Optimizing Sepsis Time Zero Definitions Through Associations between Antibiotic Timing and Mortality.
RATIONALE: Sepsis performance measures require operational definitions of "time zero", but SEP-1's heterogeneous organ dysfunction criteria may not all identify patients vulnerable to harm from antibiotic delays. OBJECTIVES: To determine which SEP-1 organ dysfunctions identify patients in whom antibiotic delays are associated with increased mortality. METHODS: We analyzed adults with suspected infection using electronic health record data from 145 U.S. hospitals (2018-2023). We estimated associations between antibiotic timing (1-3 vs. 0-1h and 3-6 vs. 0-3h) and in-hospital mortality for each SEP-1 organ dysfunction criterion, alone and in combination: hypotension, elevated lactate, respiratory failure, renal dysfunction, hyperbilirubinemia, thrombocytopenia, and coagulopathy. We used inverse probability weighting to adjust for baseline and time-varying confounders. MEASUREMENTS AND MAIN RESULTS: Among 1,849,729 encounters with suspected infection, 624,416 had ≥1 organ dysfunction and received antibiotics within 6h of time zero. Antibiotics at 1-3 vs. 0-1h were associated with higher mortality when time zero was defined by hypotension (aRR 1.07; 95% CI, 1.03-1.10) or lactate ≥4.0mmol/L (aRR 1.10; 1.05-1.15), but not for other dysfunctions alone or in combination. Among patients without hypotension or lactate ≥4.0mmol/L, antibiotics at 3-6 vs. 0-3h were only associated with higher mortality in patients with acute respiratory failure (aRR 1.11; 1.05-1.18). CONCLUSIONS: Mortality associations with antibiotics beyond 1 hour were most evident when sepsis time zero was defined by hypotension or lactate ≥4.0mmol/L, and with antibiotics beyond 3 hours when defined by acute respiratory failure. These findings could inform refinements to sepsis time zero definitions to better guide antibiotic urgency and performance measurement.
Authors
- Kenneth Sands (ORCID: https://orcid.org/0000-0002-1379-9793)
- Laura K. DelloStritto (ORCID: https://orcid.org/0000-0003-0295-0010)
- Michael Klompas (ORCID: https://orcid.org/0000-0001-8641-4498)
- Theodore R. Pak (ORCID: https://orcid.org/0000-0002-1676-2523)
- Henry Masur (ORCID: https://orcid.org/0000-0002-3688-021X)
- Russell E. Poland (ORCID: https://orcid.org/0000-0003-4395-5811)
- Jessica G. Young (ORCID: https://orcid.org/0000-0002-2758-6932)
- Anna A. Agan (ORCID: https://orcid.org/0009-0002-7035-9601)
- Chanu Rhee (ORCID: https://orcid.org/0000-0002-9537-4245)
- Sameer S. Kadri (ORCID: https://orcid.org/0000-0002-4420-9004)
- Sayon Dutta (ORCID: https://orcid.org/0000-0002-0381-6860)
- Michael R. Filbin (ORCID: https://orcid.org/0000-0002-2588-7504)
- Caroline S. McKenna (ORCID: https://orcid.org/0009-0000-1093-3080)
- Jeffrey S. Guy
- Claire N. Shappell (ORCID: https://orcid.org/0000-0002-8316-1496)
- Edward Joel Septimus (ORCID: https://orcid.org/0000-0002-0102-3330)
- Robert Jin
Institutions
- Brigham and Women's Hospital (US)
- National Institutes of Health (US)
- HCA Healthcare (US)
- Massachusetts General Hospital (US)
- National Heart, Lung, and Blood Institute (US)
- Harvard Pilgrim Health Care (US)
- National Institutes of Health Clinical Center (US)
- Belmont University (US)
- Texas A&M University (US)
Publication Details
- Journal
- PubMed
- Published
- 2026-11-12
- DOI
- https://doi.org/10.1093/ajrccm/aamag610
- Primary Topic
- Sepsis Diagnosis and Treatment
- Type
- article
- Field-Weighted Citation Impact
- 0.00