodd D, Whildin S. Rapid response team: challenges, solutions, benefits. Crit Care Nurse. 2007;27(1):20–27. quiz 28. 28. Daffurn K, Lee A, Hillman KM, Bishop GF, Bauman A. Do nurses know when to summon emergency assistance? Intensive Crit Care Nurs. 1994;10(2):115–120. 29. Hodgetts TJ, Kenward G, Vlachonikolis IG, Payne S, Castle N. The identification of risk factors for cardiac arrest and formulation of activation criteria to alert a medical emergency team. Resuscitation. 2002;54(2):125–131.

428

D. Edelson and R. Bellomo

30. Hillman K, Parr M, Flabouris A, Bishop G, Stewart A. Redefining in-hospital resuscitation: the concept of the medical emergency team. Resuscitation. 2001;48(2):105–110. 31. Goldhill DR, Worthington L, Mulcahy A, Tarling M, Sumner A. The patient-at-risk team: Identifying and managing seriously ill ward patients. Anaesthesia. 1999;54(9):853–860. 32. Smith GB, Prytherch DR, Schmidt PL, Featherstone PI. Review and performance evaluation of aggregate weighted “track-and-trigger” systems. Resuscitation. 2008;77(2):170–179. 33. Morgan RJM, Williams F, Wright MM. An early warning scoring system for detecting developing critical illness. Clin Intensive Care. 1997;8(2):100. 34. Subbe CP, Kruger M, Rutherford P, Gemmel L. Validation of a modified Early Warning Score in medical admissions. Qjm-Monthly J Assoc Phys. 2001;94(10):521–526. 35. Duckitt RW, Buxton-Thomas R, Walker J, et al. Worthing physiological scoring system: derivation and validation of a physiological early-warning system for medical admissions. An observational, population-based single-centre study. Br J Anaesth. 2007;98(6):769–774. 36. Meltzer D. Accounting for future costs in medical cost-effectiveness analysis. J Health Econ. 1997;16(1):33–64. 37. Chen J, Flabouris A, Bellomo R, Hillman K, Finfer S. The medical emergency team system and not-for-resuscitation orders: results from the MERIT study. Resuscitation. 2008;79(3):391–397. 38. Bristow PJ, Hillman KM, Chey T, Daffurn K, Jacques TC, Norman SL, Bishop GF, Simmons EG. Rates of in-hospital arrests, deaths and intensive care admissions: The effect of a medical emergency team. Med J Aust. 2000;173(5):236–240. 39. Buist MD, Moore GE, Bernard SA, Waxman BP, Anderson JN, Nguyen TV. Effects of a medical emergency team on reduction of incidence of and mortality from unexpected cardiac arrests in hospital: preliminary study. BMJ. 2002;324(7334):387–390. 40. Bellomo R, Goldsmith D, Uchino S, et al. A prospective before-and-after trial of a medical emergency team. Med J Aust. 2003;179(6):283–287. 41. Jones D, Opdam H, Egi M, et al. Long-term effect of a medical emergency team on mortality in a teaching hospital. Resuscitation. 2007;74(2):235–241. 42. Bellomo R, Goldsmith D, Uchino S, et al. Prospective controlled trial of effect of medical emergency team on postoperative morbidity and mortality rates. Crit Care Med. 2004;32(4):916–921. 43. Gage H, Kenward G, Hodgetts TJ, Castle N, Ineson N, Shaikh L. Health system costs of inhospital cardiac arrest. Resuscitation. 2002;54(2):139–146. 44. Ward WJ. The business case for implementing rapid response teams: RRT Benefit Calculator. Available at: http://www.ihi.org/NR/rdonlyres/DEAD74FE-089B-4954-9A84E40B538FD79B/3670/RRTROIPresentation.ppt#427,26,RRT. Accessed February 26, 2009 45. Administration UDoHaHSHRaS. National Practitioner Data Bank 2000: Annual Report. Rockville, MD: Department of Health and Human Services. 2000 46. Fenn P. Counting the cost of medical negligence – NHS litigation authority will be able to report on costs and high risk procedures. Br Med J. 2002;325(7358):233–234. 47. Tillinghast. U.S. Tort Costs and Cross-Border Perspectives: 2005 Update. Towers Perrin. 2005 48. Winters BD, Pham JC, Hunt EA, Guallar E, Berenholtz S, Pronovost PJ. Rapid response systems: a systematic review. Crit Care Med. 2007;35(5):1238–1243. 49. Sharek PJ, Parast LM, Leong K, et al. Effect of a rapid response team on hospital-wide mortality and code rates outside the ICU in a children’s hospital. JAMA. 2007;298(19):2267–2274. 50. Tibballs J, Kinney S, Duke T, Oakley E, Hennessy M. Reduction of paediatric in-patient cardiac arrest and death with a medical emergency team: preliminary results. Arch Dis Child. 2005;90(11):1148–1152. 51. DeVita MA, Braithwaite RS, Mahidhara R, Stuart S, Foraida M, Simmons RL. Use of medical emergency team responses to reduce hospital cardiopulmonary arrests. Qual Saf Health Care. 2004;13(4):251–254. 52. Chen J, Bellomo R, Flabouris A, Hillman K, Finfer S. The relationship between early emergency team calls and serious adverse events. Crit Care Med. 2009;37(1):148–153.

Index

A Acute care education, Ward staff. See also Ward staff costs effectiveness, 402 evidence for benefit in, 407 improvement, evidence, 399–400 response team role, 406–407 resuscitation training, 406 roles and responsibilities, 402 short courses, 406 simulation-based courses, 405 Acute Care Undergraduate Teaching (ACUTE), 400 Acute hospitalist medicine. See Hospitalist Acute Life-threatening Events: Recognition and Treatment (ALERT) course, 406 Administrative limb, 14 aims and objectives, 328–329 clinical governance unit, 333 components, 329–330 costs and savings, 437–438 defibrillators, 16 deteriorating ward patient, chain of events, 328 MET administrative group component interaction, 330 future aspects, 332 management training, 332 members of, 330 nurse unit manager (NUM), 331 outcomes, monitoring, 332 senior hospital administration group, 329, 330 vital sign monitoring, 16–17 Adult crash cart, 316 Afferent limb, 14, 69–70 components, 15 costs and savings, 435–436

deteriorate patient identification clinical observations, 186–187 continuous monitoring advantages, 187–188 track-and-trigger systems, 184–186 regular monitoring and assessment improvement, 182–183 selection/diagnostic/triggering criteria, 15 steps involved, 15 summoning protocol, 188 technologic monitoring, 15 technology role, 188–189 triggering response mechanism, 15 vital signs measurements, 183 Aggregate weighted track-and-trigger systems (AWTTS), 184–185 Aiken, L.H., 152 Aiken’s conceptual model, 152 Allgower, M., 118 Area-under-the-receiver-operating curves (AUROC), 125, 149 AWTTS. See Aggregate weighted track-andtrigger systems B Bader, M.K., 227 Baker, R.B., 248 Baldisseri, M., 277 Barriers, RRS implementation demanding analysis, 174 impediments cost cutting, 171 individual errors, 173–174 medical and nursing education, 173 patients intervention, 172 professional differences, 172–173 patients safety improvement, 176 residency education program, 175

M.A. DeVita et al. (eds.), Textbook of Rapid Response Systems: Concepts and Implementation, DOI 10.1007/978-0-387-92853-1, © Springer Science+Business Media, LLC 2011

429

430 Barriers, RRS implementation (cont.) social changes, 170–171 sources cohort comparison study, 167 entrenched professional silos, 169 general reluctance, 169 healthcare education, 170 MET implementation, 168 staff education, 176 team-triggering criteria, 177 unconsulted doctors, 175 Bayesian approach, 393 Bell, M.B., 69, 149, 347 Bellomo, R., 105, 157, 195, 231, 327, 431 Berg, R.A., 227 Bessant, J., 95 Blood administration team (BAT), 270 Blum, R.H., 382 Bradycardia, 116 Braithwaite, J., 83 Braithwaite, R.S., 42 Brewer, T.L., 248 Brilli, R.J., 243, 245, 247 Buist, M., 99 Buist, M.D., 195, 198 Buri, C., 118 C Cardiac arrests evidence based studies deteriorating patient, 69–70 incidence, aggregate analysis, 73 RRS effectiveness, 71–73 iatrogenic patient death prediction electronic observation data capture, 106 Harvard Medical Practice Study (HMPS), 103 hazards, 99–100 individual and organizational accidents, 100–103 latent conditions, 99–100 nested case control study, 104 patient crisis, risk factors for, 107 prospective cohort studies, 105 QAHCS, 103 retrospective case study, 103 risk of mortality, independent predictors, 106 scoring systems, 106 soft defenses, 99–100 organizational crises theory, 99–100 Carroll, J.S., 42 Casamento, A.J., 198

Index Castle, N., 236 Chan, P.S., 227 Chen, J., 72, 83, 387, 421 Chest pain team, 270–271 Clinical governance unit, 333 Coba, V., 257 Complex system interventions characteristics, 387–388 components, 388–389 cost and cost-effectiveness, 392–393 critical linkages and process, RRS outcomes, 389 health services research, 388 research methodology ANZ Feeding Guideline Trial, 391 incomplete implementation, 391 intra-class correlation coefficients (ICCs), 391 large sample size, 391 MERIT study intervention phase, 391 randomized controlled trial (RCT), 390 time frames, 392 result interpretation, 393–394 sub-system interactions after, 392 Comprehensive Unit Safety Program (CUSP), 30 Condition M, 273–274 Condition O usage goal, 285 medical crisis team activation, indications, 283–284 parameters of, 285–286 patient safety initiatives, 286 reasons for, 283–284 Cooper, S., 380 Cost and saving implications adverse events, 431–432 rapid response system administrative arm costs, 437–438 afferent arm costs, 435–436 efferent arm costs, 434–435 hidden costs, 438–439 quality improvement arm costs, 436–437 societal costs, 438 hospital savings, 439–441 “societal” savings, 441–442 template for calculating, 441 Cox, K., 335 Crash cart process. See Equipments, crisis response; Medication and supplies Cretikos, M., 104

Index Crisis response team. See also Equipments, crisis response; Obstetric-specific crisis team A-B-C (airway-breathing-circulation) ad hoc nature of, 371 ad hoc team being “all-capable” concept, 297 communication, 297 MET responders, roles and goals, 298 operating room crisis teams, 300–302 resource management, 301 structure, 25 designated roles assignment role, 374–375 human-simulator crisis team-training program, 375–376 pediatric crisis response, 377, 379 performance improvement, role related task, 376 pre-set task-completion goals, 375 responsibilities, 376–377 resuscitation training sessions, 379 educational objectives for, 373–374 goals of, 374 herd immunity, 290 hierarchical system, 289 human resources, 294–295 inexperienced anesthesiology residents, 291 medical education model, 289 non-educational healthcare setting, 290 organization fixation error, 292 hospital-crisis team training, 293 institutional or system advantage, 292 latent error correction, 292 role-oriented goals and objectives, 292 team member’s tasks, 293 respiratory distress, 290 roles and goals of, 378 RRS activation, 295 RRT implementation, 290 simulation of, 372–373 structure, 294 teamwork training, 291 time-honored method, 290 time-inefficient process, 290 unique aspects of, 371 untrained, 291 Critical care physicians, 1–2 Critical Care Response Team Project (CCRT), 92 Critical incident stress management (CISM), 341

431 D Dalby, P., 277 Deane, B., 115 Decision making, “Not-for-Resuscitation” status critical-care physicians, 442–443 delegation, variability and change, 422 hospital in-patient population, chronic disease and co-morbidities, 423 junior medical staff and nursing staff, 422 senior clinician, 442 Defibrillators, 16, 324–325 Delayed RRS activation avoidance of, 199 causes of, 198–199 classification, 197–198 consequences of, 197, 198 objective calling criteria, 198 principles of, 195–196 retrospective studies, 196 Delgado, E., 305 Demeritt, B., 248 Denham, C., 342 Deteriorate patient identification clinical observations, 186–187 continuous monitoring advantages, 187–188 track-and-trigger systems aggregate weighted, 184–185 efficiency of, 186 single parameter, 185–186 Deteriorations, 1 DeVita, M.A., 1, 13, 149, 267, 289, 305, 371 Difficult airway team (DAT), 272 Diffusion of innovation, 94 Dodgson, M., 95 Donabedian’s approach, 22 Dongilli, T., 372 Downey, A.W., 197 Duke, T., 245, 247, 248 Duncan, K.D., 215 E Early goal-directed therapy (EGDT) definition, 258 entry criteria, 258 implementation barriers, 261 components, 260 ED-based model, 260 ICU-based model, 261 mobile ICU, 261 protocol, 259

432 Early warning score (EWS) pediatric RRS Bristol tool, 250 Cardiff and Vale Paediatric Early Warning System, 248–249 development, 248 scoring system, 249 vital signs, 125, 149 Edelson, D., 431 Edwards, E.D., 248 Efferent limb, 14–16, 70–71 basic condition response, 268 blood administration, 270 chest pain, 270–271 condition M, 273–274 cost and savings, 434–435 difficult airway, 272 lost patient condition, 271–272 pediatric response, 272–273 stroke, 268–269 trauma, 269–270 Equipments, crisis response airway equipment, 314–315 annual competency evaluations, 306 crash carts adult crash cart, 316 crash code and supply contents, 308–309 fully-stocked, 306 house-wide, 318 pediatric crash cart, 316–317 replacement process, 311 selection of, 315–316 standardization, 315 syringe drawer contents list, expiration dates, and billing, 310 syringe drawer layout, 311 vial drawer layout, 312–313 vial tray contents, medication list, expiration dates, and billing, 310–311 crisis event management, training, 305–306 emergency airway bag contents, 307 implementation, barriers, 322–323 intubating equipment, 306–307 “mock code” scenarios, 306 nursing responder equipment, 314 types of, 306 in University of Pittsburgh Medical Center Presbyterian Hospital, 306–307 Evidence based studies guiding principles, 68 levels of evidence, 68 patient deterioration

Index barriers, 70 cardiorespiratory arrest, 69 physiological limits, 69 simulation, 70 staffing model, 70 vital sign criteria, 70 quality and quantity, 68 RRS effectiveness cardiorespiratory arrest, 72–73 cost, 75 Hawthorne effect, 71 historical control methodology, 71 hospital mortality, 72 long-term mortality, 73 MERIT study data, 71–72 nursing and hospital satisfaction, 72 septic and hypovolemic shock, 72 unanticipated ICU admission, 72 F Failure mode and effect analysis (FMEA), 43–44 Failure-to-rescue (FTR) cardiorespiratory instability, 151 complications, 147–148 dynamic patient-level factors early warning scores (EWS), 149 integrated monitoring systems (IMS), 150 medical emergency teams (METs), 148 personal digital assistant (PDA), 150 sign parameters and thresholds, 150 track-and-trigger mechanisms, 150 hospital-level factors, 152–153 static patient-level factors, 149 system-level factors, 152–153 Family-activated rapid response systems. See Patient-and family-activated rapid response systems Family Activated Safety Team (FAST) program, 205 Flabouris, A., 57, 421 Foraida, M.I., 198 G Gaba, D.M., 382 Galhotra, D., 150 Gao, H., 150 Gibson, R., 245, 247 Gleeson, M., 115 Goals and benefits, RRS implementation, 3 Goldman, L., 50

Index Goldsmith, D., 355 Gosbee, J., 39 Gosman, G.G., 277 Grbach, W.J., 305 H Haase, M., 195 Hahn-Cover, K., 335 Hall, L.W., 335 Hamilton, M.F., 371 Haskell, H., 203 Havstad, S., 258 Healthcare environment, 58 Healthcare systems and integration complex surgery, 84 education programs, 87–88 essential components, 86 evaluation data, 88 hospital system, 84 medical training, doctor/patient relationship, 84 multiple co-morbidities, 84 public hospitals, 83 seriously ill patient response to, 87 at risk, 87 teaching and research, 84 silo-based care, 86 support system, 88–89 Hillman, K., 83 Hirschinger, L.E., 335 Hodgetts, T., 236 Hooker, E.A., 121 Hospitalist as acute providers, 52–53 benefits dedicated physician caring, 51 hospital medicine, 51 literature and journals, 52 history, 49–50 models hospitalist models, 51 hospital physician, 50 PCP model, 50 private practice group, 50 Society of Hospital Medicine (SHM), 50 Hospital quality improvement (QI) process, 3–4 Hospital size and localization adverse events and unexpected deaths, 157 district general hospitals, 162 intensive care unit (ICU), 163 Medical Emergency Team (MET) reviews, 158, 159

433 operative death risk, 158 secondary referral centers, 161–162 small city hospitals, ICU, 163 teaching hospitals, 159–161 Hospital-wide systematic approach failure, 423 Hravnak, M., 147 Hunt, E.A., 21, 371 Hutchinson, J., 115 Hypotension, 117 I Ikeda, M., 118 Institute for Healthcare Improvement (IHI), 2, 88, 216 Integrated monitoring systems (IMS), 150 Intra-class correlation coefficients (ICCs), 391 J Jäderling, G., 413 Jenkins, W., 91 Johnson, L., 227 Johnston, L., 245, 247, 248 Jones, D., 157, 195, 231, 327, 349 K Kaplan–Meier graph, patient survival, 351 Kause, J., 131 Kellett, J., 115 Kenward, G., 236 Kerridge, R.K., 204 Khalid, A., 227 King, S., 204 Kinney, S., 245, 247, 248 Konrad, D., 347, 413 Kosiborod, M., 227 Kowiatek, J., 305 L Lam, S.W., 57 Lawless, B., 91 Leeson-Payne, C.G., 132 Leong, K., 247, 248 Levels of care acute illness, response to, 135 care and resuscitation limitation, 139 definitions, 133 early therapy, emergency department, 137 incorrect placement of patients, 132–133

434 Levels of care (cont.) new generalist physician on-call admission, 138 physician-of-the-week concept, 138 responsibilities, 137–138 new patient admission processes, 137 patient illness identification clinical signs, 134 smart alarms, 135 “track-and-trigger” systems, 134–135 rapid response and medical emergency teams, 138 staffing levels and patient flow, clinical outcomes, 136 ward staff, 135–136 Levine, C., 215 Lighthall, G.K., 361 Longmire, L.S., 227 Lost patient condition, 271–272 Luria, J.W., 245, 247 M Mackowiak, P.A., 120 Magee-Womens Hospital, 278 Marsch, S.C.U., 380 Mason, B.W., 248 McAdams, D.J., 49 McCoig, M., 335 McFadden, J.P., 121 McQuillan, P., 362 Medical emergency team (MET), 221 activation criteria, 233 administrative group component interaction, 330 future aspects, 332 management training, 332 members of, 330 nurse unit manager (NUM), 331 outcomes, monitoring, 332 barriers, 167–177 cost and saving implications administrative arm costs, 437–438 afferent arm costs, 435–436 efferent arm costs, 434–435 hidden costs, 438–439 quality improvement arm costs, 436–437 societal costs, 438 efferent limb teams, 267–275 MET call A G approach, 235–236 causes, 236 standards, 235 nurse-led rapid response team, 215–228

Index physician-led MET, 231–238 structure and roles, 234 Medical trainees patient safety acute medical emergencies, 60–61 provision of care, 59–60 postgraduate clinical training healthcare facilities, 57 hospitalist, 61 integral components, 61 requirements for, 61 specialization, 61–62 undergraduate training, 58–59 Medication and supplies, 318 airway management bags, 321–322 controlled substances, 320 emergency cart defibrillators, 324 exchange process, 321 restocking medications, 321 supply standardization, 323–325 implementation, barriers, 322–323 pediatric crash cart layout of, 320 syringe medication list, 319 vial medication list, 319 warning labels, 320 Miller, M., 21 Mininni, N.C., 355 Mobile sepsis team, 261 Monaghan, A., 248 Morgan, R.J., 125 Murray, A.W., 289 N National Institute of Health and Clinical Excellence (NICE), 400 Neal, B., 227 Needleman, J., 152 Nguyen, B., 258 Non-physician-led rapid response teams (RRTs). See Physician-led MET “Not-for-Resuscitation” status decision making critical-care physicians, 442–443 delegation, variability and change, 422 hospital in-patient population, chronic disease and co-morbidities, 423 junior medical staff and nursing staff, 422 senior clinician, 442 rapid response teams critical care-based staff, 422 evidence for, 424–427

Index hospital in-patients indentification program, 423 hospital-wide systematic approach failure, 423 Nurse-led rapid response team benefits goal, 223 rapid response evaluation, 225 role, 223 staff member feedback, 226 components command process, 220–222 communication tools, 218–219 specific protocols, 219–222 data collection mnemonic tools, 227 quality improvement initiatives, 225–226 tool, 224–225 efficacy, 227 hospital resources identification, 217 leadership role, 217–218 mentoring, 224 Nurse unit manager (NUM), 331 Nurse view, RRS critical thinking, 355 MERIT study intervention survey, 357 nurse-led RRT, 357 physician-led MET, 356–357 Robert Wood Johnson Foundation, RRT evaluation project, 356 RRT activation criterion, 356 successful RRS communication, 357–358 education, 358 empowerment, 358 experience, 358 “not-an-option-to-not-call” practice, 358 O Observed structured clinical examination (OSCE), 59 Obstetric-specific crisis team background, 278 condition O usage goal, 285 medical crisis team activation, indications, 283–284 parameters of, 285–286 patient safety initiatives, 286 reasons for, 283–284 data collection, 282 design core skills, 280–281

435 medical crisis vs. obstetric crisis, 278–279 neonatal resuscitation team, 279–280 nursing, 279 response team training, 282 staff education, 281 Oliver, A., 248 One-tiered system, 244–245 Ontario’s critical care strategy, 92 Otis, A.B., 121 Ott, L., 147 Oximetry, 115, 121 P Parast, L.M., 247, 248 Patient-and family-activated rapid response systems accreditation and safety organizations, 206 features of administration and design, 206–207 data collection, 208 follow-up, 208 patient education, 207 screening, 207–208 team composition, 208 triggering criteria, 207 gauging success, 208–209 legislation, 205 origins of, 204–205 Patient education, 207 Patient safety complexity normalization, 45 creating and sustaining, 40 defect investigation caregiver factors, 32 institutional environment factors, 33 local environment factors, 33 patient factors, 32 process of, 31 task factors, 32 team factors, 32 training and education factors, 32–33 deviance normalization, 45 medical trainees (see also Medical trainees) acute medical emergencies, 60–61 provision of care, 59–60 MET, driving force broad-sweeping, 41 difficulties and limitations, 41 failure mode and effect analysis, 43–44 functions, 41 root cause analysis, 41–43

436 Patient safety (cont.) safety culture and high-reliability organizations, 44–45 organizational evaluation of adverse drug events, 23 components, 24 Donabedian’s approach, 22 institutional scorecard, framework for, 24 intervention, impact of, 24 measuring defects, 25–28 medication safety, 23 organization’s learning and culture, 23 process and outcome measures, 22 safety improvement, reliability checklists, 34–35 evidence-based therapies, 34 physician autonomy, 34 standardization aspects, 29–30 transfusion medicine, 34 Pediatric crash cart layout of, 320 syringe medication list, 319 vial medication list, 319 Pediatric response team, 272–273 critical illness recognition, 245 development, 244 early warning scores Bristol tool, 250 Cardiff and Vale Paediatric Early Warning System, 248–249 development, 248 scoring system, 249 institutional barriers, 253–254 one-tiered system, 244–245 outcomes cardiac and respiratory arrest, prevention, 251–253 rapid response system, origin, 251 two-tiered system, 251 triggers/calling criteria MET activation criteria, 246 surrogate marker, 247–248 types, 245 two-tiered systems, 244–245 Personal digital assistant (PDA), 150 Personnel response, crisis event, 307 Pham, J.C., 67 Physician-led MET advantages, 237 definition, 232–234 disadvantages, 237–238 key members, 234 management plan

Index deteriorating patient, 235 MET call, 235–236 prompt patient review, 231 need for, 236–237 principles, 231–233 Pinsky, M.R., 147 Policy creation, RRS communication strategy, 95 Critical Care Response Team Project (CCRT), 92 diffusion of innovation, 94 healthcare innovation, 94 late majority adopters, 95 Ontario’s critical care strategy, 92 opinion leaders, 94 social science paradigm, 94 Powell, C.V., 248 Primary Care Physician (PCP), 49 Pronovost, P.J., 21 Q Quach, J.L., 197 Quality improvement arm costs, 436–437 Quality improvement limb, 16 R Rao, A.D., 13 Reason, J., 99 Resident training, 361, 363–364 cardiac arrest, 362–363 opportunities for cross-discipline training, 364 residents’ perceptions, 364–365 trainees, 364 patient safety error analysis, 367 patient outcome monitoring, 366–367 practice-based learning, 366 socio-medical interactions, 367 suboptimal care, 362 US Accreditation Council of Graduate Medical Education, 365–366 Reynolds, S.F., 91 Rivers, E.P., 257 Rodgers, E., 44 Rogers, E.M., 94 Root cause analysis (RCAs), 40 S Safar, P., 89 Santiano, N., 15

Index Schmid, A., 147 Scott, S.D., 335 Second victim recovery trajectory case study, 335–336 critical incident stress management (CISM), 341 emotional vulnerability, 336–337 healthcare institutions, 343 high-risk clinical events, 339, 340 impact realization emotional response, 338–339 patient stabilization, 338 stages of, 338 traumatic clinical event, 339 professional colleagues and peers, 340–341 support network, 342 team debriefings, 341–342 Senior hospital administration group, 329, 330 Sepsis response team. See Early goal-directed therapy (EGDT) Seriously ill patient, healthcare systems response to, 87 at risk, 87 Shaefer III, J.J., 289 Shaikh, L., 236 Sharek, P.J., 245, 247, 248 Shearn, D., 267 Shock Index, 118–119 Silber, J.H., 147, 148, 152 Simhan, H.N., 277 Single parameter track-and-trigger systems (SPTTS), 185–186 Smith, G.B., 131, 149, 218, 397 Societal costs, 438 Society of Hospital Medicine (SHM), 50 Spertus, T.A., 227 Spodick, D.H., 116 SPTTS. See Single parameter track-and-trigger systems (SPTTS) Staff education, 281 Staff member feedback, 226 Standardization tool and assessment, RRS process cluster-randomized controlled trial, 413 conventional grading system, 413 data collection, 415–416 evaluation, 416–417 fundamental system change, 414 initiation criteria validation, 414 protocol implementation, 415 trial design, 415 outcomes, 417 single-center studies, 414

437 Stein, K., 277 Stroke team, 268–269 Sund-Levander, M., 120 Surrogate marker, 247–248 T Teaching hospitals, 347 adverse events, 348 afferent limb arm, 349 benefits of, 348 calls distribution, 160 composition of, 159 culture and management, 350 efferent limb arm, 350 eminence-based medicine, 349 Kaplan–Meier graph, patient survival, 351 medical management, 348 medical staff., 352 role of, 161 RRS implementation, 349 Terminology, RRS afferent limb, 5–6 critical care outreach teams, 5 efferent limb, 5–6 history, 4–5 Track-and-trigger systems, 134–135 deteriorate patient identification aggregate weighted, 184–185 efficiency of, 186 single parameter, 185–186 failure-to-rescue (FTR), 150 Trauma team, 269–270 Tucker, K.M., 248 U University of Pittsburgh Medical Center (UPMC). See also Equipments, crisis response Presbyterian hospital differences, 204 team members responsibilities, 268 teams, 267 Shadyside hospital assessment process, 208 condition C (MET) program, 204 condition H, 205 patient advise, 207 V Vital signs age, mortality, 122–124 blood pressure, 115, 117–118

438 Vital signs (cont.) EWS systems, 125 patients proportion, 116, 117 pulse oximetry, 115, 121 pulse rate, 115–117 respiratory rate, 115, 121 Shock Index, 118–119 temperature, 115, 119–120 Vital signs measurements, 183 Vossmeyer, M.T., 248 W Wachter, R.M., 50 Wakelam, A., 380 Ward staff acute care education costs effectiveness, 402 evidence for benefit in, 407 improvement, evidence, 399–400

Index response team role, 406–407 resuscitation training, 406 roles and responsibilities, 402 short courses, 406 simulation-based courses, 405 Acute Care Undergraduate Teaching (ACUTE), 400 DoH competencies grades, 401 organization, 402 recognizers, 401 responders, 401 level 1 care, 398 National Institute of Health and Clinical Excellence (NICE), 400 patient deterioration, 397 Ward, W.J., 75 Welch, J.R., 397 Winters, B.D., 1, 21, 67, 72