Timeliness of Broad-Spectrum Antibiotic De-escalation and Mortality in Emergency Department Patients With Sepsis: A Retrospective Hierarchical Logistic Regression Study

Abstract

Background Early initiation of broad-spectrum antibiotics (BSA) is a cornerstone of sepsis management; however, prolonged use without timely de-escalation to narrow-spectrum antibiotics (NSA) can contribute to adverse outcomes, including higher mortality and antimicrobial resistance. While intensive care unit (ICU) studies support the benefits of antibiotic de-escalation, its timing and impact in emergency department (ED) settings remain underexplored. Objective To examine the association between delayed de-escalation from BSA to NSA and in-hospital mortality among ED patients with sepsis, adjusting for the Charlson Comorbidity Index (CCI), age, and sex. Methods We conducted a retrospective multicenter cohort study of adult ED patients (≥18 years) hospitalized with sepsis between January 2021 and December 2023. All received empiric BSA followed by de-escalation to NSA. De-escalation timing was measured in 6-hour intervals starting 61 minutes after BSA initiation. Hierarchical binary logistic regression assessed associations with in-hospital mortality, controlling for CCI, age, and sex. Results Among 2,906 eligible patients, delayed de-escalation was independently associated with increased mortality (OR per 6-hour delay = 1.006; 95% CI, 1.001-1.012; p = .023). Higher CCI and age were also significant predictors (CCI OR = 1.175; age OR = 1.014; both p ≤ .001). Sex was not significantly associated with mortality (p = .502). Each six-hour delay beyond the first hour increased the odds of death by approximately 0.6%. Conclusion Delayed de-escalation of empiric BSA in ED sepsis patients is linked to higher in-hospital mortality, reinforcing the need for prompt reassessment and narrowing of antibiotic therapy as part of ED antimicrobial stewardship.