P-888. Variability in Durations of Therapy for Gram-Negative Bloodstream Infections Across U.S. Hospitals

Abstract Background There is a paucity of data on durations of antibiotic therapy utilized in clinical practice for patients with Gram-negative bloodstream infections (GN-BSI). The objectives of this study were to describe treatment durations for patients with GN-BSI and evaluate variability across U.S. hospitals. Histogram of antibiotic treatment durations Methods This was a retrospective cohort study of patients hospitalized at 24 hospitals (16 academic, 4 community, and 4 Veterans Affairs) with a GN-BSI between January 1, 2019 and December 31, 2019. Episodes of GN-BSI from a urinary, intra-abdominal, hepatobiliary, vascular catheter, skin and soft tissue, or respiratory source were included while those associated with a bone, joint, central nervous system, or endovascular infection or prostatitis were excluded. Subgroups of interest were cases of uncomplicated bacteremia and those with underlying immunosuppression. Uncomplicated bacteremia was defined as: (1) achievement of source control, when applicable; (2) absence of immunosuppression; and (3) resolution of fever and hypotension within 3 days. Variability in median duration of therapy across hospitals Results 3,956 episodes of GN-BSI were included, of which 2,007 were uncomplicated bacteremia and 1,202 involved underlying immunosuppression. Urinary tract (55%), intra-abdominal (14%), and hepatobiliary (12%) sources accounted for the majority of the infections, while E. coli (53%), K. pneumoniae (12%), and P. aeruginosa (8%) were the most common pathogens. The median duration of therapy in the overall cohort and uncomplicated and immunocompromised subgroups was 12 (interquartile range [IQR] 9 - 15) days, 11 (IQR 8 - 14) days, and 11 (9 - 14) days, respectively (Figure 1). In the overall cohort, there was marked variation in treatment durations across hospitals, with the median duration as low as 5 (IQR 3 - 7) days and as high as 14 (IQR 10 - 17) days (Figure 2). Similar variability was observed in the uncomplicated bacteremia and immunocompromised subgroups. Conclusion There is marked variation in treatment durations for GN-BSI across U.S. hospitals. Dissemination of interventions to promote evidence-based short durations of therapy for uncomplicated bacteremia and elucidating the optimal duration of therapy in immunocompromised patients will be necessary to standardize care and reduce unnecessary antibiotic use. Disclosures Sara M. Karaba, MD, PhD, MHS, Entasis: Advisor/Consultant Kelly E. Pillinger, PharmD, AHFS: Contractor Judianne C. Slish, Pharm.D., BCPS, Glaxosmithkline Pharmaceuticals: Honoraria