Proprietary machine-learning model built into the Epic electronic health record that scores sepsis risk in emergency and inpatient adults and alerts clinicians, as a decision aid.
Not directly applicable — an embedded software module, not a separately marked/cleared device.
Clinical evidence
Vendor-supported
No peer-reviewed vendor study found. Epic’s own accuracy figures (AUROC 0.76-0.83 for version 1, 0.83-0.86 for version 2) come from vendor documents, as cited by independent authors.
Independent
Independent study by the University of Michigan (Michigan Medicine) documenting materially poor real-world predictive performance, despite very wide deployment via the Epic EHR — the best-documented cautionary example in this scan.
Real-world deployment
Epic reports 95 organisations (731 hospitals) using version 2 as of August 2025, a vendor figure. Published US users include Michigan Medicine, OHSU, Emory Healthcare and MetroHealth.
Interoperability
European status
Editorial verdict
Limits and uncertainties
Independent evaluation found materially poor real-world performance — this must be the headline finding on any public profile, not a footnote, if this entry is published at all.
Sources
https://www.michiganmedicine.org/health-lab/epic-sepsis-models-ability-predict-depends-hospital-factors
Study — Wong A et al., External validation of a widely implemented proprietary sepsis prediction model (version 1), JAMA Internal Medicine, June 2021 (PubMed)
https://pubmed.ncbi.nlm.nih.gov/34152373/
Study — Wong A et al., Multicenter prospective validation of an updated proprietary sepsis prediction model (version 2), JAMA Network Open, February 2026 (PubMed)
https://pubmed.ncbi.nlm.nih.gov/41758510/
This profile is not medical advice or an endorsement. Hospitals should perform their own due diligence (regulatory compliance, GDPR, cybersecurity, integration, local clinical validation). How we evaluate
Every profile documents the precise indication, regulatory status, vendor-supported and independent evidence, real-world deployment, limits and dated sources.
A independent systematic review or prospective multicentre study · B at least one independent peer-reviewed study · C peer-reviewed, vendor-supported · D vendor claims only.