A case-mix classification system for explaining healthcare costs using administrative data in Italy

Area of Science:

• Health Services Research
• Health Economics
• Public Health

Background:

• The Italian National Health Service (NHS) offers universal healthcare coverage with copayments for certain services.
• Local Health Trusts (LHTs) are financed via capitation adjusted for age, gender, and residence.
• A need exists to better explain healthcare costs beyond basic demographic adjustments.

Purpose of the Study:

• To evaluate the effectiveness of the Johns Hopkins Adjusted Clinical Groups (ACG) System for risk adjustment in predicting healthcare costs.
• To compare the predictive power of the ACG System against a simple age-gender model using administrative data.

Main Methods:

• The ACG System was applied to all residents of the Veneto Region in 2015.
• Data sources included hospital discharges, emergency room visits, chronic disease registries, ambulatory visits, medications, home care, and drug prescriptions.
• Linear regressions compared an age-gender model with more comprehensive ACG-based risk adjustment models.

Main Results:

• The age-gender model explained only 8% of the variance in total healthcare costs.
• Incorporating diagnosis-related variables increased explained variance by 23%.
• Pharmacy-based variables added another 17% to the explained variance, with the comprehensive model's R-squared being 6 times higher than the basic model.

Conclusions:

• The ACG System offers a substantial improvement in predicting healthcare costs over simple age-gender adjustments.
• Healthcare costs are more strongly influenced by the accumulation of chronic conditions and multimorbidity than by aging alone.