Introduction
Defined as ‘the comparative analysis of alternative courses of action in terms of both their costs and consequences’ [
1], economic evaluations have become an important tool to inform decision makers about both the costs and effects of health care technologies. It is typically recommended to adopt a comprehensive societal perspective in these evaluations [
1,
2]. This implies that all costs and effects should be taken into account, regardless of who experiences them. As several authors have pointed out, this means that both the costs and effects experienced by significant others, such as informal caregivers, should be incorporated in economic evaluations [
2,
3]. Even in jurisdictions where a more restrictive health care perspective is adopted, such as England and Wales, it has been argued that it would be appropriate to at least incorporate health effects in carers in these evaluations [
4].
In recent years, much progress has been made regarding how to value the costs of informal care, even though several controversial issues remain, for example regarding how to measure the time spent on informal care (separating it from time spent on other activities) and how exactly to attach a monetary value to the time inputs by caregivers [
5,
6]. In this paper, we concentrate on measuring the health-related quality-of-life (HRQoL) effects in informal caregivers. Measuring such effects could be used to complement methods that measure and value the costs of informal care (without simultaneously valuing effects) in order to include the full impact of informal care in economic evaluations.
In the literature, a distinction has been made between ‘family effects’ and ‘caregiving effects’ [
3,
7‐
10]. The former refers to the effects (on HRQoL, or another outcome measure) of the fact that someone within the caregiver’s social environment is ill. These effects are not specifically related to caregiving: they may also exist when no caregiving tasks have to be performed and may therefore occur in a broader group than caregivers only. The latter relates to the effects of performing caregiving tasks. By definition, these effects are only present in informal caregivers. Of note, depending on the context, these two types of effects may both occur in informal caregivers, since caregivers are often family members of patients.
Effects in caregivers can be assessed not only in terms of HRQoL changes but also in terms of burden or well-being [
8,
9]. Indeed, there has been extensive investigation of the objective and subjective burden of informal care [
11‐
13]. Yet, such burden assessments in itself do not comprise a valuation of some type, and therefore, they are not directly useful for economic evaluations. Also, it is likely that the caregivers’ general well-being, in a broad sense, may be affected, given that caregiving may involve different implications such as having to perform unpleasant tasks, interruption of daily activities, social isolation, and adverse financial consequences. Methods to measure and value the general well-being effects on caregivers have been developed [
14,
15]. However, the incorporation of such effects in economic evaluations may be difficult for two reasons. First, in jurisdictions where a strict health care perspective is prescribed, effects on well-being may be considered beyond the scope of the analysis. Second, since ‘even’ in patients usually only effects on HRQoL—which refers to the aspects of quality of life that relate specifically to a person’s health—are measured, measuring broader effects in caregivers may be considered inconsistent and may be practically difficult (creating the problem of how to combine two distinct outcome measures) [
3].
Therefore, a first step to include the effects on informal caregivers in economic evaluations may be to measure HRQoL changes as a result of caregiving. It has been documented before that caregiving has health effects for the caregivers, both physically and mentally [
16‐
18]. HRQoL effects may be measured by using instruments similar to the ones used to assess patients’ HRQoL and could, in principle, be easily combined with outcomes of patients’ HRQoL in economic evaluations. Such HRQoL effects in carers are not only relevant when adopting a societal perspective but also when taking a more narrow health care perspective. Note again that these effects in caregivers may comprise both the caregiving and the family effect (raising the obvious question of whether HRQoL effects in non-caregiving family members should also be measured).
Still, until now, relatively few attempts to measure HRQoL effects in carers within the context of economic evaluations have been reported. To give some examples, Mohide et al. used the time trade-off technique to assess HRQoL of caregivers. The authors found this method, normally used to assess patient’s HRQoL, to be feasible, reliable, and valid [
19]. Dixon and colleagues carried out a study into the relationships between patient HRQoL, carer HRQoL, and time spent on caring. They concluded that improving patient HRQoL may reduce the need for carer time and improve carer HRQoL [
20]. Davidson et al. tried to capture the effects of caring for older people by adjusting the carers’ QALY weights. They found that caring for a relative had a negative effect on these QALY weights [
21]. Drawing on a theoretical framework for incorporating the effects of patients’ health on their family members [
7], Basu et al. applied a time trade-off technique to elicit HRQoL weights of partners of patients with prostate cancer. There appeared to be a significant effect of the patient’s health condition on the partner’s HRQoL [
22]. Finally, Bobinac et al. studied the effects of caregiving on well-being or HRQoL in a large sample of Dutch caregivers and tried to separate family effects and caregiving effects [
8,
9].
In conclusion, the HRQoL of caregivers is likely to change due to caregiving and family effects, and it is worthwhile to further develop valid and practical methods to measure this change in the context of economic evaluations. With this in mind, we set up a study in a group of informal caregivers consisting of parents of children with major congenital anomalies. Several studies—albeit not within the context of economic evaluations—revealed that parents of children with congenital and/or chronic conditions have a relatively low HRQoL [
23‐
26]. Our study aimed to identify what these parents of children with major congenital anomalies do for their children and what effects caregiving has on their HRQoL and to compare their HRQoL with that of the general population. The underlying aim was to find, in an exploratory way, suitable methods to assess informal caregivers’ HRQoL, to be applied in economic evaluations.
Discussion
When adopting a societal perspective or a more narrow health care perspective in economic evaluations of health care interventions, the HRQoL effects in informal caregivers need to be taken into account. To date, few attempts to quantify these effects have been made. In this paper, we report on an evaluation done in a population of parents caring for children with major congenital anomalies, in an attempt to assess changes in caregivers’ HRQoL specifically in relation to caregiving. Moreover, we investigated the impact of informal caregiving by comparing HRQoL of caregivers with that of the general population. Our findings emphasize the importance of investigating the position of informal caregivers when conducting economic evaluations and provide important lessons for the future.
Caregivers were shown to have a lower HRQoL than their counterparts in the general population. However, this result did not hold for some subgroups investigated, which implies that in this study, with the methods that we have used, effects on carers (either ‘family effects’ or ‘caregiving effects’) were not universal. In this respect, it should be noted that we included, where relevant, both parents of each child. They may not both be providing care (and thus experiencing the caregiver effect) to an equal extent, yet both parents may be expected to experience the family effect. The analysis might have shown stronger results if we had concentrated only on primary caregivers providing high levels of caregiving, thus experiencing both effects simultaneously. Besides, some parents may have come to perceive informal caregiving tasks as normal tasks and have difficulty to distinguish between their role as caregiver and a ‘normal’ parenting role. Furthermore, coping resources of the parents are likely to have affected the reported HRQoL [
49,
50].
In view of these reflections, our findings suggest that the impact on primary caregivers may indeed be significant. Therefore, our study underlines the importance of considering caregivers, also in the context of economic evaluations. Moreover, it indicates that general HRQoL measures, as used in patients, may be able to detect HRQoL effects in caregivers. Using such outcome measures also facilitates the incorporation of HRQoL effects in common economic evaluations by, in principle, allowing the addition of HRQoL effects in caregivers to those in patients. Analysts and policy makers should therefore be aware that if HRQoL improvement is an important aim, they should register HRQoL changes not only in patients but also in their caregivers and perhaps even in their broader social environment. This, in turn, may alter the cost-effectiveness of treatments, as was also discussed by Basu and Meltzer—who also called attention to possible equity concerns, in the sense that including effects on others (such as caregivers and family members) may lead to differences between patients who have or do not have a partner, for example [
7].
In studies on the effects in caregivers, it should not be ignored that the results may partly depend on the setting. For example, in a parent–child relationship, caregiving is normal to some extent, although it diminishes with time, while this is less so in a partner relationship or a child–parent relationship. Moreover, in the context of parents caring for their children, caregivers are typically relatively young, which may result in other effects on HRQoL (and well-being) than when caregivers are older. Furthermore, it is good to emphasize the broader social context of caregiving. For example, in a parent–child relationship, caregiving for one child may lead to the deprivation of other children in the family because of parental inattention or too much early responsibility [
51,
52]. Finally, the decisions to provide care or to cease caregiving (or: ‘the endogeneity of caregiving and health’ [
53]) will be made for different reasons in different settings, caused by different underlying mechanisms. Considering all this, an important area for future studies is to further unravel the influence of the relationship between caregiver and care receiver on the impact of caregiving. These issues have received little attention in the literature so far.
The EQ-5D was the instrument of choice in this study. Among its advantages are that it is a short, easily understandable questionnaire, and suitable for mail administration. Another advantage is that it includes all basic domains of HRQoL—psychological, physical, and social as well. So far, many studies have mainly focused on psychological health of caregivers. Studies of the physical health effects are less conclusive but suggest increased physical vulnerability [
54‐
56]. Thus, it is desirable to adhere to a broadly defined HRQoL measure, without eliminating any domains in advance. Moreover, the EQ-5D, comprising the EQ-5D descriptive system and the EQ-VAS, has proven a reasonably valid instrument, also in people suffering from health problems that mainly affect specific dimensions of HRQoL [
57‐
60]. The EQ-5D thus appeared an excellent starting point for attempting to quantify the HRQoL effects in caregivers. Nevertheless, more investigation into for example the sensitivity of different HRQoL instruments to these effects seems warranted. In that context, it is also important to note the social dimension of HRQoL. The EQ-5D captures (elements of) this social dimension by asking about problems with performing usual activities, such as work, study, housework, family, or leisure activities. This can be an important problem for caregivers as well. This is also recognized in the CarerQol instrument [
14,
15], which for example asks about the quality of the relationship with the patient and the support from one’s social environment in providing care. Finally, it needs emphasizing that by concentrating on HRQoL effects other, broader quality-of-life effects on caregivers have remained unmeasured, though they likely exist [
8,
14].
HRQoL effects in caregivers may be due to caregiving (the ‘caregiving effect’) and to the presence of illness within their social environment (the ‘family effect’). To assess the exact magnitude of HRQoL changes specifically due to caregiving, we asked informal caregivers to express how their HRQoL would be if they would not have to provide informal care for their ill child anymore and compared this to their actual HRQoL. Several explanations for the fact that most people did not indicate a HRQoL change in the hypothetical scenario need to be considered:
(i)
Responders may find it difficult to recognize that HRQoL problems are related to caregiving and to estimate how their HRQoL would be affected if they did not have to provide care tasks. Some people may find it difficult to comprehend such a hypothetical question, anyhow. Alternatively, there simply may be no significant effects on HRQoL (apart from ‘family effects’);
(ii)
The fact that quite a number of parents (9%) indicated that their HRQoL would decline in the hypothetical scenario could mean that the responders feel uncomfortable about the questions and give protest answers. Also, some parents may not want others (perhaps strangers) caring for their child’s health. Furthermore, it may suggest positive process utility derived from caregiving, such as feelings of gratification, self-confidence, and finding meaning in the care [
61‐
63]. It is unclear to what extent this influences HRQoL, but one might argue that such a positive externality might be especially present in caring for one’s child, although negative effects may still predominate.
(iii)
The instrument appears to detect changes in HRQoL particularly in situations in which caregiving is very demanding, as shown by Table
5, which largely confirms our hypotheses on the combination of the two EQ-VAS scenarios. Only in these straining circumstances carers may actually experience HRQoL problems due to caregiving and be able and willing to report them.
These issues also seem to have played a role in two studies from the literature that took a similar approach to assessing HRQoL or well-being effects in carers [
21,
64]. Even though there were differences with our study: Davidson and colleagues asked carers about their HRQoL under the assumption that the health of their relative was so good that he or she would not need care, while Brouwer et al. asked the caregivers to rate their happiness under the assumption that all caregiving tasks would be taken over.
In general, while our results appear to indicate that directly asking responders to estimate their HRQoL in the hypothetical scenario of not providing informal care may provide some relevant information such as revealing relatively strained caregivers, it also indicates that it is unlikely to provide exact measurements of the HRQoL effects of providing informal care. For the latter purpose, direct measurement of HRQoL (changes) and comparisons with relevant counterparts in the general population seem more appropriate.
Interestingly, there was a moderately strong correlation between the EQ-VAS scores of the caregivers themselves and the EQ-VAS scores of the child as indicated by the caregivers (e.g., for all female caregivers: r = 0.42; P < 0.001). This provides an important indication for a health-related ‘family effect’, which is worth pursuing in future studies. Obviously, the close relationship between how the parent perceives his or her child feels and how the parent experiences his or her own HRQoL may also be due to concurrent effects on the parent’s and the child’s HRQoL (e.g., environmental and genetic influences) and to the fact that parents in being proxies for their children’s HRQoL may have been influenced by their own HRQoL.
Future research may furthermore be aimed at better explaining the differences in terms of HRQoL between caregivers and the general population. It should be realized, for example, that even caring for a healthy child could disrupt HRQoL. Ideally, therefore, reference values are derived from parents with healthy children, not the general population. In this experiment using the EQ-5D, such data were not available. Moreover, HRQoL of non-caregiving parents or relatives of ill individuals could be measured as an alternative way to distinguish ‘family effects’ from ‘caregiving effects’ on HRQoL. Worthy of note are two other studies that attempted to separate these two types of effects. Bobinac et al. used regression techniques (taking an approach similar to our analysis presented in Table
7) to investigate to what extent the HRQoL or the well-being of caregivers can be explained by the patient’s health (which would be indicative of the family effect) or the objective burden of caregiving (indicative of the caregiving effect) [
8,
9]. Finally, it would be interesting to see whether informative subjective burden measures would better explain HRQoL differences than the more objective explanatory variables used here. It must be noted, however, that this can induce problems of endogeneity in the analysis. In terms of studying the influence of coping and adaptation, one could register people’s ability to cope and coping strategies to estimate the influence on HRQoL in relation to caregiving. Aspects of caregiving such as benefits of caring and the influence of support can be measured through the CarerQoL instrument and linked to well-being effects as well as HRQoL effects [
14,
15].
To conclude, present understanding of HRQoL effects of informal caregiving is still rudimentary. In this paper, we tested a simple and straightforward method to measure the HRQoL impact in informal caregivers for use in economic evaluations of health care interventions. We also identified several issues that could usefully be addressed by future research. The demand for informal care and its impact on families are expected to rise, because several acute diseases have increasingly become chronic diseases (with long-term morbidity and a continuing need for care) and because of current trends toward early hospital discharge and outpatient treatment. These trends hold true for pediatrics as well as for other branches of medicine [
24]. This makes it all the more important not to disregard the position of informal caregivers in future research efforts.