This Ohio Health System Tested an AI Tool to Predict Sepsis. Here’s How It Went

“IT teams need to rebuild paths to feed live patient data — not old medical records — into the model. ‘Those data sources are totally different when they’re moved to real time,’ said Weissman. Clinical leaders need to create workflows to make sure sepsis alerts are shared with the right people in the right way, and decide exactly how they would want clinicians to act on a red card. […]

Without significant investment, especially at community health systems like Summa, it’s difficult to prove how early alerts change clinical decisions, the time it takes to deliver antibiotics, or patient outcomes. ‘Those are really still the essential missing questions in the world of sepsis early warning systems,’ said Weissman.”

Read more at STAT News.