DataWatch Paying For Mental Health And Substance Abuse Care by Richard G. Frank, Thomas G. McGuire, Darrel A, Regier, Ronald Manderscheid, and Albert Woodward Abstract: Fifty-four billion dollars was spent on alcohol/drug abuse and mental health treatment in 1990. These expenditures were concentrated in the area of inpatient psychiatric care and on persons with severe mental health and substance abuse problems. The data on expenditure patterns for mental health and substance abuse care suggest that successful health care reform in this area must implement mechanisms for controlling inpatient utilization and managing the care of persons with the most severe disorders. President Clinton's proposal for health reform contains a plan for changing dramatically the manner in which mental health and substance abuse services are financed and delivered in the United States. The plan calls for eventual full integration of mental health/substance abuse coverage with the rest of health care. Assessing the impact of such changes is complicated and depends on critical implementation decisions to be made over the next several years during a phase-in of the president's health reform plan. Thus, we cannot yet make comprehensive projections regarding the impact of the fully integrated system. The purpose of this DataWatch is twofold: (1) to provide a framework for undertaking a quantitative impact analysis in the future; and (2) to provide data that underscore the challenges for the phase-in period of health reform. We focus our attention specifically on the challenges for payment policy and public financing of mental health/substance abuse care within a reformed health care system. Financing Mental Health/Substance Abuse Care: The Baseline We organize the data by population segments and by the type of insurance coverage that currently is in place. Since the coverage provisions differ Richard Frank is aprofessor in the School of Public Health, The Johns Hopkins University. Thomas McGuire is a professor in the Department of Economics, Boston University. Darrel Regier is director of the Division of Epidemiology and Services Research, National Institute of Mental Health. Ronald Manderscheidis chief of the Survey and Analysis Branch, Center for Mental Health Services. Albert Woodward is acting deputy director of the Office of Applied Study, Substance Abuse and Mentol Health Services Administration.
338 HEALTH AFFAIRS Spring (I) 1994 dramatically across population groups, we expect the impact of health reform to vary also. Baseline data. The basic building block of this analysis is the estimates of Dorothy Rice and colleagues of the direct costs of alcohol/drug abuse and mental (ADM) disorders. 2 These estimates are based on utilization and expenditure data and thus exclude administrative costs. We did not count support costs and 90 percent of nursing home costs (which are those accounted for by people age sixty-five and older, who are unlikely to be included in health care reform). We also deducted from the numbers the prevention and data development set-asides in the federal ADM Block Grant. These deductions were $20.9 million for mental health and $225.6 million for substance abuse. 3 The allocations between alcohol and drug abuse were made according to the distributions reported in the Rice study. We also eliminated $43 billion in ADM costs stemming from comorbid physical conditions (for example, a portion of liver transplants can be linked to alcohol abuse). Costs associated with forensic hospital care in state mental hospitals also were deducted from the state mental health expenditures, $0.7 billion in 1990. 4 After these adjustments, ADM costs totaled $54 billion in 1990 (see Exhibit 1). We then divided the total expenditures among sources of payment. Using data from the Center for Mental Health Services (CMHS) inventory and the National Drug and Alcoholism Treatment Unit Survey (NDATUS), we allocated $2.2 billion to Medicare. Medicaid dollars were based on the high estimate of the range identified by George Wright and Jeffrey Buck, trended forward to 1990 dollars. 5 The high end was chosen because since 1984 the advent of disproportionate-share rules has increased Medicaid payments associated with ADM disorders. The Medicaid total was $9.5 billion in 1990. Medicaid dollars were allocated to ADM accord- Exhibit 1 Baseline Funding For Mental Health And Substance Abuse, 1990 Billions Of Dollars, Source of funds Private Medicaid State and local Mental health $18.8 8.1 11.7 Drug $0.9 0.3 1.1 Alcohol $2.5 1.0 3.1 $22.2 9.5 15.9 Department of Veterans Affairs Other federal Medicare 1.5 0.8 1.5 0.2 0.6 0.0 0.6 0.6 0.7 8.5 2.3 1.9 2.2 All sources 42.4 3.0 54.0 Sources: Authors' calculations based on D.P. Rice et al., The Economic Costs of Alcohol and Drug Abuse and Mental Illness: 1985; Center for Mental Health Services, unpublished data; and NDATUS data. Note: s may differ due to rounding.
DATAWATCH 339 ing to the aggregate proportions found by Rice and colleagues. Privatesource expenses consist of total spending, out-of-pocket payments, insurance expenses, and philanthropy. Some adjustment was made to the private-source expenses reported by the Rice study based on more recent data from the Substance Abuse and Mental Health Services Administration (SAMHSA). 6 The Department of Veterans Affairs (VA) share was based on 1990 data from the CMHS inventory and data on substance abuse care from the VA. State and local data also were derived from data from the CMHS inventory and NDATUS. The "other federal" category is defined as a residual of federal funds. Exhibit 1 summarizes the distribution of baseline costs by payer source. Population groups. In 1990 there were 22.1 million Medicaid eligibles and 31.2 million people age sixty-five and older who were assigned to Medicare. 7 We assume that there were thirty-five million uninsured persons, 16 percent of whom had a serious mental health/substance abuse problem. 8 The private-source population constitutes the remainder. Baseline per capita estimates. We used the four population groups defined above to estimate baseline per capita costs for each group (Exhibit 2). For the privately insured and Medicaid and Medicare population groups, we divided expenditures by the size of the group. Costs from the VA, costs of state and local governments, and other federal costs were allocated to the uninsured. 9 Observations And Lessons Observations. Exhibit 3 shows that for the entire U.S. population, we Exhibit 2 Spending Per Capita On Mental Health And Substance Abuse, 1990 Population (millions) Mental health $117 368 49 Alcohol $ 16 46 22 Drug $ 138 430 71 Private sources Medicaid Medicare 160.4 22.1 31.2 $ 5 16 0 Other and uninsured Other severely mentally ill (SMI) Other non-smi 35.0 5.6 29.4 399 123 53 575 3,205 75 All non-medicare 248.7 171 188 34 36 12 14 217 238 Sources: Authors' calculations based on D.P. Rice et al., The Economic Costs of Alcohol and Drug Abuse and Mental Illness: 1985; Center for Mental Health Services, unpublished data; and NDATUS data. Note: s may differ due to rounding.
340 HEALTH AFFAIRS Spring (I) 1994 Exhibit 3 Aggregate Distribution Of Baseline Funds For Supporting The Comprehensive Mental Health Benefit, Billions Of Dollars, 1990 And 1994 Source of funds Private insurance Medicaid Federal State Federal sources Department of Veterans Affairs Block grant Other Medicare (excluded) State and local State Local 1990 $22.2 9.5 5.3 4.2 4.2 2.3 1.0 1.0 2.2 15.9 13.0 2.9 54.0 1994 $28.0 12.0 6.7 5.3 5.3 2.9 1.2 1.3 2.8 20.1 16.5 3.6 68.2 Sources: Authors' calculations based on D.P. Rice et al., The Economic Costs of Alcohol and Drug Abuse and Mental Illness: 1985; Center for Mental Health Services, unpublished data; and NDATUS data. Note: s may differ due to rounding. spent $54 billion or $217 per capita on mental health and substance abuse care in 1990. A great deal of this total paid for inpatient treatment. The literature confirms this trend, showing that private insurance plans spend about 60 percent of their mental health/substance abuse dollars on inpatient treatment; Medicaid spent 65 percent; and state mental health agencies devoted an average of 59 percent of their budgets to hospital-based mental health care in 1990. 10 Exhibit 2 shows that mental health/substance abuse services to the uninsured accounted for roughly $20.1 billion in 1990 ($575 times thirty-five million people). A substantial portion of this (about 89 percent) represents expenditures made on behalf of persons with serious mental health/substance abuse problems. 11 A related analysis of direct costs for patients with the most severe mental illnesses in the public and private sectors estimated that $20 billion was spent in 1990 for their treatment. 12 Exhibit 3 shows that state and local governments through their direct support of mental health/substance abuse care and via contributions to the Medicaid program spent $20.1 billion in 1990. The variation in spending across the states is dramatic. The mean per capita level of state spending on mental health in 1990 was $43.11. The top five states in terms of per capita expenditures all spent more than $65 per capita. The five states with the lowest levels of expenditures each spent less than $25 per capita. Thus there is more than a twofold difference between the groups with the highest and lowest 10 percent of outlay for public
DATAWATCH 341 mental health care. In 1990 the federal government spent about $53 billion on mental health/substance abuse care in the Medicaid program, $23 billion in support of the VA mental health/substance abuse treatment program, $1.0 billion on the ADM Block Grant, and $1.0 billion on miscellaneous other programs such as Title XX. In addition, the federal government spent $2.2 billion on mental health/substance abuse treatment for Medicare beneficiaries. Challenges. Three important challenges for health care reform emerge from the data presented. The first stems from the composition of expenditures. The heavy reliance on inpatient mental health/substance abuse care suggests that successful cost control depends on mechanisms to manage and control inpatient use. Health plans that have successfully found alternatives to inpatient care and that have created financial incentives for decreased inpatient utilization have been able to pay for increased use of community-based services through savings from inpatient psychiatric care. 13 Failure to control inpatient use threatens the creation of an integrated system of mental health/substance abuse care given current budgetary politics. The second important challenge stems from the fact that individuals with severe mental illnesses and substance abuse disorders consume a large share of total mental health/substance abuse resources. We estimate that two-fifths of all mental health/substance abuse expenditures are made on behalf of roughly 2 percent of the population. Full integration of mental health/substance abuse care means developing systems within health plans that can efficiently manage the complex and extensive treatment needs of people with severe mental health/substance abuse illnesses. Moreover, it means developing a method of paying health plans that allocate dollars to mental health programs for the treatment of the severely ill and guarding against incentives to undertreat these vulnerable populations. The third challenge is to implement an integrated mental health/ substance abuse benefit within the framework of managed competition that is fair to all citizens. How can the existing variation in public financing of mental health/substance abuse care be accommodated within a national uniform benefit? A financing mechanism must be created to allocate fairly the burden of moving to an integrated mental health/substance abuse system across the entire polity. The opinions expressed in this paper are those of the authors and do not represent those of any of the institutions with which they are associated.
342 HEALTH AFFAIRS Spring (I) 1994 NOTES 1. See B.S. Arons et al, "Mental Health and Substance Abuse Coverage under Health Reform," Health Affairs (Spring 1994): 192-205. 2. The original work is D.P. Rice et al., The Economic Costs of Alcohol and Drug Abuse and Mental Illness: 1985, ADAMHA Contract 283-87-0007 (Washington: U.S. Government Printing Office, 1990). We use estimates prepared by Rice and colleagues that update their 1985 estimates to 1990. 3. These totals were the 1990 actual allocations under the block grant. 4. These data were obtained from the National Association of State Mental Health Program Directors. 5. G.E. Wright and J.A. Buck, "Medicaid Support of Alcohol, Drug Abuse, and Mental Health Services," Health Care Financing Review (Fall 1991): 117-128. 6. These data were reported in a memorandum from Albert Woodward, acting director, Substance Abuse and Mental Health Services Administration, to Richard Frank and Darrel Regier, 16 April 1993. 7. Current Population Survey, 1991; and unpublished calculations from the Congressional Budget Office. 8. G. Norquist and K. Wells, "Mental Health Needs of the Uninsured," Archives of General Psychiatry 48 (1991): 476. This number is consistent with data from the National Comorbidity Survey, which reports that approximately 23 percent of those without insurance report any mental disorder. These data were obtained via personal correspondence with Ron Kessler and are available in a draft manuscript. R. Kessler et al., "Psychiatric Disorders in the U.S.: Prevalence, Impairment, and Frequency of Treatment" (Unublished paper, Institute for Social Research, University of Michigan, April 1993). See also W.E. Narrow et al., "Use of Services by Persons with Mental and Addictive Disorders: Findings from the NIMH Epidemiological Catchment Area Program," Archives of General Psychiatry 50, no. 2 (1993): 95-107. 9. We recognize that this allocation procedure represents a substantial simplification of the complex relationships between insurance coverage and the use of public programs. 10. R.G. Frank, D.S. Salkever, and S.S. Sharfstein, "A New Look at Rising Mental Health Insurance Costs," Health Affairs (Summer 1991 ): 116-123; Wright and Buck, "Medicaid Support of Alcohol, Drug Abuse, and Mental Health Services;" and T.C. Lutterman and V.L. Hollen, "Change in State Mental Health Agency Revenues and Expenditures between Fiscal Years 1981 and 1990," in Mental Health, United States, 1992, ed. R. Manderscheid and M. Sonnenschein (Washington: U.S. GPO, 1992). 11. For a definition of serious mental disorders, see Norquist and Wells, "Mental Health Needs of the Uninsured." 12. National Advisory Mental Health Council, "Health Care Reform for Americans with Severe Mental Illness: Report of the National Advisory Mental Health Council," American Journal of Psychiatry 150 (1993): 1447-1465. 13. R. Coulam and J. Smith, Evaluation of the CPA-Norfolk Demonstration: Final Report, U.S. Department of Defense Contract MDA-907-87-C-0003 (Cambridge, Mass.: Abt Associates, 1990).