It is not uncommon to see cost savings of 30 percent of mental health costs in connection with the introduction of managed care. Providers respond to managed care by doing less in ways that cannot be explained by utilization review or supply-side price incentives. It appears that the opportunity to be included within the managed care works a powerful incentive. We define a """"""""network"""""""" as a subset of the providers in a market selling services to a managed care plan on favorable terms (to the plan). In exchange, the plan directs business to the providers in the networks. We propose to study the effect of networks on the quantity of care supplied to patients, using data spanning three rounds of network creation within a state-level market for hospital services. We will also study the problem of network design. We will develop and test a theory of an """"""""optimal network"""""""" that balances the gain in bargaining power and information to the payer against the loss in choice for consumers. The market for hospital care for mental health and substance abuse in Massachusetts is an excellent site for the study of network effects. Hospitals have undergone three rounds of network formation in connection with carve-out plans, two carve outs in Medicaid and one in the largest private employer in the state, the Group Insurance Commission. Furthermore, hospitals have experienced the growth of HMOs using networks of hospitals to care for their enrollees. This setting thus offers an opportunity to test the impacts of network creation on patient care, health plan costs, and the provider industry, and to study the problem of design of the """"""""optimal"""""""" network. The proposal has four specific aims: 1) Describe the changes in average discharge cost and length of stay (LOS) at the hospital level in Massachusetts in association with the introduction of three rounds of networks, and the growth of HMOs. Compare these changes to trends. Describe the relation between hospital behavior and payer, hospital characteristics, and market-level information. 2) Estimate the magnitude of the """"""""network"""""""" effect by testing a model of hospital response in cost and LOS per discharge to network incentives. Test subsidiary hypothesis about hospital response to managed care. 3) Develop a theory of an """"""""optimal network"""""""" from the point of view of the enrollees in a managed care plan. Use this to describe how a plan should decide on the structure of the network and on the pattern of redirections in the network. 4) Test the behavior of the managed care plans against the predictions of the optimal network theory.
| Fleming, Elaine; Lien, Hsienming; Ma, Ching-To Albert et al. (2003) Managed care, networks and trends in hospital care for mental health and substance abuse treatment in Massachusetts: 1994-1999. J Ment Health Policy Econ 6:3-12 |
