Neighborhood- and State-Level Characteristics Associated with Racial/Ethnic Disparities in Mental Health Care

Racial/ethnic disparities in mental health care exceed disparities in many other areas of health care services.  For example, Blacks and Latinos access mental health care at half the rate of non-Latino Whites after adjustment for mental health status (Cook, McGuire, and Miranda 2007). These disparities in service use likely contribute to the greater persistence, severity, and disease burden of mental disorders among Blacks and Latinos (Alegria et al. 2008; Breslau et al. 2005; U.S. Department of Health and Human Services 2001; Wells et al. 2001; Williams et al. 2007), but few policy handles have been identified to reduce these disparities.

We seek to empirically identify potential policy levers for reducing mental health care disparities in two ways.  First, we focus on episodes of care, as opposed to cross-sectional group averages, to better target analyses of the underlying care-seeking behaviors that drive disparities.  For example, using cross-sectional data, it is difficult to know the extent to which racial/ethnic disparities in any use of services are due to whites’ greater initiation of care or to whites’ longer treatments. The policy implications differ dramatically depending on the answer, with the former calling for improvements in access to care while the latter calls for improvements in retention once in care for minorities. Second, we focus on potential mechanisms operating at multiple levels, including 1) individual level characteristics (health insurance, language proficiency, citizenship status, and patient perception/stigma of mental health care); 2) area-level characteristics (e.g., geographic differences in insurance status, poverty, provider supply, and population density of racial/ethnic minorities); and 3) state-level policies (e.g., Medicaid program eligibility and mental health care reimbursement policies).

We link two-year panel data of white, black, and Latino adults from panels 9-14 (2004-2010) of the Medical Expenditure Panel Survey (MEPS) to area-level indicators from the Area Resource File, the Kaiser Family Foundation Medicaid Benefits Online Database, and census tract and block group level information from the 2005-2010 American Community Survey.  We use multi-level linear regression modeling to provide a comprehensive look at the interrelation of geographic and racial/ethnic disparities, and use cross-level interaction terms to identify how significant area-level correlates of mental health care vary by race/ethnicity. We use these models to assess geographic and racial/ethnic disparities in initiation and intensity of episodes of mental health treatment for adults with psychiatric illness.

We have two research objectives: 1) Understanding the contribution of geographical differences to overall racial/ethnic disparities in access to mental health care; and 2) identifying whether state and local-area correlates of mental health care vary by racial/ethnic group. The first objective provides policymakers with a quick understanding of how much geography matters in mental health care disparities and offers preliminary guidance on whether it is more optimal to allocate resources to reducing geographic differences or racial/ethnic disparities, or a combination of both.  The second objective assists policymakers in identifying specific area-level policy levers for reducing disparities and could potentially reveal how area-level policies targeting the general population have unintended negative consequences on racial/ethnic disparities.