STRUCTURING HOSPITAL MARKETS--EFFECTS ON SERVICE/COSTS

Harvard University (Medical School), Boston MA

Investigators

Linked publications & trials

Abstract

Description: This proposed project will examine the formation and performance of a network model to provide health care. Here the plan contracts with a subset of the providers in the market. A brief version of the underlying economic model is in the initial part of the proposal, and a more complete version in Appendix 2-1. Rather than relying on provider incentives or UR, the plan grants access to its enrollees on terms more favorable than could be obtained otherwise. The model will be implemented in both a descriptive and analytical manner using data on hospital use for ten years in Massachusetts. In addition to a description of the networks, they will examine the extent that the plans redirect patients to hospitals relative to what would have occurred in the absence of such incentives, based on the geographic admission patterns of non-managed care plans. Apparently the state files have data on the system type and the actual plan's identity, for a total of about 15 plans during the period. At a later date in the project they will add national and California data. It is not clear what level of plan or system type detail is available on those files. They will also examine the cost and length of stay as a function of year, plan, MDC and some severity measures. Apparently some of this will be averaged at the hospital level and some will be patient specific, using log regression models. They also will test the predictions of their theory for an optimal network. The description of the approach is too terse to ascertain what exactly they will do. Finally, they will extend the approach they develop to both data from California and from a national data source.