Mental health and substance abuse parity: a case study of Ohio's state employee program

BACKGROUND: In the United States, insurance benefits for treating alcohol, drug abuse and mental health (ADM) problems have been much more limited than medical care benefits. To change that situation, more than 30 states were considering legislation that requires equal benefits for ADM and medical care ("parity") in the past year. Uncertainty about the cost consequences of such proposed legislation remains a major stumbling block. There has been no information about the actual experience of implementing parity benefits under managed care or the effects on access to care and utilization. AIMS OF THE STUDY: Document the experience of the State of Ohio with adopting full parity for ADM care for its state employee program under managed care. Ohio provides an unusually long time series with seven years of managed behavioral health benefits, which allows us to study inflationary trends in a plan with unlimited ADM benefits. METHODS: Primarily a case study, we describe the implementation of the program and track utilization, and costs of ADM care from 1989 to 1997. We use a variety of administrative and claims data and reports provided by United Behavioral Health and the state of Ohio. The analysis of the utilization and cost effect of parity and managed care is pre-post, with a multiyear follow-up period. RESULTS: The switch from unmanaged indemnity care to managed carve-out care was followed by a 75% drop in inpatient days and a 40% drop in outpatient visits per 1000 members, despite the simultaneous increase in benefits. The subsequent years saw a continuous decline in inpatient days and an increased use of intermediate services, such as residential care and intensive outpatient care. The number of outpatient visits stabilized in the range of 500-550 visits per 1000. There was no indication that costs started to increase during the study period; instead, costs continued to decline. A somewhat different picture emerges when comparing utilization under HMOs with utilization under a carve-out with expanded benefits. In that case, the expansion of benefits led to a significant jump in outpatient utilization and intermediate services, while there was a small decrease in inpatient days. Insurance payments in 1996/1997 were almost identical to the estimated costs under HMOs in 1993. CONCLUSIONS: In contrast to the emerging inflation anxiety regarding overall health care costs, managed care can provide long-run cost containment for ADM care even when patient copayments are reduced and coverage limits are lifted. This may differentiate ADM care from medical care and reasons for this difference include the state of management techniques (more advanced for ADM care), complexity of treatments (much higher technology utilization in medical care) and demographic factors (medical, but not behavioral health, costs increase as the population ages). IMPLICATIONS FOR HEALTH POLICY: The experience of the state of Ohio demonstrates that parity level benefits for ADM care are affordable under managed care. It suggests that the concerns about costs that have stymied ADM policy proposals are unfounded, as long as one is willing to accept managed care. IMPLICATIONS FOR RESEARCH: The continuing decline in costs raises concerns that levels of care may become insufficient. While concerns about costs being too high dominate the policy hurdle for parity legislation at this moment, the next step in research is to address quality of care or health outcomes, areas about which even less is known than about costs.