Bayesian Health Receives First-Ever FDA Clearance for Continuous AI Sepsis Monitoring
Tags AI · Enterprise
Bayesian Health received FDA 510(k) clearance for the first continuous AI-powered sepsis monitoring system, developed by Johns Hopkins researchers. The system detects sepsis 2–48 hours earlier than traditional methods by analyzing EHR data including chief complaints, lab measurements, vital signs, procedures, and medications. A prospective study published in Nature found that patients whose alerts were confirmed within three hours had reduced in-hospital mortality, organ failure, and length of stay. Deployed across dozens of US hospitals including Cleveland Clinic and MemorialCare, the system reduced sepsis mortality by 18%. Sepsis claims over 250,000 lives annually in the US. FDA clearance also opens the door to Medicare/Medicaid reimbursement under the New Technology Add-on Payment program.
Technical significance
This clearance sets a regulatory precedent for continuous AI monitoring in acute care, moving beyond point-in-time diagnostic tools. The 18% mortality reduction is clinically significant at scale given 250,000 annual US sepsis deaths. For the health tech industry, the NTAP reimbursement pathway means hospitals can recover costs for adopting the technology, accelerating clinical AI deployment beyond pilot programs.