Reducing blood culture contamination in a pediatric emergency department.

BACKGROUND: Blood cultures (BCs) are used to diagnose bacteremia in febrile children. False-positive BCs increase costs because of further testing, longer hospital stays, and unnecessary antibiotic therapy. Data from a study at our hospital showed the emergency department consistently exceeded established guidelines of 2% to 4%. A phlebotomy policy change was made whereby BC had to be obtained by a second venipuncture and no longer obtained during insertion of intravenous catheters.
METHODS: A descriptive study compared preintervention and postintervention blood culture contamination (BCC) rates. A BC was considered contaminated if a single culture grew coagulase-negative staphylococci, diphtheroids, Micrococcus spp, Bacillus spp, or viridans group streptococci. Patients with indwelling central lines or who grew pathogenic bacteria were excluded.
RESULTS: Preintervention BCC was 120 (6.7% [SD, 2.3%]) of 1796. Postintervention BCC was 29 (2.3%, [SD, 0.8]) of 1229 with odds ratio of 2.96 (confidence interval, 1.96-4.57; P = 0.001). The most common contaminant was coagulase-negative staphylococcus, 21 (72%) of 120, followed by viridans streptococcus, 3 (10%) of 29, which was not significantly different between intervention periods. Before intervention, 44 patients were called back to the emergency department, and 25 were admitted because of BCC. After intervention, a total of 9 patients were called back, and 5 were admitted. The decrease in unnecessary hospitalization was statistically significant (P < 0.05).
CONCLUSIONS: The new policy significantly reduced BCC rates, thereby decreasing unnecessary testing and hospitalizations. Coagulase-negative staphylococci and viridans streptococci remain the most common BC contaminants. Further research should focus on additional interventions to reduce BCC.