Machine-learning-derived sepsis bundle of care
Authors' conclusion
Our bundle of six recommendations is associated with a dramatic reduction in mortality in sepsis and septic shock.
Study context
- Population
- Intensive care unit (ICU) patients with sepsis/septic shock from two US databases (Medical Information Mart for Intensive Care-IV (MIMIC-IV); EICU Collaborative Research Database (eICU-CRD))
- Participants
- 42,735 patients
- Exposure
- Bundle vs no bundle
- Outcome
- All-cause 28-day mortality
- Follow-up
- 28 days
Methodological detailsmachine learning (least absolute shrinkage and selection operator), propensity-score matched, internal and external validation
- Machine learning (least absolute shrinkage and selection operator)
- Propensity-score matched
- Internal and external validation
Main finding
As reported in the abstract
28-day mortality
Bundle vs No-bundleOdds ratio (CI)P value
PS-matched internal validation cohort
0.41(0.33–0.53)
P < 0.001PS-matched external validation cohort
0.75(0.60–0.94)
p 0.02Observational data: associations, not effects.
Original publication
Machine-learning-derived sepsis bundle of care. Intensive Care Med. Published 29 November 2022. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →