Cumitech 1C: Blood Cultures IV

1C Blood Cultures IV ELLEN JO BARON, MELVIN P. WEINSTEIN, W. MICHAEL DUNNE, JR., PABLO YAGUPSKY, DAVID F. WELCH, AND DONNA M. WILSON COORDINATING EDITOR

ELLEN JO BARON

Cumitech CUMULATIVE TECHNIQUES AND PROCEDURES IN CLINICAL MICROBIOLOGY

Cumitech 1C

Blood Cultures IV

Cumitech 2B

Laboratory Diagnosis of Urinary Tract Infections

Cumitech 3A

Quality Control and Quality Assurance Practices in Clinical Microbiology

Cumitech 5A

Practical Anaerobic Bacteriology

Cumitech 6A

New Developments in Antimicrobial Agent Susceptibility Testing: a Practical Guide

Cumitech 7B

Lower Respiratory Tract Infections

Cumitech 12A

Laboratory Diagnosis of Bacterial Diarrhea

Cumitech 13A

Laboratory Diagnosis of Ocular Infections

Cumitech 16A

Laboratory Diagnosis of the Mycobacterioses

Cumitech 18A

Laboratory Diagnosis of Hepatitis Viruses

Cumitech 19A

Laboratory Diagnosis of Chlamydia trachomatis Infections

Cumitech 21

Laboratory Diagnosis of Viral Respiratory Disease

Cumitech 23

Infections of the Skin and Subcutaneous Tissues

Cumitech 24

Rapid Detection of Viruses by Immunofluorescence

Cumitech 25

Current Concepts and Approaches to Antimicrobial Agent Susceptibility Testing

Cumitech 26

Laboratory Diagnosis of Viral Infections Producing Enteritis

Cumitech 27

Laboratory Diagnosis of Zoonotic Infections: Bacterial Infections Obtained from Companion and Laboratory Animals

Cumitech 28

Laboratory Diagnosis of Zoonotic Infections: Chlamydial, Fungal, Viral, and Parasitic Infections Obtained from Companion and Laboratory Animals

Cumitech 29

Laboratory Safety in Clinical Microbiology

Cumitech 30A

Selection and Use of Laboratory Procedures for Diagnosis of Parasitic Infections of the Gastrointestinal Tract

Cumitech 31

Verification and Validation of Procedures in the Clinical Microbiology Laboratory

Cumitech 32

Laboratory Diagnosis of Zoonotic Infections: Viral, Rickettsial, and Parasitic Infections Obtained from Food Animals and Wildlife

Cumitech 33

Laboratory Safety, Management, and Diagnosis of Biological Agents Associated with Bioterrorism

Cumitech 34

Laboratory Diagnosis of Mycoplasmal Infections

Cumitech 35

Postmortem Microbiology

Cumitech 36

Biosafety Considerations for Large-Scale Production of Microorganisms

Cumitech 37

Laboratory Diagnosis of Bacterial and Fungal Infections Common to Humans, Livestock, and Wildlife

Cumitech 38

Human Cytomegalovirus

Cumitech 39

Competency Assessment in the Clinical Microbiology Laboratory

Cumitech 40

Packing and Shipping of Diagnostic Specimens and Infectious Substances

Cumitech 41

Detection and Prevention of Clinical Microbiology Laboratory-Associated Errors

Cumitech 42

Infections in Hemopoietic Stem Cell Transplant Recipients

Cumitechs should be cited as follows, e.g.: Baron, E. J., M. P. Weinstein, W. M. Dunne, Jr., P. Yagupsky, D. F. Welch, and D. M. Wilson. 2005. Cumitech 1C, Blood Cultures IV. Coordinating ed., E. J. Baron. ASM Press, Washington, D.C. Editorial board for ASM Cumitechs: Alice S. Weissfeld, Chair; Maria D. Appleman, Vickie Baselski, B. Kay Buchanan, Mitchell l. Burken, Roberta Carey, Linda Cook, Lynne Garcia, Mark LaRocco, Susan L. Mottice, Michael Saubolle, David L. Sewell, Daniel Shapiro, Susan E. Sharp, James W. Snyder, Allan Truant. Effective as of January 2000, the purpose of the Cumitech series is to provide consensus recommendations regarding the judicious use of clinical microbiology and immunology laboratories and their role in patient care. Each Cumitech is written by a team of clinicians, laboratorians, and other interested stakeholders to provide a broad overview of various aspects of infectious disease testing. These aspects include a discussion of relevant clinical considerations; collection, transport, processing, and interpretive guidelines; the clinical utility of culture-based and non-culture-based methods and emerging technologies; and issues surrounding coding, medical necessity, frequency limits, and reimbursement. The recommendations in Cumitechs do not represent the official views or policies of any third-party payer. Copyright © 2005 ASM Press American Society for Microbiology 1752 N Street NW Washington, DC 20036-2904 All Rights Reserved 10 9 8 7 6 5 4 3 2 1

Blood Cultures IV Ellen Jo Baron Stanford University Medical Center, Stanford, CA 94305

Melvin P. Weinstein Robert Wood Johnson University Hospital, New Brunswick, NJ 08903

W. Michael Dunne, Jr. Washington University School of Medicine, St. Louis, MO 63110

Pablo Yagupsky Soroka University Medical Center, Ben Gurion University of the Negev, Beer-Sheva 84101, Israel

David F. Welch Medical City Dallas and North Texas Children’s Hospital, Dallas, TX 75230

Donna M. Wilson California Department of Health Services, Medi-Cal Benefits Branch, Medical Policy Section, Sacramento, CA 94234

COORDINATING EDITOR: Ellen Jo Baron Stanford University Medical Center, Stanford, CA 94305

Introduction and Rationale for Performing Blood Cultures . . . . . . . . . . . . . . 2 Characteristics of Bacteremia and Fungemia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

Clinical Use of Blood Cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Recommended Volume of Blood To Be Cultured Recommended Number of Blood Cultures . . . . . Timing of Blood Cultures . . . . . . . . . . . . . . . . . . Importance of Separate Blood Cultures . . . . . . . Limitations of Blood Cultures . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

3 4 5 5 5

Collection of Blood Cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Skin Antisepsis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Methods of Obtaining Blood for Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Transport and Initial Processing of Blood Culture Bottles . . . . . . . . . . . . . . . 8 Checklist for Blood Cultures before Leaving the Patient’s Bedside . . . . . . . . . . . . . . . . . . . 8 Transport to the Laboratory and Handling and Moving within the Laboratory . . . . . . . . . . . 8 Bottle Examination, Processing Protocols, and Rejection Criteria . . . . . . . . . . . . . . . . . . . . 8 Safe Handling of Blood Cultures in the Laboratory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Media and Incubation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Culture Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Blood-to-Broth Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Atmosphere of Incubation and Use of Anaerobic Blood Culture Vials . . . . . . . . . . . . . . . . 12 Length of Incubation of Blood Cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Commercially Available Manual Blood Culture Systems . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Automated Blood Culture Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Incubation and Examination of Blood Cultures . . . . . . . . . . . . . . . . . . . . . . 13 General Concepts for Detecting and Initial Handling of Positive Blood Cultures . . . . . . . . 13 Incubation Durations with Automated Blood Culture Systems . . . . . . . . . . . . . . . . . . . . . 13 Incubation Duration and Detecting Positive Blood Cultures in Nonautomated Systems . . . 13 Visual Examination of Smears from Positive Blood Cultures . . . . . . . . . . . . . . . . . . . . . . . 13 Initial Subcultures of Positive Blood Cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Direct Identification and Susceptibility Testing from Blood Culture Broth . . . . . . . . . . . . . . 14 False-Positive Blood Cultures (“Contaminants”) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Polymicrobic Bacteremia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Saving Blood Culture Isolates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Reporting Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16

CPT-4 Coding and Billing Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

1

2

Baron et al.

CUMITECH 1C

Laboratory Diagnosis of Sepsis Caused by a Colonized Indwelling Vascular Catheter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Conventional Methods for Detecting Pathogens That Fail To Grow in Standard Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Fungi . . . . . . . . . . . . . . . . . . . . . Bartonella . . . . . . . . . . . . . . . . . Brucella . . . . . . . . . . . . . . . . . . . Campylobacter and Helicobacter Legionella . . . . . . . . . . . . . . . . . Mycoplasma . . . . . . . . . . . . . . . Leptospira . . . . . . . . . . . . . . . . . Mycobacterium . . . . . . . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

21 21 21 22 22 22 22 22

Molecular Methods for Detecting BSIs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Quality Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Quality Assurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Specimen Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Numbers of Cultures Drawn per Patient . . . . . . . . . . . . . . . . . . . . . . Positive Culture Rate and Number of Cultures per 1,000 Patient Days Technologist Competency and Result Reporting . . . . . . . . . . . . . . . . . Ordering, Billing, and Reimbursement . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . . . . . . . . . 24 . . . . . . . . . . . . 24 . . . . . . . . . . . . 25 . . . . . . . . . . . . 25 . . . . . . . . . . . . 25

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

INTRODUCTION AND RATIONALE FOR PERFORMING BLOOD CULTURES

T

he presence of living microorganisms in the blood of a patient is an event of major diagnostic and prognostic importance (16, 163, 165). When microorganisms multiply at a rate that exceeds the capacity of the reticuloendothelial system to remove them, bloodstream infection (BSI) is the result (13). If there is failure of the host’s defenses to localize infection at its primary focus or failure of a physician’s attempts to remove, drain, or adequately treat localized infection, persistent BSI may result. When blood cultures yield a clinically important pathogen, not only is an infectious cause established for the patient’s illness, but the etiologic agent also becomes available for antimicrobial susceptibility testing and optimization of therapy. Identification of an etiologic agent of septicemia results in more appropriate reimbursement for the resources needed to maintain quality laboratory services, an important issue for health care institutions.

Characteristics of Bacteremia and Fungemia Microorganisms usually enter the blood from extravascular sites via lymphatic vessels. Direct entry of bacteria and fungi into the bloodstream occurs with intravascular infections such as infective endocarditis, infected arteriovenous fistulas, mycotic aneurysms, suppurative thrombophlebitis, and colonized intravascular devices (e.g., intravenous catheters, arterial lines, subcutaneous ports, and vascular grafts).

A sudden influx of bacteria ordinarily is cleared from the bloodstream in minutes to hours, except in the case of overwhelming infection or an intravascular focus (41). The fixed macrophages in the liver and spleen play major roles in clearing bacteria from the blood. Bacterial capsules and other virulence factors delay clearance; specific antibodies enhance removal. Although polymorphonuclear leukocytes (PMNs) are crucial for localizing infections in extravascular sites, intravascular PMNs play a minor role, if any, in clearance. When the body responds to the presence of an infectious agent in the bloodstream with systemic signs and symptoms of illness (e.g., a systemic inflammatory response syndrome), the condition is called septicemia. An understanding of the circumstances in which different types of bacteremia and fungemia are likely to occur is helpful in planning diagnostic studies and interpreting results of blood cultures. The common sources of BSIs are intravascular devices (19%), the genitourinary tract (17%), the respiratory tract (12%), the bowel and peritoneum (5%), skin (5%), the biliary tract (4%), intra-abdominal abscesses (3%), other known sites (8%), and unknown sites (27%) (166). The clinical pattern of BSIs can be transient, intermittent, or continuous. Transient bacteremia is most common and occurs after manipulation of infected tissues (e.g., abscesses, furuncles, and cellulitis), instrumentation of contaminated mucosal surfaces (dental procedures; urologic manipulations such as cystoscopy, urethral dilatation, and catheterization;

CUMITECH 1C

suction abortion; and upper and lower gastrointestinal endoscopic procedures), and surgery involving contaminated areas (transurethral resection of the prostate, vaginal hysterectomy, and debridement of infected burns). Bacteremia also occurs early in the course of many systemic and localized infections and has been reported in various studies in 50 to 80% of patients with meningitis, 5 to 30% of patients with pneumonia, 20 to 70% of patients with pyogenic arthritis, 30 to 50% of patients with osteomyelitis, and 5 to 90% of patients with gonococcal and meningococcal infections (121). Intermittent bacteremia most often is associated with undrained intra-abdominal, pelvic, perinephric, hepatic, prostatic, and other abscesses. Such abscesses are a common cause of fever of unknown origin. Continuous bacteremia is a cardinal feature of endovascular infections, most notably acute and subacute infective endocarditis (119, 168). This pattern may also be seen during the first few weeks of typhoid fever and brucellosis. “Breakthrough” bacteremia occurs while a patient is receiving systemic therapy with antimicrobial agents to which the infecting microorganism is susceptible (4, 164). Breakthrough bacteremia that occurs early in therapy is usually due to inadequate concentrations of the antimicrobial agent, whereas breakthrough episodes that occur later usually are due to inadequate drainage of a focus of infection or impairment of host defenses (4).

CLINICAL USE OF BLOOD CULTURES Based on the discussion above, recommendations for the clinical use of blood cultures can be made. Ideally, blood should be obtained for culture prior to the administration of systemic antimicrobial therapy from any patient who is sufficiently ill to be hospitalized and who has fever (
38°C) or hypothermia (36°C); leukocytosis (total peripheral leukocyte count of more than 10,000 leukocytes per liter), especially with a left shift toward immature or band forms; absolute granulocytopenia (less than 1,000 mature PMNs per liter); or a combination of the above (153). Blood cultures complement urine and cerebrospinal fluid cultures in the evaluation of neonates with suspected sepsis whose only clinical findings, in addition to fever or hypothermia, may be poor feeding and failure to thrive (46). Young children, especially those 2 years old or less, with pneumococcal and Haemophilus influenzae bacteremia may present as outpatients, with a marked fever (
39.4°C) and leukocytosis (total peripheral leukocyte count often
20,000 leukocytes per liter) (39). Nondescript complaints, such as myalgia, malaise, or fatigue, may result from bacteremia in elderly patients

Blood Cultures IV

3

who may remain afebrile (53, 119). A low-grade fever in elderly patients may signal infective endocarditis, especially when accompanied by malaise, myalgia, or stroke (119, 167). Finally, a critically important and fundamental concept is that a blood culture, often referred to as a culture set, is defined as a volume of blood obtained under aseptic conditions (preferably by venipuncture) that is inoculated to one or more bottles or vials (usually containing broth culture medium). A culture, whether it consists of only one bottle or several bottles inoculated from the same venipuncture or line draw, is considered to be positive if one or more than one of the bottles demonstrates growth. The individual bottles from a single venipuncture or line draw are not considered separately. However, for organisms other than putative “contaminants,” discussed in more detail later, higher numbers of circulating organisms per ml of blood result in more positive blood culture bottles among those obtained. Recommended Volume of Blood To Be Cultured The volume of blood that is obtained for each blood culture (culture set) is the single most important variable in recovering microorganisms from patients with BSIs (7, 27, 55, 60, 77, 93, 110, 129, 146, 158). Adult Patients Several studies using conventional manual blood culture methods and one using an early-generation semiautomated blood culture system demonstrated a direct relationship between the diagnostic yield from a blood culture and the volume of blood cultured from adults (55, 60, 77, 146). When the volume of blood was increased from 2 to 20 ml, the yield of positive cultures increased by 30 to 50%. More recently, Cockerill and colleagues reassessed this relationship using the BACTEC 9240 (Becton Dickinson Diagnostic Instrument Systems, Sparks, Md.) blood culture system (27). These authors used an unconventional definition of a blood culture to assess the volume versus yield relationship. In adult patients, separate 20-ml blood samples obtained within a 30min period, each divided equally between an aerobic and anaerobic blood culture vial in 10-ml aliquots, were considered to represent a single culture. Using this definition, the investigators then assessed the incremental yield over a volume range from 10 to 40 ml. In patients without infective endocarditis, volumes of 20 ml increased the yield by 30% compared with volumes of 10 ml, and volumes of 30 ml increased the yield 47% compared with volumes of 10 ml. Although there was an additional increase in yield when 40 ml of blood was cultured, the increment compared with 30 ml was only 7%. Based on all of the currently available data, the recommended

4

Baron et al.

CUMITECH 1C

volume of blood to obtain from adults per culture is 20 to 30 ml. Pediatric Patients The optimal volume of blood that should be obtained from infants and children has not been defined with certainty, but the available data suggest that a direct relationship between the volume of blood cultured and detection of BSIs also exists in this patient population (67, 145). An early study documented that in infants and young children, blood samples of
1 ml detected more bacteremias than samples of 1 ml (145). More recently, Kellogg et al. documented that low-level bacteremia (10 CFU/ml) occurred in 68% of infants up to 2 months of age (66) and 60% of children from birth to 15 years of age (67). Twenty-three percent of episodes had 1 CFU/ml of blood (67). Although the authors did not quantify the incremental yield as a function of blood volume, they concluded logically that volumes larger than the 1 to 5 ml recommended previously will improve detection of BSIs in pediatric patients and improve patient care (38, 161). Based on the premise that it is safe to obtain as much as 4 to 4.5% of a patient’s total blood volume for culture and on the known relationship between total blood volume and patient weight, the recommendations of Kellogg et al. (65, 67), which we have modified slightly, are shown in Table 1. As a practical matter, however, it may be impossible to obtain the recommended volumes from tiny, premature infants. Recommended Number of Blood Cultures Several studies have addressed the number of blood cultures needed to detect BSIs in adults. In 1975, using a manual system with basal broth culture media and 20 ml of blood per culture, Washington reported sequential results from 80 patients with bacteremia at the Mayo Clinic (157). The first blood culture detected 80% of episodes, two blood cultures detected 88%, and three blood cultures detected 99% (79 of 80 bacteremias studied). Subsequently, Weinstein et al. reported results from 282 patients

with bacteremia and fungemia from the University of Colorado (165); these results were based on use of a manual blood culture system with basal broth culture media and culture volumes of 15 ml. In this study, 91% of episodes were detected with the first culture, and 99% (281 of 282 bacteremias) were detected with two blood cultures. Cockerill et al. recently reported results from the mid-1990s using the BACTEC 9240 continuous monitoring blood culture instrument with aerobic resin and anaerobic lytic broth culture media, 20 ml per culture (27). In 163 BSIs without endocarditis, 65% were detected with the first blood culture, 80% with two blood cultures, 96% with three blood cultures, and the remainder with four blood cultures. In patients with endocarditis, the first blood culture was positive in approximately 90% of episodes (27). Cockerill et al. speculated as to the reason why more rather than fewer blood cultures were needed despite using a newer culture system with enhanced culture media. One possibility was that lower levels of bacteremia and fungemia were being detected but that detection of low numbers of organisms required more blood cultures. Another possibility was that more patients were receiving broad-spectrum antimicrobial therapy (an unmeasured variable in their study), with a concomitant impairment of bacterial growth that increased the number of blood cultures required for optimal diagnostic sensitivity. It can be concluded from the available data that two to four blood cultures are necessary to optimize detection of bacteremia and fungemia. Although the traditional recommendation in routine circumstances has been two or three blood cultures of at least 20 ml, the recent report by Cockerill et al. (27), if corroborated for other systems and media, may establish four blood cultures (80-ml total volume) as the optimum, although this volume of blood may lead to nosocomial anemia and may not be safe to remove from some severely ill and septic patients. Given that a substantial number of bacteremias would be missed if only one blood culture were obtained, laboratory policy with appropriate medical board approval should facilitate including a

Table 1. Blood volumes suggested for cultures from infants and childrena Wt of patient

a

Total blood vol (ml)

kg

lb

1 1.1–2 2.1–12.7 12.8–36.3 36.3

2.2 2.2–4.4 4.5–27 28–80 80

Modified from reference 67.

50–99 100–200 200 800 2,200

Recommended vol of blood for culture (ml) Culture no. 1

Culture no. 2

Total vol for culture (ml)

2 2 4 10 20–30

2 2 10 20–30

2 4 6 20 40–60

% of total blood vol 4 4 3 2.5 1.8–2.7

CUMITECH 1C

second blood culture as a reflexive test, even if the physician has specified only one blood culture. A mechanism for physicians to opt out of this standard protocol must, of course, be available. Documentation of the medical necessity of the second blood culture requires that the billing system recognizes that it is not a duplicate order, but a distinct procedural service. This is usually done by adding a modifier (-59) to the Current Procedural Terminology 4 (CPT-4) code (i.e., 87040-59). Timing of Blood Cultures Few studies have systematically evaluated the timing of blood cultures and the optimum interval between successive blood cultures. Experimental studies have shown that after an influx of bacteria into the bloodstream, there is a lag time of approximately 1 h, after which chills (rigors) occur (13). Fever then follows. However, Thomson et al. observed no significant differences in positivity rates of blood cultures obtained in relation to the fever spikes of patients (147). As a practical matter, blood cultures are usually obtained after the onset of fever; however, by this time the blood may be sterile due to the efficiency of clearance mechanisms. Thus, blood cultures should be obtained as soon as possible after the onset of fever or chills. Some authorities have recommended arbitrary intervals of 30 to 60 min between blood cultures, except in critically ill septic patients, from whom specimens should be obtained rapidly, prior to initiation of antibiotic therapy (142). However, Li et al. demonstrated no difference in yield whether blood samples for cultures performed within a 24-h period were obtained simultaneously or at spaced intervals (77). For these reasons, the clinical status of the patient should be the primary guide to the timing of blood cultures. In urgent situations where prompt administration of antimicrobial therapy is mandated, blood cultures should be obtained simultaneously or over a short time frame (e.g., less than 1 h). In less urgent situations in which the patient is relatively stable, drawing blood at spaced intervals, such as 1 to 2 h apart, may be indicated. This may be especially helpful if the clinician wishes to document continuous bacteremia in patients with suspected endovascular infections. Recommendations for the timing of blood cultures are shown in Table 2. Importance of Separate Blood Cultures Obtaining more than a single blood culture is important both for diagnostic sensitivity and for interpretation of the clinical significance of a positive result. As indicated previously, studies have shown that two to four blood cultures may be needed for optimum diagnostic sensitivity (27, 157, 165). Equally impor-

Blood Cultures IV

5

Table 2. Recommendations for the timing of blood cultures in different clinical conditions and syndromes Condition or syndrome Suspected acute primary bacteremia or fungemia, meningitis, osteomyelitis, arthritis, or pneumonia

Fever of uncertain origin (e.g., occult abscess, typhoid fever, brucellosis or other undiagnosed febrile syndrome)

Suspected bacteremia or fungemia with persistently negative blood cultures

Recommendations Obtain two or three blood cultures, one right after the other, from different sites following the clinical events that precipitated the blood culture Obtain two or three blood cultures, one right after the other, from different sites initially. If these are negative after 24–48 h of incubation, obtain two more blood cultures, one right after the other, from different sites Consider alternative blood culture methods designed to enhance recovery of mycobacteria, fungi, and rare or fastidious microorganisms (see text)

tant, the number of blood culture sets that grow microorganisms, especially when measured as a function of the total number of sets obtained, has proved to be a useful aid in interpreting the clinical significance of positive blood cultures (83, 165, 166). In contrast to patients with infective endocarditis (continuous bacteremia or fungemia) or other true positive BSIs (transient or intermittent bacteremia or fungemia), patients whose blood cultures grow contaminants usually have only a single blood culture (when two or more are obtained) that is positive (Fig. 1). This information has great practical value for both clinicians and microbiologists. Limitations of Blood Cultures Blood culture methods and techniques, as described in this Cumitech, represent the current state of the art and optimal practice. However, a true gold standard for the diagnosis of BSIs does not exist. Current systems require hours to days of incubation until microbial growth is detected. No single commercially available system or culture medium has been shown to be best suited for the detection of all potential blood pathogens. Some microorganisms grow poorly, if at all, using conventional methods and reagents and require special diagnostic techniques as detailed elsewhere in this document. In the future, it is possible that current culture-based systems will be replaced or supplemented with molecular techniques that are not only more sensitive clinically and analytically, but also faster. Whether any new system will be more cost-effective than cultures remains to be determined.

6

Baron et al.

CUMITECH 1C

Figure 1. Patterns of positivity in successive blood cultures: evidence of the diagnostic importance of separate cultures. Reprinted from reference 165 with permission from the University of Chicago Press.

COLLECTION OF BLOOD CULTURES Skin Antisepsis The likelihood that a positive blood culture represents infection rather than contamination is, at least in part, a function of skin antisepsis at the time blood is obtained. Failure to adequately cleanse the skin using meticulous technique and an appropriate antiseptic agent increases the risk that microbial flora of the skin, such as coagulase-negative staphylococci or Corynebacterium spp., will contaminate the blood culture. Initial cleansing with 70% isopropyl alcohol, followed by use of 1 to 2% iodine tincture or an iodophor has been recommended as standard practice (38, 121). Iodophors (aqueous iodine solutions) require 1.5 to 2 min of contact time for maximum antiseptic effect, whereas iodine tincture (iodine in alcohol) exerts its effect after 30 s (71). It is the change of state from wet to dry of the agent that causes bacterial cell wall disruption, and alcoholbased antiseptics dry more quickly than water-based products. Health care workers who obtain blood cultures (many of whom have little or no formal medical education) are often in a hurry, do not understand the importance of antiseptic preparation contact time, and are less likely to wait 1.5 to 2 min as opposed to half a minute before obtaining blood. At least two studies have documented lower contamination rates using iodine tincture rather than an iodophor (80, 143). Another report compared the use of 0.2% chlorine peroxide and 10% povidoneiodine and demonstrated lower contamination rates when chlorine peroxide was used (139). A study

comparing an alcoholic solution of 0.5% chlorhexidine gluconate versus povidone-iodine for skin antisepsis prior to blood culture demonstrated reduced contamination rates with chlorhexidine (94). Finally, a recent report assessed contamination rates when skin was prepared with iodine tincture versus a commercial product containing 2% chlorhexidine gluconate and 70% isopropyl alcohol (11); in this study, there was no significant difference in the contamination rates associated with the two preparation methods. Thus, the available data suggest that iodine tincture and chlorhexidine products are likely to be equivalent and that both may reduce contamination rates to a greater degree than products containing povidone-iodine preparations. Chlorhexidine preparations have the advantage of being both colorless and less irritating to skin, so that their use may allow one to abandon the additional step necessary with iodine preparations of removing the iodophor using a final alcohol scrub after the venipuncture is completed. Both iodophors and chlorhexidine may have toxicity for neonates; further studies are needed (47). One study found skin preparation for catheter insertion using 0.5% chlorhexidine in 70% isopropyl alcohol was safe and more effective than 10% povidone-iodine for neonates at least 7 days old (49). Until further studies are available, it is important to wash the area with alcohol when these disinfectants are used on the skin of neonates. Some institutions substitute two separate alcohol cleansings, allowing the alcohol to dry thoroughly after each use, for the more active disinfectants if neonates are known to be sensitive to the other agents. A 2002 guideline for

CUMITECH 1C

catheter insertion suggests 2% chlorhexidine as the agent of first choice for skin antisepsis, but no recommendations were made for children 2 years old (45). The venipuncture site should be prepared by first vigorously cleansing for 30 s back and forth across the site with 70% isopropyl alcohol followed by either 1 to 2% tincture of iodine, which is allowed to dry for at least 30 s, or 2% chlorhexidine, which is allowed to dry for at least 30 s, before inserting the needle. If aqueous iodophors are used after the alcohol step, they are applied in increasing concentric circles from the actual venipuncture spot because the disinfectant takes so long to act (to dry) that viable contaminants might be reintroduced to the area that had been prepared if the swab is allowed to recontact the initial site. There are no data supporting applying alcoholic disinfectants in an outward concentric circle, but vigorous friction is important. If palpation of the vein is necessary after skin disinfection, the gloved finger should be cleansed with the antiseptic agent and allowed to dry before touching the site. Some commercial preparations incorporate alcohol and disinfectant into a sponge to allow a one-step process. Methods of Obtaining Blood for Culture Venipuncture remains the technique of choice for obtaining blood for culture (6, 161). Arterial blood cultures are not associated with higher diagnostic yields than venous blood cultures and are not recommended (121). Blood cultures obtained from indwelling intravascular access devices are associated with greater contamination rates than are blood cultures obtained by venipuncture (17, 31, 42). Although physicians and nurses may think they are saving patients the pain of an extra needle stick when blood cultures are obtained from catheters as opposed to venipuncture, they may actually be doing their patients and the health care system a disservice if contaminants are grown, resulting in the need for even more blood cultures and costly additional diagnostic studies or the immediate institution of long-term intravenous antibiotic therapy. Blood obtained through an indwelling line is twice as likely to yield a contaminant than blood obtained through a properly prepared skin site (17). So although blood occasionally may need to be obtained from intravenous lines and similar access devices, a culture of blood from such a device should be paired with another culture of blood obtained by venipuncture to assist in interpretation in the event of a positive result. In the era before human immunodeficiency virus (HIV) infection, blood cultures were obtained with a “two-needle” technique, using a sterile needle and syringe to obtain blood, then changing to a second

Blood Cultures IV

7

needle before inoculating the culture vial. This technique was used to reduce potential contamination in case any skin microorganisms might be present on the first needle following the venipuncture. With increased concern about the risk of HIV transmission to health care workers associated with needle-stick injuries, the two-needle technique was abandoned when several studies showed that contamination rates were not significantly increased when a single needle was used for both venipuncture and inoculation of blood culture vials (61, 73, 76). Although a subsequent meta-analysis demonstrated somewhat higher contamination rates with the single-needle method (3.7%) compared with the two-needle method (2.0%), the higher rates are tolerated in order to reduce the risk of occupational needle-stick injuries (118). Blood may also be inoculated directly to evacuated culture vials containing broth media through a transfer set or a double-ended needle, or into an evacuated blood collection tube containing sodium polyanetholsulfonate (SPS) used for transfer to the laboratory where the specimen is then inoculated to culture vials. If the latter method is used, SPS is the preferred anticoagulant, since citrate, heparin, EDTA, and oxalate may be toxic for some bacteria. However, SPS is also toxic to some bacteria, such as Neisseria meningitidis (140). If a direct inoculation method is used, collection bottles or tubes must be held below the level of the venipuncture needle to minimize the risk of reflux. In general, use of intermediate collection tubes is discouraged, since (i) SPS present in collection tubes will be added to that present in blood culture bottles, thereby increasing the final concentration of SPS in the blood-broth mixture; (ii) the extra step of transferring blood to ultimate bottles, tubes, or plates provides additional opportunity for contamination; (iii) transferring blood increases the risk of exposure of laboratory personnel to blood-borne pathogens; and (iv) use of collection tubes instead of direct inoculation into vials containing media may compromise cultures that are delayed or require transport to a distant laboratory. Use of direct-draw methods does not allow the amount of backpressure to be controlled as is the case when a needle and syringe method is used; the arbitrary vacuum may result in collapse of veins and inability to obtain the correct volume of blood from some patients, such as the frail elderly or those on long-term chemotherapy. In addition, the vacuum present within culture vials may not be precise, and the volume of blood obtained may not be optimal. After the appropriate volume of blood is obtained and inoculated to culture vials, the mixture should be gently agitated or the vials inverted to prevent clotting. If the vein is missed initially, a new needle (or

8

Baron et al.

transfer set) should be used for each repeat venipuncture. As with all specimens submitted to the laboratory, blood culture vials should be labeled with the appropriate identification information and accompanied by an electronic or written requisition showing date and time of collection, the identification of the person collecting the specimen, and any other information required by institutional policy.

TRANSPORT AND INITIAL PROCESSING OF BLOOD CULTURE BOTTLES Checklist for Blood Cultures before Leaving the Patient’s Bedside 1. All bottles or tubes are labeled with two unique patient identifiers, such as name and birthdate or name and Social Security number. 2. All bottles or tubes and requisitions are labeled with collector’s employee name, number, or code. 3. The specific site of collection (which vein, which arm, etc.) is recorded for each set of bottles or tubes obtained. Denoting individual lumens has no additional value, and there is no literature to support the clinical relevance of results from an individual lumen. One method of obtaining this information is to supply test codes for order entry that represent specific sites. Examples of such a code are as follows: AL, arterial line; DIALL, dialysis catheter left; and MEDL, mediport left. 4. If the volume in any bottle is less than the total amount desired, the actual volume of blood injected into the bottle should be documented on information accompanying the bottle to the laboratory. The volume of blood in collection tubes can be measured when the blood is transferred into blood culture broth. Transport to the Laboratory and Handling and Moving within the Laboratory 1. Timing. Blood should be transported as quickly as possible to the laboratory, preferably within 2 h, and should be placed into the incubator as quickly as possible. Although most modern blood culture instruments will detect growth at any stage, delay beyond 2 h in incubating cultures usually results in delay in detection of positives. Bottles from systems that depend on colorimetric detection of positives, such as BacT/ALERT (Table 3), should be examined visually for color change of the indicator before being placed into the instrument. Those with positive color changes can be managed immediately. For systems that depend on other detection methods, the manufacturer’s recommendations for incubating bottles delayed in

CUMITECH 1C

transit should be followed. For example, BACTEC (Table 3) bottle handling is based on media type. For Aerobic PLUS, Anaerobic PLUS, Peds PLUS, and lytic media, bottles may still be placed into the instrument after delays of 20 h at 35°C or 48 h at room temperature (20 to 25°C). Standard aerobic and anaerobic bottles, however, can only be placed into the instrument within 12 h at 35°C or 48 h at room temperature. Those samples not received within these time frames must be stained and subcultured and incubated manually off-line. 2. Temperature. Blood for culture should never be refrigerated or allowed to cool. Keeping the bottles warm (no warmer than 37°C) is preferable to leaving them at room temperature, although organisms in blood culture broths should remain viable at room temperature for several days. Blood collected in tubes should remain at room temperature until it is injected into culture bottles. Because of their relatively small headspace and thin walls, tubes could be stressed by gas formation from growing organisms enough that they may explode if they are incubated. 3. Safety for transport. Blood culture bottles should be carried in some sort of container that will protect them from dropping and from knocking against each other. Carrying tubes or bottles in the hands is dangerous and should not be done, even for short distances within the laboratory. If tubes or bottles must travel through a pneumatic tube system, they should be checked in advance for ability to withstand the harshest conditions of transport. Many people feel that if the container is intact after being dropped to the floor from a 4foot height, it will be strong enough for transport in a pneumatic tube. There are no published studies on container strength.

Bottle Examination, Processing Protocols, and Rejection Criteria Staff who receive blood cultures in the laboratory should check the bottles and requisitions carefully to detect a number of problems or errors, listed here. 1. Depending on laboratory and hospital guidelines, specimens with improper labeling may need to be rejected. Bottles with no labels are usually rejected at all times. Patient identification data on the culture bottles and requisition must match, and mismatched specimens may have to be rejected, depending on laboratory policy. Whenever rejection is being considered, the physician and/or nursing unit must be notified immediately so they can recollect the samples or discuss the options with a laboratory director or supervisor.

Miscellaneous

Media now available in shatter-resistant, plastic bottles that retain transparency and thermal resistance of glass. Bottles do not require venting and contain a liquid emulsion CO2 sensor and larger headspace with performance equal to traditional glass bottles. Controller module has touch-activated panel for text-free random loading and unloading of samples. Bar code scanner recognizes bottle type and patient accession data.

Mycobacterial cultures; monitoring bacterial contamination of platelets Standard aerobic (SA) and anaerobic (SN), FAN aerobic (FA) and anaerobic (FN), pediatric FAN (PF), mycobacteria process (MP) and blood (MB). FAN media formulations contain activated charcoal particles for nonspecific removal of toxic inhibitors of bacterial growth.

Additional indications

Available media formulations

Rocking, 70 cycles/min 10 ml standard; 4 ml for pediatric bottles; requires external volume control

bioMerieux, Inc., Marcy I’Etoile, France, and Durham, N.C. CO2 production Colorimetric CO2 sensor in base of bottle, LED illuminator and photodiode detector in incubator module Visual and audible alerts with remote alarm capability 120 or 240 6 1,440 Once every 10 min

BacT/ALERT 3D

Aerobic bottle agitation Blood volume

Bottle capacity/unit Unit capacity/system Maximum bottle capacity/system Bottle monitor frequency

Positive culture signal

Growth indicator Detection mechanism

Manufacturer

Characteristic

Mycobacterial cultures, possibly sterile fluid cultures BACTEC Standard/10 AEROBIC/F and Standard 10 ANAEROBIC/F (aerobic and anaerobic media without resins), BACTEC PLUS AEROBIC/F and PLUS ANAEROBIC/F (aerobic and anaerobic media with antimicrobial removal resins), BACTEC PEDS PLUS/F medium (small sample volume aerobic medium with resins), BACTEC Lytic/10 ANAEROBIC/F medium (contains a lysing agent to release intracellular organisms—supports the growth of obligate anaerobic and facultative organisms), BACTEC Myco/F Lytic for growth of mycobacteria and fungi A new system called the BACTEC LX is currently in clinical trials. This system detects CO2 production directly in the headspace of the bottle through the use of infrared laser spectroscopy.

Rocking, 30 cycles/min 10 ml standard; 3 ml for pediatric bottles; requires external volume control

CO2, H2, and N2 production, O2 consumption Pressure transducer detects increase/decrease in bottle headspace pressure

CO2 production Fluorimetric CO2 sensor in base of bottle, LED illuminator and photodiode detector in incubator module Visual and audible alerts with remote alarm capability 50, 120, or 240 depending on system choice 5 1,200 Once every 10 min

The VersaTREK system represents a total redesign of the ESP blood culture system originally developed by Difco Laboratories and provides numerous enhancements in ergonomic and mechanical design as well as software upgrades. The system LCD screen provides one-touch access to bottle information and allows new bottles to be scanned and loaded into any available culture station.

Visual and audible alerts with remote alarm capability 240 or 528 6 3,168 Once every 12 min for stirred bottles; once every 24 min for stationary bottles Vortexing with a magnetic stir bar at 3,100 rpm 5 ml maximum (40-ml bottle) or 10 ml maximum (80-ml bottle), 0.1 ml minimum; accurate vacuum draw Culture of sterile body fluids, mycobacterial cultures and susceptibility testing REDOX 1 (aerobic) and REDOX 2 (anaerobic) in 80-ml and 40-ml bottles, the latter available as direct measured draw bottles, MYCO medium with compressed cellulose sponges for culture of mycobacteria

Trek Diagnostic Systems, Cleveland, Ohio

VersaTREK

BD Diagnostic Systems, Sparks, Md.

BACTEC 9000

Table 3. Characteristics of the three major continuously monitored automated blood culture systems currently used in the United States

CUMITECH 1C Blood Cultures IV 9

10

Baron et al.

2. Bottles must be checked for leaking or blood on the outside of the bottle. All blood products should be handled with gloves to avoid skin contact with blood. If the bottle is intact but blood is found on the outside, use a disinfectant solution (10% bleach is best) to wipe off the bottle and then proceed to accession and process it. The collecting person should be notified immediately by telephone of the potential for causing risk to laboratory workers, and a note should be placed on the report stating that the sample was received contaminated, e.g., “container received leaking; specimen was accepted and nurse manager (name) was called by (name) at (time, date).” Lack of observation of laboratory rules for specimen submission by an individual or patient care unit should be monitored and appropriate educational activities or sanctions should be undertaken for continual violations. 3. In some rare instances, bottles may crack before use or in transit to the laboratory. If the crack does not destroy the integrity of the bottle, i.e., there is no detectable fluid leaking out, place a note externally to warn handlers to be careful and continue to process the bottle. 4. Specimens submitted in wrong containers or tubes, such as a purple-top EDTA Vacutainer tube, cannot be processed as blood cultures. Notify the physician or collector immediately that the specimen is being rejected and explain the proper method. Request resubmission of a sample using correct protocols. 5. Blood cultures submitted in expired tubes or bottles should be processed, but the collecting site should be notified immediately to discard all expired bottles and obtain current supplies. The report should note that media were expired when specimens were received and that results may not be reliable. Part of ongoing quality assurance is to monitor all media for expiration dates and prevent such failures from recurring. 6. Incorrect volume of blood in the bottle, either too low or too high, should be noted on the report, but the blood cultures should be processed. Annual monitoring of volumes should be used for an ongoing educational process to reinforce the importance of correct volumes for the system. 7. Too few bottles, or a single blood culture set when two sets are recommended, should be noted on the report, but the blood cultures should be processed. Continual monitoring of adherence to protocols and recommendations by units or clinics is part of annual Quality Assurance. Repeat violators should be counseled and sanctioned as needed.

CUMITECH 1C

8. More than the recommended number of blood cultures per 24-h period should be processed, but the physician should be called to discuss the lack of improved detection after a threshold volume of blood is in progress. Infectious diseases specialists can be enlisted to discuss these issues with the ordering physicians. 9. If blood cultures were inoculated more than 2 h before receipt in the laboratory, it is prudent to examine the bottles for evidence of growth (bubbles, sediment, lysis of red cells, change in color indicator, as indicated above in the Timing section) before incubating the bottles. Some rapidly growing bacteria or large numbers of organisms in the blood could be recovered at this time without further incubation. Lysis-centrifugation tubes (Isolator; Wampole Laboratories, Princeton, N. J.; see description below), because of the lysing component and other additives in the tubes, may show delayed or even decreased recovery of organisms if processing is performed more than 8 h after collection. Practically speaking, the laboratory should attempt to process all correctly labeled blood cultures, despite problems with transport or delayed delivery. Results should then include a disclaimer statement documenting the extent of the delay between collection and incubation, with a comment that results may not be as reliable as those from properly handled cultures. Safe Handling of Blood Cultures in the Laboratory 1. Blood culture bottles should be transported in a container that prevents them from falling, knocking into each other, or rolling off the surface. They should never be carried from one area of the laboratory to another in the hands of laboratorians, but should be placed in a carrier device. 2. Blood and body fluids require standard precautions for handling (48). That entails wearing gloves and working with the bottle behind a shield, or wearing goggles during performance of any manipulations that may generate droplets or aerosols. 3. Withdrawing blood from positive blood culture bottles can be dangerous. Especially if the septum is seen to be bulging or if gas is visible in the broth, the bottle can explode or pressure can cause aerosols when the septum is pierced. Such bottles should be handled in a shielded safety cabinet, and the person inserting a needle into the bottle should wear a full face shield to guard against splashes that might breach the laminar flow barrier. It is advisable to handle all positive blood cul-

CUMITECH 1C

ture bottles in a biological safety cabinet, if possible. Bottles with mycelial phase mold growth should always be handled in a safety cabinet. If the mold is white and potentially Coccidioides, the bottle should not be opened at all unless it is the only way to make the diagnosis. 4. Special blunt-ended needles should be used to insert into the septa for withdrawing samples from positive blood culture bottles, to decrease the possibility of a needle-stick injury. The bottles should be held firmly, since these needles require additional pressure to penetrate the septum. Careful attention to technique is important to avoid accidents. 5. Bottles should be inspected for leaks and cracks at the time of accessioning and again before placing them into the incubator system. Bottles that leak or those with cracks all the way through the bottle must be rejected and redrawn. 6. Used blood culture bottles, both positive and negative, should be disposed of in hard-sided “sharps” biohazard containers or other puncture-resistant infectious waste containers, rather than plastic bags, which are prone to rip and spill the bottles out. The containers should not be filled so full that disposal workers have difficulty lifting them. 7. Most laboratories either terminally decontaminate (e.g., autoclave) all disposed blood culture bottles and tubes before sending them to the landfill, or make arrangements with a licensed contractor to properly handle them as biohazardous infectious waste.

MEDIA AND INCUBATION Culture Media Multiple nutritionally enriched broth-based culture media have been used successfully; no one medium or system will detect all potential pathogens optimally. The most widely used medium is soybean-casein digest broth (118). Other media include brain heart infusion, supplemented peptone, Columbia, and brucella broths. Most manufacturers supplement their base media with proprietary additives designed to enhance microbial growth, so it cannot be assumed that common generic media (e.g., soybean-casein digest broth) from different manufacturers will perform in an equivalent manner. Decisions as to the choice of medium formulation in individual laboratories should be based on data from well-controlled field trials in which large numbers of cultures were assessed. In general, most commercially marketed blood culture media perform well for common blood pathogens. Medium formulations designed to enhance

Blood Cultures IV

11

the detection of anaerobes, fungi, and mycobacteria are also marketed commercially. A more detailed discussion of detection of fungemia and mycobacteremia can be found elsewhere in this Cumitech. Most commercially available blood culture media are able to support the growth of yeasts such as Candida spp. Although blood should be obtained for culture before the administration of antibiotics, many patients with suspected BSIs are already receiving empiric antimicrobial therapy at the time blood cultures are obtained (166), potentially reducing the sensitivity of the test. With dilution of the blood in the culture broth, the concentration of the antimicrobial agents will be reduced and the inhibitory effects minimized. Moreover, the amount of SPS in most commercial blood culture media will inactivate some aminoglycosides and polymyxins. Nevertheless, because of concerns about microbial recovery from blood cultures in patients receiving systemic antimicrobials, some manufacturers have marketed products to enhance detection in this common clinical situation. These include media containing antibioticbinding resins (Plus/F, BD Diagnostic Systems, Sparks, Md.) and activated charcoal (FAN, FA, FN; bioMerieux, Durham, N.C.). Data from large, multicenter studies comparing media with activated charcoal versus basal media formulations suggested that the medium formulations with activated charcoal (i) had improved yield of microorganisms, including staphylococci, enteric gram-negative rods, and yeasts (162, 170); (ii) had improved yields in patients receiving theoretically effective antimicrobial therapy (90); and (iii) detected more contaminants, especially coagulase-negative staphylococci (162). Comparative studies of resin media versus charcoal-containing media have suggested generally equivalent performance overall (64, 111). Interpretation of Gram-stain results from positive cultures may require more technical expertise and experience when the charcoalcontaining medium formulations are used (144). Blood-to-Broth Ratio Human blood contains substances (e.g., complement, lysozyme, and phagocytic leukocytes) capable of inhibiting microbial growth. Diluting blood in culture broth reduces the concentrations of these inhibitory substances, as well as the concentrations of any antimicrobial agents that may have already been administered to patients. Studies that have addressed the blood-to-broth ratio have recommended 5- to 10-fold dilution (8, 120, 128). Dilutions less than 1:5 may result in reduced yield and may increase the chances that blood will clot, thereby trapping potential pathogens within the clot; thus,

12

Baron et al.

filling blood culture bottles with more than the recommended volume of blood should be avoided. Atmosphere of Incubation and Use of Anaerobic Blood Culture Vials Traditional two-vial blood culture systems have included one aerobic and one anaerobic vial, with blood distributed equally to the two vials. Since the 1970s, the proportion of BSIs due to anaerobes has decreased, whereas the proportion due to fungi has increased (35, 54, 81, 99). Several studies in the 1990s of adults (95, 99, 107, 135) and children (178) concluded that routine use of anaerobic blood culture vials was not necessary and recommended that they be used only selectively for patients who are at high risk for anaerobic bacteremia. By contrast, a recent study comparing two FAN aerobic bottles with activated charcoal versus one FAN aerobic and one FAN anaerobic bottle, both with activated charcoal, demonstrated increased yield of staphylococci, Enterobacteriaceae, and anaerobes when the aerobic/ anaerobic combination culture set was used (124). As very few fungemias were present in this study, the potential reduction in detection of these obligate aerobes with use of the combination set could not be assessed. It is difficult to make conclusive recommendations regarding routine versus selective use of anaerobic blood culture vials for several reasons. Patient populations differ according to the type of institution or laboratory, and not all blood culture systems and medium combinations have been, or likely ever will be, studied. Moreover, selective use of anaerobic vials requires that certain logistical issues be addressed. These include adequate notification of physicians and staff of a change in laboratory procedure and the reasons the change was implemented; determining which, if any, patient care locations (e.g., gynecologic or colon and rectal surgery units) should be stocked routinely with anaerobic vials; and developing algorithms for the physicians ordering blood cultures and/or the individuals who obtain blood for culture (i.e., phlebotomists, nurses, clinical care technicians, house officers, and medical students). A sequential plan that includes in-service education and quality assurance after implementation should be undertaken to confirm that the protocols are being followed. Length of Incubation of Blood Cultures In routine circumstances, using automated continuous monitoring systems, blood cultures need not be incubated for longer than 5 days (27, 40, 58, 87, 169). For laboratories using manual blood culture systems, 7 days should suffice in most circumstances (38). A recent study at the Mayo Clinic, in which one

CUMITECH 1C

of the widely used continuous monitoring blood culture systems was used, demonstrated that 99.5% of nonendocarditis BSIs and 100% of endocarditis episodes were detected within 5 days of incubation (27). These data suggest that the extended incubation periods recommended in the past for detection of fastidious microorganisms that sometimes cause endocarditis, including Brucella, Capnocytophaga, and Campylobacter spp. (see the section on “difficult to grow” organisms later in this document) and the HACEK group (Haemophilus, Actinobacillus, Cardiobacterium, Eikenella, and Kingella spp.), are usually not necessary for microbiology laboratories that use modern continuous monitoring blood culture systems. Commercially Available Manual Blood Culture Systems The Isolator System The Isolator system (Wampole Laboratories), both adult- (Isolator 10) and pediatric-size tubes, contains a lysing agent that lyses red and white blood cells, thus releasing intracellular bacteria. For the adult Isolator, blood is drawn directly into a vacuumsuction tube containing a lytic compound. The wellmixed adult-size 10-ml maximum volume tube of blood is centrifuged in a fixed-angle rotor centrifuge, and a specially designed pipette is used to withdraw the supernatant containing the plasma, human cellular debris, and any antimicrobial materials (such as antibiotics or complement) that had been in the blood. A second pipette is used to mix the sediment containing lysed white blood cells, red cell membranes, and microbes and withdraw the inoculum. The sediment containing the etiologic agent can then be inoculated to media and handled in a process that maximizes recovery of the organism being sought. The blood drawn into a Pediatric Isolator 1.5-ml tube is lysed in the tube and the entire volume is subcultured directly. A direct colony count of growth on the media can be extrapolated to yield a quantitative blood culture result. The bacteria-containing portion can be inoculated onto any medium, including media specific for the pathogen being sought, such as buffered charcoal-yeast extract agar (BCYE) for Legionella, appropriate media for mycobacteria, or fungal media for isolation of molds. Limitations include increased detection of some contaminants due to manipulation of the sample, and inhibition or decreased recovery of some pathogens, such as Streptococcus pneumoniae (26, 59, 75, 156). It is, however, a useful and perhaps essential method for recovery of some molds and fastidious organisms that fail to grow in routine blood culture broths, as detailed in the fastidious organism section below.

CUMITECH 1C

The Septi-Chek System The Septi-Chek system (BBL; BD Diagnostic Systems, Sparks, Md.) consists of either an adult bottle containing 70 ml of broth (tryptic soy broth, Columbia broth, brain heart infusion broth, and thioglycolate broth are available) for inoculation of 8 to 10 ml of blood, or a pediatric bottle containing 20 ml of broth (tryptic soy or brain heart infusion broth) for inoculation of 1 to 3 ml of blood. After inoculation, the cap is removed and replaced with a device containing an agar-coated slide (three agars are included: chocolate, MacConkey, and malt). Inverting the bottle allows inoculation of the agar, so that the system becomes a closed biphasic blood culture (26). Aerobic media only are available. Automated Blood Culture Systems There are three common continuously monitored blood culture systems used in the United States at this time. Summaries of key features are seen in Table 3.

INCUBATION AND EXAMINATION OF BLOOD CULTURES General Concepts for Detecting and Initial Handling of Positive Blood Cultures In an ideal world, blood cultures (either visually or automatically monitored) would be inspected several times on all three shifts to ensure the rapid identification of a potentially life-threatening condition. In light of personnel limitations and budget reductions, however, this goal is not often achieved. Blood cultures represent one of the two most important diagnostic culture procedures performed in the clinical microbiology laboratory (the other being cerebrospinal fluid cultures) and, therefore, they deserve a higher degree of attention than other analyses. Automated blood culture instruments should be inspected for signal-positive cultures at least once per 8-h shift. Ideally, bottles should be removed and evaluated as soon as the instrument indicates growth. A recent study has shown that a telephone report of a positive blood culture with Gram stain results has a greater influence on antimicrobial therapy than does the final issue of antimicrobial susceptibility results (97). If sufficient microbiology staff are lacking, technical staff in other round-the-clock laboratories (e.g., a rapid response laboratory) could be cross-trained to perform this task. All signal-positive bottles that occur during the primary daily working shift should be subcultured, smears prepared for Gram stain analysis, and the results of the Gram stain reported by telephone to the patient’s health care team. The amount of blood culture monitoring and processing of positives on additional shifts must be determined based on staffing availability and the priorities of the

Blood Cultures IV

13

institution. If an automated blood culture system breaks down or fails, all bottles must be manually subcultured. Incubation Durations with Automated Blood Culture Systems All manufacturers of continuously monitored blood culture systems received Food and Drug Administration clearance based on performance that included a 5-day incubation cycle. However, two studies using the ESP blood culture system, which is no longer marketed, indicated that either a 4-day or a 3-day incubation cycle did not reduce the detection of clinically significant isolates (32, 56). Laboratories wishing to deviate from the manufacturer’s product insert must validate their modified procedure in their own setting with sufficient data to justify a change, and must notify their physicians of their protocol. For all blood culture systems, the final report for a negative culture should state the number of days in which the bottle itself was incubated, and not the total number of incubation days of the bottle plus subcultures, if any. Incubation Duration and Detecting Positive Blood Cultures in Nonautomated Systems In the absence of an automated detection system, bottles should be visually examined for signs of bacterial growth initially after 6 to 12 h of incubation and daily thereafter for 7 days (12, 37). Growth is enhanced by shaking the bottle, but visual inspection is then precluded and blind subcultures are essential. An initial blind subculture should be performed after 12 to 18 h. Gram or acridine orange fluorescent staining can be performed during the initial examination to aid in the early detection of microbial growth (37, 89, 148), but blind or terminal subcultures are of little use thereafter, particularly if bottles are allowed to settle and can be visually inspected (18, 50). Nonautomated cultures in anaerobic media are incubated without shaking, and anaerobic bacteria growing in anaerobic blood culture media are almost always detectable by visual signs, either turbidity, lysis of red blood cells, visible microcolonies, or gas in the broth. Blind subcultures from anaerobic bottles are not necessary (98, 108). Visual Examination of Smears from Positive Blood Cultures The single most important test to perform on any visual- or signal-positive blood culture is the Gram stain. It is highly likely that the information provided by this test when coupled with pertinent patient information will dictate the choice of primary antimicrobial therapy. Terminology used to report Gram stain results should be as descriptive as possi-

14

Baron et al.

ble without being misleading or ambiguous (37). For example, a report of “gram-positive cocci” could indicate a variety of microorganisms including staphylococci, streptococci, and enterococci, but a report of “gram-positive cocci in pairs” narrows the field considerably. This information can be detailed further by denoting whether cocci are present in pairs and short chains (suggestive of pneumococci, enterococci, or group B streptococci) or long chains (suggestive of other beta-hemolytic streptococci and viridans streptococci) for better differentiation. The use of a term like “diphtheroid” should be avoided because it could be misinterpreted by the clinician to mean a contaminant, when many potentially devastating microorganisms such as Rhodococcus or Mycobacterium species can share the morphological characteristics of coryneform gram-positive bacteria. These organisms could be described as “coryneform gram-positive bacilli” or “pleomorphic gram-positive rods not resembling Bacillus or Clostridium.” It is often useful to examine Gram-stain-negative, signal-positive blood cultures microscopically using an acridine orange stain visualized with an ultraviolet microscope using a fluorescein isothiocyanate filter. While this method does not distinguish between gram-positive and gram-negative organisms (it is a nonspecific nucleic acid fluorescent stain), it is often useful to identify bacteria with poor counterstaining qualities such as Mycoplasma, Campylobacter, and Brucella spp. (22, 23) or to reduce the need to extensively subculture truly false-positive cultures (1). Initial Subcultures of Positive Blood Cultures Independent of the microscopic results, all signal or visually positive blood cultures require an initial subculture to an appropriate selection of agar-based growth media. Positive blood cultures may appear cloudy, show bubbles of gas formation, or change color from red to brownish. Other positive signs are the presence of granular or spherical structures (tiny floating colonies or colonies at the edge of the meniscus), fluffy structures (potential mold colonies), or lines of turbidity resembling comet tails. The subculture media should consist, at the very least, of Trypticase soy agar with 5% sheep blood as an all-purpose medium and a chocolate agar plate for the isolation of nutritionally fastidious organisms. These two media are inoculated with a few drops of wellmixed blood culture broth which is streaked for isolation, and they are incubated at 35 to 37°C in 3 to 5% carbon dioxide (CO2) for a minimum of 48 h and examined daily for the appearance of distinct colonies. If growth occurs in the anaerobic bottle only, or if organisms seen in both bottles morphologically suggest anaerobes, a supplemented (vitamin K and hemin) anaerobic blood agar plate should be

CUMITECH 1C

inoculated and incubated in an anaerobic atmosphere for 48 to 72 h. Selective or specific media such as MacConkey agar (for gram-negative bacilli), colistin-nalidixic acid agar (for mixed infections or streptococci), or mycologic agar (for yeasts and molds) can be added based on Gram-stain results. The second edition of the Clinical Microbiology Procedures Handbook contains suggested media based on Gram stain criteria (177). Certain yeasts (Malassezia species) may be visible as small budding yeasts that resemble bowling pins in blood culture smears, but the yeasts will not grow on routine fungal media without the addition of a thin film of sterile olive oil as a growth factor. If the Gram stain of the blood culture bottle indicates the possibility of staphylococci or enterococci, vancomycin and methicillin screening agar can be added to the regimen of subculture media when resistance is a consideration, a form of direct susceptibility testing further described in the next section. Supplemental media such as BCYE or rabbit blood agar for the isolation of opportunistic pathogens can be used for subculture when dealing with immunocompromised individuals or when infections caused by fastidious organisms (e.g., Bartonella spp.) are considered and this information is conveyed to the laboratory (36). Recovery of fastidious organisms is discussed more extensively in the next section. Some rapidly growing mycobacteria grow well on routine media within a few days and can be mistaken for coryneform rods. The organisms may stain as gram-positive, slightly pointed rods, or they may not take the stain well at all and appear as clear “ghosts.” Mycobacteria should be ruled out by acid-fast stain for any small colonies of gram-positive or variable rods that cannot be otherwise identified. Direct Identification and Susceptibility Testing from Blood Culture Broth For Gram-stain-positive blood culture bottles, a number of protocols have been designed to facilitate rapid identification and susceptibility testing of bacteria directly from the blood culture medium. These protocols have included direct identification using commercial biochemical panels (57, 79, 155), detection of bacterial-specific enzymes such as tube coagulase (20, 152), immunologic detection using antibody assays (62, 112), probe hybridization (20), protein-nucleic acid fluorescence in situ hybridization (Food and Drug Administration cleared for Staphylococcus aureus, Enterococcus faecalis, and Candida albicans) (20, 106), and PCR (115), to name a few. In a similar fashion, protocols have been designed for susceptibility testing of microorganisms directly from positive blood culture bottles (21, 29, 57, 74, 114) or PCR detection of resistance genes (82). One method involves centrifugation of 5 ml of

CUMITECH 1C

the broth at 160  g for 5 min to pellet the blood cells, followed by centrifugation of the supernatant at 650  g for 10 min to pellet bacteria. The bacterial pellet is adjusted to a McFarland 0.5 turbidity and handled as for standard susceptibility testing (79). To facilitate inoculation into growth-requiring systems, the organism may also be enriched from the blood culture broth by adding 1 drop of the broth to 0.5 ml of brain heart infusion broth, incubating (with shaking if possible) for 3 to 4 h to achieve a McFarland 0.5 turbidity, and adding the entire volume to the inoculation liquid for microtiter MIC trays (177). A modification of this protocol uses a serum separator tube (BD Vacutainer Systems, Franklin Lakes, N.J.) to perform a single centrifugation at 1,300 to 1,400  g for 10 min (152). A number of derivations of this process have been used (21, 57, 74, 79, 114). Because this is not a standardized means of susceptibility testing, a confirmatory antimicrobial susceptibility test is generally recommended once the organism has been recovered in pure culture; however, a recent article reported 97.6% correlation with very few major errors when a direct susceptibility method was compared to standard susceptibility testing of the isolate (79). From an economic standpoint, rapid identification and susceptibility testing has been shown to reduce the length of hospitalization of patients with sepsis, which translates into reductions in cost of patient care (10, 12). However, the laboratory cannot bill for two susceptibility tests, so if the direct and standard susceptibilities are both performed, only one of them will receive reimbursement. The logistics, cost, and performance of an assay must be carefully weighed when the laboratory is considering the feasibility of incorporating identification and susceptibility testing of microorganisms directly from positive blood cultures. Further, it should be noted that with certain organism-drug combinations, false susceptibility or resistance results can be generated using direct testing (21, 82, 114). Laboratories should validate their method against the standard results. False-Positive Blood Cultures (“Contaminants”) One of the most perplexing problems associated with blood cultures is establishing a means of avoiding or dealing with microorganisms that enter the bottle during procurement but were not actually circulating in the patient’s bloodstream. Differentiating “contaminants” from true pathogens is extremely difficult, especially because the same species can easily be found in either situation. Policies for processing and reporting possible/likely blood culture contaminants should be in place to standardize the minimal laboratory evaluation and to avoid unnecessary therapy for the patient. Obviously, avoidance is the best pol-

Blood Cultures IV

15

icy and this can be accomplished most effectively by paying strict attention to the process of skin antisepsis, venipuncture, and specimen transfer to blood culture bottles, as described previously (37). Beyond preemptive techniques, however, it is considered difficult to reduce the overall contamination rate below 2% (38). Further, the organisms commonly associated with contaminated blood cultures (Bacillus [not B. anthracis] spp., Corynebacterium spp., Propionibacterium spp., coagulase-negative staphylococci, viridans group streptococci, Aerococcus spp., Micrococcus spp., and many others) are perfectly capable of causing serious infection in the appropriate setting. Coagulase-negative staphylococci, because of their ability to colonize and form a biofilm on indwelling and prosthetic devices and their ubiquitous presence on the human skin, are the primary agents of both catheter-associated septicemia and false-positive blood cultures (122). In many cases when only one blood culture is received from a given patient, a potential contaminant is recovered from one or both bottles and without a second blood culture for comparison, interpretation of the clinical relevance of that positive culture is impossible. Physicians may be forced to initiate treatment for practical and legal reasons if susceptibilities are reported by the laboratory. Therefore, the evaluation of an isolate with low virulence potential recovered from a single blood culture set (one or both bottles) should be limited to the extent to which phenotypically similar but medically important organisms can be safely excluded from the identification. Some examples include the performance of a bile solubility test to differentiate S. pneumoniae from viridans group streptococci; an agglutination assay or coagulase test to distinguish S. aureus from coagulase-negative staphylococci; and a motility assay or demonstration of hemolysis to separate non-anthracis Bacillus spp. from B. anthracis. Routine susceptibility testing is not necessary for suspected contaminants, but all isolates should be saved so that additional studies can be performed if an identical organism is recovered from a subsequent blood culture from the same patient. At that point, full identification of both isolates along with susceptibility testing should be initiated. More than one publication provides a detailed description of a laboratorybased algorithm for minimizing the extent of evaluation of blood culture contamination (122, 150, 160). Polymicrobic Bacteremia The incidence of true polymicrobic bacteremia in the absence of contamination is currently unknown, since most blood cultures are not reincubated after the initial positive subculture. A recent report on the

16

Baron et al.

epidemiology of bloodstream infections in the United States over a 20-year period indicates that polymicrobic bacteremia is relatively rare, accounting for only 4.7% of all septic episodes (86). However, in select patient populations, blood cultures positive for two or more organisms can range from 10% in children (171) to nearly 30% in immunocompromised patients (70, 84). Dental extractions represent one known risk factor for polymicrobic bacteremia (5). Polymicrobic bacteremia in association with Pseudomonas aeruginosa carries a higher risk for mortality and is more frequently seen among older patients (2). Polymicrobic bacteremia can be an indicator of underlying disease and represent translocation of microorganisms from the gut, skin, and mucosal surfaces (84). Even so, the laboratory approach toward the evaluation of blood cultures positive for multiple organisms should include a fair degree of common sense. Whether isolated singly or in conjunction with other organisms, bacteria with high pathogenic potential such as P. aeruginosa, S. aureus, or Escherichia coli require complete identification and susceptibility testing. When potential contaminating bacteria such as coagulase-negative staphylococci, Bacillus species, coryneform grampositive aerobic bacilli, or Propionibacterium acnes accompany the mix of recovered bacteria, it throws the validity of a positive blood culture in doubt, and clinical relevance of the presence of true pathogens is difficult to interpret. In this instance, the rules described above under the heading of “contaminants” come into play. All potential contaminants are identified to a limited extent and saved pending future isolation of the same organism from a subsequent blood culture. Saving Blood Culture Isolates All blood culture isolates should be maintained as frozen stock cultures for a minimum of 6 months pending the possibility of additional studies. Centers with sufficient resources should attempt to save isolates much longer, for as long as storage space is available. Possible/probable contaminants are saved for comparison purposes in case a similar organism is recovered from the same patient in subsequent blood cultures, albeit for a shorter time frame (10 days to 2 weeks). If freezer space is limited, laboratories should negotiate with their infection control practitioners to arrive at a mutually acceptable list of isolates to save for specified time periods longer than that used for isolates from other sources. A variety of techniques are available for the preparation of frozen stock cultures of bacteria, including the use of 10% skim milk or Trypticase soy broth supplemented with 10% glycerol as freezer media (117). One convenient method is to place approximately 1 ml of freezer

CUMITECH 1C

medium in a sterile freezer vial. A sterile swab is used to obtain growth from a fresh subculture of the blood culture isolate. The swab is immersed into the freezer medium and broken off below the cap of the vial. If subcultures are required, the vial needs only to be thawed and the swab removed with a flamed forceps and transferred to solid growth medium for isolation. Commercial products containing plastic beads with linear holes for preserving frozen suspensions of bacteria are also available (Cryocare; Key Scientific Products, Round Rock, Tex.). After making a suspension of the organism in the broth, excess liquid is removed and the beads are frozen. To reconstitute the culture, a single bead is removed and rolled on an agar plate or dropped into a broth and incubated. Reporting Results Unlike many other laboratory results, positive blood cultures can have an immediate impact on patient care decisions, and clinically relevant results must be reported to caregivers as quickly as they are available. It is often helpful to review other cultures, such as urine or sputum, received previously from the same patient for possible clues to the identity of the isolate before making the call. The number of positive sets (all bottles from a single phlebotomy) and the total number of sets received, the time from collection until positivity, and a description of the organism seen in the smear should be reported. Several studies have shown broadly improved outcomes and efficiencies when such reports are delivered quickly (10, 12, 33, 97). As much information as possible should be conveyed to the clinician, although the microbiologist may have to explain the nuances of these interpretations to non–infectious diseases specialist physicians. For example, “tiny gram-negative coccobacilli” is more useful than “gram-negative bacilli.” A report such as “gram-positive, boxcarshaped rods, no visible spores, and abundant gas in the blood culture bottle” can alert a physician to possible clostridial sepsis, whereas “gram-positive rods” raises no alarms. Using cellular arrangement to report “gram-positive cocci resembling staphylococci” is more meaningful than reporting “gram-positive cocci in clusters.” In some circumstances, not all results must be called in to physicians, but the algorithms of which results require a phone call and the person to whom positive blood culture results can be delivered are unique to each health care institution. For example, in some institutions, a single coagulase-negative staphylococcal isolate from one blood culture set is not reported to the caregiver, but if a second blood culture becomes positive for coagulasenegative staphylococci, the result is reported and further workup is initiated. Conversely, any coagulasenegative staphylococci isolated from the blood of a

CUMITECH 1C

pediatric patient, even when only one bottle has been sent to the laboratory, is called in immediately. Whether positive results can be delivered to a nurse, a unit clerk, or to a physician only should be dictated by the laboratory’s institutional policy. How such information is conveyed can also vary. Some busy clinical services, such as the emergency department, may prefer faxed results, or they may monitor the hospital information system closely enough to preclude any additional report of positive culture results. Clinically relevant positive blood cultures are considered “critical values” and reported according to guidelines developed to satisfy the Joint Commission on Accreditation of Healthcare Organizations 2003 Patient Safety Goals (see http://www.jcaho.org/ accreditedorganizations/patient+safety/npsg.htm). The receiver of a critical value is expected to write down the information and read it back to the caller; the laboratorian making the call must ask the receiving person to “please read back the results.” Regardless of who receives the report, the microbiologist must document to whom the report was given and read back by, and the date and time of report delivery. Reports should be updated on a regular basis, preferably daily, as to whether there is growth in the blood culture or not. This is to document that the specimen was received in the laboratory. A report such as “no growth detected so far” is acceptable. Once an organism has been detected, a preliminary report should be issued, either on paper or electronically into the laboratory information system. Reports should include the total number of days that the bottle was incubated before being detected as positive, the type of bottle(s) such as aerobic or anaerobic, and the source of the sample (such as “PICC line” or “right hand”). As new information is determined, the report should be updated. At the end of the prescribed incubation period, negative reports should state the number of days the bottle was incubated. Any comments about the specimen or its handling that could compromise results, such as extended delays before the bottle was incubated in the instrument, or suboptimal volume fills, should be added to the report. Interpretive comments, such as the probable clinical significance of the isolate, if added at all, should be carefully worded and developed in partnership with the infectious disease specialists or other interested medical specialists at the institution. One example of a comment that was developed with infectious diseases physician input is used when Enterococcus species are isolated from blood cultures: “For treatment of endocarditis, the combination of penicillin or ampicillin and an appropriate aminoglycoside (streptomycin or gentamicin) is indicated. Vancomycin is usually reserved for penicillin-

Blood Cultures IV

17

allergic patients.” It is prudent to avoid use of the word contaminant because of the ease with which it can be misinterpreted. An alternative comment for reporting a potential contaminant is “This isolate is possibly a collection-associated skin organism; notify the microbiology laboratory if further studies are indicated.” As with all laboratory reports, the format should be clear and easy to interpret, and the physician should find all relevant information easily. It is best to work with physician groups to jointly agree on a reporting format. Part of an ongoing quality assurance program includes auditing reports to ensure that what the physician sees at the user end is what the laboratory intended and that no information is missing or confusing.

CPT-4 CODING AND BILLING ISSUES One bacterial blood culture billed with CPT-4 code 87040 (aerobic and anaerobic bottles filled from a single phlebotomy) includes the incubation, initial smear of a positive bottle, and subculture to appropriate media. A second blood culture obtained on the same day, which is clearly the standard of practice, should be billed in a manner that allows recognition of the fact that it represents a distinct procedural service (e.g., by attaching a modifier, 87040–59). However, some payers may prefer use of modifier -91 (repeat service on the same date). It is best to check with your specific local reimbursement organization about appropriate CPT-4 coding if you have been denied reimbursement with what you consider to be correct coding, especially since some payers have developed inappropriate payment policies limiting blood cultures to one unit of service per day. Blood cultures for isolation of yeasts or molds are billed with CPT-4 code 87103, and blood cultures for isolation of mycobacteria are billed with CPT-4 code 87116. When isolates are identified using a combination of cellular and colonial morphology and up to three tests (e.g., “spot” biochemicals), the identifications are considered presumptive for reimbursement purposes and are included in the primary culture code described above. However, when organisms are identified using more than three biochemical tests or by other complex processes, an additional definitive identification code may be added (e.g., 87077 for aerobic isolates, 87076 for anaerobic isolates, 87106 for yeasts, 87107 for molds, and 87118 for mycobacteria). In addition, identifications using molecular methods, immunofluorescent stains, or other immunologic methods such as agglutination assays are also billed with the appropriate CPT-4 codes. Each antiserum used for an agglutination grouping, for

18

Baron et al.

example, is billed separately with CPT-4 code 87147. Susceptibilities are billed separately. If an organism is tested for beta-lactamase using a cefinase disk, that test can be billed with CPT-4 code 87185 in addition to the routine method, which is billed with CPT-4 code 87186 for each microdilution plate or 87184 for each disk diffusion plate (12 or fewer disks). If the antibiotic gradient strip system Etest (AB Biodisk N. A., Inc., Piscataway, N.J.) is used for some antibiotics, such as when testing is requested by a physician but the agents are not included in the laboratory’s standard automated panel, each antibiotic strip can be billed separately with CPT-4 code 87181. The lysis-centrifugation system (Isolator; Wampole Laboratories, Princeton, N.J.) calls for different billing. An adult-size Isolator tube is billed for concentration with CPT-4 code 87015 in addition to the blood culture code 87040, and positive plates can also be billed for identification and susceptibilities as above. There is no CPT-4 code for the quantitative aspect of an Isolator blood culture. However, if the Isolator blood culture is ordered for isolation of fungi only, the CPT-4 code is 87103 (plus 87015 if it is centrifuged). In all matters pertaining to correct coding and billing for blood cultures, clinical microbiologists should be aware of reimbursement policies issued by payers for which the laboratory is an approved provider of services.

LABORATORY DIAGNOSIS OF SEPSIS CAUSED BY A COLONIZED INDWELLING VASCULAR CATHETER Indwelling vascular catheters are a cornerstone of modern medical care, enabling administration of fluids, parenteral nutrition and drugs, easy blood sampling, and monitoring of a patient’s physiological parameters. However, even when careful antisepsis is followed during the catheter insertion and maintenance, these devices tend to become colonized by commensal indigenous and pathological skin flora over time, and catheter-related infections (CRI) are now among the most common sources of nosocomial bacteremia. Microorganisms attach to the inner and outer surfaces of the intravenous device, and bacteria such as Staphylococcus epidermidis and other species of coagulase-negative staphylococci secrete an excess of extracellular matrix (“slime”) and form a tenacious biofilm that protects the organism from the host immunological response. Flushing of the colonized catheter by intravenous fluid solutions provides nutritional support to residing organisms and contributes to dispersal of the infection through the bloodstream. Studies performed in patients as well as animal models have demonstrated that the

CUMITECH 1C

risk for CRI, either local infection or bacteremia, correlates with the number of organisms colonizing the catheter (137). Although purulence or frank cellulitis at the catheter exit site makes the diagnosis of CRI obvious in some cases, local signs of inflammation are absent in more than 70% of catheter-related bloodstream infections, and the disease usually manifests as onset of a new febrile episode in a hospitalized patient (138). The ideal laboratory method for diagnosing CRI should meet all of the following criteria. 1. It should be rapid enough to enable timely therapeutic decisions such as catheter removal and/or institution of antimicrobial therapy. 2. It should be technically simple. 3. It should be able to detect intraluminal organisms as well as those attached to the external catheter surface. 4. It should be highly sensitive. 5. It should be highly specific (i.e., able to differentiate between colonization and true infection). 6. It should be able to confirm or rule out the diagnosis of CRI without the need to remove the indwelling vascular device. 7. It should enable isolation of the infecting bacteria or fungi to enable complete identification, antibiotic susceptibility testing, and eventual typing. 8. It should be safe for the patient. Although a long list of different laboratory methods has been used over the years to confirm or rule out CRI, unfortunately none of them meet all the aforementioned requirements. In addition, most evaluations of methods to diagnose indwelling vascular catheter infections have been hampered by lack of a precise “gold standard” for defining CRI, enrollment of a small number of patients with culture-proven infection, different case mixes (immunocompetent and immunocompromised patients, children and adults), multiple catheter types, lack of uniformity of the culture techniques, and use of different criteria for positivity, precluding valid comparisons of results and making the laboratory diagnosis of CRI a muchdebated issue. The traditional and simple methods of performing microscopy and nonquantitative culture of the exit site or the catheter tip are clearly unsatisfactory because they do not allow quantification of the number of microbes and thus are not able to differentiate between true CRI, skin colonization around the catheter, and contamination of the external catheter

CUMITECH 1C

surface by a small number of organisms during withdrawal of the intravascular device through the skin tunnel (78). Accepted laboratory methods for diagnosing CRI can be divided, for practical purposes, into two broad categories according to the need to remove the indwelling catheter in order to make the diagnosis. This issue is of clinical importance for patients with suspected CRI and no signs of life-threatening infection because, if the diagnosis of CRI is not confirmed, an alternative source of infection should be sought, and there is no indication for removing the catheter. However, in a patient with a central line and signs of overwhelming infection such as septic shock but no apparent clinical focus, removal of the catheter under suspicion of infection without waiting for laboratory confirmation of CRI is probably indicated. Diagnostic methods that leave the catheter in place are based on the rationale that if the central line is the focus of bacteremia, because of the in situ replication of organisms, a high microbial concentration should be present in blood samples taken from the infected catheter. This high microbial load can be readily detected by examining an acridine orangestained or Gram-stained sample of the blood cell layer obtained by the cytocentrifugation of 5
l of blood drawn from the catheter (127). In one experimental method, the sensitivity was enhanced by pushing a sterile endoluminal brush through the catheter and then sending the brush for culture (72). Although there were no adverse events in this study of 230 catheters, the safety of this method has not been evaluated in a large trial and it remains controversial. The high concentration of organisms present in intravascular catheters of patients with CRI can also be compared to that found in blood specimens obtained from distant peripheral veins. The difference in the microbial concentration between blood specimens obtained from the infected central line and from a peripheral site can be measured using a quantitative blood culture system such as the pour plate or the lysis-centrifugation method (43). A colony count ratio greater than 4 to 10:1 between the central venous blood and a peripheral vein blood specimen has been found to have 78 to 94% sensitivity and 99 to 100% specificity for diagnosing CRI (19). When only a quantitative central venous blood sample but no peripheral vein culture is obtained, the diagnosis of CRI can still be confirmed by finding more than 100 CFU of bacteria/ml of blood or 25 CFU of fungi/ml, because bacteremia and fungemia caused by infections other than CRI in adult patients tend to yield lower CFU counts (19). Because the lysis-centrifugation method is not readily available in all clinical microbiology labora-

Blood Cultures IV

19

tories and has the disadvantage of being manual, time-consuming, and prone to contamination, an alternative approach for diagnosing CRI based on the use of broth-based blood culture systems has been recently developed. Continuous monitoring of blood culture vials by modern automated instruments enables determining and recording the precise time-to-positivity in each vial. Finding of a differential time to positivity greater than 2 h between vials inoculated with blood drawn through the catheter and those obtained from a peripheral vein has been found to be reliable for the diagnosis of CRI by more than one series of investigations (14, 116), although some others have not found such positive results (123). Diagnostic laboratory methods that require removal of the central catheter have the obvious disadvantage that the diagnosis is always made retrospectively. In cases where the diagnosis of CRI is ruled out, insertion of a new blood vessel access device would still be required, adding hospital costs and potential risks to the patient. Methods that require catheter removal for making the diagnosis of CRI include (alone or in combination) (i) use of an endoluminal brush (72); (ii) sonication or flushing of the catheter lumen to remove adherent organisms (136), followed by a staining procedure and/or a quantitative culture; and (iii) a semiquantitative culture of the catheter tip (85) and/or hub (44). Of all these procedures, the semiquantitative culture of the catheter tip (Maki method) is the most widely used and evaluated (138). This method involves culturing the external surface of the removed catheter tip by rolling a 3- to 4-cm segment four times over the surface of a blood-agar plate with the aid of a sterile forceps. After incubation in aerobic conditions, the number of growing colonies is counted. The original cutoff value to diagnose CRI recommended by Maki (15 CFU of bacteria) has been lowered by some investigators to 5 to improve sensitivity (28). Numbers for yeast have not been as well studied. It should be noted, however, that the method does not detect indwelling vascular catheters with exclusive endoluminal colonization. This problem has been addressed by flushing the catheter as described by Cleri et al. (25) or by sonication (68). The Cleri method consists of cutting and separating the first 1-cm-long intradermal segment of the removed catheter from the intravascular portion. A needle is inserted into the proximal end of the intravascular segment, which is then immersed in 2 or 10 ml of Trypticase soy broth, depending on the length of the segment to be cultured, and flushed three times. The broth is serially diluted 100-fold, and 0.1 ml of each dilution is streaked onto blood agar plates. In the sonication

20

Baron et al.

procedure, the catheter is placed into a tube containing 4 ml of brain heart infusion (68). The tube is then sonicated for 1 min followed by vortexing for 15 s. Aliquots of 0.1 and 0.001 ml of the sonicated broth are then plated onto solid media and incubated. When sonication methods are evaluated, cutoff values of 102 CFU (136) or 103 CFU (15, 68) have been shown to yield 94 to 97% sensitivity and 75 to 88% specificity for diagnosing CRI. The recently introduced brush technique appears to improve the diagnosis of CRI by mechanically removing organisms attached to the endoluminal biofilm, which are then suitable for detection by microscopy and culture (72). Although the early experience with this novel method appears promising, concerns regarding its safety when used while the catheter is still in place have been raised because of the theoretical possibility of blood-borne dissemination of dislodged organisms and distal embolization (72). More widespread studies using this technique are needed. The question of which is the most appropriate laboratory method for diagnosing CRI remains unsettled. Despite the lack of consensus, the following recommendations are made. 1. If septic shock of undetermined source develops in a patient with an indwelling vascular catheter, or when local signs of infection such as purulence or cellulitis are present, the catheter should be removed and a semiquantitative culture (by the Maki method) or quantitative culture (by the Cleri or sonication methods) of the catheter tip should be performed in addition to at least two blood culture sets obtained peripherally. 2. If the patient with suspected CRI is in a stable clinical condition, confirmation of the diagnosis by ruling out other sources, for example, prior to administration of antibiotics and removal of the catheter, should be attempted. 3. When quantitative blood cultures are available, blood specimens should be drawn through the catheter and from a peripheral vein and concentration of microorganisms should be compared. A catheter-blood to peripheral-blood ratio 4 is indicative of CRI (19). 4. When an automated broth-based blood culture system is used, a differential time to positivity of 2 h between hub blood and peripheral blood can be used to confirm CRI (116). 5. Complementary methods such as the Gram stain or acridine orange cytospin tests and use of the endoluminal brush seem promising, but additional experience with these techniques is needed prior to their recommendation. 6. Only aerobic cultures of the intravascular device should be performed.

CUMITECH 1C

CONVENTIONAL METHODS FOR DETECTING PATHOGENS THAT FAIL TO GROW IN STANDARD MEDIA A number of microorganisms present in blood, either transiently or as agents of intravascular infections including endocarditis, cannot be recovered with standard routine or automated blood culture protocols. Some bacteria traditionally thought to require longer incubation times for recovery, including the HACEK agents and Brucella (see specifics in this section), are now routinely recovered by the usual automated blood culture procedures and incubation times. Nutritionally deficient streptococci, Abiotrophia species, and Granulicatella species grow well in blood culture media and exhibit positive signals in the bottles. They are readily visualized by Gram staining but fail to grow on blood agar plates, although they should be viable on chocolate agar. Subculturing to blood agar with a Staphylococcus aureus cross-streak also reveals the satelliting colonies typical for these species. Each new lot of chocolate agar should be tested with a nutritionally variant streptococcus to ensure that blood culture subcultures will reveal the agent. If such an added quality assurance activity is not feasible, S. aureus should be cross-streaked on blood agar subcultures for gram-positive cocci that fail to grow on initial subcultures. Francisella tularensis grows sufficiently in commercial systems to be routinely recovered from blood. Not only is this organism a potential bioterrorism agent, but it is considered a biosafety level 3 organism and should be handled minimally in routine clinical laboratories. For example, if an isolate exhibits extremely tiny, pale-staining gram-negative or gram-variable coccobacilli on the smear, all plates should be sealed immediately and all handling should be performed in a biological safety cabinet by using Sentinel Laboratory Response Network protocols (see http://www.bt.cdc.gov/labissues/#testing). If the isolate is found to be catalase negative or weak positive, urease negative, oxidase negative, and nitrate negative, it should be referred to the appropriate Response Network Reference Level Laboratory to rule out Francisella. Some agents, including Coxiella, Chlamydia, Rickettsia, and Tropheryma spp., do not grow on artificial media and are best diagnosed using serological tests (slow) or molecular amplification methods (available at limited reference laboratories). Although serology is virtually the only method for diagnosis of Chlamydia psittaci endocarditis, cross-reactivity with antibodies directed against Bartonella has been problematic (88). Blood for molecular amplification testing may require EDTA tubes, plasma preparation tubes, or acid-citrate tubes, and reference laborato-

CUMITECH 1C

ries often request that the sample be frozen at 70°C and shipped on dry ice. The performing laboratory should be consulted before obtaining the blood to learn which type of tube, preservative, and temperature should be used for handling. Musso and Raoult have isolated Coxiella burnetii from the blood of 53% of untreated patients with Q fever by inoculation of human fibroblasts grown in shell vials (100). Immunofluorescent staining of the monolayer is required to detect infection. This method has not been validated in other laboratories and is generally not available elsewhere, but laboratories should again follow Sentinel Laboratory Response Network protocols for referral to a network confirmatory laboratory for testing. Another group of organisms may be present in blood and viability is maintained in standard blood culture media, but either the organisms do not grow sufficiently to produce the metabolic end products that serve as growth indicators in the system or their growth is too scant to be visible in the broth. Cryptococcus neoformans, Legionella species, some Helicobacter species, and molds, including the systemic dimorphic species, fall into this category. Fungi Although systemic cryptococcal disease is most rapidly diagnosed with the serum cryptococcal antigen test (either agglutination or ELISA format), the organism may be flagged by automated systems and can certainly be recovered from blood culture broths by performing blind subcultures to fungal media at the end of a standard 5- or 7-day incubation protocol. Other yeasts do grow well in commercially available blood culture broths and do not need special handling, with the exception of Malassezia furfur, which requires the addition of olive oil to the subculture agar. The best recovery of M. furfur is achieved with a very thin coating of olive oil (extra virgin, sterilized before use) applied to the fungal plate medium (brain heart infusion, Sabouraud dextrose, or potato flake agar) with a swab before inoculation. Infants undergoing lipid emulsion therapy and neonates in the intensive care unit are most at risk for M. furfur sepsis, and special handling should be requested if this organism is suspected. Because the pH, nutrients, and oxygen partial pressure of standard blood culture broths have been optimized for bacteria, molds are rarely recovered. If circulating molds, such as Fusarium or Histoplasma, are suspected, either a special fungal liquid medium, such as MYCO/F Lytic medium (BACTEC), BacT/ALERT MB, or the Isolator lysis-centrifugation system should be used (96). For Cryptococcus and molds, inoculation onto media suited for culture of systemic fungi such as

Blood Cultures IV

21

brain heart infusion, Sabouraud glucose and brain heart infusion, or potato dextrose agar affords the best recovery (26). Bartonella Antibody testing is the most reliable method for diagnosis of Bartonella bacteremia (45). Failing this, nucleic acid amplification methods are also quite sensitive (179). Blood cultures may be attempted using lysis centrifugation, with the sediment plated onto freshly prepared Columbia agar base blood, chocolate, or BCYE agar plates incubated at 35°C in a moist 5 to 7% CO2 atmosphere for 14 to 21 days. Bartonella bacilliformis, unlike B. henselae and B. quintana, grows best at 25 to 30°C. Tierno and colleagues have isolated Bartonella species from BacT/ALERT blood culture bottles that signaled positive by injecting 7.5 ml of the “positive” broth into lysis-centrifugation tubes and processing the sediment as above (149). Brucella Depending on the automated system, Brucella can be isolated using standard or slightly prolonged incubation protocols. A review of methods reported that continuous monitoring systems were faster than lysis-centrifugation, although all isolates detected by the Isolator grew within 7 days (173). However, the Isolator system failed to detect a substantial number of positives (6 of 28) detected by BACTEC (174). Some studies reported 100% recovery within 4 days for either BACTEC or BacT/ALERT (125, 174). Although variable results were reported for different commercial systems, only 1 of 41 true-positive cultures was detected by a blind subculture at 7 days versus automated detection in the BACTEC (175). One study reported detection of 3 of 97 positive cultures by the BACTEC between days 8 and 9 (9); thus, a 10-day total incubation in an automated blood culture system should approach 100% sensitivity for detection of Brucella. However, with sufficient volume of blood, it is reasonable to expect almost all Brucella to be detected within the standard 5-day protocol. This genus is among the most common laboratory-acquired infectious agents as well as a potential bioterrorism agent, and all manipulations should be performed in a biosafety level 2 laboratory setting within a biological safety cabinet. Plates should be taped shut, and as soon as preliminary tests suggest Brucella (including small gram-negative coccobacilli, rapid positive urease, positive oxidase, and nitrate), the isolate should be referred to the appropriate Confirmatory Laboratory Response Network laboratory. Two recent articles reported Brucella appearing as

22

Baron et al.

gram-positive or gram-variable coccobacilli in the initial Gram stain from broth, and that has been seen in other laboratories as well, so all small gram-variable coccobacilli should be handled as if they were bioterrorism agents until proven otherwise (91, 104). Campylobacter and Helicobacter These two genera consist of small, thin, curved rods that may require acridine orange for visualization in instrument-flagged positive bottles. Several species of Campylobacter, including C. jejuni, C. lari, and C. fetus, are isolated from blood occasionally, and they usually grow within a 5-day protocol. Subculture plates should be incubated to Columbia blood-agar base, specific Campy media (without antibiotics, if possible), and incubated in a microaerophilic atmosphere at 35 to 37°C for up to 3 days. Helicobacter cinaedi and other species (H. westmeadii, for example) have been recovered from blood, usually from immunocompromised patients (69, 151). These isolates were detected after a 7-day incubation, but standard Gram stains failed to reveal organisms and acridine orange staining was required. Legionella Although legionellae survive fairly well in commercial blood culture broth, they do not multiply (24). In fact, bacteremia does occur during systemic illness. Detection requires either specialized BCYE broth (no longer available commercially) or lysis-centrifugation and plating onto BCYE followed by incubation in a moist atmosphere for up to 5 days. In a seeded blood culture study, recovery of the organism decreased within 30 min of inoculation of the Isolator tube, and only 10% of the original inoculum was recovered if Isolator tubes were processed after 15 h (24). Mycoplasma Mycoplasma hominis has been implicated as a cause of postpartum sepsis, but its clinical significance is still in doubt. It is recovered in blood cultures occasionally and may be subcultured on standard sheep blood-agar plates. M. pneumoniae may also cause septicemia. Standard blood culture systems rarely support growth of these cell-wall-absent organisms without the addition of gelatin to neutralize the inhibitory effect of SPS and arginine for enhancement of the growth of M. pneumoniae, but the organisms can be visualized in “positive” bottles if acridine orange stain is used. Subculture to specialized Mycoplasma medium such as Shepard’s A-7 or SP4 biphasic medium and incubation for up to 7 days at 35°C in 5% CO2 or anaerobically (best) is essential, although occasional colonies may grow on Columbia CNA agar plates. Physicians should explain how results of cultures positive for Mycoplasma will alter

CUMITECH 1C

patient treatment as a prerequisite for the laboratory attempting to isolate this difficult genus. Leptospira Blood cultures should be obtained during the first week of illness, as the bacteremia is no longer present beyond 7 days. The organism does not survive at 35°C, so cultures must be held and incubated at 30°C. Ideally, two drops of blood should be inoculated directly into 10 ml of semisolid oleic acid albumin medium such as Ellinghausen-McCulloughJohnson-Harris broth or PLM-5 broth (Intergen Co., Purchase, N.Y.) at the bedside. These media or others that contain bovine serum albumin and Tween 80 are recommended (176). This medium is incubated at 28 to 30°C for up to 13 weeks and examined under darkfield microscopy weekly for the presence of the tightly coiled, hooked spirochetes. Alternatively, blood may be collected into heparin, oxalate, or citrate tubes and maintained at ambient temperature for shipment to a reference laboratory for culture. Mycobacterium In most cases, mycobacterial blood cultures are requested for detection of disseminated Mycobacterium avium-M. intracellulare complex in patients with AIDS. Commercial broth systems for mycobacterial cultures have specialized broths for isolating mycobacteria from blood. In a study reported in 2001, the MYCO/F Lytic medium for the BACTEC instrument performed better than other media and worked well in combination with an Isolator tube (154). A more recent controlled study found that mycobacteria (predominantly M. avium complex) were recovered equally well in both MYCO/F Lytic and BacT/ ALERT MB media in their respective continuously monitored systems with comparable sensitivity, approximately 81 to 85%, but faster time to positivity than the previous standard Isolator 10 tube (30). When commercial mycobacterial blood culture media are used, the manufacturer’s recommendations for handling and processing should be followed. In some cases, such as for the BacT/ALERT MB system, blood is inoculated directly into the bottle at the bedside and a special enrichment fluid is added to those bottles in the laboratory before incubation. Blood can also be collected in a heparin tube (not EDTA or acid-citrate tubes) and submitted to the laboratory for inoculation directly onto plate media (7H-11) and automated instrument broths. If blood collected in Isolator tubes is then inoculated into BACTEC 12B bottles, only 0.2 ml of the bacterial pellet should be used, as a component in the Isolator was found to be inhibitory (34, 159). Inhibitory effects of Isolator samples should be checked before using with other systems. Blood may be obtained in Isolator tubes,

CUMITECH 1C

processed, and plated directly onto mycobacterial media for colony count determination. Several earlier studies showed the Isolator to be effective for recovery of mycobacteria in blood (51, 52, 134, 154). Although high numbers of circulating M. avium-M. intracellulare complex in patients with AIDS is rarely seen today, the success of treatment can be monitored by performing quantitative cultures with the Pediatric 1.5-ml Isolator plated directly onto Middlebrook 7H11 or 7H12 agar plates (172).

MOLECULAR METHODS FOR DETECTING BSIs For certain organisms (more commonly viruses), molecular amplification methods including PCR, nucleic acid sequence-based amplification assays, and occasionally other protocols are the state of the art. This is true today for cytomegalovirus, where the presence and quantity of circulating virus are used to aid diagnosis of exacerbating disease, particularly in patients on immunosuppressive therapies or who are otherwise immunocompromised, or neonates with congenital disease (92). Quantitative tests for EpsteinBarr virus are used to monitor transplant recipients for posttransplantation lymphoproliferative disease (126). And for several years, quantitative molecular tests have been the standard for monitoring viral loads in staging and therapy of hepatitis C and HIV (109, 132, 141). As discussed above with reference to fastidious agents of septicemia, molecular amplification may be the only method for detecting some bacterial agents, such as Rickettsia, Bartonella, Chlamydia, Borrelia burgdorferi, Ehrlichia, and Anaplasma, as well as blood-borne parasites, including Plasmodium, Babesia, Leishmania, and Toxoplasma spp. If the agents cause endocarditis, blood may not be the best specimen; the valve tissue removed during surgery would be the preferred sample for detection of the organism (e.g., Coxiella). When whole blood or white blood cell components only are tested, there is a concentration or extraction step followed by the amplification step, and finally a detection step to identify and in some cases quantify the amplified products. However, for infectious agent molecular testing, a single analyte- and method-specific code is used to represent the entire procedure. The most specific CPT-4 codes available should be used, and nonspecific codes should be used only when an analytespecific code does not yet exist. Several manufacturers are currently evaluating broad-spectrum universal primer amplification for detection of groups of organisms known to cause septicemia, followed by more specific tests when a positive result is obtained. A few leading-edge labo-

Blood Cultures IV

23

ratories are also evaluating their own internally developed molecular amplification methods (102, 115). Unfortunately, the new molecular methods will likely not preclude use of conventional blood cultures for a long time to come, as the organisms must be isolated for susceptibility testing and epidemiological typing studies. In addition, there are too many chances of false-negative or false-positive molecular tests unless a traditional method is used to corroborate molecular results (103). However, as a supplemental procedure, nucleic acid amplification may allow more rapid recognition of septicemia and the earlier use of appropriate antimicrobials, thus enhancing patient outcomes, but it will certainly not decrease overall laboratory costs in the short term.

QUALITY CONTROL Meeting standards for the performance of media and reagents, monitoring temperatures and instrument performance parameters, susceptibility results with organisms of known susceptibility, and many other activities are part of the routine quality control (QC) performed by all laboratories. National regulations promulgated by the Clinical Laboratory Improvement Amendments (CLIA) and published in the Federal Register (42a; also available to download at http://www.asm.org/ASM/files/LeftMarginHeader List/DOWNLOADFILENAME/0000000809/cliafin reg[1].pdf) should be followed unless state guidelines supersede the national guidelines. The College of American Pathologists (CAP) has deemed status within CLIA and publishes checklists available to all laboratories, even those not accredited by CAP, that contain excellent criteria for quality control (http:// www.cap.org/apps/docs/laboratory_accreditation/ checklists/checklistftp.html). Another excellent resource is the chapters on quality assurance and QC in the Clinical Microbiology Procedures Handbook, second edition (63, 130). A future Cumitech on quality systems will also discuss monitoring activities for the blood culture process (D. L. Sewell et al., unpublished work). CLIA and the Clinical and Laboratory Standards Institute (CLSI; formerly NCCLS) both state that commercially produced blood culture media do not require additional in-laboratory QC testing beyond visual inspection if the manufacturer follows CLSI guidelines (101). Users should, however, keep accurate records of lot numbers, dates received and expired, and the product inserts certifying the proper QC by the manufacturer. Isolator tubes do require some in-laboratory QC, as outlined by the manufacturer, including monitoring of vacuum draw volume, breakage in the centrifuge, lysis of blood cells, and appropriate organism recovery.

24

Baron et al.

CUMITECH 1C

QUALITY ASSURANCE Blood culture accessioning, processing, incubating, and all the activities that occur within the laboratory for identification of isolates, susceptibility testing, and results formatting are activities for which QC parameters should be measured and monitored. Quality assurance activities include attention to preanalytical stages, which include venipuncture, filling and labeling the bottles, and transport to the laboratory; and postanalytical activities, which include results reporting to physicians and patient outcomes. The laboratory’s responsibilities include all aspects of testing, and for those functions performed outside the laboratory, communication and cooperation with other caregivers are essential. Some suggested quality assurance activities are summarized in Table 4. Specimen Collection The laboratory must provide protocols for proper collection of blood cultures from all sites, including those drawn through indwelling lines, and specific guidelines on numbers of cultures to obtain, volume Table 4.

Quality assurance parameters to monitor

Factors to monitor

Timing

“Contamination” rate based on laboratory-specific parameters. Separate audits for patient care units, different phlebotomists (nurses vs laboratory, for example), and linedrawn vs. peripheral cultures. Volume of blood obtained by unit. If problems are found, individual phlebotomist monitoring may be necessary. Single blood cultures

Monthly, with timely reporting to units that exceed standards

Too many blood cultures Percent of positive blood cultures Number of blood cultures/ 1,000 patient days Correlation between smear result and culture results Time to calling results to caregiver from time of detection of positive blood culture Direct susceptibility results compared with definitive susceptibility results from pure isolates Paperwork, clerical errors, billing, reimbursement

Annually, for a limited time

Monthly at first, longer increments if this is not a problem Constantly Monthly, per unit and patient type Annually Annually, per technologist For some limited time period, annually or more often if necessary (physician complaints) Constantly

Periodically, such as quarterly

of blood based on weight of patient (for infants and children), and how to handle a low-volume specimen. Criteria for rejection are essential. To monitor that all workers who collect blood are following proper procedures, the laboratory should monitor monthly the contamination rate using parameters such as those developed for the CAP Q-Probe series (131). One suggested definition of contamination is “a single culture positive for coagulase-negative staphylococci or coryneform gram-positive rods or Micrococcus or Propionibacterium, or a single bottle with Bacillus species, not anthracis.” Because some hospital care units (typically the emergency department or a pediatric oncology ward) have higher contamination rates, units should be monitored separately. Contamination rates for blood cultures obtained through lines should be monitored separately from those obtained percutaneously. If rates do not fall below national standards, which are currently placed at 2 to 3% (131, 139), monitoring of an individual phlebotomist may be instituted with counseling and retraining where necessary. Neonatal blood cultures pose a special problem in monitoring for contamination, as it is common practice to obtain only a single blood culture. It has been suggested that the C-reactive protein value may be used to distinguish contamination from infection for blood cultures yielding coagulase-negative staphylococci (113). However there is no current consensus and the relative clinical relevance of coagulase staphylococci in such cultures must be determined by clinicians and infection control practitioners. The volume of blood obtained should also be monitored, at least annually, for a period of time sufficient to get information on the areas where the majority of blood cultures are drawn. One method is to weigh bottles before they are sent to the wards, and to reweigh them once they have been inoculated. One milliliter of blood weighs approximately 1 g. Different patient parameters, such as pediatric versus adult, must be taken into consideration when averages are computed. The percentage of blood samples 2 ml over or below the recommended volume for the vial type should be determined and appropriate inservice activities presented to those phlebotomy groups with problems. False-negative results can be attributed to both too little blood and too much blood, which may allow clotting and trapping of organisms inside the clot. Numbers of Cultures Drawn per Patient Given that a single blood culture is never appropriate, if a laboratory does not have an automatic reflex policy for a second blood culture, the number of single blood cultures should be monitored. Some patient care areas, such as the neonatal intensive care nurs-

CUMITECH 1C

ery, may need to be exempt from the “two-bottle minimum” rule due to the difficulty of collecting multiple cultures, although in truth, coagulase-negative staphylococci are both the most common pathogen and most common contaminant in that setting as well (65). Physicians or patient care areas where single-bottle cultures are more commonly obtained should be counseled. A CAP Q-Probe study evaluated this parameter (105). The authors found that in the 10% lowest-performing hospitals, solitary blood cultures comprised 42.5% of all blood cultures, whereas the top 10% of hospitals had only 3.4% solitary blood cultures. Another quality assurance monitor is for the occasional occurrence of too many blood cultures ordered. Given that increased detection of additional positive cultures with volumes of 80 ml is extremely unlikely (27), blood cultures beyond this total volume are probably unnecessary and the physician should be informed of other potential testing methods. This audit should be ongoing, so the laboratory director, infectious diseases specialist, or clinical pathologist can discuss diagnostic options with the ordering physicians in real time. Positive Culture Rate and Number of Cultures per 1,000 Patient Days If the positive culture rate is too high or too low, physicians may not be ordering blood cultures appropriately. A national audit comprising 649 laboratories found an average of 7.7% positives by laboratory criteria and 8.2% by clinical criteria (131). Depending on the type of hospital (primary care versus tertiary, academic versus community), the rate could be higher or lower. But if positivity drops below 5% or rises above 15%, an investigation should be initiated into whether physicians are ordering blood cultures appropriately. Another monitor of appropriate ordering practices is to audit the number of blood cultures ordered per 1,000 patient days. A sampling of a number of hospitals found values between 103 and 188 (L. Peterson and J. M. Miller, ClinMicroNet Microbiology Laboratory Directors World Wide Web-based forum, 1999). A number between these two extremes is recommended. Technologist Competency and Result Reporting Correlation of smear results (which should be reported immediately to the caregiver) and culture results is one measure of the technologist’s ability to properly interpret smear results. Lack of correlation has patient care consequences, as physicians may begin inappropriate antibiotics if the smear interpretation is incorrect. This should be monitored individually for the technologist and overall, and the results can become part of each technologist’s annual competency assessment. Another aspect of the communica-

Blood Cultures IV

25

tion between laboratory and physician is turnaround time from first detection of positive to informing the caregiver. This should be part of ongoing quality assurance monitoring. Both Doern and Barenfanger have shown major improvement in patient outcomes with faster results reporting (10, 12, 33, 97). For those laboratories that perform immediate susceptibility tests from the blood culture broth directly, the results of the preliminary tests should be compared with those of the definitive test performed on the pure culture isolates. Cumitech 41 has additional suggestions for quality assurance monitoring and other activities to prevent laboratory-associated errors (3). Ordering, Billing, and Reimbursement Finally, laboratories deserve to be fairly compensated for the effort and materials expended on the care of patients. If possible, laboratories should periodically audit a series of patient results to determine whether the CPT-4 coding was correct, the billing was correct, and whether the payment was actually received. The information about specific reimbursement may not be available without special requests, but billing errors and incorrect coding can result in huge fines, as well as lost revenues, so these functions merit scrutiny. One strategy is to pick randomly five patient results to audit quarterly, including the transfer of orders to the computer system, the correct translation of laboratory results to the hospital information system, and the appropriate billing. Another strategy is to request the rate of reimbursement for a particular CPT code. REFERENCES 1. Adler, H., N. Baumlin, and R. Frei. 2003. Evaluation of acridine orange staining as a replacement of subcultures for BacT/ALERT-positive, Gram stain-negative blood cultures. J. Clin. Microbiol. 41:5238–5239. 2. Aliaga, L., J. D. Mediavilla, J. Llosa, C. Miranda, and M. Rosa-Fraile. 2000. Clinical significance of polymicrobial versus monomicrobial bacteremia involving Pseudomonas aeruginosa. Eur. J. Clin. Microbiol. Infect. Dis. 19:871–874. 3. Amsterdam, D., J. Barenfanger, J. Campos, N. Cornish, J. A. Daly, P. Della-Latta, L. D. Gray, G. S. Hall, H. Holmes, and R. L. Sautter. 2004. Cumitech 41, Detection and Prevention of Clinical Microbiology Laboratory-Associated Errors. Coordinating ed., J. W. Snyder. ASM Press, Washington, D.C. 4. Anderson, E. T., L. S. Young, and W. L. Hewitt. 1976. Simultaneous antibiotic levels in “breakthrough” gram-negative bacteremia. Am. J. Med. 4:493–497. 5. Appleman, M. D., V. L. Sutter, and T. N. Sims. 1982. Value of antibiotic prophylaxis in periodontal surgery. J. Periodontol. 53:319–324. 6. Aronson, M. D., and D. H. Bor. 1987. Blood cultures. Ann. Intern. Med. 106:246–253.

26

Baron et al.

CUMITECH 1C

7. Arpi, M., M. W. Benzton, J. Jensen, and W. Fredriksen. 1989. Importance of blood volume cultured in the detection of bacteremia. Eur. J. Clin. Microbiol. Infect. Dis. 8:838–842.

21. Chapin, K. C., and M. C. Musgnug. 2003. Direct susceptibility testing of positive blood cultures by using Sensititre broth microdilution plates. J. Clin. Microbiol. 41:4751–4754.

8. Auckenthaler, R. D., D. M. Ilstrup, and J. A. Washington II. 1982. Comparison of recovery of organisms from blood cultures diluted 10% (volume/volume) and 20% (volume/volume). J. Clin. Microbiol. 15: 860–864.

22. Chusid, M. J., R. W. Perzigian, W. M. Dunne, and E. A. Gecht. 2003. Brucellosis: an unusual cause of a child’s fever of unknown origin. Wis. Med. J. 88: 11–13.

9. Bannatyne, R. M., M. C. Jackson, and Z. Memish. 1997. Rapid diagnosis of Brucella bacteremia by using the BACTEC 9240 system. J. Clin. Microbiol. 35: 2673–2674. 10. Barenfanger, J., C. Drake, and G. Kacich. 1999. Clinical and financial benefits of rapid bacterial identification and antimicrobial susceptibility testing. J. Clin. Microbiol. 37:1415–1418. 11. Barenfanger, J., C. Drake, J. Lawhorn, and S. J. Verhuist. 2004. Comparison of chlorhexidine and tincture of iodine for skin antisepsis in preparation for blood sample collection. J. Clin. Microbiol. 42:2216– 2217. 12. Beekmann, S. E., D. J. Diekema, K. C. Chapin, and G. V. Doern. 2003. Effects of rapid detection of bloodstream infections on length of hospitalization and hospital charges. J. Clin. Microbiol. 41:3119–3125. 13. Bennett, I. L., Jr., and P. B. Beeson. 1954. Bacteremia: a consideration of some experimental and clinical aspects. Yale J. Biol. Med. 26:241–262. 14. Blot, F., G. Nitenberg, B. Chachaty, B. Raynard, N. Germann, S. Antoun, A. Laplanche, C. Brun-Buisson, and C. Tancrede. 1999. Diagnosis of catheter-related bacteremia; a prospective comparison of the time to positivity of hub-blood versus peripheral-blood cultures. Lancet 354:1071–1077. 15. Brun-Buisson, C., F. Abrouk, P. Legrand, Y. Huet, S. Larabi, and M. Rapin. 1987. Diagnosis of central venous catheter-related sepsis. Critical level of quantitative tip cultures. Arch. Intern. Med. 147:873–877. 16. Bryan, C. S. 1989. Clinical implications of positive blood cultures. J. Clin. Microbiol. 2:329–353. 17. Bryant, J. K., and C. L. Strand. 1987. Reliability of blood cultures collected from intravascular catheter versus venipuncture. J. Clin. Microbiol. 88:113–116. 18. Campbell, J., and J. A. Washington II. 1980. Evaluation of the necessity for routine terminal subcultures of previously negative blood cultures. J. Clin. Microbiol. 12:576–587. 19. Capdevila, J. A., A. M. Planes, M. Palomar, I. Gasser, B. Almirante, A. Pahissa, E. Crespo, and J. M. Martinez-Vazquez. 1992. Value of differential quantitative blood cultures in the diagnosis of catheter-related sepsis. Eur. J. Clin. Microbiol. Infect. Dis. 11:403–407. 20. Chapin, K. C., and M. C. Musgnug. 2003. Evaluation of three rapid methods for the direct identification of Staphylococcus aureus from positive blood cultures. J. Clin. Microbiol. 41:4324–4327.

23. Chusid, M. J., D. W. Wortmann, and W. M. Dunne. 1990. Campylobacter upsaliensis sepsis in a boy with acquired hypogammaglobulinemia. Diagn. Microbiol. Infect. Dis. 13:367–369. 24. Cirillo, D. M., E. J. Baron, and G. Marchiaro. 2000. Legionella pneumophila fails to multiply in blood culture broths but can be recovered from Isolator tubes, abstr., C-255, p. 189. Abstr. 100th Gen. Meet. Am. Soc. Microbiol. 2000. American Society for Microbiology, Washington, D.C. 25. Cleri, D. J., M. L. Corrado, and S. J. Seligman. 1980. Quantitative culture of intravenous inserts. J. Infect. Dis. 141:781–786. 26. Cockerill, F. R., III, C. A. Torgerson, G. S. Reed, E. A. Vetter, A. L. Weaver, J. C. Dale, G. D. Roberts, N. K. Henry, D. M. Ilstrup, and J. E. Rosenblatt. 1996. Clinical comparison of Difco ESP, Wampole Isolator, and Becton Dickinson Septi-Chek aerobic blood culturing systems. J. Clin. Microbiol. 34:20–24. 27. Cockerill, F. R., J. W. Wilson, E. A. Vetter, K. M. Goodman, C. A. Torgerson, W. S. Harmsen, C. D. Schleck, D. M. Ilstrupt, J. A. Washington II, and W. R. Wilson. 2004. Optimal testing parameters for blood cultures. Clin. Infect. Dis. 38:1724–1730. 28. Collignon, P. J., N. Soni, I. Y. Pearson, W. P. Woods, R. Munro, and T. C. Sorrel. 1986. Is semiquantitative culture of central vein catheter tips useful in the diagnosis of catheter-associated bacteremia? J. Clin. Microbiol. 24:532–535. 29. Coyle, M. B., L. A. McGonagle, J. J. Plorde, C. R. Clausen, and F. D. Schoenknecht. 1984. Rapid antimicrobial susceptibility testing of isolates from blood cultures by direct inoculation and early reading of disk diffusion tests. J. Clin. Microbiol. 20:473–477. 30. Crump, J. A., D. C. Tanner, S. Mirrett, C. M. McKnight, and L. B. Reller. 2003. Controlled comparison of BACTEC 13A, MYCO/F LYTIC, BacT/ALERT MB, and ISOLATOR 10 systems for detection of mycobacteremia. J. Clin. Microbiol. 41:1987–1990. 31. DesJardin, J. A., M. E. Falagas, R. Ruthazer, J. Griffith, D. Wawrose, D. Schenkein, K. Miller, and D. R. Snydman. 1999. Clinical utility of blood cultures drawn from indwelling central venous catheters in hospitalized patients with cancer. Ann. Intern. Med. 131:641–647. 32. Doern, G. V., A. B. Brueggemann, W. M. Dunne, S. G. Jenkins, D. C. Halstead, and J. C. McLaughlin. 1997. Four-day incubation period for blood culture bottles

CUMITECH 1C

processed with the Difco ESP blood culture system. J. Clin. Microbiol. 35:1290–1292. 33. Doern, G. V., and N. M. Gantz. 1983. Detection of bacteremia in patients receiving antimicrobial therapy; an evaluation of the Antimicrobial Removal Device and 16B medium. J. Clin. Microbiol. 18:43–48. 34. Doern, G. V., and J. A. Westerling. 1994. Optimum recovery of Mycobacterium avium complex from blood specimens of human immunodeficiency virus– positive patients by using small volumes of Isolator concentrate inoculated into BACTEC 12B bottles. J. Clin. Microbiol. 32:2576–2577. 35. Dorscher, C. W., J. E. Rosenblatt, W. R. Wilson, and D. M. Ilstrup. 1991. Anaerobic bacteremia: decreasing rate over a 15-year period. Rev. Infect. Dis. 13:633– 636. 36. Dougherty, M. J., D. H. Spach, A. M. Larson, T. M. Hooton, and M. B. Coyle. 1996. Evaluation of an extended blood culture protocol to isolate fastidious organisms from patients with AIDS. J. Clin. Microbiol. 34:2444–2447. 37. Dunne, W. M., Jr., and M. LaRocco. 2001. Blood culture systems, p. 189–209. In N. Cimolai (ed.), Laboratory Diagnosis of Bacterial Infections. Marcel Dekker, Inc., New York, N.Y. 38. Dunne, W. M., Jr., F. S. Nolte, and M. L. Wilson. 1997. Cumitech 1B, Blood Cultures III. Coordinating ed., J. A. Hindler. ASM Press, Washington, D.C.

Blood Cultures IV

27

neous segment for conservative diagnosis of intravascular catheter-related infection. J. Parent. Ent. Nutr. 24:210–214. 45. Foucault, C., K. Barrau, P. Brouqui, and D. Raoult. 2002. Bartonella quintana bacteremia among homeless people. Clin. Infect. Dis. 35:684–689. 46. Franciosi, R. A., and B. E. Favara. 1972. A single blood culture for confirmation of the diagnosis of neonatal septicemia. Am. J. Clin. Pathol. 57:215–219. 47. Garland, J. S., K. Henrickson, and D. G. Maki. 2002. The 2002 Hospital Infection Control Practices Advisory Committee Centers for Disease Control and Prevention guideline for prevention of intravascular device-related infection. Pediatrics 110:1009–1013. 48. Garner, J. S. 1996. Guideline for isolation precautions in hospitals. The Hospital Infection Control Practices Advisory Committee. Infect. Control Hosp. Epidemiol. 17:53–80. 49. Garner, J. S., R. K. Buck, P. Maloney, D. M. Durkin, S. Toth-Lloyd, M. Duffy, P. Szocik, T. L. McAuliffe, and D. Goldmann. 1995. Comparison of 10% povidone-iodine and 0.5% chlorhexidine gluconate for the prevention of peripheral intravenous catheter colonization in neonates: a prospective trial. Pediatr. Infect. Dis. J. 14:510–516. 50. Gill, V. J. 1981. Lack of clinical relevance in routine terminal subculturing of blood cultures. J. Clin. Microbiol. 14:116–118.

39. Durbin, W. A., E. G. Szymzcak, and D. A. Goldmann. 1978. Quantitative blood cultures in childhood bacteremia. J. Pediatr. 92:778–780.

51. Gill, V. J., C. H. Park, F. Stock, L. L. Gosey, F. G. Witebsky, and H. Masur. 1985. Use of lysis centrifugation (Isolator) and radiometric (BACTEC) blood culture systems for the detection of mycobacteremia. J. Clin. Microbiol. 22:543–546.

40. Evans, M. R., A. L. Truant, J. Kostman, and L. Locke. 1991. The detection of positive blood cultures by the BACTEC NR660. The clinical importance of four-day versus seven-day testing. Diagn. Microbiol. Infect. Dis. 14:107–110.

52. Gill, V. J., and F. Stock. 1987. Detection of Mycobacterium avium-Mycobacterium intracellulare in blood cultures using concentrated and unconcentrated blood in conjunction with a radiometric detection system. Diagn. Microbiol. Infect. Dis. 6:119–123.

41. Everett, E. D., and J. V. Hirschmann. 1984. Transient bacteremia and endocarditis prophylaxis. Medicine 56:61–77.

53. Gleckman, R., and D. Hilbert. 1982. Afebrile bacteremia. A phenomenon in geriatric patients. JAMA 248:1478–1481.

42. Everts, R. J., E. N. Vinson, P. O. Adholla, and L. B. Reller. 2001. Contamination of catheter-drawn blood cultures. J. Clin. Microbiol. 39:3393–3394.

54. Gomez, J., V. Banos, J. Ruiz, F. Herrero, M. Perez, M. Canteras, and M. Valdes. 1993. Clinical significance of anaerobic bacteremias in a general hospital. A prospective study from 1988 to 1992. Clin. Investig. 71:595–599.

42a. Federal Register. 2003. Medicare, Medicaid, and CLIA programs; laboratory requirements relating to quality systems and certain personnel qualifications; final rule. Fed. Regist. 68:3639–3714.

55. Hall, M. M., D. M. Ilstrupt, and J. A. Washington II. 1976. Effect of volume of blood cultured on detection of bacteremia. J. Clin. Microbiol. 3:643–645.

43. Flynn, P. M., J. L. Shenep, and F. F. Borret. 1988. Differential quantitation with a commercial blood culture tube for diagnosis of catheter-related infection. J. Clin. Microbiol. 26:1045–1046.

56. Han, X. Y., and A. L. Truant. 1999. The detection of positive blood cultures by the AccuMed ESP-384 system: the clinical significance of three-day testing. Diagn. Microbiol. Infect. Dis. 33:1–6.

44. Fortun, J., J. A. Perez-Molina, A. Asensio, C. Calderon, J. L. Casado, N. Mir, A. Moreno, and A. Guerrero. 2000. Semiquantitative culture of subcuta-

57. Hansen, D. S., A. G. Jensen, N. Norskov-Lauritsen, R. Skov, and B. Bruun. 2002. Direct identification and susceptibility testing of enteric bacilli from positive

28

Baron et al.

blood cultures using VITEK (GNI/GNS-GA). Clin. Microbiol. Infect. Dis. 8:38–44. 58. Hardy, D. J., B. B. Hulbert, and P. C. Migneault. 1992. Time to detection of positive BacT/Alert blood cultures and lack of need for routine subculture of 5to 7-day negative cultures. J. Clin. Microbiol. 30:2743–2745. 59. Henry, N. K., C. A. McLimans, A. J. Wright, R. L. Thompson, W. R. Wilson, and J. A. Washington II. 1983. Microbiological and clinical evaluation of the Isolator lysis-centrifugation blood culture tube. J. Clin. Microbiol. 17:315–317. 60. Ilstrupt, D. M., and J. A. Washington II. 1983. The importance of volume of blood cultured in the detection of bacteremia and fungemia. Diagn. Microbiol. Infect. Dis. 1:107–110. 61. Isaacman, D. J., and R. B. Karasic. 1990. Lack of effect of changing needles on contamination of blood cultures. Pediatr. Infect. Dis. 9:274–278. 62. Janyapoon, K. S., S. Korbsrisate, H. Thamapa, S. Thongmin, S. Kanjanahareutai, N. Wongpredee, and S. Sarasombath. 2000. Rapid detection of Salmonella enterica serovar choleraesuis in blood cultures by a dot blot enzyme-linked immunosorbent assay. Clin. Diagn. Lab. Immunol. 7:977–979. 63. Jenkins, S. G., and D. L. Sewell. 2004. Quality control, p. 14.2.1–14.2.34. In H. D. Isenberg (ed. in chief), Clinical Microbiology Procedures Handbook, 2nd ed. ASM Press, Washington, D.C. 64. Jorgensen, J. H., S. Mirrett, L. C. McDonald, P. R. Murray, M. P. Weinstein, J. Fune, C. W. Trippy, M. Masterson, and L. B. Reller. 1997. Controlled clinical laboratory comparison of BACTEC Plus Aerobic/F resin medium with BacT/Alert Aerobic FAN medium for detection of bacteremia and fungemia. J. Clin. Microbiol. 35:53–58. 65. Kellogg, J. A., F. L. Ferrentino, J. Liss, S. I. Shapiro, and D. A. Bankert. 1994. Justification and implementation of a policy requiring two blood cultures when one is ordered. Lab. Med. 25:323–330. 66. Kellogg, J. A., F. L. Ferrentino, J. Liss, S. I. Shapiro, and D. A. Bankert. 1997. Frequency of low-level bacteremia in infants from birth to two months of age. Pediatr. Infect. Dis. 16:381–385. 67. Kellogg, J. A., J. P. Manzella, and D. A. Bankert. 2000. Frequency of low-level bacteremia in children from birth to fifteen years of age. J. Clin. Microbiol. 38:2181–2185. 68. Kelly, M., R. Wunderlich Wciorka, S. McConico, and L. R. Peterson. 1996. Sonicated vascular catheter tip cultures. Quantitative association with catheter-related sepsis and the non-utility of an adjuvant cytocentrifuge Gram stain. Am. J. Clin. Pathol. 105:210–215. 69. Kiehlbauch, J. A., R. V. Tauxe, C. N. Baker, and I. K. Wachsmuth. 1994. Helicobacter cinaedi–associated bacteremia and cellulitis in immunocompromised patients. Ann. Intern. Med. 121:90–93.

CUMITECH 1C

70. Kiehn, T. E., and D. Armstrong. 1990. Changes in the spectrum of organisms causing bacteremia and fungemia in immunocompromised patients due to venous access devices. Eur. J. Clin. Microbiol. Infect. Dis. 9:869–872. 71. King, T. C., and P. B. Price. 1963. An evaluation of iodophors as skin antiseptics. Surg. Gynecol. Obstet. 116:361–365. 72. Kite, P., B. M. Dobbins, M. H. Wilcox, W. N. Fawley, A. J. L. Kindon, D. Thomas, M. J. Tighe, and M. J. McMahon. 1997. Evaluation of a novel endoluminal brush method of in situ diagnosis of catheter-related sepsis. J. Clin. Pathol. 50:278–282. 73. Krumholz, H., S. Cummings, and M. York. 1991. Blood culture phlebotomy: switching needles does not prevent contamination. Arch. Intern. Med. 14:111– 118. 74. Kubina, M., B. Jaulhac, X. Delabrance, C. Lindenman, Y. Piemont, and H. Monteil. 1999. Oxacillin susceptibility testing of staphylococci directly from BACTEC plus blood cultures by the BBL Crystal MRSA ID system. J. Clin. Microbiol. 37:2034–2036. 75. LaScolea, L. J., D. Dryja, T. D. Sullivan, L. Mosovich, N. Ellerstein, and E. Neter. 1990. Diagnosis of bacteremia in children by quantitative direct plating and a radiometric procedure. J. Clin. Microbiol. 13:478– 482. 76. Leisure, M. K., D. M. Moore, J. D. Schwartzman, G. F. Hayden, and L. G. Donowitz. 1990. Changing the needle when inoculating blood cultures: a no-benefit and high-risk procedure. JAMA 264:2111–2112. 77. Li, J., J. J. Plorde, and L. C. Carlson. 1994. Effects of volume and periodicity on blood cultures. J. Clin. Microbiol. 32:2829–2831. 78. Linares, J., A. Dominquez, and R. Martin. 1997. Current laboratory techniques in the diagnosis of catheterrelated infections. Nutrition 13:10S–14S. 79. Ling, T. K. W., Z. K. Liu, and A. F. B. Cheng. 2003. Evaluation of the VITEK 2 system for rapid direct identification and susceptibility testing of gram-negative bacilli from positive blood cultures. J. Clin. Microbiol. 41:4705–4707. 80. Little, J. R., P. R. Murray, P. Traynor, and E. Spitznagel. 1999. A randomized trial of povidone-iodine compared with iodine tincture for venipuncture site disinfection: effects on rates of blood culture contamination. Am. J. Med. 107:119–125. 81. Lombardi, D. P., and N. C. Engleberg. 1992. Anaerobic bacteremia: incidence, patient characteristics, and clinical significance. Am. J. Med. 92:53–60. 82. Louie, L., J. Goodfellow, P. Mathieu, A. Glatt, M. Louie, and A. E. Simor. 2002. Rapid detection of methicillin-resistant staphylococci from blood culture bottles by using a multiplex PCR assay. J. Clin. Microbiol. 40:2786–2790.

CUMITECH 1C

83. MacGregor, R. R., and H. N. Beaty. 1972. Evaluation of positive blood cultures. Arch. Intern. Med. 130: 84–87. 84. Madani, T. A. 2000. Clinical infections and bloodstream isolates associated with fever in patients undergoing chemotherapy for acute myeloid leukemia. Infection 28:367–373. 85. Maki, D. G., C. E. Weise, and H. W. Sarafin. 1977. A semiquantitative culture method for identifying intravenous-catheter-related infection. N. Engl. J. Med. 296:1305–1309. 86. Martin, G. S., S. Mannino, S. Eaton, and M. Moss. 2003. The epidemiology of sepsis in the United States from 1979 through 2000. N. Engl. J. Med. 348:1546– 1554. 87. Masterson, K. C., and J. E. McGowan, Jr. 1988. Detection of positive blood cultures by the BACTEC NR660: the clinical importance of five versus seven days of testing. Am. J. Clin. Pathol. 90:91–94. 88. Maurin, M., F. Eb, J. Etienne, and D. Raoult. 1997. Serological cross-reactions between Bartonella and Chlamydia species: implications for diagnosis. J. Clin. Microbiol. 35:2283–2287.

Blood Cultures IV

29

and recovery of Histoplasma capsulatum from blood, abstr. C-101, p. 140. Abstr. 104th Gen. Meet. Am. Soc. Microbiol. 2004. American Society for Microbiology, Washington, D.C. 97. Munson, E. L., D. J. Diekema, S. E. Beekmann, K. C. Chapin, and G. V. Doern. 2003. Detection and treatment of bloodstream infection: laboratory reporting and antimicrobial management. J. Clin. Microbiol. 41:495–497. 98. Murray, P. R., and J. E. Sondag. 1978. Evaluation of routine subcultures of macroscopically negative blood cultures for detection of anaerobes. J. Clin. Microbiol. 8:427–430. 99. Murray, P. R., P. Traynor, and D. Hopson. 1992. Critical assessment of blood culture techniques: analysis of recovery of obligate and facultative anaerobes, strict aerobic bacteria, and fungi in aerobic and anaerobic blood culture bottles. J. Clin. Microbiol. 30:1462– 1468. 100. Musso, D., and D. Raoult. 1995. Coxiella burnetii blood cultures from acute and chronic Q-fever patients. J. Clin. Microbiol. 33:3129–3132.

89. McCarthy, L. R., and J. E. Senne. 1980. Evaluation of acridine orange stain for detection of microorganisms in blood cultures. J. Clin. Microbiol. 11:281–285.

101. National Committee for Clinical Laboratory Standards. 2004. Quality Assurance for Commercially Prepared Microbiological Culture Media. Approved standard M22-A3. National Committee for Clinical Laboratory Standards, Wayne, Pa.

90. McDonald, L. C., J. Fune, L. D. Guido, M. P. Weinstein, L. G. Reimer, T. M. Flynn, M. L. Wilson, S. Mirrett, and L. B. Reller. 1996. Clinical importance of the increased sensitivity of BacT/Alert FAN aerobic and anaerobic blood culture bottles. J. Clin. Microbiol. 34:2180–2184.

102. Nikkari, S., R. Gotoff, P. P. Bourbeau, R. E. Brown, N. R. Kamal, and D. A. Relman. 2002. Identification of Cardiobacterium hominis by broad-range bacterial polymerase chain reaction analysis in a case of culture-negative endocarditis. Arch. Intern. Med. 162:477–479.

91. Memish, Z. A., and M. W. Mah. 2001. Brucellosis in laboratory workers at a Saudi Arabian hospital. Am. J. Infect. Control 29:48–52.

103. Nikkari, S., I. J. McLaughlin, W. Bi, D. E. Dodge, and D. A. Relman. 2001. Does blood of healthy subjects contain bacterial ribosomal DNA? J. Clin. Microbiol. 39:1956–1959.

92. Mengoli, C., R. Cusinato, M. A. Biasolo, S. Cesaro, C. Parolin, and G. Palu. 2004. Assessment of CMV load in solid organ transplant recipients by pp65 antigenemia and real-time quantitative DNA PCR assay: correlation with pp67 RNA detection. J. Med. Virol. 74:78–84. 93. Mermel, L. A., and D. G. Maki. 1993. Detection of bacteremia in adults: consequences of culturing an inadequate volume of blood. Ann. Intern. Med. 119:270–272.

104. Noviello, S. 2004. Laboratory-acquired brucellosis. Emerg. Infect. Dis. 10:1848–1850. 105. Novis, D. A., J. C. Dale, R. B. Schifman, S. G. Ruby, and M. K. Walsh. 2001. Solitary blood cultures: a College of American Pathologists Q-probes study of 132,778 blood culture sets in 333 small hospitals. Arch. Pathol. Lab. Med. 125:1290–1294.

94. Mimoz, O., A. Karim, A. Mercat, M. Cosseron, B. Falissard, F. Parker, C. Richard, K. Samii, and P. Nordmann. 1999. Chlorhexidine compared with povidone-iodine as skin preparation before blood culture: a randomized controlled trial. J. Clin. Microbiol. 131:834–837.

106. Oliveira, K., S. M. Brecher, A. Durbin, D. S. Shapiro, D. R. Schwartz, P. C. De Girolami, J. Dakos, G. W. Procop, D. Wilson, C. S. Hanna, G. Haase, H. Peltroche-Llacsahuanga, K. C. Chapin, M. C. Musgnug, M. H. Levi, C. Shoemaker, and H. Stender. 2003. Direct identification of Staphylococcus aureus from positive blood culture bottles. J. Clin. Microbiol. 41: 889–891.

95. Morris, A. J., M. L. Wilson, S. Mirrett, and L. B. Reller. 1993. Rationale for selective use of anaerobic blood cultures. J. Clin. Microbiol. 31:2110–2113.

107. Ortiz, E., and M. A. Sande. 2000. Routine use of anaerobic blood cultures: are they still indicated? Am. J. Med. 108:445–447.

96. Munier, G., D. Black, and J. Snyder. 2004. Evaluation of BacT/ALERT blood culture media for the detection

108. Paisley, J. W., J. E. Rosenblatt, M. Hall, and J. A. Washington. 1978. Evaluation of a routine anaerobic

30

Baron et al.

subculture of blood cultures for detection of anaerobic bacteremia. J. Clin. Microbiol. 8:764–766. 109. Panel on Clinical Practices for Treatment of HIV Infection. 2004. Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected Adults and Adolescents, p. 1–110. Department of Health and Human Services, Bethesda, Md. 110. Plorde, J. J., F. C. Tenover, and L. G. Carlson. 1985. Specimen volume versus yield in the BACTEC blood culture system. J. Clin. Microbiol. 22:292–295. 111. Pohlman, J. K., B. A. Kirkley, K. A. Easley, B. A. Basille, and J. A. Washington. 1995. Controlled clinical evaluation of BACTEC Plus Aerobic/F and BacT/Alert aerobic FAN bottles for detection of bloodstream infections. J. Clin. Microbiol. 33:2856– 2858. 112. Pongsunk, S., N. Thirawattanasuk, N. Piyasangthong, and P. Ekpo. 1999. Rapid identification of Burkholderia pseudomallei in blood cultures by a monoclonal antibody assay. J. Clin. Microbiol. 37:3662– 3667. 113. Pourcyrous, M., H. S. Bada, S. B. Korones, V. Baselski, and S. P. Wong. 1993. Significance of serial Creactive protein responses in neonatal infection and other disorders. Pediatrics 92:431–435. 114. Putnam, L. R., W. J. Howard, M. A. Pfaller, F. P. Koontz, and R. N. Jones. 1997. Accuracy of the Vitek system for antimicrobial susceptibility testing Enterobacteriaceae bloodstream infection isolates: use of “direct” inoculation from Bactec 9240 blood culture bottles. Diagn. Microbiol. Infect. Dis. 28:101–104. 115. Qian, Q., Y.-W. Tang, C. P. Kolbert, C. A. Torgerson, J. G. Hughes, E. A. Vetter, W. S. Harmsen, S. O. Montgomery, F. R. Cockerill III, and D. H. Persing. 2001. Direct identification of bacteria from positive blood cultures by amplification and sequencing of the 16S rRNA gene: evaluation of BACTEC 9240 instrument true-positive and false-positive results. J. Clin. Microbiol. 39:3578–3582.

CUMITECH 1C

120. Reller, L. B., K. A. Lichtenstein, S. Mirrett, and W.-L. L. Wang. 1978. Controlled evaluation of the ratio of blood to broth in the detection of bacteremia by blood culture, p. 306. Abstr. Annu. Meet. Am. Soc. Microbiol. 1978. American Society for Microbiology, Washington, D.C. 121. Reller, L. B., P. R. Murray, and J. D. MacLowry. 1982. Cumitech 1A, Blood Cultures II. Coordinating ed., J. A. Washington II. American Society for Microbiology, Washington, D.C. 122. Richter, S. S., S. E. Beekmann, D. J. Croco, F. P. Koontz, M. A. Pfaller, and G. V. Doern. 2002. Minimizing the workup of blood culture contaminants: implementation and evaluation of a laboratory-based algorithm. J. Clin. Microbiol. 40:2437–2444. 123. Rijnders, B. J. A., C. Verwaest, W. E. Peetermans, A. Wilmers, S. Vandescasteele, J. Van Eldere, and E. Van Wijngaerden. 2001. Difference in time to positivity of hub-blood versus nonhub-blood cultures is not useful for the diagnosis of catheter-related bloodstream infection in critically ill patients. Crit. Care Med. 29:1399–1403. 124. Riley, J. A., B. J. Heiter, and P. P. Bourbeau. 2003. Comparison of recovery of blood culture isolates from two BacT/ALERT FAN aerobic blood culture bottles with recovery from one FAN aerobic bottle and one FAN anaerobic bottle. J. Clin. Microbiol. 41:213–217. 125. Roiz, M. P., F. G. Peralta, R. Valle, and R. Arjona. 1998. Microbiological diagnosis of brucellosis. J. Clin. Microbiol. 36:1819. 126. Rowe, D. T., L. Qu, J. Reyes, N. Jabbour, E. Yunis, P. Putnam, S. Todo, and M. Green. 1997. Use of quantitative competitive PCR to measure EpsteinBarr virus genome load in the peripheral blood of pediatric transplant patients with lymphoproliferative disorders. J. Clin. Microbiol. 35:1612–1615. 127. Rushforth, J. A., C. M. Hoy, P. Kite, and J. W. L. Puntis. 1993. Rapid diagnosis of central venous catheter sepsis. Lancet 342:402–403.

116. Raad, I., H. A. Hanna, B. Alakech, I. Chatzinikolaou, M. M. Johnson, and J. Tarrand. 2004. Differential time to positivity: a useful method for diagnosing catheter-related bloodstream infections. Ann. Intern. Med. 140:18–25.

128. Salventi, J. F., T. A. Davies, E. L. Randall, S. Whitaker, and J. R. Waters. 1979. Effect of blood dilution on recovery of organisms from clinical blood cultures in medium containing sodium polyanethol sulfonate. J. Clin. Microbiol. 9:248–252.

117. Reimer, L. G., and K. C. Carroll. 2003. Procedures for the storage of microorganisms, p. 67–73. In P. R. Murray, E. J. Baron, J. H. Jorgensen, M. A. Pfaller, and R. H. Yolken (ed.), Manual of Clinical Microbiology, 8th ed. ASM Press, Washington, D.C.

129. Sandven, P., and A. Hoiby. 1981. The importance of blood volume cultured on detection of bacteraemia. Acta. Pathol. Microbiol. Scand. 89:149–152.

118. Reimer, L. G., M. P. Weinstein, and M. L. Wilson. 2003. Update on the detection of bacteremia and fungemia. Clin. Microbiol. Rev. 10:444–465.

130. Schifman, R. B., and S. G. Jenkins. 2004. Quality assessment and improvement (quality assurance), p. 14.1.1–14.1.29. In H. D. Isenberg (ed. in chief), Clinical Microbiology Procedures Handbook, 2nd ed. ASM Press, Washington, D.C.

119. Reller, L. B. 1981. Laboratory procedures in the management of infective endocarditis, p. 235–267. In A. L. Bisno (ed.), Treatment of Infective Endocarditis. Grune & Stratton, New York, N.Y.

131. Schifman, R. B., C. L. Strand, F. A. Meier, and P. J. Howanitz. 1998. Blood culture contamination: a College of American Pathologists Q-Probes study involving 640 institutions and 497,134 specimens

CUMITECH 1C

from adult patients. Arch. Pathol. Lab. Med. 122: 216–221. 132. Schutten, M., and H. G. Niesters. 2001. Clinical utility of viral quantification as a tool for disease monitoring. Expert Rev. Mol. Diagn. 1:153–162. 133. Reference deleted. 134. Shanson, D. C., and M. S. Dryden. 1988. Comparison of methods for isolating Mycobacterium aviumintracellulare from blood of patients with AIDS. J. Clin. Pathol. 41:687–690. 135. Sharp, S. E., J. C. McLaughlin, J. M. Goodman, J. Moore, S. M. Spanos, D. W. Keller, and R. J. Poppiti. 1993. Clinical assessment of anaerobic isolates from blood cultures. Diagn. Microbiol. Infect. Dis. 17:19–22. 136. Sherertz, R. J., S. O. Heard, and I. I. Raad. 1997. Diagnosis of triple-lumen catheter infection: comparison of roll plate, sonication, and flushing methodologies. J. Clin. Microbiol. 35:641–646. 137. Sherertz, R. J., I. I. Raad, A. Belani, L. C. Koo, K. H. Rand, D. L. Pickett, S. A. Straub, and L. L. Fauerbach. 1990. Three-year experience with sonicated vascular catheter cultures in a clinical microbiology laboratory. J. Clin. Microbiol. 28:76–82. 138. Siegman-Igra, Y., A. M. Anglim, D. E. Shapiro, K. M. Adal, B. A. Strain, and B. M. Farr. 1997. Diagnosis of vascular catheter-related bloodstream infection: a meta-analysis. J. Clin. Microbiol. 35:928–936. 139. Souvenir, D., D. E. Anderson, S. Palpant, H. Mroch, S. Askin, J. Anderson, J. Claridge, J. Eiland, C. Malone, M. W. Garrison, P. Watson, and D. M. Campbell. 1998. Blood cultures positive for coagulase-negative staphylococci: antisepsis, pseudobacteremia, and therapy of patients. J. Clin. Microbiol. 36:1923– 1926. 140. Staneck, J. L., and S. Vincent. 1981. Inhibition of Neisseria gonorrhoeae by sodium polyanetholesulfonate. J. Clin. Microbiol. 13:463–467. 141. Strader, D. B., T. Wright, D. L. Thomas, and L. B. Seeff. 2004. Diagnosis, management, and treatment of hepatitis C. Hepatology 39:1147–1171. 142. Strand, C. L. 1988. Blood Cultures: Consensus Recommendations in 1988. Microbiology no. MB 88-1 (MB-172). American Society for Clinical Pathologists Check Sample Continuing Education Program. American Society of Clinical Pathologists, Chicago, Ill. 143. Strand, C. L., R. R. Wajsbort, and K. Sturmann. 1993. Effect of iodophor vs. iodine tincture skin preparation on blood culture contamination rate. JAMA 269: 1004–1006. 144. Sumerkan, B., S. Gokahmetoglu, and D. Esel. 2001. Brucella detection in blood: comparison of the BacT/Alert standard aerobic bottle, BacT/Alert FAN aerobic bottle and BacT/Alert enhanced FAN aerobic bottle in simulated blood culture. Clin. Microbiol. Infect. 7:369–372.

Blood Cultures IV

31

145. Szymzcak, E. G., J. T. Barr, W. A. Durbin, and D. A. Goldman. 1979. Evaluation of blood culture procedures in a paediatric hospital. J. Clin. Microbiol. 9:88–92. 146. Tenney, J. H., B. Reller, S. Mirrett, W.-L. L. Wang, and M. P. Weinstein. 1982. Controlled evaluation of the volume of blood cultured in detection of bacteremia and fungemia. J. Clin. Microbiol. 15:558– 561. 147. Thomson, R. B., Jr., B. L. Evans, and J. L. Sourtheland. 1991. Collecting Blood for Culture. Generalist Microbiology Tech Sample No. G-1. American Society of Clinical Pathologists, Northfield, Ill. 148. Tierney, B. M., N. K. Henry, and J. A. Washington II. 1983. Early detection of positive blood cultures by the acridine orange staining technique. J. Clin. Microbiol. 18:830–833. 149. Tierno, P. M., Jr., K. Inglima, and M. T. Parisi. 1995. Detection of Bartonella (Rochalimaea) henselae bacteremia using BacT/Alert blood culture system. Am. J. Clin. Pathol. 104:530–536. 150. Tokars, J. I. 2004. Predictive value of blood cultures positive for coagulase-negative staphylococci: implications for patient care and health care quality assurance. Clin. Infect. Dis. 39:333–341. 151. Trivett-Moore, N. L., W. D. Rawlinson, M. Yuen, and G. L. Gilbert. 1997. Helicobacter westmeadii sp. nov., a new species isolated from blood cultures of two AIDS patients. J. Clin. Microbiol. 35:1144– 1150. 152. van Griethuysen, A., A. Buiting, W. Goessens, R. van Keulen, R. Wintermans, and J. Kluytmans. 1998. Multicenter evaluation of a modified protocol for the RAPIDEC Staph system for direct identification of Staphylococcus aureus in blood cultures. J. Clin. Microbiol. 36:3707–3709. 153. Veterans Administration Ad Hoc Interdisciplinary Advisory Commitee on Antimicrobial Usage. 1977. Audits of antimicrobial usage: guidelines for peer review. 11. Routine blood cultures. JAMA 237:1723. 154. Vetter, E., C. Torgerson, A. Feuker, J. Hughes, S. Harmsen, C. Schleck, C. Horstmeier, G. Roberts, and F. Cockerill III. 2001. Comparison of the BACTEC MYCO/F Lytic bottle to the Isolator tube, BACTEC Plus Aerobic F/bottle, and BACTEC Anaerobic Lytic/10 bottle and comparison of the BACTEC Plus Aerobic F/bottle to the Isolator tube for recovery of bacteria, mycobacteria, and fungi from blood. J. Clin. Microbiol. 39:4380–4386. 155. Waites, K. B., S. A. Brookings, S. A. Moser, and B. L. Zimmer. 1998. Direct bacterial identification from positive BacT/ALERT blood cultures using MicroScan overnight and rapid panels. Diagn. Microbiol. Infect. Dis. 32:21–26. 156. Walker, R. C., N. K. Henry, J. A. Washington II, and R. L. Thompson. 1986. Lysis-centrifugation blood culture technique: clinical impact in Staphylococcus

32

Baron et al.

aureus bacteremia. Arch. Intern. Med. 146:2341– 2343. 157. Washington, J. A. 1975. Blood cultures: principles and techniques. Mayo Clin. Proc. 50:91–95. 158. Washington, J. A., II. 1978. Conventional approaches to blood culture, p. 41–88. In J. A. Washington II (ed.), The Detection of Septicemia. CRC Press, West Palm Beach, Fla. 159. Wasilauskas, B., and R. J. Morrell, Jr. 1994. Inhibitory effect of the Isolator blood culture system on growth of Mycobacterium avium-M. intracellulare in BACTEC 12B bottles. J. Clin. Microbiol. 32:654– 657. 160. Weinstein, M. P. 2003. Blood culture contamination: persisting problems and partial progress. J. Clin. Microbiol. 41:2275–2278. 161. Weinstein, M. P. 1996. Current blood culture methods and systems; clinical concepts, technology, and interpretation of results. Clin. Infect. Dis. 23:40–46. 162. Weinstein, M. P., S. Mirrett, L. G. Reimer, M. L. Wilson, S. Smith-Elekes, C. R. Chuard, K. L. Joho, and L. B. Reller. 1995. Controlled evaluation of BacT/ Alert standard aerobic and FAN aerobic blood culture bottles for detection of bacteremia and fungemia. J. Clin. Microbiol. 33:978–981. 163. Weinstein, M. P., J. R. Murphy, L. B. Reller, and K. A. Lichtenstein. 1983. The clinical signficance of positive blood cultures; a comprehensive analysis of 500 episodes of bacteremia and fungemia in adults. II. Clinical observations with special reference to factors influencing prognosis. Rev. Infect. Dis. 5:54–70. 164. Weinstein, M. P., and L. B. Reller. 1984. Clinical importance of “breakthrough” bacteremia. Am. J. Med. 76:175–180. 165. Weinstein, M. P., L. B. Reller, J. R. Murphy, and K. A. Lichtenstein. 1983. The clinical signficance of positive blood cultures; a comprehensive analysis of 500 episodes of bacteremia and fungemia in adults. I. Laboratory and epidemiologic observations. Rev. Infect. Dis. 5:35–53. 166. Weinstein, M. P., M. L. Towns, S. M. Quartey, S. Mirrett, L. G. Reimer, G. Parmigiani, and L. B. Reller. 1997. The clinical significance of positive blood cultures in the 1990s; a prospective comprehensive evaluation of the microbiology, epidemiology, and outcome of bacteremia and fungemia in adults. Clin. Infect. Dis. 24:584–602. 167. Weiss, H., and R. Ottenberg. 1932. Relation between bacteria and temperature in subacute bacterial endocarditis. J. Infect. Dis. 50:61–68.

CUMITECH 1C

168. Werner, A. S., C. G. Cobbs, D. Kaye, and E. W. Hook. 1967. Studies on the bacteremia of bacterial endocarditis. JAMA 202:199–203. 169. Wilson, M. L., S. Mirrett, L. B. Reller, M. P. Weinstein, and B. W. Reary. 1993. Recovery of clinically important microorganisms from the BacT/Alert blood culture system does not require 7-day testing. Diagn. Microbiol. Infect. Dis. 16:31–34. 170. Wilson, M. L., M. P. Weinstein, S. Mirrett, L. G. Reimer, R. J. Feldman, C. R. Chuard, and L. B. Reller. 1995. Controlled evaluation of BacT/Alert standard anaerobic and FAN anaerobic blood culture bottles for the detection of bacteremia and fungemia. J. Clin. Microbiol. 33:2265–2270. 171. Wisplinghoff, H., H. Seifert, S. M. Tallent, T. Bischoff, R. P. Wenzel, and M. B. Edmond. 2003. Nosocomial bloodstream infections in pediatric patients in United States hospitals: epidemiology, clinical features and susceptibilities. Pediatr. Infect. Dis. 22:686–691. 172. Wong, B., F. F. Edwards, T. E. Kiehn, E. Whimbey, H. Donnelly, E. M. Bernard, J. W. Gold, and D. Armstrong. 1985. Continuous high-grade Mycobacterium avium-intracellulare bacteremia in patients with the acquired immune deficiency syndrome. Am. J. Med. 78:35–40. 173. Yagupsky, P. 1999. Detection of brucellae in blood cultures. J. Clin. Microbiol. 37:3437–3442. 174. Yagupsky, P., N. Peled, J. Press, O. Abramson, and M. Abu-Rashid. 1997. Comparison of BACTEC 9240 Peds Plus medium and Isolator 1.5 microbial tube for detection of Brucella melitensis from blood cultures. J. Clin. Microbiol. 35:1382–1384. 175. Yagupsky, P., N. Peled, J. Press, M. Abu-Rashid, and O. Abramson. 1997. Rapid detection of Brucella melitensis from blood cultures by a commercial system. Eur. J. Clin. Microbiol. Infect. Dis. 16:605–607. 176. York, M. K. 2004. Leptospira culture, p. 3.14.1– 3.14.5. In H. D. Isenberg (ed. in chief), Clinical Microbiology Procedures Handbook, 2nd ed. ASM Press, Washington, D.C. 177. York, M. K., N. K. Henry, and P. Gilligan. 2004. Blood cultures, p. 3.4.1.1–3.4.1.19. In H. D. Isenberg (ed. in chief), Clinical Microbiology Procedures Handbook, 2nd ed. ASM Press, Washington, D.C. 178. Zaidi, A. K. M., A. L. Knaut, S. Mirrett, and L. B. Reller. 1995. Value of routine anaerobic blood cultures in pediatric patients. J. Pediatr. 127:263–268. 179. Zeaiter, Z., P. E. Fournier, G. Greub, and D. Raoult. 2003. Diagnosis of Bartonella endocarditis by a realtime nested PCR assay using serum. J. Clin. Microbiol. 41:919–925.