Impact of BioFire FilmArray blood culture identification panel 2 (BCID2) and antimicrobial stewardship interventions on time to optimal antimicrobial therapy in patients with positive blood cultures

Abstract

Background: Delayed treatment of bloodstream infections (BSI) is associated with increased morbidity and mortality. Conventional methods for organism identification and susceptibility data from blood cultures can take on average 2-5 days, with pathogen identification taking longer. Technological advancements in gene-based polymerase chain reaction (PCR) tests amplify DNA targets from positive blood cultures which can shorten the identification time of certain organisms and resistance genes, aiding in optimization of antimicrobial therapy. Our goal is to evaluate the impact of the Biofire FilmArray Blood Culture Identification Panel 2 (BCID2) combined with real-time ASP intervention on time to optimal antimicrobial therapy and clinical outcomes.

Methods: This study has a letter of determination from the Oklahoma State University (OSU) Center for Health Sciences IRB as a multidisciplinary quality improvement initiative. This pre/post quasiexperimental study conducted at OSU Medical Center in Tulsa, Oklahoma, will include adult inpatients with at least 1 positive blood cultures from a 6-month period prior to implementation of the BCID2 panel and for 6 months post-implementation. Patients less than 18 years of age, those with identical positive blood cultures within previous 7 days, positive blood cultures at an outlying facility, transitioned to CMO or died within 24 hours of positive blood cultures and patients with BSI with an organism not included on the BCID2 panel were excluded. Demographic information will be collected including comorbid health conditions, risk factors for BSI, infection source, and clinical status. The primary endpoints include time to effective therapy and time to optimal therapy. Secondary outcomes include: 30-day all-cause mortality, hospital length of stay, and microbiologic clearance and 30-day readmission with bacteremia.

Results: The pre-intervention group has a final included patient population of 125 for demographic, primary, and secondary endpoints, while the post-intervention group has 89 included for demographic and primary endpoints and 62 included in secondary outcomes with final number pending. Primary endpoint data between the two groups shows a 22-hour difference in average time to pathogen ID (51 hours (pre) vs. 29 hours (post)), a 1.6-hour difference in average time to effective therapy (17.4 hours (pre) vs. 15.8 hours (post)), and a 15.1-hour difference in average time to optimal therapy (61.4 hours (pre) vs. 45.3 hours (post)). Secondary endpoint data has shown decreased time to microbiologic cure (55.6 hours (pre) vs. 49.1 hours (post)) and decreased length of stay in the intensive care unit (ICU) (3.9 days (pre) vs. 2.4 days (post)) and in the hospital total (10.8 days (pre) vs. 9 (post)). Adverse event data has shown similar rates of 30-day mortality (11.2% (pre) vs. 12.9% (post)), but a possible increase in 30- day readmission with bacteremia (0.8% (pre) vs. 6.5% (post)).

Conclusions: Based on preliminary data the implementation of the BCID2 panel has decreased time to pathogen identification and optimal therapy, but has similar time to effective therapy. Data also shows a 6.5-hour difference in microbiologic clearance time with over 1.5 and 1.8-day difference in ICU and total length of stay. Further investigation into statistical significance of final data set needs to be completed.