Study finds AI warning system linked to fewer deaths among high-risk hospital patients
Researchers at RWJBarnabas Health and Rutgers say the technology helped identify deteriorating patients sooner across 11 New Jersey hospitals.
More Stories
A study by researchers at RWJBarnabas Health and Rutgers Robert Wood Johnson Medical School found that an AI-powered early warning system was associated with fewer deaths among high-risk hospitalized patients across 11 New Jersey hospitals.
The study, published in NEJM AI, evaluated outcomes for 23,132 high-risk adult patients treated at academic medical centers, community teaching hospitals and community hospitals within the RWJBarnabas Health system.
Researchers found the in-hospital death rate among high-risk patients fell from 23.1% before the system was implemented to 18.6% afterward, representing an 18% reduction in the risk-adjusted odds of death.
The system relies on the Epic Deterioration Index, which continuously analyzes information already stored in patients' electronic health records, including vital signs, laboratory results, nursing assessments and age, to estimate a patient's risk of serious clinical decline. The tool recalculates risk scores every 15 minutes and automatically alerts rapid response teams when patients reach the highest-risk category.
"Our goal was to identify patients earlier, before they reached a point where intervention becomes much more difficult," lead author Dr. Thomas Nahass, an intensive care and clinical informatics physician at RWJBarnabas Health and Rutgers Robert Wood Johnson Medical School, said in a statement. "If we can get a critical care eye on the patient sooner, we can change the course of their outcome."
After the warning system was introduced, rapid response team evaluations of high-risk patients increased from 25.3% to 37.5% of hospital stays. Researchers said transfers to intensive care units did not significantly increase despite the higher number of rapid response evaluations.
The authors said the reduction in deaths likely reflected the broader care process surrounding the technology, including clinician training, automated alerts and faster responses by care teams, rather than the AI tool alone.
(Photo Credit: Business Wire)
