Method: Caregivers (N = 637) were from the Resources for Enhancing Alzheimer’s Caregivers Health (REACH II) study. Service use data were from the 11-item formal care and services measures assessing whether a service was used during the past month. Among these items, using homemakers, home health aids, home delivered meals, transportation, senior daycare, or support group was considered as discretionary services. Services that are largely unavoidable were considered as non-discretionary, such as having nurse visit home, visiting physician, visiting counselor or psychologist, visiting emergency rooms, and receiving inpatient care. Having at least one service use was recoded as service utilization (1 = yes, 0 = no). Frequency of each service was also asked with larger number indicating more frequent use. Binary logistic regression was used to predict whether participants had service utilization and linear regression was conducted to predict frequency of service utilization.
Results: Results of binary logistic regression showed that predisposing factor of being educated (OR = 1.10, p < .05), enabling factors of full employment (OR = 0.45, p < .05) and social network (OR = 1.09, p < .05), needs factors of caregiving burden (OR = 1.03, p < .05) and caregivers’ functional symptoms (OR = 1.08, p < .05) were significantly associated with discretionary service utilization. For non-discretionary service utilization, enabling factor of help from other family members (OR = 0 .81, p < .01), and need factors of caregiving burden (OR = 1.04, p < .01) and caregivers’ functional symptoms (OR = 1.19, p < .001) were significant factors. For frequency of service utilization, need factors of care recipients’ functional limitation and behavioral problems were significant factors of both discretionary and non-discretionary service use. A predisposing factor of being Africa America (β = 0.17, p < .001) was significantly associated with discretionary service use. In addition, our study found that being Hispanic or African American moderated the association between some of the enabling or needs factors and service utilization.
Conclusion and implications: The findings support our hypotheses that predisposing, enabling, and need factors were better predictors of discretionary than nondiscretionary service use, and non-discretionary service utilization was largely determined by need factors. For interventions or programs that aim to help dementia caregivers by providing services, we need to pay attention to those factors that limit their access to service utilization. More attention is needed to several enabling and needs factor among Hispanic and African American dementia caregivers, including social network, social support, and caregiving burden, to encourage their service utilization.