Willingness to pay for informal care in France: the value of funding support interventions for caregivers

Total costs of dementia in Europe vary markedly depending on the underlying assumptions that are used in their calculation, for instance in regard to the ‘rather uncertain estimates of informal care costs’ [1]. Costs of dementia were estimated at €160.3 billion in the EU27 in 2008 by Wimo et al. [2] and at €105 billion (€PPP 2010) by Olesen et al. [1]. In Alzheimer’s disease (AD), the most common form of dementia, caregiving costs are primarily accounted for by informal caregivers whose time accounts for most hours of care and assistance [1],[3]-[5].

Because appropriate support is often lacking, caregivers can experience “burden of caring” [6] especially when the need for care provision has not been anticipated [7]. This burden, that can result in impaired caregiver health or opportunity costs, has been defined as a spillover effect by Bobinac et al. [8], as when caring effects are treated as externalities in health economics [9],[10]. From a welfare economic perspective, externalities can lead to Pareto inefficiency in AD resource allocation and thus to suboptimal utilization of informal care [11]. Such a situation could be highly detrimental at a time when the population with a strong preference for home care, as in the EU, is ageing and when informal caregiving may become the cornerstone of AD care because of scarce public resources [12],[13].

For long-term AD caregiver policies to be effective, the needs of carers for support require better recognition as these needs influence both the carer’s well-being and the resources used in AD care arrangement [14]. For Rosa et al. [14], carers of people with dementia and related disorders experience a large spectrum of needs that include needs for medical and psychological care, social support and education [14]. The policies implemented in most EU countries to support long-term carers are of two main types: financial support measures (aka cash-for-care measures) and in-kind services (e.g. home-based professional services, respite care, counselling, training or support groups) [15]-[18]. In France, cash-for-care targeting the needs for support of care recipients rather than those of carers prevails [19]; in-kind services are underused [20].

There are effective policies for “alleviating caregiver’s perceived burden and helping prolong the caregiving task” [21], but the evidence for optimal support interventions for carers of people with dementia is scant [22]-[24]. Moreover, few studies have addressed how carer needs for support affect the economic value of informal care. Because informal care encompasses the intangible value of carer need for support, including both the positive and negative externalities of caring, the estimation method that is used has to be sensitive to carer preferences [25],[26]. The contingent valuation method (CVM), which is based on stated preferences, consists in valuing the informal care hour at the ‘price’ the carer would be willing to pay (WTP) to reduce caregiving time by one hour - or would be willing to accept for increasing this time by one hour [26],[27]. By finding out carer preferences in relation to caregiving costs, CVM associates carer utility to factors such as their need for support.

The present study highlights funding of care training programmes as a policy driver among support interventions for caregivers in order to improve the benefits of informal care. We asked to what extent an informal carer’s WTP is influenced by their need for in-kind support services (i.e. respite, care training and support groups). Respite can be offered by “taking over informal care temporarily in order to give the caregiver a break and/or enabling the caregiver to participate in other activities” [21]; training covers practical and educational support and information aimed at improving caregiving capabilities; support groups offer collective emotional support.

According to logistic regression analyses, each type of need for support services was significantly and positively associated with one caregiver health-related issue such as sleep disorders or depression (Table 2). In a Venn diagram (Figure 2), there was only a small overlap between the three different types of support needs, indicating few collinearity problems.

Carer health status	Need for respite care			Need for care training			Need for support group
	OR	SE	SD	OR	SE	SD	OR	SE	SD
General health (5 categoriesa)	1.74**	.320565	-.44013 1.0428	.681	.1516	-.81944 .05254	.87	.18052	-.53753 .2709
Depression (yes vs no)	1.35	.511376	.19798 .91741	1.96	.8309	-.1573 1.5039	2.44*	.97593	.11381 1.6766
Sleep disorders (yes vs no)	1.27	.439664	-.26351 1.1123	2.22*	.8640	-.44018 1.1765	1.97	.73080	-.36302 1.1905
Anxiety (yes vs no)	1.52	.53654	-.43124 .91913	1.44	.5960	.03581 1.5605	1.51	.59943	-.04514 1.4057

Abbreviations: OR Odds ratio, SE standard error, SD standard deviation; *p < 0.05; **p < 0.01.

^aHighest value represents poorest health.

No multicollinearity was observed in the model (Variance Inflation Factors were below 2), confirming the results of the Venn diagram. The WTP was positively influenced by need for care training (p < 0.001) but not by need for respite care or for a support group. However, the need for a support group positively affected the probability of reporting WTP (p < 0.05). All three exclusion restrictions, ‘benefiting from other informal support’, ‘already benefiting from professional support’ and ‘lacking time for self because of caregiving activities’ had a significant impact on caregiver probability of reporting their own WTP.

The positive Inverse Mills Ratio (p < 0.05) suggested positive data selection, i.e. carers who estimated their WTP gave a higher WTP than a random population of caregivers with comparable characteristics would have done.

In this first exploration of the influence of informal caregiver needs for support services on the monetary value of informal care, as given by their WTP to reduce their caregiving time, the need for care training was found to increase the monetary value of informal care unlike the need for respite care or participation in support groups.

The divergence in needs may arise from the informal caregiver’s perception of their role. The utility they derive from caring depends on cultural and social dimensions, and this may influence the monetary value they place on their caregiving activity, their choice of being an informal carer, and their use of services. The need for care training probably arises from a keen sense of responsibility, and from a desire for skill development, coping strategies for use in burdensome situations, and competencies for reducing perceived strain. Caregivers with a ‘higher level of mastery of care perceive themselves as able to better answer care demand’ [33]-[35]. Quality of life, in particular sleep time which was associated with need for care training in our study, is improved [36]. The association between ‘WTP to reduce caregiving’ and ‘care training programmes’ suggests that care training as a support intervention might empower caregivers as reported in the literature on care recipients with cognitive and physical impairment [37]-[39]. On the other hand, a need for respite or support groups might emphasize the burden of caring including feelings of being overwhelmed or wanting to give up. Provision of respite does not necessarily imply a feeling of respite [40]. Informal carers who are overprotective or take time off may, for instance, feel guilty [41]. In addition, unlike training programmes which not only improve quality of care but also the carer’s process utility, respite care, which just replaces informal care by professional care, or support groups, in which no active individual help is offered to carers, have no direct impact on the caring process [42].

Our informal carers spent on average 4 hours per day (±4.6) on ADL and IADL, supervision (which was associated with PRWTP, p < 0.05), and on managing contacts with formal care. This is in line with the more than 20 hours of care per week reported by Parker for ‘heavily involved caregivers’ [43]. Carer perceived utility and well-being are related to duration and intensity of caregiving, and may thus depend on interventions for improving the overall care process. Care training might help caregivers achieve some “reward” or positive utility and might even improve recipient well-being as coping skills are improved within an altruistic carer-recipient relationship [44]-[46]. Good carer-recipient relationships (as reported by 65% of our carers) or a positive association between recipient and carer health illustrate this altruistic relationship [44],[47].

Our research has economic and policy implications as individual WTPs for informal care were collected in a sample representing 339,340 French caregivers. WTP value is known to be sensitive to total economic value of non-market services, including non-use and intangible values [25],[48]. Our estimated monetary value could thus be included in cost-benefit analyses looking for the most efficient AD care arrangement. Our work could lead to replicable studies that include the informal care-related cost and outcomes allowed by WTP elicitation that are usually absent [11]. The marginal effects of need for care training on WTP could inform public decision makers on possible incentives or compensation for improving the benefits of informal care. In line with previous findings [48, 57], our results highlight the value of funding training programmes as drivers for improving the benefits of informal care for both carer and recipient [42],[49].

A strong point of our study is that the estimated value of informal care represents the overall value, thus respecting caregiver plurality. Both primary and secondary carers as well as carers who are already supported by professional carer were considered in the DHCS survey and potential interaction effects were controlled via exclusion restriction variables ‘benefits from other informal help’ and “benefits from professional help”. This is important because caring for a single recipient does not necessarily imply a need for the same kind or level of support - especially where two caregivers are siblings, as suggested by Fontaine et al. [50] or because filial norms about care provision differ between daughters and sons of a single care recipient [51]. Another strong point is that heterogeneity in caregiver preferences was taken into account thanks to the sample selection model, as in previous models of preference uncertainty and heterogeneity [27],[52].

Our study has several limitations. First, like Koopmanschap et al. [21], we used dichotomous variables to represent support needs. Because intensity of carer preference could not be measured, we could not qualify results in terms of level of caregiver needs. Second, the use of CVM to elicit caregiver preferences in the DHCS is contentious as it supposes that a societal perspective can be adopted by aggregating individual WTPs. Although CVM can be justified by the similarity of WTP elicitation to a market demand schedule, CVM could generate both anchoring effects and opportunistic behaviours [48],[53]. Furthermore, because the WTP is estimated for the marginal caregiving hour, it may not be adequate for estimating the total informal care burden. Caution is needed when CVM is the sole method used to advise policy decision makers, with mixed methods using qualitative interviews to capture caregiver motives for refusing to answer the WTP question probably being preferable [54]. Third, provision of support to just a single member of a care recipient - caregiver dyad might not be efficient and the effectiveness of care training thus should be addressed [15],[55]. Fourth, the questionnaire did not ask ‘when’ the measure should be implemented, for instance at the onset of caregiving as a burden prevention programme in early-stage AD as suggested by Ryan et al. [56] or only when carers are already experiencing a burden. Lastly, our work focuses exclusively on French caregivers for people with AD living in the community. This might limit its policy implications given the specific context of support service use by caregivers.

Identifying support interventions for informal caregivers can pave the way for future research that evaluates their cost-effectiveness in a context of a growing number of people with AD and ever scarcer resources. Our findings showed a significant association between the need for care training and the monetary value placed on informal care by caregivers, highlighting caregiver needs to develop their coping skills and competencies and to improve their caring capabilities. Implementing such a policy of care training should lead to a two-fold positive effect, satisfaction of care recipient to be able to stay at home and satisfaction of carers in providing better help to care recipients. Further studies are warranted to ascertain whether these interventions might be both cost-effective and cost-efficient in the long run and relevant to the carer-recipient dyad. Nevertheless, future health economics studies should not overlook needs for respite care and support groups as these are associated with caregiver health. A contingent method to value informal care might not be adequate to reveal their impact on caregiver utility as higher moral constraints might govern these needs than the need for care training.