2431. Evaluation of Clinical Outcomes in Bacteremia Due to AmpC β-Lactamase Producing Organisms Stratified by Treatment

Abstract

BackgroundEnterobacteriaceae and Pseudomonas aeruginosa are common bloodstream pathogens with variable AmpC β-lactamase (AmpC) incidence. The clinical utility of treatment with non-carbapenem/cefepime options remains unclear. The objective of this study was to compare the clinical outcomes for patients receiving a carbapenem or cefepime (CC) and alternative therapy (AT) for bacteremia caused by organisms known to produce AmpC.MethodsHospitalized adults with a confirmed mono-microbial bacteremia admitted from June 2016 to December 2017 were included. Patients were stratified by definitive therapy (DT) with CC or AT. The AT group was treated with fluoroquinolones, third-generation cephalosporins, piperacillin–tazobactam, aztreonam, or tobramycin. The primary outcome was in-hospital mortality. Secondary outcomes included treatment failure, microbiological failure, hospital length of stay (LOS), and intensive care unit LOS. Multiple regression analysis was used to adjust for potential confounding variables.ResultsOf 68 patients meeting eligibility criteria, 46% received CC for DT. Enterobacteriaceae were isolated in 45% of patients in the CC group. In-hospital mortality was 32% and 3% (P = 0.0017) in the CC and AT groups, respectively. Source control, APACHE II score on the date of index culture, and immune status did not differ between groups. Definitive CC therapy was independently associated with mortality (odds ratio, 15.17; 95% confidence interval, 1.69–135.76; P = 0.0150). Only 6 (9%) patients received AT as empiric and DT. Those who received definitive AT received a median of 5 days (interquartile range, 3–9 days) of CC prior to being switched to AT.ConclusionWhile most patients received empiric CC, definitive treatment with CC was found to be an independent predictor of in-hospital mortality. These findings suggest that AT may be a de-escalation treatment strategy for clinicians to consider. However, these results should be confirmed in a larger population.Disclosures All authors: No reported disclosures.