Trends in Alcohol Treatment Use from 1999-2013: Differential Effects of State Parity Laws on Racial/Ethnic Groups?

Background: This study examines alcohol treatment utilization during a 14-year period spanning the Great Recession and the passage of the Mental Health Parity and Addiction Equity Act (MHPAEA), from 1999-2013. Given prior evidence of racial/ethnic disparities in treatment use, we assessed racial/ethnic differences in treatment utilization trends during this period. Our analysis took into account state heterogeneity in the existence and strength of parity laws for alcohol treatment services prior to 2009.

Methods: Data on alcohol treatment admissions, population size, and state parity laws were merged from SAMHSA’s Treatment Episodes Data-Admissions (TEDS), US Census, and NIAAA’s Alcohol Policy Information System (APIS) into a state- and year-level database from 1999 to 2013. The study outcome was the treatment admissions rate per 10,000 state residents overall, and White, Black and Latino residents separately. Based on APIS data, states were categorized as strong or weak according to their number of mandated parity requirements (0-4) related to deductibles, copays, and service and financial limits. State- and year-specific unemployment, poverty and uninsured rates were included as covariates. Generalized Estimating Equations (GEE) by state-year were used to model treatment rates over time, and to examine racial/ethnic differences.

Results: We found a significant curvilinear (cubic) trend in US treatment utilization, with rates decreasing in the early 2000s, increasing in the mid-2000s, and decreasing again starting 2009. Accounting for state parity laws, we found that strong parity states had a significant increase in alcohol treatment rates post-2009, while weak parity states showed little change during the latter period. Race-stratified models indicated that the treatment rate increase in strong parity states post-2009 was most pronounced for Latinos. Adjustment for state poverty and particularly uninsured rates (but not state unemployment) further increased the Latino treatment rate post-2009 in strong parity states, while eliminating the marginal increase in treatment use among Whites and Blacks in these states.

Conclusions: Latinos appeared to show the largest increase in treatment use post-2009, a period encompassing both a severe economic recession and the implementation of the MHPAEA. Results suggest that lack of insurance and poverty suppress treatment utilization, even in states with strong parity laws. Additional, planned analyses will aim to better understand these preliminary findings of differential treatment effects across racial/ethnic groups.