Impact of a Rapid Diagnostic for Bloodstream Infections with Antimicrobial Stewardship Intervention at a Comprehensive Cancer Center

Abstract Background Molecular based assays reduce time to organism identification for bloodstream infections (BSI) but have limited impact without Antimicrobial Stewardship (AS) intervention. The benefit of pairing molecular based assays with AS in immunocompromised hosts is unknown. Immunocompromised patients present unique challenges to AS efforts and evaluations of traditional AS interventions are needed in this patient population. The purpose of this analysis was to evaluate the utility of a molecular based assay for BSI with and without AS intervention at a cancer hospital. Methods A retrospective quasi-experimental pre-post study was performed to evaluate the impact of the FilmArray© Blood Culture Identification (BCID) panel with and without AS intervention on time to appropriate antimicrobial therapy defined as de-escalation to the narrowest spectrum agent taking into account need to cover concomitant infections and antibiotic allergies or intolerances. We included inpatients with positive blood cultures between 2014 and 2016 in three separate 100-day cohorts: prior to BCID implementation (pre); after BCID implementation without AS intervention (post); after BCID implementation with AS intervention (ASP) involving blood culture review and antimicrobial treatment recommendations. Results 130 of 155 subjects with a BSI during the study period were included. The pre (n = 52), post (n = 43), and ASP (n = 35) cohorts were balanced with the exception of more immunocompromised patients in the ASP compared with pre (91% vs 65%; P < 0.01) and post cohorts (91% vs 72%; P = 0.04). Time to appropriate antimicrobial therapy, although not statistically different, was shorter in the post and ASP groups as compared with the pre-BCID group [40.2 hours (pre) vs 24.6 hours (post) vs 25.9 hours (ASP); P = 0.46]. Conclusion Implementation of the BCID in a cancer hospital was associated with reduced time to appropriate antimicrobial therapy; however, additional reductions were not seen when coupled with AS intervention. Further large-scale evaluation is warranted due to unbalanced study groups and small study size to understand the role of rapid diagnostics and AS interventions for BSIs in immunocompromised populations. Disclosures R. Benefield, Merck: Grant Investigator, Research grant