| Abstract: | Abstract This study examines the assumption that including Black-Americans with larger samples of White-Americans produce an accurate picture of the predictors of nursing home use by Black-Americans. Using the Longitudinal Study of Aging (1984-1990), Cox proportional hazard models are estimated for Black-Americans and White-Americans. This study finds that in fact the models for Black-Americans are not the same as for White-Americans. For Black-Americans, living below the poverty line, problems of functional status, age, and not having the potential for care for a few weeks from relatives (in or out of the home) are predictive of a higher risk of nursing home use while, for White-Americans, age, lower versus higher income, being female versus male, poor evaluation of self health care, poor functional status, lower levels of community activity, and living alone compared to living with a spouse or children were predictive of a higher risk of nursing home admission. The implications for policy and case management practice are discussed regarding these findings. |
