Abstract
Background
Coagulase-negative staphylococci (CoNS) are common blood culture (BCx) contaminants, but can also be causes of true blood stream infection (BSI). As a result, the clinical interpretation of CoNS positive BCx poses a significant challenge for providers and drives unnecessary antibiotic use, extended lengths of stay, and increased hospital costs. Despite these challenges, little is known about whether the number of positive BCx bottles within a set can be used to predict contamination vs. true BSI.
Methods
This study was conducted in an 865-bed tertiary care academic medical center in Richmond, VA. A retrospective chart review of CoNS-positive BCx from October to December 2018 was performed. Data collection included patient demographics, number of positive bottles within a set (i.e., were 1 or 2 bottles positive), care setting, antibiotic use, clinical judgement of contamination, and additional workup following the positive BCx result. Polymicrobic BCx were excluded.
Results
50 patients (mean age 58.2 years, 60% male) with CoNS-positive BCx were included in this study. Forty (80%) of the cultures had only 1 of 2 BCx bottles positive within a set. 10 (20%) were positive from both bottles in the set. All patients were drawn in the Emergency Department and 90% were subsequently admitted to the hospital. Upon chart review, 47 (94%) and 3 (6%) of cultures were considered to be contaminants and real BSI, respectively. Of those judged to be contaminants, 10 (20%) were positive in both bottles within a set, and thus falsely suggested true BSI. Of the 3 judged to be true BSI, 2 (66%) were positive in 1 out of 2 bottles, and thus falsely suggested contamination.
42 (84%) patients had repeat BCx drawn following the initial positive culture, and 26 (52%) were continued on IV antibiotics. Forty (80%) of the cultures were judged contaminants by the primary medical service, and 77% stopped antibiotics (20/26) when CoNS was identified.
Conclusion
These data show that reporting the number of bottles which are positive within a set provides misleading information and should not be used to determine whether a culture result represents contamination or true BSI.
Disclosures
All authors: No reported disclosures.
Session: 37. Bacteremia, CLABSI, and Endovascular Infections
Thursday, October 3, 2019: 12:15 PM