Access to Medicaid dental services and distant-stage oral cancer diagnoses in Utah: A quasi-experimental approach.

Abstract

106 Background: Late detection may contribute to disparities in oral cancer outcomes, as most low-income oral cancer patients are diagnosed at distant stages. Because dentists are the most likely provider to screen for oral cancer, access to Medicaid dental services could improve outcomes. However, states are not mandated to cover adult Medicaid dental benefits. In 2010, Utah’s Medicaid program added dental benefits. My study leverages this policy change to test if dental coverage is associated with distant stage diagnoses and if this association changes with the number of dentists per capita. Methods: I analyzed SEER data for all oral cancers diagnosed between 2007-2013. The outcome is a binary variable, indicating if the oral cancer patient was diagnosed at a distant stage. All analyses control for age, race, and gender as well as state and year fixed-effects. Under a Difference-in-Differences design, where DD = Post2010*Utah, I estimate the effect of adding dental coverage for Medicaid enrollees in Utah compared to Medicaid enrollees in states that did not change their Medicaid program. Then, to account for differential trends, I implement a Triple Differences design, where DDD = Post2010*Utah*Medicaid, and estimate the effect of adding dental coverage for Medicaid enrollees in Utah compared to adults in Utah who are not on Medicaid. Finally, I interact the DDD variable with a variable measuring Utah's dentists per 100,000 population (county-level data from AHRQ). Results: Compared to Medicaid enrollees in other states, adding Medicaid dental coverage in Utah was associated with a decline in the probability of a distant-stage oral cancer diagnosis (DD = -0.046; ci = -0.078, -0.014). The Triple Difference result was similar (DDD = -0.074; ci = -0.107, -0.041). The heterogeneity analysis reveals that the association between adding Medicaid dental coverage and distant stage diagnoses are small and insignificant in Utah counties without dentists. However, the interaction term between adding dental coverage and dentists per capita yields a significant decline for every additional dentist in the county (DDD*Dentists/100,000Capita. = -.001 ci = [-0.002, -0.0009]. Putting these results into context, the marginal effect of adding dental coverage in counties with 80 dentists/100,000 is associated with a 0.104-percentage point decline in distant diagnoses: a 20% decline relative to pre-2010 rates. Conclusions: Access to dental services could improve outcomes for low-income oral cancer patients, but the persistent barriers to dental care may perpetuate disparities.