P-1677. Assessing An Institution’s Approach to Management of Gram-Negative Bacteremia in High-Risk Populations

Abstract

Listen

Translate article

Abstract Background Immunocompromised populations present unique antimicrobial management and stewardship challenges. These challenges can persist despite existing stewardship policies. This study examined management of gram-negative bacteremia (GN-BSI) in a diverse immunocompromised population at the National Institutes of Health Clinical Center (NIH CC). Methods Episodes of GN-BSI admitted to the NIHCC from January 1 to December 31, 2023 were identified through positive blood culture results. Charts were abstracted for baseline demographics, episode-specific data, antimicrobial prescribing, and clinical outcomes. Results Seventy-five episodes of GN-BSI were identified. Thirty-six (48%) were undergoing cellular therapy and 54 (58.7%) were neutropenic. About 50.7% identified translocation as etiology of infection. Eleven (14.7%) were ESBL-positive by BCID2 panel. Seventy-two (96%) involved infectious disease (ID) consultation. Empiric choice of gram-negative antibiotics included ceftazidime/avibactam in 9 (12%) episodes, meropenem in 38 (50.7%) episodes, piperacillin/tazobactam in 15 (20%) episodes, and an anti-pseudomonal cephalosporin in 32 (42.7%). Twenty (26.7%) received 2+ empiric anti-pseudomonal antibiotics. After 72 hours, 37 (49.3%) still received meropenem with most isolates susceptible to narrower spectrum (Table 1). Similar results were seen with final ID recommendations (Table 2). Seven (10%) had transitioned to an oral agent by the seventh day. Comparing 72-hour antibiotic (Meropenem vs anti-pseudomonal cephalosporin), ICU admission (32.4% v 25%, p=0.62) and 30-day mortality (5.4% v 6.3%, p=0.87) were similar. Median therapy duration was 14 days (IQR 14-18.5). Of the 65 episodes with indwelling central venous access (CVAD), only 16 had removal of all CVADs. Those with CVAD removal had lower 30-d mortality (0% v 11.9%, p=0.22) though not statistically significant. Conclusion In a high-risk immunocompromised population, there is a low rate of narrowing gram-negative coverage at 72 hours although options for narrowing exist. Choice of empiric antimicrobial and targeted therapy for these GN-BSI episodes are the two major areas of improvement in our population. Primary teams and ID consultants should be involved in ongoing stewardship interventions. Disclosures All Authors: No reported disclosures