strophic Coverage Act (MCCA) of 1988, in conjunction with an administrative directive, or transmittal, from the Health Care Financing Administration (HCFA), changed the Medicare SNF benefit structure substantially. In this study, we specified a simultaneous equation system to explain the effects of the benefit changes on Medicare use. The results suggest that the two policychanges increased Medicare use in Pennsylvania SNFs substantially; however, the increase was associated with facility and case mix characteristics, which suggest that the increase was largely attributable to reclassification of current patients from other payer categories, Medicaid and self-pay, rather than new admissions. The effects of the MCCA and the HCFA transmittal on increased Medicare use were unanticipated and have important implications for the way in which subacute care is defined and financed infuture benefit discussions.

27:T41e,

As employers have turned to managed care to curtail the rising cost of health care benefits, the number of HMO enrollees has proliferated. Between 1984 and 1994, HMO enrollment incrased from approximately 15 million to over 49 million individuals. Although research has indicated that HMOs have been effective in limiting medical costs, there is mixed evidence in the literature on how they achieve these savings. This article uses data from the 1987 National Medical Expenditure Survey to examine one hypothesis for these patterns: that HMOs enroll a healthier population than fee-for-service plans. To test this hypothesis we examine HMO andjee-for-evice enrollees with respect to sxioeconomic variables such as age, race, sex, income, education, health status, and location. Our results indicate that HMOs tend to enroll a younger but not much healthier population than traditional fee-for-service plans, suggesting that self-selection is not a major contriutor to HMO cost savings.

28:T412,

This article presents the results ofa study to design and assess the feasibility of conducting a national Medicaid Access Survey to generate timely, state-specific estimates of access to care for Medicaid enrollees. State-specific data on Medicaid access is especially relevant because state programs and environments vary considerably and are changing rapidly in ways that could influence access. We analyze (1) basic survey design parameters and instrument content, (2) alternative sampling approaches and their feasibility, (3) pilot test results, (4) the feasibility of using existing national surveys to generate comparison estimates for state-based surveys, and (5) estimates of the required sample size and costs for such a national Medicaid Access Survey. We conclude that a survey generating timely, state-speafic estimates of Medicaid access is both feasible and affordable if attention is paid to key design challenges while keeping objectives and design simple.

29:T836,