The benefits of transferring to another nursing home are potentially large for residents whose care needs have changed or whose initial placement was Suboptimal. Knowledge of transfer rates is also necessary to defining episodes of long-term care. The literature contains few estimates of transfer prevalence and the lack of large data sets tracking residents across facilities has precluded analyses of the timing or predictors of transfers. This study's goals are the development of measures of both the prevalence and timing of nursing home transfers and the identification of resident and facility characteristics associated with transferring. The analyses combine previously unavailable resident level data (using the Minimum Data Set) with facility information and county level characteristics to examine transfer rates and their predictors for six states from 1994 through 1996. Preliminary estimates indicate that transfer rates during the first six months of the study period were 3.35 percent in New York and 7.28 percent in Maine. That estimate includes both prevalent and incident patients, and the prevalence of transfers will be higher for patients who are observed in the first six months of their nursing home stay. The transfer rate declined the longer a resident was under observation, such that between months 24 and 30 of the study, 1.6 percent of New York and 3.5 percent of Maine residents transferred. The underlying behavior model posits that the likelihood of a transfer increases with factors that contribute to higher expected benefits and lower explicit and implicit costs of transferring. Factors contributing to the benefits and costs of transferring include resident's sociodemographics, payment source, levels and changes in functional status, involvement in social activities, family involvement, facility ownership and size, services provided, and quality of care, and county market-level characteristics, including rurality and distance to other providers. This project will contribute to appropriate care as designing interventions that better match individuals and providers and ensure access to appropriate care as resident's needs evolve and to policy discussions about the availability of appropriate services to vulnerable populations and the significance of barriers to choice of provider.
