207. Ineffective Initial Antibiotics for Enteric Bacteremia: Does Every Hour Count with Rapid Diagnostics and Stewardship Intervention?

Abstract

Abstract Background It has been a long-standing practice to administer broad-spectrum antibiotics early for sepsis as each hour delay is associated with increase in mortality. With increasing rates of antibiotic resistance fueled by unnecessary use of antibiotics, it is delicate to balance the benefits vs consequences of empiric carbapenem therapy. With rapid molecular blood culture diagnostics available, identification of extended-spectrum beta-lactamase (ESBL) producing bacteremia can occur within hours and therapy optimized with active stewardship intervention. With rapid diagnostics, does each hour of ineffective antibiotic therapy really count? Methods This multicenter, retrospective, cohort study compared adult inpatients with E. coli bacteremia from a urinary source who received initial effective (EA) vs ineffective antibiotics (IA). The primary outcome was clinical treatment success at day 4. Secondary endpoints included length of stay (LOS), infection-related mortality, incidence of C. difficile infection (CDI), and subgroup analysis of outcomes by ESBL (CTX-M type) vs non-ESBL. Associations with endpoints were assessed using Fisher’s Exact tests using R v. 4.0.3. Results Clinical treatment success at day 4 was higher in the EA (n = 488) vs IA (n = 119) groups (93.7% vs 86.6%, p = 0.01) and median LOS was shorter (5 [IQR 4-6] vs 5 [IQR 5-7] days, p < 0.01). There were no differences in infection-related mortality (3.1% vs 3.4%, p = 0.8), 30-day mortality (2.5% vs 2.5%, p > 0.9), or incidence of CDI (1.8% vs 0%, p = 0.3) in the EA vs IA groups, respectively. For patients on IA < 24 h vs > 24 h, there was no difference in clinical improvement at day 4 (86.7% vs 90.5%, p > 0.9) nor 30-day mortality (2.4% vs 4.8%, p = 0.4). Clinical treatment success at day 4 was higher among non-CTX-M (n = 476) vs CTX-M (n = 131) patients (93.9% vs 86.3%, p = 0.01) even among those that received initial EA (94.5% vs 83.3%, p = 0.02). Median LOS was also shorter in CTX-M vs non CTX-M (5 [IQR 4-6] vs 5 [IQR 4-8] days, p < 0.01). Conclusion There was no mortality difference among patients receiving initial EA vs IA for E. coli bacteremia with rapid molecular blood culture diagnostics with active stewardship. Therapy for patients on IE is rapidly corrected and stewardship programs can use this intervention to promote judicious use of carbapenems. Disclosures All Authors: No reported disclosures