4034.0: Tuesday, October 23, 2001 - 8:30 AM

Abstract #26202

A comparison of probability of Medicare skilled nursing facility use in a Social HMO and a Medicare risk (TEFRA) HMO

Bruce Friedman, PhD1, Bryan E. Dowd, PhD2, and Robert L. Kane, MD2. (1) Department of Community and Preventive Medicine, University of Rochester, 601 Elmwood Avenue, Box 644, Rochester, NY 14642, (716) 273-2618, Bruce_Friedman@urmc.rochester.edu, (2) Division of Health Services Research and Policy, University of Minnesota, 420 Delaware Street S.E., Mayo Mail Code 729, Minneapolis, MN 55455

Objective. To estimate, separately for the functionally impaired and non-impaired, the effect on probability of Medicare skilled nursing facility (SNF) use of being enrolled in a Social HMO rather than in a Medicare risk (TEFRA) HMO. Methods. The study sample consisted of 5,110 enrollees in a TEFRA HMO and 2,446 members of a Social HMO sponsored by the same health insurance/delivery organization. Separate analyses were carried out for functionally impaired and non-impaired enrollees using two definitions of impairment/non-impairment. Bivariate probit regression with correlated error terms was used to test for selection bias. Because evidence of sample selection bias was detected in only one of 12 equations, probability of Medicare SNF use was estimated using “ordinary” probit. Results. Probability of Medicare SNF use was significantly higher for the Social HMO irrespective of whether enrollees were functionally impaired or not. Conditional on some hospital use during the study year, no difference was found in probability of use between the two HMOs for both functionally impaired and intact enrollees. Conditional on no hospital use, among functionally impaired enrollees probability of Medicare SNF use was significantly higher for the Social HMO. This was also observed for one of the two definitions of functional non-impairment. Conclusions. The addition of case management and a limited amount of long term care to a Medicare HMO benefits package was associated with higher probability of Medicare SNF use among both functionally impaired and non-impaired enrollees. This was due to higher likelihood of use among those who were not hospitalized.

Learning Objectives: To gain understanding of the impact on probability of Medicare SNF use of adding case management/long term care to a Medicare HMO benefits package.

Keywords: Medicare, Long-Term Care

Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: Group Health (Minneapolis, MN)
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.

The 129th Annual Meeting of APHA