Improving risk equalization using information on physiotherapy diagnoses.

Area of Science:

• Health economics
• Insurance risk modeling
• Public health

Background:

• Risk-equalization (RE) models in health insurance have advanced from demographic to diagnosis and pharmacy-based indicators.
• Current RE models have limitations; incorporating physiotherapy treatment data can enhance their accuracy.
• A novel risk-adjuster based on prior-year physiotherapy costs was implemented in the Dutch RE model in 2016.

Purpose of the Study:

• To evaluate the effectiveness of the current physiotherapy cost-based risk-adjuster.
• To assess the impact of replacing the cost-based adjuster with diagnosis-based risk-adjusters.
• To identify the optimal modality for incorporating physiotherapy diagnoses into the Dutch RE model.

Main Methods:

• Utilized physiotherapy claims data (2012) and administrative data (2013) for 15.8 million individuals in the Netherlands.
• Evaluated a cost-based risk-adjuster and four diagnosis-based modalities, including separate diagnoses, pathology-based clusters, residual cost clusters, and age-interaction clusters.
• Filtered 89 diagnoses to 27 relevant ones for model inclusion.

Main Results:

• The cost-based risk-adjuster improved predictive power and reduced undercompensation for physiotherapy users.
• All four diagnosis-based modalities outperformed the cost-based adjuster.
• A modality with 4 risk-classes based on clustered residual costs demonstrated superior predictive power and simplicity.

Conclusions:

• Incorporating physiotherapy information substantially enhances sophisticated RE models.
• For the Dutch RE model, a risk-adjuster utilizing 4 risk-classes for clustered diagnoses based on residual costs is recommended.
• Diagnosis-based risk adjusters offer improved accuracy in health insurance risk equalization.