This study will use a natural experiment to assess whether community-based long-term care is effective in preventing institutionalization, preserving functional health, and reducing hospitalizations. The research will use retrospective and longitudinal data on panels of geriatric enrollees in two comparable community-based long-term care programs. The termination of the program at one site offers a unique opportunity to study outcomes for at-risk elderly associated with community-based care. The social health maintenance organization (S/MHO) at HealthPartners in Minnesota was terminated in December 1994; the S/HMO at Kaiser-Permanente Northwest in Oregon continues to operate. A S/HMO provides community based long-term care services, including care coordination, home health care, homemaking and other supplemental services for at-risk elderly, within an integrated model of health care serving a broad-based population of Medicare HMO enrollees. The S/HMO demonstration which was established in 1985, was designed to address both chronic and acute health care needs in a geriatric population, to reduce the risk of institutionalization, to optimize health and functional status, and to provide a cost-effective mix of health care services. The primary outcome of interest is time to permanent nursing home placement. As secondary outcomes, we will track time-to-death, adverse change in functional health status, and hospitalization. The proposed study will use a non-equivalent group design without controlled selection. By studying the ending of a program, our research design eliminates many threats to validity inherent in other pre- post quasi-experimental designs, including history, selection, and instrumentation. The research population will be comprised of S/HMO cohorts at both sites, along with matched samples of non-S/HMO geriatric enrollees. We will use survival analysis techniques to compare the adjusted time until permanent nursing home admission. The secondary analyses will use distribution-appropriate analysis of co-variance (ANCOVA).
| Fischer, Lucy Rose; Green, Carla A; Goodman, Michael J et al. (2003) Community-based care and risk of nursing home placement. Med Care 41:1407-16 |
