Allina Health saves $13.4M by optimizing length of stay with analytics platform

Allina Health could not accurately identify LOS improvement opportunities because its traditional reporting calculated stays in whole days, lacked acuity adjustment, and couldn't estimate financial impact. Performance data was delivered weeks or months after discharge, preventing real-time clinical decision-making. Most providers had no access to LOS data, limiting engagement and improvement.

Allina built an inpatient data mart on a Late-Binding Enterprise Data Warehouse (EDW) and Health Catalyst Analytics Platform to measure LOS by the minute and benchmark against MS-DRG geometric mean LOS (GMLOS). The platform delivered near-real-time dashboards to clinicians, surfacing discharge order timing, practice pattern variation, and weekend delay patterns. A standardized discharge planning process and SNF transfer agreements were implemented based on analytic findings.

In the first two years, Allina saved 26,000+ inpatient days and $13.4 million in direct operating expenses. Hospital capacity was created for 5,000+ additional admissions—equivalent to 90 new beds. Adverse events were also reduced, including 90+ averted patient falls (20 with harm) and prevention of 25+ C. diff infections.

• Measuring LOS by the minute rather than whole days revealed meaningful variation that coarse reporting masked, enabling targeted interventions.
• Acuity-adjusted benchmarking using MS-DRG GMLOS gave providers credible, non-accusatory performance comparisons that drove buy-in and behavior change.
• Making data available in near-real-time at the provider and unit level eliminated the staffing bottleneck of manual reporting and created self-sustaining demand for improvement.