the selection incentive. However, risk adjustment systems in general explain only about 7 percent of variations in spending. Ettner and colleagues compared the performance of several risk adjustment systems for M/SU services in private insurance and Medicaid. They found that no existing risk adjustment classification system displayed strong predictive ability for spending on M/SU care (Ettner, 2001; Ettner et al., 1998). Thus existing risk adjustment approaches appear to be limited in their ability to stem selection-driven distortions in quality affecting those at risk for using M/SU treatment services. Organizational strategies can also be used to address selection incentives.

Direct Public Purchase of Behavioral Carve-out Services in Medicaid

Currently, approximately 16 states and territories (or substate units) contract directly with carve-out vendors for some M/SU health care services. As discussed earlier, this approach to the purchase of coverage and managed care for M/SU treatment offers the advantage of attenuating distortions in quality of care that may result from selection-related incentives. Yet the procurement process for these contracts is frequently structured so that performance with respect to contract costs may be overemphasized relative to performance on quality of care. If costs and quality are positively correlated, the implication of this procurement structure is that there will be a tendency to choose lower-cost, lower-quality contract proposals. One is reminded of the remarks of astronaut Allen Shepard, who just prior to being launched into space expressed concern that he would be aboard the spacecraft that had been built by the lowest-cost bidder.

One approach to addressing this concern would be to reorient the procurement process toward a two-step procedure. In the first step, states would engage in a rate-finding process. They would collect information on the outcomes of contracting processes in other states and in the private sector and interview potential vendors concerning the costs of their services. They would then analyze these data to determine the “reasonable” costs of providing MBHO services to the state’s Medicaid population. A level of reasonable costs would be selected and announced as the contract “price.” In the second stage of the procurement, interested vendors would submit competing proposals. The proposals would be judged solely on the proposed quality of services, access to care, and ability to coordinate with other systems of care (general medical care, social services, housing, and income support). With the procurement process structured in this way, competition would be reoriented toward offering innovative approaches to improving the quality of clinical services and developing mechanisms for coordination of care with relevant components of the health and human service sectors that offer complementary services. The incentive would be



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement