JSON twin: https://www.healthaidb.com/software/epic-sepsis-model.json
Epic
https://www.epicshare.org/perspectives/streamlining-sepsis-prevention-and-care
The Epic Sepsis Model (ESM) is a proprietary predictive algorithm embedded in Epic's EHR that generates alerts to identify hospitalized patients at risk of developing sepsis.
A machine‑learning–based sepsis risk score and alerting tool built into Epic Systems' electronic health record that uses routinely collected EHR data (vitals, labs, demographics, medications, flowsheet data) to predict sepsis risk and notify clinical teams; deployed and configurable by Epic customers, with published validations showing variable performance across sites and subsequent vendor updates to the model.
yes
1979
Customer-controlled (hosted on customer infrastructure or customer-selected cloud region; regional hosting per health system policy)
P0553
2025-09-07
proprietary (Epic commercial license)
Contact Epic for licensing and deployment pricing; typically included as an Epic EHR module/feature or as a configurable enhancement for Epic customers.
enterprise_quote
2018
Epic server stack (Chronicles/InterSystems Caché or IRIS), Windows clients for Hyperspace; otherwise customer-hosted infrastructure
product
1.0
{
"company_name": "Epic",
"company_url": "https://www.epic.com/",
"company_offices": [
"United States"
],
"company_founding": "1979",
"product_url": "https://www.epicshare.org/perspectives/streamlining-sepsis-prevention-and-care",
"categories": [
"clinical",
"diagnostic",
"AI Inpatient Risk Monitoring",
"predictive analytics",
"clinical decision support"
],
"market_segment": [
"enterprise"
],
"links": [
"https://www.epic.com/",
"https://www.epicshare.org/perspectives/streamlining-sepsis-prevention-and-care",
"https://elion.health/products/epic-sepsis-model",
"https://pmc.ncbi.nlm.nih.gov/articles/PMC10317482/",
"https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2781307",
"https://www.statnews.com/2022/10/03/epic-sepsis-algorithm-revamp-training/",
"https://news.umich.edu/widely-used-ai-tool-for-early-sepsis-detection-may-be-cribbing-doctors-suspicions/",
"https://journal.chestnet.org/article/S0012-3692(22)01971-7/fulltext"
],
"summary": "The Epic Sepsis Model (ESM) is a proprietary predictive algorithm embedded in Epic's EHR that generates alerts to identify hospitalized patients at risk of developing sepsis.",
"description": "A machine‑learning–based sepsis risk score and alerting tool built into Epic Systems' electronic health record that uses routinely collected EHR data (vitals, labs, demographics, medications, flowsheet data) to predict sepsis risk and notify clinical teams; deployed and configurable by Epic customers, with published validations showing variable performance across sites and subsequent vendor updates to the model.",
"target_users": [
"clinicians",
"nurses",
"rapid response teams",
"hospitalists",
"emergency medicine clinicians",
"ICU clinicians",
"quality improvement teams",
"health IT/clinical informatics",
"hospital administrators"
],
"specialties": [
"critical care",
"emergency medicine",
"internal medicine",
"infectious disease",
"hospital medicine"
],
"regions_available": [
"United States",
"International (available to Epic customer sites)"
],
"languages_supported": [
"English"
],
"pricing_model": "enterprise_quote",
"pricing_details": "Contact Epic for licensing and deployment pricing; typically included as an Epic EHR module/feature or as a configurable enhancement for Epic customers.",
"license": "proprietary (Epic commercial license)",
"deployment_model": [
"on_prem",
"hosted/cloud-by-customer",
"hybrid"
],
"os_platforms": [
"Web (Epic Hyperspace)",
"Windows (client)",
"iOS (Rover/Haiku)",
"Android (Haiku/Rover Android)"
],
"features": [
"Real-time sepsis risk prediction using proprietary model (Epic Sepsis Model/ESM)",
"Continuous monitoring of vitals and labs for early warning",
"Automated clinician alerts/notifications and configurable thresholds",
"Clinical decision support/workflow recommendations",
"Order set and protocol triggering (eg. sepsis bundles, antibiotics, labs)",
"Trend visualization of risk score and contributing variables",
"Integration with nursing and physician workflows",
"Alert suppression and tuning to reduce alert fatigue",
"Reporting and analytics for performance and outcomes",
" EMR-embedded documentation templates and documentation prompts"
],
"optional_modules": [
"Advanced analytics/dashboarding",
"Remote inpatient monitoring/telemonitoring integration",
"Customizable alert tuning modules",
"Sepsis registry/surveillance module"
],
"integrations": [
"Epic EHR (Hyperspace/Chronicles/Caché/InterSystems)",
"Bedside/physiologic monitors (via middleware)",
"Laboratory information systems (labs/results)",
"ADT/patient demographics",
"Pharmacy/medication ordering modules",
"Microbiology/Culture systems",
"Clinical messaging/notification services",
"Telehealth/remote monitoring platforms"
],
"data_standards": [
"FHIR",
"HL7 v2",
"LOINC",
"SNOMED CT",
"ICD-10"
],
"api_available": "yes",
"system_requirements": "Epic server stack (Chronicles/InterSystems Caché or IRIS), Windows clients for Hyperspace; otherwise customer-hosted infrastructure",
"compliance": [
"HIPAA",
"HITECH"
],
"certifications": [],
"security_features": [
"Role-based access control (RBAC)",
"Audit logging",
"Encryption in transit",
"Encryption at rest",
"Single sign-on/SAML integration",
"Multi-factor authentication (optional)"
],
"privacy_features": [
"Business Associate Agreement (BAA) available",
"Patient consent management (via EHR)",
"Data de-identification/limited datasets for analytics"
],
"data_residency": "Customer-controlled (hosted on customer infrastructure or customer-selected cloud region; regional hosting per health system policy)",
"customers": [
"Saint Luke's Health System",
"Prisma Health",
"Lee Health",
"Ochsner Health",
"Novant Health",
"MetroHealth"
],
"user_reviews": [
"Everybody gets a sepsis alert! Mild fever? Sepsis. Pale and sweaty? Sepsis. Lactate marginally up? Sepsis. I've given up arguing about it. It's like shouting into a hurricane.",
"To be fair to Epic, my hospital already had a useless sepsis alert system in place even before we implemented our new EHR.",
"Study published in JAMA suggests that the Epic Sepsis Model may perform poorly at external sites — we saw mixed results when implemented locally.",
"I'm a professor interested in how nurses perceive, use, or ignore the EPIC sepsis model — many nurses say they often ignore alerts due to alert fatigue."
],
"ratings": [],
"support_channels": [
"email",
"phone",
"ticketing",
"community",
"live_support"
],
"training_options": [
"documentation",
"webinars",
"live_online_training",
"onsite_training",
"computer_based_training_modules"
],
"release_year": "2018",
"integration_partners": [
"Galaxy (Epic document/knowledge distribution)",
"Cosmos (Epic analytics platform)",
"MyChart",
"open.epic"
],
"id": "P0553",
"slug": "epic-sepsis-model",
"type": "product",
"version": "1.0",
"last_updated": "2025-09-07",
"links_json": {
"self": "https://www.healthaidb.com/software/epic-sepsis-model.json"
}
}