The Effect of Insurance Expansions on Smoking Cessation Medication Use: Evidence from Recent Medicaid Expansions

Smoking is the largest preventable cause of morbidity and mortality in the U.S., leading to more than 480,000 deaths each year. Despite these health effects and numerous anti-smoking campaigns undertaken by all levels of government over the past several decades, the smoking rate is 17%. The smoking rate is particularly high, 29%, within the Medicaid population. This high prevalence rate within the Medicaid population is particularly troubling from a public finance perspective as smoking-attributable Medicaid costs will be borne predominately by taxpayers. Importantly, research suggests that that expanding coverage for cessation medications can lead to increases in the utilization of these medications and reductions in smoking within the traditional Medicaid population – poor parents, pregnant women, and adults with disabilities.

We examine the impact of expanding coverage for evidence-based cessation prescription medications (Chantix, Nicotrol, and Zyban) to the non-traditional Medicaid population – low-income nondisabled childless adults. This population had little access to insurance prior to the passage of the Affordable Care Act (ACA). Although these medications are efficacious, they require a prescription from a healthcare provider, and are costly for low-income and insured smokers. We leverage rich administrative data from the Centers for Medicare and Medicaid Services State Drug Utilization Database (SDUD) 2011-2015. These data cover the universe of prescription drug claims purchased through retail and online pharmacies for which Medicaid is a third-parity payer. We combine these unique data with (1) the quasi-experiment offered by the ACA, which requires public insurance plans to cover these medications at low patient cost-sharing, and (2) states’ decisions to expand Medicaid to non-traditional populations.

This combination of policy changes and differences-in-differences methods allow us to shed new light on the small literature that examines the impact of expanding access to effective cessation medications within Medicaid in several ways. First, we are able to examine the effect of expanding coverage to non-traditional Medicaid populations. Second, our use of the SDUD offers us access to the universe of Medicaid-financed prescription medications purchased at retail and online pharmacies, while previous studies have relied on survey data or claims data that lack detailed payer information. Third, our data contain information on the exact Medicaid payments which we leverage to estimate program costs. Fourth, we are able to study heterogeneity across prescription cessation medications in Medicaid effects. Fifth, we explore medication financing; that is we are able to address the following question: did state Medicaid programs or patients shoulder the financial burden of any increased medication utilization?

Our findings suggest that expansions increased smoking cessation prescriptions by 36% and total payments for these medications by 28%. Our effects are driven by increased prescriptions and payments for Zyban. We provide evidence that these medications were financed by state Medicaid programs and not patients. Our findings are highly robust to numerous robustness checks, including policy endogeneity and control for between-state heterogeneity. Overall our findings suggest that the recent Medicaid expansions allowed low-income smokers to access efficacious cessation medications.