Seminario Fragmented Formularies and Patient Outcomes: Evidence from the US and Implications for Evaluating Unified Drug Coverage

Abstract

Competitive, multi-payer health insurance markets allow plans to design their own drug formularies, generating potential cost-containment benefits but also frequent and fragmented coverage changes across plans. Evaluating the trade-offs of this approach relative to unified formulary systems requires credible evidence on how formulary restrictions affect patient outcomes. We provide such evidence using a difference-in-differences design that exploits year-to-year variation in formulary exclusions across competing insurance plans in the U.S. Medicare prescription drug benefit. Focusing on branded diabetes medications, we find that formulary exclusions increased plan switching by 4.6 percentage points, reduced use of the excluded drug by 37 percentage points, reduced use of any drug in the class by 4 percentage points, and produced a temporary 5–8% increase in diabetes-related emergency department visits in the two quarters following exclusion. Effects were substantially larger for low-income, minority, disabled, and urban patients, who were also least able to mitigate exclusions by switching plans. We discuss how these estimates can inform a broader framework for comparing the welfare implications of fragmented versus unified formulary design.