VIEWS: 362 PAGES: 53 POSTED ON: 8/18/2012
Bacteremia, Fungemia, and Blood Cultures Dr. John R. Warren Department of Pathology Northwestern University Feinberg School of Medicine June 2007 Essential Elements of Blood Cultures • Pathophysiology of bacteremia and fungemia • Microbiology of bacteremia and fungemia • Determinants of mortality in bacteremia and fungemia • Clinical signs and symptoms of septicemia (predictors) Essential Elements of Blood Cultures • Technical variables in the collection of blood cultures • Incubation conditions • Blood culture systems • Quality management monitors Pathophysiology of Bacteremia and Fungemia • Bacteria and fungi normally cleared from blood by the mononuclear phagocyte system (MPS) (reticuloendothelial system) • If entry of bacteria or fungi into the circulation exceeds MPS clearance capacity, bacteremia or fungemia results Pathophysiology of Bacteremia and Fungemia • Encapsulated bacteria and yeast poorly cleared from the circulation by fixed macrophages of the MPS (parasite factor) especially in the absence of opsonizing antibody (host factor) • Diminished numbers of neutrophils impedes ability of the host to contain encapsulated bacteria and yeast at sites of tissue infection Pathophysiology of Bacteremia and Fungemia • Bacteremia or fungemia represents a failure of host defenses to localize an infection at its primary tissue site • Bacteremia or fungemia also reflects a failure of a physician to remove, drain, or otherwise sterilize sites of infection • Clinical patterns of bacteremia or fungemia: transient, intermittent, and continuous Pathophysiology of Bacteremia and Fungemia • Transient bacteremia or fungemia lasts minutes to a few hours, is the most common type, and occurs after manipulation of infected tissue (abscesses, furuncles, cellulitis), instrumentation of colonized mucosal surfaces (dental procedures, cystoscopy, sigmoidoscopy), and surgery in contaminated areas (prostate resection, debridement of infected burns, vaginal hyterectomy), reflecting release of organisms into the circulation secondary to tissue trauma resulting from medical procedures Pathophysiology of Bacteremia and Fungemia • Transient bacteremia or fungemia also occurs early in acute bacterial infection including pneumonia, meningitis, septic arthritis, and hematogenous osteomyelitis, reflecting release of organisms through lymphatics draining infected tissue into the circulation • Intermittent bacteremia or fungemia occurs, clears, then recurs with the same organism, and develops with undrained closed-space abscesses (intra- abdominal, pelvic, perinephric, hepatic, prostatic), and also focal infections that fail to resolve (pneumonia, osteomyelitis), reflecting irregular cycles of release into and clearance from the circulation of organisms infecting tissue Pathophysiology of Bacteremia and Fungemia • Continuous bacteremia or fungemia is a cardinal feature of endocarditis and other types of endovascular infections (suppurative thrombophlebitis, infected aneurysms), reflecting continuous shedding of organims from endovascular foci into the circulation • Continuous bacteremia also occurs early (initial few weeks) in typhoid fever and brucellosis Microbiology of Bacteremia and Fungemia Etiological Significance of Gram-Negative Bacteria Escherichia coli 142/143 (99.3%) Klebsiella pneumoniae 65/65 (100%) Enterobacter cloacae 25/25 (100%) Serratia marcescens 22/22 (100%) Proteus mirabilis 16/16 (100%) Other Enterobacteriaceae 41/45 (91%) Pseudomonas aeruginosa 53/55 (96%) Acinetobacter baumannii 13/16 (81%) Stenotrophomonas maltophilia 5/7 (71%) Weinstein et al, 1997 Microbiology of Bacteremia and Fungemia Etiological Significance of Gram-Positive Bacteria Staphylococcus aureus 178/204 (87%) Coagulase Θ Staphylococcus 87/703 (12%) Streptococcus pneumoniae 34/34 (100%) Group A Streptococcus 3/3 (100%) Group B Streptococcus 10/15 (67%) Enterococcus 65/93 (70%) Viridans streptococci 27/71 (38%) Bacillus 1/12 (8%) Corynebacterium 1/53 (2%) Lactobacillus 6/11 (55%) Weinstein et al., 1997 Microbiology of Bacteremia and Fungemia Etiological Significance of Anaerobic Bacteria Bacteroides fragilis group 16/18 (89%) Other gram-negative’s 2/5 (40%) Clostridium perfringens 3/13 (23%) Clostridium species 12/15 (80%) Propionibacterium 0/48 (0%) Other gram-positive’s 4/7 (57%) Weinstein et al., 1997 Microbiology of Bacteremia and Fungemia Etiological Significance of Yeast Candida albicans 27/30 (90%) Candida glabrata 14/15 (93%) Other Candida species 15/15 (100%) Cryptococcus neoformans 8/8 (100%) Weinstein et al., 1997 Microbiology of Bacteremia and Fungemia Blood Culture Contaminants1 • Propionibacterium (0%) • Corynebacterium (2%) • Bacillus (8%) • Coagulase-negative Staphylococcus (12%) • Viridans streptococci2 (38%) 1Inorder of decreasing probability of being a contaminant; % of blood isolates considered etiological in parentheses (Weinstein et al., 1997) 2Excluding Streptococcus pneumoniae Microbiology of Bacteremia and Fungemia Association of Organisms with Neutropenia1 <1,000 >1,000 S. aureus 9 (6%) 169 (21%) Yeasts 17 (12%) 48 (6%) Polymicrobial 24 (30%) 55 (13%) 1No.(%) of blood culture isolates from patients with indicated neutrophil count. Only differences statistically significant (p<.05) by χ2 are included (Weinstein et al., 1997). Microbiology of Bacteremia and Fungemia Association of Organisms with Shock1 Hypotensive Normotensive Polymicrobial2 16 (14%) 63 (9%) 1No. (%) of blood culture isolates from hypotensive and normotensive patients (Weinstein et al., 1977). 2p<.05 (χ2). Significant differences not observed with individual organisms, nor for unimicrobial bacteremia and fungemia. Microbiology of Bacteremia and Fungemia Sources of Bacteremia and Fungemia No. of episodes % of total episodes1 Unknown 216 26% Vascular Catheter 161 19% Genitourinary Tract 147 18% Respiratory Tract 104 12% Abdominal2 102 12% Other3 113 13% 1Total no. of episodes = 843 (Weinstein et al., 1997) in which source confirmed by culture and/or clinical evidence 2Bowel and peritoneum, biliary tract, intra-abdominal abscess 3Skin, bone and joint, surgical wound, and other Microbiology of Bacteremia and Fungemia Sources of Gram-Positive Bacteremia1 Source & No. (%) of Episodes2 S. aureus (159) IV=56 (35%) Skin=19 (12%) Respiratory=18 (11%) Bone/joint=10 (6%) Coagulase Θ Staph (73) IV=59 (81%) Skin=5 (7%) Strep. pneumoniae (34) Respiratory=26 (76%) Enterococcus (38) GU=15 (39%) IV=3 (8%) 1Numbers in parentheses indicate bacteremic episodes for each organism. Weinstein et al., 1997 2(%)=% of the indicated source for the organism Microbiology of Bacteremia and Fungemia Sources of Bacteremia due to Enterobacteriaceae1 Source & No. (%) of Episodes2 E. coli (116) GU=67 (58%) Biliary=11 (9%) Peritoneal=10 (9%) K. pneumoniae (48) Biliary=10 (21%) GU=8 (17%) Peritoneal=5 (10%) S. marcescens (20) GU=4 (20%) Respiratory=4 (20%) IV=3 (15%) Prot. mirabilis (13) GU=9 (69%) 1Numbers in parentheses indicate bacteremic episodes for each organism. Weinstein et al., 1997 2(%)=% of the indicated source for the organism Microbiology of Bacteremia and Fungemia Sources due to Gram-Negative Non-Fermenters and Yeast1 Source & No. (%) of Episodes2 P. aeruginosa (48) Respiratory=19 (40%) GU=9 (19%) A. baumannii (12) Respiratory=3 (25%) C. albicans (21) IV=5 (24%) Peritoneal=2 (10%) C. glabrata (12) GU=4 (33%) 1Numbers in parentheses indicate bacteremic or fungemic episodes for each organism. Weinstein et al., 1997 2(%)=% of the indicated source for the organism Microbiology of Bacteremia and Fungemia No. (%) deaths/ Relative risk No. episodes of death Yeast 19/53 (36%) 6.54 Polymicrobial 27/79 (34%) 2.16 Enterobacteriaceae2 31/125 (25%) 4.53 S. pneumoniae 6/34 (18%) 3.22 P. aeruginosa 8/48 (17%) 3.04 Unimicrobial 120/764 (16%) 1.00 Enterococccus 5/38 (13%) 2.40 E. coli 14/116 (12%) 2.20 Staph. aureus 19/159 (12%) 2.18 CoagΘ Staph. 4/73 (6%) 1.00 1Associated mortality for individual organisms in unimicrobial bacteremia or fungemia, and unimicrobial vs. polymicrobial bacteremia. Weinstein et al, 1997 2Enterobacteriaceae other than E. coli Microbiology of Bacteremia and Fungemia No. (%) deaths/ No. episodes1 Hypotensive Normotensive Staph. aureus 5/18 (28%) 14/141 (10%) E. coli 4/13 (31%) 10/103 (10%) Enterobac. (other) 12/23 (52%) 19/104 (18%) P. aeruginosa 3/6 (50%) 5/42 (12%) Yeast 5/7 (71%) 14/46 (30%) Unimicrobial 35/96 (36%) 85/668 (13%) Polymicrobial 10/16 (63%) 17/63 (27%) 1Onlydifferences statistically significant (p<.05) by χ2 are included (Weinstein et al., 1997). No significant differences in mortality were observed for coagulase-negative staphylococci, S. pneumoniae, other streptococci, Enterococcus, other gram- negative non-fermenters, and anaerobic bacteria. Determinants of Mortality in Bacteremia and Fungemia Multivariate Relative Risk of Death (<.05, Weinstein et al, 1997) Respiratory tract, 2.86 bowel, peritoneum, or unknown source Inappropriate antibiotic 2.72 treatment Hypotension 2.29 Enterobacteriaceae and 2.27 yeast Absence of fever 2.04 Malignancy, AIDS, or 1.98 renal failure Age (>70 years) 1.80 Clinical Indicators of True Bacteremia or Fungemia1 • Peripheral leukocyte count >20,000 or <4,000 • Neutropenia (neutrophil count < 1,000) • Hypotension • Hypothermia (<36oC) or hyperthermia (>40oC) 1At the time of first positive blood culture (Weinstein et al., 1997) Clinical Manifestations of Bacteremia Significant Independent Multivariate Predictors of Positive Blood Culture Results1 Variable OR (95% CI) P2 Temp >37.8oC 2.42 (1.41-4.14) .001 WBC >12,000 2.40 (1.41-4.10) .001 Hospital >10 d 2.02 (1.25-3.24) .004 Age >30 y 2.07 (1.19-3.60) .010 Heart rate >90 1.90 (1.13-3.17) .015 Central venous lines 1.89 (1.02-3.50) .043 1Jaimes et al, 2004; n=89 patients with positive blood cultures, and n=411 patients with negative cultures 2Likelihood ratio statistic Clinical Manifestations of Bacteremia Insignificant Predictors of Positive Blood Culture Results1 Variable Negative Positive P2 Comorbities2 159 (38.7) 32 (36) .630 Chills 43 (10.4) 9 (10) .769 Antibiotic use 210 (51.1) 45 (50.6) .785 1No. (%) of patients with negative blood cultures (n=411) and positive blood cultures (n=89) subjected to univariate analysis for each variable by Student’s t test (Jaimes et al., 2004) 2HIV infection, chronic renal failure, diabetes mellitus, immunosupressive chemotherapy, systemic cortico- steroid, and malignant disease Technical Variables in the Collection of Blood Cultures • Skin antisepsis • Volume of blood • Ratio of blood to broth • Number of blood cultures • Sites of blood collection • Blood culture sets Skin antisepsis • Preparation of skin with an agent bactericidal for surface bacterial commensals • Commensals (especially coagulase-negative staphylococci) residing deep within sebaceous glands evade the bactericidal action of skin preparation agents • Preparation agents include povidone-iodine (skin contact killing time of 1.5-2 min), tincture of iodine (contact killing time of 0.5 min), and recently chlorhexidine (ChloraPrep) (NMH: application to skin venepuncture site for 30 sec by back and forth friction scrub, followed by 30 sec drying time) • Blood culture contamination rate reflects effectiveness of antisepsis with lowest rates obtained by laboratory phlebotomy teams Volume of blood • Bloodstream infections frequently caused by relative few organisms in a given volume of blood (<1- 10 colony forming units/mL of blood) • Sensitivity of blood cultures thus directly proportional to the volume of blood cultured • Optimal volume for adults: 20-30 mL of blood per culture set • Blood volumes >30 mL do not enhance the sensitivity of blood cultures for adults and contribute to nosocomial anemia • Optimal volume for children (birth-15 y): 4 to 4.5% of patient’s total blood volume (Kellogg et al., JCM 38:2181-2185, 2000) Ratio of blood to broth • Balance between dilutional effect of broth on antibiotics, complement, lysozyme, and phagocytic white cells on the one hand and inadequate volume of blood on the other • Optimal dilution of blood to broth is 1:5 to 1:10 (v/v) Number of blood cultures • Collection of multiple blood culture sets at different venipuncture sites • Contamination by skin organism or environmental spores indicated if only one of multiple sets positive for coagulase- negative Staphylococcus, Bacillus, Corynebacterium, viridans streptococci, or Propionibacterium. • Acute febrile episode: 2 sets from separate sites within 10 minutes before antimicrobial • Nonacute disease: 2 or 3 sets from separate sites at > 3 hr intervals within 24 hr before antimicrobial • Acute endocarditis: 3 sets from separate sites within 1-2 hr before antimicrobial • Subacute endocarditis: 3 sets from separate sites at >1 hr intervals within 24 hr; if these sets culture negative obtain 2-3 additional sets • Fever of unknown origin: 2 or 3 sets from separate sites at >1 hr intervals within 24 hr; if these sets culture negative, obtain 2- 3 additional sets Optimal Testing Parameters for Blood Cultures1 • 37,568 blood cultures tested with automated BACTEC 9240 instrument at Mayo Medical Center June 1996-October 1997 • 20 mL blood inoculated in equal volumes to two culture bottles, first to BACTEC Plus Aerobic/F resin bottle, and then to BACTEC Lytic/10 Anaerobic bottle • 20 mL blood obtained separately for a total of 40 mL within 30 min • Blood cultures incubated 7 days on BACTEC 9240 instrument 1Cockerill, III et al., Clin Inf Dis 38:1724-1730, 2004 Optimal Testing Parameters for Blood Cultures1 No. pathogens recovered 10 mL 20 mL 30 mL 40 mL S2 235 3053 3464 3715 EC2 13 14 14 14 1Cockerill, III et al., Clin Inf Dis 38:1724-1730, 2004 2S = sepsis without endocarditis, E = endocarditis 320 mL vs. 10 mL = 29.8% increase in yield 430 mL vs. 20 mL = 13.4% increase in yield 540 mL vs. 30 mL = 7.2 % increase in yield Optimal Testing Parameters for Blood Cultures1,2 Cons 1st + 1st+ 1st + 1st + Cult3 >1 >2 >3 >4 1 615(81) 497(77) 106(65) 62(61) 2 116(15) 116(18) 25(15) 17(17) 3 25(3) 25(4) 25(15) 15(15) 4 7(1) 7(1) 7(5) 7(7) 5 - - - - 6 - - - - 1Cockerill, III et al., Clin Inf Dis 38:1724-1730, 2004. 2Pathogen recovery without endocarditis (n=763); 1st + bottle indicated for patients with 1 or more, 2 or more, 3 or more, and 4 or more (up to 6) collected. 3Consecutive blood culture drawn; 1 culture (n=118), 2 cultures (n=482), 3 cultures (n=62), 4 cultures (n=98), 5 cultures (n=0), and 6 cultures (n=3) drawn for the indicated number of patients. Optimal Testing Parameters for Blood Cultures1 1st + 1st+ 1st + 1st + >1 >2 >3 >4 1 37(93) 31(91) 16(89) 9(82) 2 2(5) 2(6) 1(6) 1(9) 3 0 0 0 0 4 1(3) 1(3) 1(6) 1(9) 5 - - - - 6 - - - - 1Cockerill,III et al., Clin Inf Dis 38:1724-1730, 2004 2Pathogen recovery with endocarditis (n=40) and consecutive blood cultures (1-6) over 24 hours. 1st+ bottle indicated for patients with 1, 2, 3, or 4 or more bottles (up to 6) collected. Optimal Testing Parameters for Blood Cultures1 • Twenty mL of blood inoculated in equal volumes to an aerobic and anaerobic broth bottle (one blood culture set) • Second blood culture set inoculated immediately after the first from a different venipuncture site • Two additional blood culture sets inoculated over remaining 24 hr if sepsis persist 1Cockerill, III et al., Clin Inf Dis 38:1724-1730, 2004 Sites of blood collection • Venipuncture at separate skin sites method of choice • Blood should not be obtained from an indwelling intravenous or intra-arterial catheter unless catheter-related infection suspected, or skin venipuncture sites not available • For the evaluation of catheter-associated bloodstream infection, a concomitantly drawn venipuncture specimen should be paired with a catheter specimen Blood culture sets • Two aerobic bottles versus a pair of aerobic and anaerobic bottles (?) (Decreased frequency of anaerobic bacteremia) • Three bottle versus two bottle sets (?) (Increased sensitivity of automated continuously monitoring blood culture instruments) Incubation conditions Without EC With EC 1d 2,052 (76.5) 144 (77.8) 2d 393 (91.2) 26 (91.9) 3d 123 (95.8) 8 (96.2) 4d 61 (98.1) 1 (96.8) 5d 35 (99.4) 2 (97.8) 6d 14 (99.9) 4 (100) 7d 3 (100) ------------- 1Positivity of blood cultures for pathogens by incubation day (BACTEC 9240), EC=endocarditis 2Cockerill, III et al., Clin Inf Dis 38:1724-1730, 2004 Blood Culture Systems Manual • Septi-Chek (Becton Dickinson) • Isolator (Wampole) Semi-Automated • BACTEC 460 (Becton Dickinson) Automated • BacT/ALERT (bioMérieux) • BACTEC 9000 System (Becton Dickinson) • VersaTREK (formerly ESP System) (Trek) Septi-Chek (Becton Dickinson) • Aerobic broth bottle with attached plastic paddles containing agar medium (chocolate, MacConkey, malt) • Inoculated broth bottle initially inverted to allow blood-broth mixture to flood the agars, and then each time bottle is inspected for growth • Paddles are visually examined once or twice daily for colony formation, and broth for evidence of microbial growth (hemolysis, turbidity, gas production, chocolatization of blood, visible colonies or layer of growth on fluid meniscus) • Separate anaerobic bottle without agar paddles • Not practical for larger laboratories due to manual steps required for monitoring and processing Isolator (Wampole) • Isolator tube contains a blood cell lysing solution consisting of saponin and a fluorocarbon cushion that captures organisms during centrifugation • Following centrifugation supernatants are discarded, and pellets resuspended for inoculation to solid medium appropriate for type of culture being performed (sheep blood, chocolate, MAC, CNA, BCYE, IMA, Middlebrook agar) • Highly versatile and sensitive (except for anaerobic bacteria), but labor intensive for routine work • Excellent system for isolation from blood of Mycobacterium avium or M. tuberculosis complex, dimorphic fungi, Bartonella, Legionella, and other fastidious pathogens BACTEC 460 (Becton Dickinson) • Blood culture broth medium (aerobic and anaerobic) contains 14C labeled carbohydrate substrate • With microbial growth carbohydrate substrate metabolized and 14C labeled CO gas released into head space of bottles 2 • Needles perforate rubber septum of each bottle and headspace gas withdrawn for radiometric measurement of 14C labeled CO2 • Bottles reaching threshold 14C levels flagged as positive for Gram’s stain and subculture of broth to solid medium • A semi-automated system that requires placement of racks with bottles on a shuttle of the BACTEC 460 radiometer for testing • Replaced in routine work by automated blood culture systems • Remains an excellent system for recovery of mycobacteria from a wide variety of specimens, including sputum, where broth contains 14C labeled palmitic acid as substrate for cell wall mycolic acid synthesis, with release of 14C labeled CO2 into headspace of bottles containing Middlebrook broth BacT/ALERT (bioMérieux) • The first continuous-monitoring blood culture system (CMBCS), introduced in early 1990’s • Broth bottles monitored continuously (every 10-15 min) for growth in self-contained modular units and bottles require no manipulation until flagged as positive for growth • Growth monitored by a colorimetric CO2 sensor at the base of each bottle • Computer algorithms interpret CO2 production as microbial growth when arbitrary thresholds exceeded, minimal linear increases of CO2 occur, or there is change in the rate of CO2 production BACTEC 9000 System (Becton Dickinson) • Three instruments in the BACTEC 9000 CMBCS series: 9050 system (monitors 50 bottles), 9120 system (120 bottles), and 9240 system (240 bottles) • Growth monitored by a fluorescent CO2 sensor at the base of each broth bottle • Computer algorithms interpret CO2 production as microbial growth by linear increases in fluorescence, and an increase in the rate of flouorescence VersaTREK (ESP System) (Trek) • Third CMBCS introduced commercially • Differs from BacT/ALERT and BACTEC 9000 series in that direct measurement of CO2 production not utilized to monitor microbial growth • Bottles fitted with pressure transducers to monitor gas pressure changes within bottle headspaces • Computer algorithms interpret the consumption and/or production of gas as microbial growth in which pressure changes are plotted against time to yield growth curves • Positive cultures are signaled in accordance with these proprietary algorithms Characteristics of CMBCS’s • Microbial growth detected 1-2 days earlier than with manual systems (total incubation time of 5 days) • Provides 24/7 service with prompt reporting of Gram’s stain results for positive blood cultures • Decreases laboratory workload Quality management monitors • Contamination rate (should not exceed 3% of blood culture sets accessioned) • Identification and susceptibility testing of coagulase-negative stapylococci (CNS) (should not exceed 12-30% of blood cultures sets positive for CNS) • Blood volume (determined by broth bottles and blood culture systems utilized by the laboratory) Probability of Same Contaminant in Two Blood Culture Sets • 0.03 X 0.03 = 0.0009 (<1 per 1,000 sets) 36 episodes/40,000 culture sets • 0.05 X 0.05 = 0.0025 (2-3 per 1,000 sets) 100 episodes/40,000 culture sets • 0.10 X 0.10 = 0.01 (1 per 100 sets) 400 episodes/40,000 culture sets • 0.15 X 0.15 = 0.02 (2 per 100 sets) 800 episodes/40,000 culture sets CNS Algorithm • Single blood culture + for CNS, no other blood cultures collected +/- 48 hr: Evaluation of patient for sepsis • Single blood culture + for CNS, additional blood cultures collected +/- 48 hr negative for CNS: No ID or susceptibility unless physician request • Single blood culture + for CNS, additional blood cultures collected +/- 48 hr positive for CNS: Evaluation of patient for sepsis Richter et al. JCM 40:2437-2444, 2002. CNS Algorithm • If 2 or more blood cultures submitted and only one + for CNS, reported as likely contaminant (no species ID, no susceptibility) • If 1 of 1 blood culture submitted and is positive for CNS, reported as of indeterminate significant and physician advised to contact laboratory if ID and susceptibility needed • If 2 or more blood cultures submitted + for CNS, ID and susceptibility determined; if both isolates same species, ID and susceptibility reported; if different species, reported only as CNS without species and susceptibility Weinstein JCM 41:2275-2278, 2003. References • Weinstein et al. 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 Inf Dis 24:584-602, 1997. • Munson et al. Detection and treatment of bloodstream infection: Laboratory reporting and antimicrobial management. J Clin Micro 41:495-497, 2003. • Jaimes et al. Predicting bacteremia at the bedside. Clin Inf Dis 38:357-362, 2004. • Mohr et al. Manual and automated systems for detection and identification of microoranisms. Manual of Clinical Microbiology, Volume 1:185-191, 2003. • Reimer et al. Update on detection of bacteremia and fungemia. Clin Micro Rev 10:444-465. • Weinstein. Current blood culture methods and systems: Clinical concepts, technology, and interpretation of results. Clin Inf Dis 23:40- 46, 1996. • Dunne et al. Blood Cultures III. Cumitech 1B, 1997. References • Kellogg et al. Frequency of low-level bacteremia in children from birth to fifteen years of age. J Clin Micro 38:2181-2185. • Cockerill III et al. Optimal testing parameters for blood cultures. Clin Inf Dis 38:1724-1730. • Weinstein et al. Remarks concerning testing parameters for blood cultures. Clin Inf Dis 40:202, 2005. • Cockerill III. Reply to Weinstein and Reller. Clin Inf Dis 40:202-203, 2005. • Richter et al. Minimizing the workup of blood culture contaminants: Implementation and evaluation of a laboratory-based algorithm. J Clin Micro 40:2437-2444, 2002. • Weinstein. Blood culture contamination: Persisting problems and partial progress. J Clin Micro 41:2275-2278, 2003. • Mirrett et al. Relevance of the number of positive bottles in determining clinical significance of coagulase-negative staphylococci in blood cultures. J Clin Micro 39:3279-3281. • Bates et al. Contaminant blood cultures and resource utilization. J Am Med Assoc 265:365-369, 1991.
"Bacteremia_ Fungemia_ and Blood Cultures"