Influence of Comorbidities on Mental Health Care and Mental Health Care Disparities

This paper assesses the role of comorbidities in recognizing need for mental health care and how this influences mental health care disparities. We test a crowd-out effect (comorbidities decrease the probability that a mental health disorder will be recognized and treated during a visit) and an exposure effect (comorbidities cause more PCP visits, increasing the probability that a mental health disorder will be recognized).

Data: Non-Latino White, Latino, and African-American adults from panels 5-11 (years 2000-2007) of the Medical Expenditure Panel Survey (MEPS) in the lowest tertile of mental health in Time 2 and without mental health care in Time 1 (n=8,162). The MEPS contains the dates of PCP and mental health care visits, the round of prescription drug use, and measures of socioeconomic status, demographics, and mental and physical health status.

Methods: We split longitudinal data into Time 1 and Time 2, assessing the probability of a mental health care event after a primary care provider (PCP) visit in Time 2 among those with no mental health care in Time 1.

We test the crowd-out hypothesis by assessing whether probability of treatment following the first primary care visit in Time 2 is lower among patients with comorbidities. To test the exposure hypothesis, we estimate the probability of treatment as a function of number of Time 2 PCP visits. To assess the influence of comorbidity on disparities, we compare disparity estimates with and without adjustment for comorbid physical conditions.

Principal Findings: Persons with comorbidities were less likely to be recognized and treated for mental health care around the first two PCP visits, confirming a crowdout effect. We also find strong evidence for an exposure effect based on the positive impact of predicted Time 2 PCP visits on recognition and treatment. Regarding comorbidities as affecting treatment through crowdout and exposure decreases measured disparities, though they remain positive and significant.

Conclusions: Comorbidities reduce the probability of recognition in the short run because of time constraints, but then increase recognition because of greater exposure over time. Conceptualization of comorbidities in a disparities study will influence both the magnitude of measured disparities and the understanding of factors responsible. Identifying disparities in recognition at both initial and subsequent visits reveals the need for improving communication across all PCP visits. The issues and methods we explore have relevance to the study of other chronic and acute conditions for which comorbidities play a role in affecting treatment.