Date of Award

Winter 2007

Document Type


Degree Name

Doctor of Philosophy (PhD)


Health Services Research

Committee Director

James Alan Neff

Committee Member

Stacey B. Plichta

Committee Member

Karen Karlowicz

Committee Member

Bonnie L. Lind


The purpose of this study is to determine the extent to which the Andersen and Aday Model of Health Services Use to predict cross-sectional and longitudinal differences in the caregiving source among stroke survivors (n=477) over a six-year period. This study involves a secondary analysis of a longitudinal panel study focusing upon three waves of interview data from the HAS/AHEAD study spanning 1998 through 2002. Multinomial logistic regression was utilized to examine caregiving source cross-sectionally at each wave. Clustered multinomial logistic regression was used to examine relationships between predictor variables in the model and caregiving source across waves while holding time constant. The results of this study indicate that cross-sectionally and longitudinally the need domain accounted for the largest proportion of the explained variance. Within the need domain the most consistent predictors of caregiving were the number of activities of daily living and instrumental activities of daily living impairments. The predisposing domain became less significant once enabling and need variables were entered in the cross-sectional models. In the longitudinal models, all of the domains in the Andersen-May Model were significantly related to caregiving type. These findings provide information about the caregiving situation post stroke, but also post rehabilitation, which is often the last contact with formal care services. The results of this study have two health policy implications for survivors of stroke. First, the limited number of transitions between different types of caregiving services indicates that there is likely be a fairly large length of commitment for informal caregivers. Second, discharge planning in the hospital or acute rehabilitation provides a 'teachable' moment or the key moment for intervention where realistic options for long-term care could be discussed and planned knowing that whatever is chosen will be the arrangement for the caregiver and care receiver for a considerable amount of time.

Beyond individual and provider implications these findings pose serious questions and opportunities for national and state long-term care policy. Understanding the dynamic process of caregiving will inform federal and state policymakers on the type and amount of care desired by older Americans. Ultimately the question is not what type of care is used, but how will the cost of care be distributed and in the long run how can the federal government in partnership with the states and individuals plan for this type of care while creating a sustainable system. Studies about the continuum of care, like this one, provide details about long-term care such as the use of formal and informal caregiving that can be utilized by policymakers to design systems to enhance both formal and informal caregiving.