Epic, Microsoft A.I. Tools Show Promise Predicting Health Problems
A Louisiana-based hospital network’s recent study of machine learning tools from Epic Systems provides the latest glimpse of healthcare’s possible future, if advocates of artificial intelligence technologies have their way.
Ochsner Health System, which operates 30 hospitals and 80-plus clinics in Louisiana and Mississippi, says it’s among the dozens of healthcare organizations in the U.S. currently using a machine learning platform developed by Verona, WI-based Epic that is designed in part to help healthcare providers intervene more quickly when a patient’s health is deteriorating. Epic is one of the country’s leading vendors of electronic health records software; Ochsner has used the company’s tools to manage patient information since 2011.
Ochsner says it recently conducted a 90-day pilot study in which a “rapid response team” at the organization used a predictive software model developed by Epic that combs through patients’ clinical data and alerts a doctor or a nurse when the software predicts they’re likely to suffer “adverse health events,” like cardiac arrest. Ochsner says it observed a 44 percent decrease in the number of adverse health events during the study. (The study excluded events that occurred in the intensive care unit, Ochsner says.)
According to Ochsner, the Epic software it used in the pilot is powered by Microsoft’s (NASDAQ: MSFT) Azure cloud computing service.
“By utilizing Epic’s machine learning platform and Microsoft Azure, we can detect health patterns, learn from these insights, and develop a more aggressive treatment plan as a preventative measure,” says Laura Wilt, Ochsner’s chief information officer, in a prepared statement
In a press release, Seth Hain, Epic’s director of analytics and machine learning, says the company’s A.I. tools hold promise not only for improving patient care, but also helping organizations operate more efficiently through precise staffing models, for example.
The Ochsner trial of Epic’s and Microsoft’s software comes as blue-chip businesses like IBM (NYSE: IBM) and GE (NYSE: GE) have flocked to the intersection of healthcare and artificial intelligence, a pocket of the industry that has generated lots of hype but is only now starting to deliver on that promise.
While the pilot was relatively short in duration and its findings haven’t been peer reviewed, Ochsner was apparently satisfied with the results and says it plans to broaden its use of Epic’s machine learning platform. Used in tandem with Microsoft Azure, Epic’s tools could help Ochsner lower the rates at which its patients contract pressure ulcers and hospital-acquired infections, the healthcare provider says.