Financial Burden and Out-of-Pocket Expenditures for Mental Health across Different Socioeconomic Groups

Background: Mental health benefits have traditionally been much less generous than benefits for physical health care, with separate deductibles, higher copayments or coinsurance, and lower limits on covered services, a trend that continues despite a recent wave of 'parity' legislation. In spite of the current policy debates on mental health insurance reforms, little is known about the burden of mental health out-of-pocket expenditures.

Aims: This study examines differences in out-of-pocket expenditures and their burden across different populations, stratified by insurance status, age, ethnicity, and socioeconomic groups.

Methods: This study uses the 1998 HealthCare for Communities household survey, the latest national survey data that are currently available, to measure the burden of out-of-pocket mental health expenditures. We use several measures of burden such as total out-of-pocket expenditures, their share of total treatment costs, and their share of family income. To address the methodological issues that arise in the calculation of the relative measures of burden (e.g. outliers, measurement error, systematic underreporting) we consider three different approaches that have been suggested in the literature and discuss their relative advantages given the type of data typically available.

Results: Although there is a common perception that out-of-pocket expenditures for mental health services represent a significant burden for service users, the estimates suggest that this is not the case. In fact, across the three measures of out-of-pocket expenditures as a share of income the estimates are under 10 percent for most groups. However, there is some variation in burden across groups with people who are older, uninsured, or minority spending a larger share of their income out-of-pocket. Since many insurance plans have limits on the number of visits covered and on the total amount that the insurer will pay for mental health services, the share of total mental health expenditures that are paid by individuals is another important measure of the burden faced by people with mental health service needs. We estimate that the mean out-of-pocket share of total expenditures for the group as a whole is 25 percent. In addition, we find that the burden varies across groups with older, more educated, or privately insured individuals paying a larger share of expenditures out-of-pocket.

Discussion: Although the overall picture regarding the burden of out-of-pocket costs relative to income is encouraging, it is also important to keep in mind that individuals make treatment decisions based on their available income. The fact that the burden of actual out-of-pocket payments is relatively low may also reflect decisions to forgo potentially valuable care. Nevertheless, the results for mental health do not suggest that out-of-pocket costs are currently a major burden for most users. This situation may reflect a major change from the past given the recent shifts towards managed care; however, there are no comparable data available to test this hypothesis empirically.

Implications for Health Policy Formulation and Further research: It may be tempting to attribute the low estimates of out-of-pocket expenditures as a share of income in this paper to recent parity legislation. However, recent research shows that parity legislation has not led to significant changes in benefit design. In fact the high ratio of out-of-pocket payments relative to total mental health care expenditures presented in this paper are consistent with a limited role of parity legislation. Another possible explanation for the observed results is the growth of managed care and the shift in treatment style towards greater use of medications, which are comprehensively covered in most private insurance plans, have reduced total treatment costs and consequently the size of out-of-pocket payments.

Reprinted with permission from The Journal of Mental Health Policy and Economics, Vol. 4, 2001, pp. 141-150. Copyright © 2001 ICMPE.