ABSTRACT
BACKGROUND: The 2010 Patient Protection and Affordable Care Act reformed the individual and small group health insurance markets and established a risk adjustment program to create a level playing field for competition. A new set of predictive models for measuring enrollee risk across plans was developed for the Patient Protection and Affordable Care Act-reformed markets, referred to as the Department of Health and Human Services Hierarchical Condition Category (HHS-HCC) models. Beginning in 2018, selected prescription drug classes were added to the models as risk markers. OBJECTIVE: We describe the motivations, concerns, methodology, and results of adding prescription drug utilization to the HHS-HCC models. METHODS: Separate HHS-HCC models are estimated by enrollee age and plan actuarial value. We defined and added 10 prescription drug classes, called RXCs, to the HHS-HCC adult models. RESULTS: Using selected RXCs alongside demographic and diagnostic indicators yielded modest overall improvement in HHS-HCC models' predictive power. Also, adding RXCs captures the higher costs of enrollees taking certain expensive pharmaceuticals and allows imputation of diagnoses for enrollees utilizing a drug but lacking the associated diagnosis. CONCLUSIONS: Including selected drugs in risk adjustment improved the models' predictive power. In addition, inclusion of selected drugs may discourage insurers from using formulary and drug benefit design to avoid enrollment of patients taking high-cost drugs, such as for HIV, multiple sclerosis, and rheumatoid arthritis, and improve access for enrollees taking these drugs. Adding RXCs also may improve plan risk measurement for plans with less complete diagnosis reporting.
