ABSTRACT
OBJECTIVES: To characterize the time pattern of nursing home-to-nursing home transfers and assess which resident characteristics are associated with transfers. METHODS: Minimum Data Set assessments of all Maine and New York nursing home residents were obtained for 1994-1996. The hazard rate for nursing home transfers was estimated by nonparametric statistical techniques, censored at loss to follow-up. Comparisons of resident characteristics were made between those who transferred and those who stayed at their initial facility. RESULTS: Residents of Maine nursing homes were considerably more likely to transfer than were New York residents. Transfer rates declined during the first 2 years after admission and remained stable thereafter. Correlates of transfers were similar across states. Residents who transferred were more likely to be male, to be married, to be younger, to have better cognitive and physical health, to have Medicare or private payment sources (vs. Medicaid), and to have pressure ulcers. Rural location did not affect the likelihood of transfer. CONCLUSIONS: This study provides the most detailed information to date on the prevalence, timing, and correlates of nursing home transfers. These transitions occur most frequently early in the stay but continue at a lower rate even among long stayers. This information is useful for understanding lifetime dynamics of long-term-care utilization. Several barriers to mobility appear to be present (eg, less generous payment source, health limitations, and absence of a spouse). The higher transfer rates observed in Maine might imply that institutional or other factors limit the mobility of New York residents.
