The proposed project is a four-year longitudinal study of race and residence differences in patterns of formal and informal long-term care. There are two specific aims: (1) to determine the individual, familial, and community context variables that explain race and residence differences in the long-term care arrangements of elders; and (2) to describe patterns of long-term care as they evolve over time among both black elders and geographically isolated elders and to identify those factors that predict changes in long-term care arrangements. To achieve these goals, a multistage disproportionate sampling approach will be used to identify 1,200 elders residing in four north Florida counties. Telephone screenings will be used to stratify the sample on race, residence, and level of ADL and IADL impairment. Data will be collected using four methods: two directed at the individual elder and their family support network and two others directed at establishing the community context in which the elder lives and functions. To acquire information about the individual elder and their family support network, a series of telephone interviews will be completed with the full sample at six points in time (approximately 6-months apart). To complement this large sample data collection, a subsample of 100 elders considered to be at high-risk of entering a nursing home will be selected for more intensive monitoring. With this smaller sample, a series of in-person interviews will be conducted during years 02 and 03 of the project. To acquire information about the community context in which the elder resides, focus groups will be held to collect information about community perceptions of the need for, and accessibility of, both formal and informal sources of long-term care. Finally, for each community in which a respondent resides, a survey of the formal long-term care resources that are available in that locale will be conducted. A modification and expansion of the Andersen model of service utilization will be used as the conceptual framework. Indicators of predisposing, enabling, and need factors will be measured for the individual elder, their family support network, and the community in which they live. The dependent variable, long-term care arrangements, will be measured in two ways: the composition of the helping network, and the intensity of the help provided. The analytical strategy will include descriptive log-linear analysis and two general methods appropriate for models with multiple predictors (i.e., multinomial logistic regression and time-based analysis using a proportional hazards model). In addition, techniques appropriate for ethnographic analysis will be applied to the information from the in-person interviews and the focus groups.
Showing the most recent 10 out of 13 publications
