Health-based risk adjustment Improving the pharmacy-based cost group model to reduce gaming possibilities

Area of Science:

• Health economics
• Health services research
• Pharmacoeconomics

Background:

• The pharmacy-based cost group (PCG) model adjusts health plan capitation payments based on prior year medication use as a proxy for chronic conditions.
• While enhancing predictive accuracy, the PCG model's reliance on past utilization presents opportunities for strategic manipulation ('gaming').

Purpose of the Study:

• To investigate and identify effective strategies for mitigating potential gaming within the pharmacy-based cost group (PCG) model.
• To evaluate the predictive performance of the PCG model compared to demographic models.

Main Methods:

• The study examined various strategies to prevent gaming in the PCG model.
• Key strategies evaluated included the number of prescribed daily doses for PCG assignment, the inclusion or exclusion of comorbidity data, and the removal of PCGs with low future costs.

Main Results:

• The most effective strategies identified were utilizing a high number of prescribed daily doses for PCG assignment, excluding comorbidity data, and removing PCGs with low future costs.
• The PCG model demonstrated significantly higher predictive power, accounting for nearly twice the variance compared to traditional demographic models.

Conclusions:

• The refined PCG model offers a substantial improvement in explaining variance in healthcare costs compared to demographic models.
• Implementation of the PCG model in the Dutch sickness fund sector in 2002, covering two-thirds of the population, highlights its practical application and potential benefits for health plan payment adjustments.