Stratifying for Value: An Updated Population Health Risk Stratification Approach

Area of Science:

• Health Economics
• Population Health Management
• Clinical Informatics

Background:

• Current risk stratification models often prioritize clinical outcomes over financial value.
• Value-based contracting necessitates models that incorporate both disease burden and healthcare costs.
• Efficient care coordination programs require accurate identification of high-risk, high-cost patient populations.

Purpose of the Study:

• To develop a novel risk stratification algorithm for a large health system.
• To integrate patient medical spend with disease burden for enhanced accuracy.
• To improve the efficiency and impact of care coordination programs.

Main Methods:

• Developed a 5-category risk algorithm using the Minnesota Tiering system foundation.
• Expanded the comorbidity list for comprehensive disease burden assessment.
• Weighted risk scores by the preceding 12 months of patient medical spend.
• Applied the algorithm to a 554,805-patient Accountable Care Organization (ACO) population.

Main Results:

• The customized algorithm identified a high-risk tier comprising 5% of the patient population (27,552 patients).
• This top tier accounted for 67.9% of the total annual medical spend ($1,107,822,887).
• Focusing on the top 2 tiers (15% of patients) captured 83.2% of annual medical spend ($1,357,545,872).

Conclusions:

• The novel risk stratification approach effectively identifies patients with high disease burden and significant medical spend.
• This method allows for focused allocation of intensive care coordination resources.
• The algorithm provides greater explanatory power for value-based care initiatives by linking cost and clinical factors.