1484. The Utility of "Culture if Spikes"
Session: Poster Abstract Session: Antimicrobial Stewardship: Role of Diagnostics
Saturday, October 10, 2015
Room: Poster Hall
  • Utility of Blood Cultures.pdf (461.4 kB)
  • Background: False positive (FP) blood cultures (BCX) result in both unnecessary antibiotic use and increased hospital costs.  Optimization of true positive (TP) BCX has been studied in ER and ICU populations; however in acute care patients on a Medical service, little is known about drivers of BCX orders and how to improve the yield of TP. We sought to identify the yield and self-reported indication for BCX ordered in this population.

    Methods: All BCX orders written by inpatient providers at a large VA teaching hospital from 10/1/14-4/15/15 were reviewed for indication and positivity on a daily basis. As part of the order, providers selected from among a list of indications. Classification of BCX into TP and FP was performed as part of independent ID review. A BCX order was defined as an electronic entry including all BCX sets drawn as a result of that order. Consistent with previous literature, a BCX episode was defined as all BCX within a 48 hour period. A chart review assessing for clinical predictors of TP was performed on all patients on the Medical service.

    Results: 999 total orders were placed with 785 total episodes. The TP rate was 3.1% per order and 3.3% per episode. The FP rate was 1.8% per order and 2.3% per episode.   Indication review was limited to 545 patients on the Medical service. The most common indications were fever (48.0%) and leukocytosis (22.4%), neither of which alone was predictive of TP BCX (LR+ 0.6, 95% CI 0.2-1.7 & LR+ 1.3, 95% CI .3-4.8, respectively). The only indication significantly associated with a TP BCX was “follow-up previous positive” (11.7%, LR+ 3.3, 95% CI 1.7-6.3) although fever plus leukocytosis as an indication approached significance (8.3%, LR+ 2.3, 95% CI 0.9-5.8). The only clinical predictor of TP BCX was absence of antibiotic exposure within 72 hours of culture (7.8%, LR+ 2.3, CI 1.8-2.9).

    Conclusion: The rate of TP BCX among patients on a Medical service was notably lower than the expected rate of 10% from previous studies of primarily ER patients.  This is the first real time capture of clinical thought processes triggering BCX orders and suggests that reflex ordering of BCX based on fever or leukocytosis alone may not yield accurate information regarding bacteremia. In patients on antibiotics, the likelihood of TP is so low that the utility is questionable.  Improving the yield of TP BCX may not only contain cost but also reduce unnecessary antibiotic use.

    Katherine Linsenmeyer, MD1,2, Kalpana Gupta, MD, MPH1,2, Judith Strymish, MD1,3, Muhammad Dhanani, MD2 and Anthony Breu, MD1, (1)VA Boston HCS, West Roxbury, MA, (2)Department of Medicine/Boston University School of Medicine, Boston, MA, (3)Harvard Medical School, Boston, MA


    K. Linsenmeyer, None

    K. Gupta, None

    J. Strymish, None

    M. Dhanani, None

    A. Breu, None

    Previous Abstract | Next Abstract >>

    Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 7th with the exception of research findings presented at the IDWeek press conferences.