Jan 1, 1997
The U.S. health care system is changing rapidly. One area of especially rapid change has been mental health care, where the market has been altered by the growth of "carve-out" managed-care plans that provide separate, specialized mental health coverage. The rapidity of this change has outpaced analysts' ability to understand its effects. Adding to the climate of change is a federal law passed in 1996 mandating that mental health benefits offered by employers and insurers be on a par with physical health benefits in dollar limits and reimbursement rates. This law recently went into effect. Employers in particular now face a great deal of uncertainty about the effects of such legislation and require accurate information about the costs of mental health care in order to make informed decisions about offering mental health benefits.
RAND conducted three studies examining issues surrounding mental health benefits under managed care. The studies focused on the costs of mental health services under managed care and the implications of the 1996 parity legislation for cost and benefit design.
One study, published in the Journal of the American Medical Association, looked at the cost implications of the 1996 parity legislation. Assumptions used in the congressional debate over parity legislation had suggested that unlimited mental health care benefits would greatly increase costs. RAND researcher Roland Sturm tested these assumptions by studying managed care plans that already implemented full parity.
The study found that the assumptions used during the parity-legislation debate had substantially overstated the actual cost of mental health services under managed care. Unlimited mental health benefits under managed care cost virtually the same as capped benefits: The average increase was about $1 per employee compared with costs under a $25,000 cap, which was a typical limit in other existing plans. As Figure 1 illustrates, costs for each benefit-user group were virtually the same under unlimited plans as they were for plans at each of the typical cap levels. A major reason for the small differential in costs is the small number of high-cost episodes in the data set examined.
The study concluded that benefit caps on mental health coverage had little effect on employers' overall health care expenses.
A second study examined how the change from fee-for-service to managed care affected costs and utilization patterns of a major West Coast employer. The work was a collaborative effort between the RAND/UCLA Center on Managed Care for Psychiatric Disorders and researchers from United Behavioral Health. The study firm adopted a carve-out benefit plan that separated mental health benefits from other health coverage at the same time that it expanded mental health benefits. The study looked at claims data for the firm from 1988 through 1996. At the time of the carve-out in 1991, approximately 180,000 individuals were covered, 72 percent by managed care programs.
The study found that, despite increasing benefits, the switch to managed care led to a substantial reduction in costs for mental health care. As shown in Figure 2, costs per covered member between 1988 and 1990—during fee-for-service coverage—were high and rising at a rate of 20 percent annually.
Following the transition in 1991, costs immediately fell dramatically (40 percent) and continued to decline slowly over the following years. Because these numbers are in nominal dollars, the constant level of mental health care costs from 1991 to 1996 implies a substantial decrease in costs when adjusted for inflation.
Results also showed that the switch to managed care did not reduce access to mental health care. In fact, the number of patients receiving any mental-health specialty care increased. The main factors underlying the cost reduction have to do with declines in inpatient care: Fewer patients were hospitalized, the average length of stay was reduced, and the cost per inpatient day fell drastically. All of these trends began before the carve-out but were accelerated by managed care.
Even though parity legislation is unlikely to have a significant effect on employers' health care costs under managed care, it will require changes in benefit plan designs. A third study, also a collaboration between the RAND/UCLA Center on Managed Care for Psychiatric Disorders and United Behavioral Health, looked at the compatibility of current benefit plans with the 1996 legislation. Using data from 4,000 firms, the study analyzed the designs of carve-out mental health benefit plans and compared them with the designs of those firms' medical benefit plans.
The analysis found that almost 90 percent of the mental health plans were inconsistent with the parity legislation and required revision. The research concluded that the restructuring required by the parity act provided a useful opportunity to make benefit designs more efficient by remedying existing flaws. Many plans were unnecessarily complex and inconsistent, often reflecting a legacy of past attempts to control costs in a fee-for-service environment. Under managed care, the need for deductibles, limits, or other demand-side cost-sharing mechanisms has probably diminished. Therefore, restructuring outdated designs could benefit both enrollees and employers.
However, better information on the influence of design mechanisms on access, intensity of care, and costs is necessary before individual employers are willing to abandon traditional cost-control mechanisms in mental health plans. Future RAND research will provide this information.