Medical Record Abstraction for Quality Improvement in Sepsis Care Using Artificial Intelligence: A Cluster Randomized Trial
Authors' conclusion
In this cluster randomized trial of artificial intelligence (AI)-enabled medical record abstraction for sepsis care, rapid assessment of SEP-1 performance and targeted feedback improved overall compliance with the measure.
Study context
- Population
- ED attending physicians (>3 shifts/month) at 2 academic emergency departments (EDs), University of California, San Diego (UCSD) health system, and their patients meeting CMS SEP-1 criteria for severe sepsis/septic shock66 physicians randomized
- Participants
- 301 participants
- Comparison
- Intervention vs Control
- Interventionn = 180
- Controln = 121reference
- Primary endpoint
- Overall compliance with SEP-1
Methodological detailscluster randomised, single-blind, unstratified, 2 academic EDs
- Cluster randomised
- Single-blind
- Unstratified
- 2 academic EDs
Primary result
As reported in the abstract
SEP-1 compliance
Each group vs controlOdds ratio (95% CI)P value
Intervention
2.10(1.15–3.81)
P = 0.02- Intervention
- absolute improvement 13.0% (95% CI 2.5%–23.4%); P = 0.02
Secondary outcomes
Noncompletion of 30-mL/kg fluid bolus component · documentation-sensitive component
Agreement between LLM determination and expert review
- Agreement
- 92%
ICU admissions and 30-day mortality
- No significant differences
Original publication
Medical Record Abstraction for Quality Improvement in Sepsis Care Using Artificial Intelligence: A Cluster Randomized Trial. JAMA Netw Open. Published 1 June 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →