FDA-Cleared AI System Detects Sepsis Earlier and Reduces Mortality

Each hour of delayed recognition measurably decreases survival, yet timely diagnosis is challenging across busy inpatient settings. Hospitals therefore need earlier, reliable alerts that surface risk before clinicians already suspect infection. A newly cleared system now offers pre-suspicion detection of sepsis and earlier notification, with evidence of reduced mortality in U.S. hospitals.

The Targeted Real-Time Early Warning System, developed by Johns Hopkins University (Baltimore, MD, USA) researchers and commercialized by Bayesian Health, has received U.S. Food and Drug Administration (FDA) clearance for clinical use, making it one of the first AI-based medical tools to secure this regulatory clearance. Already deployed across dozens of hospitals, the system has been associated with reduced sepsis mortality. It is designed to address one of the leading causes of in-hospital death, a condition that claims more than 250,000 lives annually.

The technology integrates electronic health records with advanced clinical AI to provide “pre-suspicion” screening, aiming to surface risk before clinicians already suspect sepsis. In practice, the system has helped physicians identify cases nearly two to 48 hours earlier than traditional methods and can detect sepsis hours faster than doctors. By adding lead time, the platform is designed to prompt earlier clinical action when symptoms may still be nonspecific, such as fever or confusion.

Regulatory progress for the system has included Breakthrough Designation in 2023, under which it was deployed at several health systems including Cleveland Clinic, MemorialCare in California, and the University of Rochester School of Medicine. In those settings, it significantly reduced in-hospital mortality, morbidity, and length of stay for patients with sepsis. Across broader use in the United States, sepsis mortality rates declined by 18%. FDA clearance also opens the door for hospitals using the system to receive Medicare and Medicaid reimbursement under the New Technology Add-on Payment program.

“Pre-suspicion screening is what creates lead time, and lead time is what changes outcomes in sepsis. Once a clinician already suspects sepsis, the clock has been running—often for hours or even days,” said Suchi Saria, a Johns Hopkins professor and director of the AI & Healthcare Lab.

“It gives physicians an additional set of eyes and ears and could genuinely help save lives. This is a significant milestone for Johns Hopkins and Dr. Saria's team,” said Albert Wu, a Johns Hopkins expert in patient safety and a co-investigator on the work.

Related Links
Johns Hopkins University
Bayesian Health