Purpose: Though recent studies have examined the role of caregivers' attitudes, beliefs and perceptions in relation to service use outcomes, they have restricted their focus to the individual caregiver as the unit of analysis. Examining the service use attitudes of caregiver-care recipient couples at the dyadic level of analysis remains an under-explored and potentially useful line of inquiry. Towards this goal, this exploratory study examined gender differences and dyadic congruence on the community service use attitudes of 30 older spousal caregiver-care recipient dyads.

Methods: Employing Stommel, Collins, Given, & Given's (1999) framework for examining family caregivers' community service use attitudes derived from Ajzen and Fishbein's theory of reasoned action, this cross-sectional survey operationalized service use attitudes utilizing seven attitude subscales adapted from the Community Service Attitude Inventory (Stommel et al., 1999). The purposive sample consisted of community-dwelling caregiver-care receiver older married couples recruited from health and social service sites in New York City. To measure attitudinal congruence within spousal dyads on 29 items from the Community Service Attitude Inventory, intraclass correlation coefficients (ICCs) were generated for each dyad to create a composite Index of Dyadic Similarity for the total sample.

Results: For the total sample (N=60), regardless of role, older women reported a higher mean score in Wait-And-See Attitude compared to older men. Older men reported a higher level of Preference for Informal Care compared to older women. In the caregiver sub-sample (n=30), female spousal caregivers reported a higher level of Confidence in Service System compared to male spousal caregivers. Male spousal caregivers reported a higher Worry and Fear Attitude compared to female spousal caregivers. In the care recipient sub-sample (n=30), female spousal care recipients reported a higher Wait-And-See Attitude compared to male spousal care recipients. Finally, as a measure of attitudinal congruence within dyads, the composite Index of Dyadic Similarity generated a mean of .59 (SD=.23). Intraclass correlation coefficients (ICCs) for the entire sample ranged from -.26 to .84. Results indicated that dyads displayed variation in their levels of attitudinal congruence.

Implications: The findings on gender are largely consistent with prior empirical evidence, and support the importance of assessing the specific gender-related service use attitudes of older husbands and wives in both practice and research contexts. Variation in the descriptive findings on congruence suggest the need to further understand not only how older spousal couples may vary in their specific service use attitudes, but also whether congruence on specific attitude measures might predict service use outcomes.