Abstract

Background and objectives: The consequences of ill-health and advancing frailty can be stressful for the close family and friends of the individual. Long-term care services provided to compensate for the effects of impairment resulting from ill-health on quality of life (QoL) can therefore have benefits for close others. The wider benefits of services have been conceptualised as two distinct effects: a caregiving effect, reflecting the effect of services on the caregiving burden and consequently the carer’s utility; and a family effect, which suggests that as close family and friends care about the service user they will gain utility from change in service user’s utility resulting from services. Reciprocity in the relationship between the close family and friends and the service user, however, means that service users’ utility will also be affected by changes in the carers’ utility. The main aim of this study is to estimate the relative size and direction of the family and caregiving effects on carers’ care-related quality of life (CRQoL), and explore the importance of reciprocity in the estimation of these relationships. An associated aim is to compare the family and caregiving effects across three European countries.

Methods: We use data collected from a cross-sectional survey of people receiving home-based long-term care services and their informal carers in Austria, England and Finland. In total there are approximately 500 dyads. The survey included the ASCOT measure of CRQoL for service users (ASCOT-INT4) and for carers (ASCOT-Carer-INT4) along with information about caregiving (hours and tasks), the care environment and other characteristics of service users and carers.

We employ two methods to estimate the family and caregiving effects. First, we adapt the regression approach used by Bobinac et al (2010, 2011), whereby the family effect is estimated by the CRQoL status of service user, and the caregiving effect by a measure of caregiving burden (number of the care-giving tasks or hours of care provided). Second, we estimate a mutual influence model using structural equation modelling, with reciprocal effects between service user CRQoL and carer CRQoL and a path from caregiving burden to each of these outcomes. In both approaches, sample selection (endogeneity) issues are addressed by controlling for observable confounders of the caregiving-CRQoL relationship.

Results and conclusion: Analysis is ongoing, but preliminary results from the English data show that the family effect is positive and much larger than the negative caregiving effect. A reciprocal relationship between carer and service user CRQoL is found. When using the mutual influence model, the size of the family effect is larger and the caregiving effect is smaller, suggesting that it is important to model the reciprocal effects of CRQoL on each member of the dyad when estimating family and caregiving effects. In the presentation we will reflect on any differences across countries in the relative size of the family and caregiving effects. We will also reflect on the implications for LTC systems.