Is more necessarily better? The impact of prescription drug coverage on clinical outcomes in a Medicare population

This study aims to provide needed evidence on the relationship between insurance coverage for prescription drugs and health outcomes. The existing literature establishes a link between prescription coverage and utilization/adherence behavior, and there is evidence to suggest that adherence influences health status, but to date there is little data to support the often-assumed impact of drug coverage on clinical outcomes. This study uses data from the Medicare Current Beneficiary Survey linked to Medicare claims data for 1995-1999 to examine this question for Medicare beneficiaries prior to the implementation of Medicare Part D. In the study, beneficiaries are assigned to drug coverage categories (as distinct from supplemental health coverage). The study focuses on a hypertensive population—one of the most common medical conditions in this population, one for which effective medications are available, and one which accounts for a large proportion of outpatient drug expenditures. Risk-adjusted stroke rates and all-cause mortality are the outcomes of interest. Using Cox proportional hazards models, stroke risk is modeled as a function of drug coverage, controlling for clinical risk adjustors, as well as demographic and socioeconomic covariates. Findings suggest that neither the presence/absence of prescription drug coverage, nor the type of coverage affects risk-adjusted stroke rates for Medicare beneficiaries with hypertension. The model findings hold when all-cause mortality and a combined stroke-or-death endpoint are considered. Secondary analyses including the incorporation of a propensity score for drug coverage, and a bivariate probit test for the endogeneity of coverage, support these findings. It is unclear whether these findings are generalizable to other medical conditions. By investigating whether costly prescription coverage affects clinical outcomes, the study aims to inform the ongoing national policy debate about the value of large-scale prescription drug benefits programs