Predicting Readmission Risks

Align: A unified health system of physicians, hospitals and communities, wanted to reduce patient readmissions and improve patient and family education.

Appraise: Improved care by predicting readmissions across all conditions and identifying top readmission drivers, which were previously buried in physician’s notes (unstructured).  The AI solution improved prediction accuracy to 93% with AI​, a 52% improvement to population care. Margin expanded by redeploying emergency department resources to proactively engage patients and reduce readmission rates. Reduced overtime and eliminated staffing shortages resulting in a 30% reduction of annual operating cost for the ED.