Allina Health

Allina Health's 12-hospital Minnesota system faced millions in costs and Medicare penalties from preventable 30-day readmissions. Their reactive discharge model lacked the ability to identify high-risk patients before they left the hospital, leaving care coordinators without actionable signals at the point of discharge.

Implemented predictive risk scoring via Health Catalyst across all 12 hospitals in 2018, combining 47 variables — including prior admissions, medication adherence, and social determinants — into a single readmission risk score calculated at discharge. Critically, they redesigned discharge workflows around the score rather than treating it as a separate dashboard.

Achieved a 10.3% reduction in 30-day readmissions, generating $4.2 million in annual savings from avoided penalties and costs. The $890,000 implementation investment yielded a 472% ROI over three years.

• Workflow redesign is as important as the technology — embedding the risk score into discharge processes drove adoption far more than the model accuracy alone.
• Combining social determinants with clinical variables substantially improves predictive power.
• Framing the tool as a care process redesign initiative rather than an analytics project was key to clinical buy-in.