Rapid identification of pathogens from positive blood cultures may allow for more rapid optimization of antimicrobial therapy. In 2013 our children’s hospital implemented the AvanDx (PNA-FISH) test for blood cultures staining for Gram positive organisms. In 2014, we implemented the BioFire Diagnostics FilmArray blood culture identification (BCID) panel, which can identify 15 species-specific and 6 genus-level targets in an hour from aerobic positive blood cultures. Our objective was to evaluate the impact of the BCID panel on time to identification (TTI) of organisms and antibiotic use when compared to PNA-FISH and no rapid diagnostic method (NRDM).
Our study was divided into 3 consecutive 1 year time periods: NRDM, PNA-FISH, and BCID. Positive blood cultures were identified through a search of the Intermountain Enterprise Data Warehouse and associated with a patient and an episode of clinical care. The TTI, antibiotic changes made in the first 24 hrs after a positive blood culture result, and duration of select antibiotics were analyzed. Time periods were compared to determine significant differences.
1307 positive blood cultures were analyzed. Median organism TTI was 17, 26.1, and 29.3 hrs in the BCID, PNA-FISH, and NRDM time periods, respectively. BCID was run on 71% of positive blood cultures and the median TTI was 4.5 hrs. BCID TTI was significantly faster than both NRDM (p<0.001) and PNA-FISH (p=0.01). TTI of coagulase negative staphylococci and Staphylococcus aureus (SA) were significantly faster with both PNA-FISH (p<0.001, p<0.001 respectively) and BCID (p<0.001, p<0.001 respectively). TTI was faster for Candida spp. (p<0.001) and Gram negative organisms (p<0.001) when comparing BCID to NRDM. There was no difference in the number of antibiotic changes in the first 24 hours after a positive blood culture result or in the duration of vancomycin, antipseudomonal agents, or third generation cephalosporin use across time periods.
The use of rapid diagnostic methods for positive blood cultures reduced time to identification of organisms but had limited overall impact on the duration of antibiotics. Further studies may show an impact on specific antibiotic/organism combinations.
E. Thorell, None
A. L. Hersh, Pfizer: Grant Investigator , Grant recipient and Research grant
C. R. Stockmann, None
K. Korgenski, None
A. J. Blaschke, BioFire Diagnostics, LLC: Collaborator , Consultant and Scientific Advisor , Consulting fee , Licensing agreement or royalty and Research support
bioMerieux, Inc: Collaborator , Investigator and Scientific Advisor , Consulting fee and Research support
Merck: Investigator , Research grant