Clinical question: In patients hospitalized with community-onset sepsis, does empiric broad-spectrum antibiotic (BSA) de-escalation (on encounter day four), compared with continuation, affect mortality and length of stay?
Background: Current guidelines recommend early empiric BSA for suspected sepsis to avoid delays in treating multidrug-resistant organisms (MDROs). However, prolonged exposure to BSAs is associated with adverse drug events, Clostridium difficile infection, and antimicrobial resistance. Therefore, timely de-escalation within 48 to 72 hours is recommended if MDROs are not identified on initial testing, and yet real-world practice varies widely. Prior studies evaluating de-escalation in sepsis have produced mixed results, and clinicians remain concerned about the safety of narrowing therapy in patients who are acutely ill. This study sought to examine the variability in antibiotic de-escalation practices and compare outcomes of empiric BSA continuation versus de-escalation, in the absence of MDRO, with respect to 90-day mortality, length of stay, and antibiotic days.
Study design: Target trial emulation study (observational cohort with inverse probability weighting)
Setting: 67 hospitals participating in the Michigan Hospital Medicine Safety Consortium in Michigan, U.S.
Synopsis: This large, multicenter target trial emulation used data from 36,924 adults hospitalized with community-onset sepsis across 67 Michigan hospitals (June 2020 to September 2024). Two cohorts were evaluated: patients receiving empiric anti–methicillin-resistant Staphylococcus aureus (MRSA) therapy (n=6,926, 18.8%) and those receiving empiric antipseudomonal therapy (PSA) (n=11,149, 30.2%) without evidence of MDRO by hospital day three. Exposure was the de-escalation of BSA on day four versus continuation. Of these, 2,993 (43.2%) and 2,493 (22.4%) were de-escalated from anti-MRSA and anti-PSA coverage, respectively. After inverse probability weighting to balance patient and hospital characteristics, de-escalation of BSA was not associated with higher 90-day mortality (primary outcome) (anti-MRSA: odds ratio, 1.00; 95% CI, 0.88 to 1.14; anti-PSA: odds ratio, 0.98; 95% CI, 0.86 to 1.13). Among secondary outcomes, BSA de-escalation was associated with fewer antibiotic days and shorter length of stay. Limitations include observational design and potential residual confounding. These findings align with stewardship principles and support guideline-recommended reassessment and narrowing of empiric therapy in sepsis when resistant organisms are not identified.
Bottom line: In adults hospitalized with community-onset sepsis and no evidence of MDRO infection, de-escalating empiric broad-spectrum antibiotics by hospital day four appears safe and reduces antibiotic exposure and hospital length of stay.
Citation: Gupta AB, et al. Antibiotic de-escalation in adults hospitalized for community-onset sepsis. JAMA Intern Med. 2026;186(2):192-202. doi:10.1001/jamainternmed.2025.6919.
Dr. Qaiser is an academic hospitalist at the Cleveland Clinic, an associate program director in the internal medicine residency program, and clinical assistant professor of medicine at the Cleveland Clinic Lerner College of Medicine of Case Western Reserve University, both in Cleveland