Molecular typing of coagulase-negative staphylococci from blood cultures does not correlate with clinical criteria for true bacteremia

• Published in: The American journal of medicine Vol. 109; no. 9; pp. 697 - 704
• Main Authors: Seo, Susan K, Venkataraman, Lata, DeGirolami, Paola C, Samore, Matthew H
• Format: Journal Article
• Language: English
• Published: New York, NY Elsevier Inc 15.12.2000; Elsevier; Elsevier Sequoia S.A
• Subjects: Bacteremia - diagnosis; Bacteremia - drug therapy; Bacteremia - microbiology; Bacteria; Bacterial Typing Techniques - methods; Bacteriology; Biological and medical sciences; Blood; Blood - microbiology; Coagulase - metabolism; Diagnosis, Differential; DNA Primers; DNA, Bacterial - isolation & purification; Electrophoresis, Gel, Pulsed-Field; Female; General aspects; Humans; Infection pathogenesis; Infectious diseases; Male; Medical sciences; Middle Aged; Molecular biology; Odds Ratio; Polymerase Chain Reaction; Prospective Studies; Staphylococcal Infections - diagnosis; Staphylococcal Infections - drug therapy; Staphylococcus - drug effects; Staphylococcus - enzymology; Staphylococcus - genetics; Human; Bacteremia; Staphylococcus; Incidence; Infection; Polymerase chain reaction; Symptomatology; Pulsed field electrophoresis; Bacteriosis; Bacteria; Micrococcales; Micrococcaceae; Molecular biology; Blood culture; Biological contamination
• ISSN: 0002-9343; 1555-7162
• DOI: 10.1016/S0002-9343(00)00596-9

PURPOSE: Determining whether a blood culture that contains coagulase-negative staphylococci represents bacteremia or contamination is a clinical dilemma. We compared molecular-typing results of coagulase-negative staphylococcal blood culture isolates with clinical criteria for true bacteremia. SUBJECTS AND METHODS: Pulsed-field gel electrophoresis and arbitrary primed polymerase chain reaction (PCR) were used to determine whether patients with two or more blood cultures with coagulase-negative staphylococcal isolates had the same strain of organism in each culture (same strain bacteremia). We evaluated three different clinical criteria for bacteremia: whether the patient received more than 4 days of antibiotics, whether there was an explicit note in the medical chart in which the physician diagnosed a true bacteremia, and the Centers for Disease Control surveillance criteria for primary bloodstream infection. Agreement between same-strain bacteremia and each definition was examined, based on the assumption that most true infections should be the result of a single strain. RESULTS: The study sample consisted of 42 patients and 106 isolates. Nineteen of the 42 bacteremias (45%) were the same strain. Classification of bacteremias as same–strain correlated poorly with all three clinical assessments (range of percent agreement, 50% to 57%; range of kappa statistic, 0.01 to 0.15). There were both false-positive and false-negative errors. Patients with three or more positive blood cultures were more likely to have same-strain bacteremia than those with only two positive cultures [11 of 15 (73%) vs 8 of 27 (30%), P = 0.006]. Pulsed-field gel electrophoresis was more discriminating than arbitrary primed PCR (percent agreement, 83%; kappa, 0.67). CONCLUSION: Molecular typing correlated poorly with clinical criteria for true bacteremia, suggesting either that true bacteremias are frequently the result of multiple strains or that the commonly used clinical criteria are not accurate for distinguishing contamination from true bacteremia. Vancomycin treatment of clinically defined coagulase-negative staphylococcal bacteremia may frequently be unnecessary.