The purpose of this study is to examine the effect of Medicare HMO enrollment on the mortality and cost of care for Medicare beneficiaries with diabetes between 1994 and 1998. There are two major public policy concerns regarding individuals with chronic conditions that enroll in Medicare HMOs: Does Medicare contain costs by encouraging people with chronic diseases, such as people with diabetes, to join TEFRA-risk HMOs? and Are the quality of care and health outcomes provided to these enrollees comparable to those in the fee-for-service sector? This study will provide insight into both of these questions using Medicare administrative data from 1992 to 1998. This proposal will extend recent work by Dowd, et al., (1998) and Maciejewski, et al. (2001) looking at biased selection of the general Medicare population into TEFRA-risk HMOs. It will use a unique dataset (the National Medicare Diabetes Cohort), which contains 2.5 million elderly Medicare beneficiaries with diagnosed diabetes mellitus in fee-for-service plans in 1994. The following specific research questions will be addressed: 1) Do healthier beneficiaries with diabetes systematically enroll in Medicare HMOs? 2) Do HMO enrollees with diabetes have different five-year survival rates than Medicare beneficiaries with diabetes who remain in the fee-for-service (FFS) sector? 3) Do unhealthier beneficiaries with diabetes systematically disenroll from Medicare HMOs? 4) Do HMO disenrollees with diabetes have different FFS costs than Medicare beneficiaries who remain in the FFS sector? The careful analysis of enrollment and disenrollment patterns of this chronically ill population, combined with the mortality and cost analyses, will provide insight into the advantages and disadvantages of enrollment in Medicare HMOs. Analyses will be conducted on a cohort of beneficiaries with diabetes in thc fee-for-service sector in 1992-1993, so results are not generalizable to people with diabetes who enrolled in Medicare HMOs prior to 1994.
