Providing managed care options for a large population: Evaluating the CHAMPUS reform initiative

Objectives: To evaluate a managed care demonstration project in CHAMPUS (Ci vilian Health and Medical Program of the Uniformed Services), the insurance program covering physical and mental health care services for the dependen ts of active duty military personnel, military retirees, and the retirees' dependents, The demonstration project added a health maintenance organizati on (HMO) option and a preferred provider organization (PPO) option to the s tandard CHAMPUS coverage and allowed beneficiaries to select the coverage o ption they preferred. Data Sources: Utilization, costs, access, and benefic iary satisfaction were measured using data from CHAMPUS claims records, the Defense Enrollment Eligibility Reporting System, the demonstration project contractor's HMO enrollment file, the contractor's list of network hospita ls, and two surveys of CHAMPUS beneficiaries. Study Design: Changes in util ization at 11 demonstration sites were compared with changes in utilization at 11 matched control sites. The effect of the demonstration project on co sts was evaluated by estimating the costs for the demonstration sites both with and without the managed care options based on data from the control si tes. Access to care and satisfaction were compared between the demonstratio n sites and control sites based on beneficiary surveys. Data Collection: Al l claims in both demonstration and control sites were used in estimating ut ilization changes. Two mailed surveys were sent to a randomly selected samp le of active duty and retiree households with CHAMPUS beneficiaries; the sa mple was stratified by beneficiary type (active duty or retiree) and site. Principle Findings: Overall utilization in the CHAMPUS system decreased at the demonstration sites but stayed approximately the same at the control si tes. Utilization among the enrollees in the HMO demonstration option, howev er, increased dramatically. Patient access to care and satisfaction general ly remained at the same levels at both demonstration and control sites, but enrollees in the HMO option reported higher satisfaction. Costs to the gov ernment at the demonstration areas, based on regression estimates from the control sites, were about the same or slightly higher than what they would have been under the standard CHAMPUS system. Conclusions: Managed care plan s for large government-sponsored insurance programs can reduce utilization and maintain patient access and satisfaction, Careful structuring of such p lans is needed, however, if they are to reduce costs.