Background
In health care, decisions are made about treatment at the level of individual patients, of patient groups (guideline development), and at the societal level [
1]. Decisions about guideline development and decisions at the societal level are often guided by cost-utility analyses. In these analyses the gain in health obtained by treatment is compared with the costs that have to be made in order to obtain this gain [
2]. To assess the value of this gain, cost-utility analyses make use of health state valuations, i.e. health state utilities.
A health state utility is a preference for a particular health state compared with perfect health and immediate death. Utilities can be seen as a global valuation of health related quality of life (HRQL) [
3] and can be expected to show a strong relationship with health status. Nevertheless, only between 18% and 43% of the variance in health state utilities can be explained by HRQL. Most of the variance still remains unexplained [
4].
Health state utilities can be elicited from members of the public and from patients. Members of the public tend to give lower health state valuations, compared to patients [
5]. This discrepancy in health state valuations has, among others, been suggested to be explained by the failure of members of the public to anticipate on their ability to adapt. Patients adapt to the physical and psychological challenges of their illness [
6]. When health state valuations are elicited from patients, some of the variance in health state utilities might be explained by this adaptation [
7‐
9].
Tentative support has been found for the effect of adaptation on health state valuations. Members of the public who were made aware of their ability to adapt gave higher valuations on a person trade-off (PTO) and on a visual analogue scale (VAS) measuring quality of life [
10,
11], but not on the time trade-off (TTO) nor on the standard gamble (SG) [
12]. Whether health state utilities given by patients are actually correlated with adaptation has not been topic of study yet.
Adaptation can be defined as a response that diminishes or remains the same despite constant or increasing stimulus levels [
13]. The outcome of adaptation can be measured by change over time, such as change in well-being [
14] or life satisfaction [
15,
16]. If researchers aim to gain more insight in the process of adaptation itself, adaptation can be conceptualised through certain coping-strategies [
17,
18]. These coping-strategies are, among others, enabled by personal resources.
By studying adaptation Taylor [
19] developed the Cognitive Adaptation Theory (CAT) which is based on cognitive interviews with chronically ill persons. This theory is one of the dominant theories in health psychology and has often been used to empirically test adaptation. Research using this theory suggests that psychological adjustment to an illness occurs around four themes; a search for meaning in the experience, an attempt to regain mastery over the event and over one's life more generally, an effort to enhance one's self-esteem, and the ability to find positive illusions, i.e. optimism. These concepts as suggested in the CAT are further described below.
After a threatening event, people often cannot find a sense of meaning in the experience and lose their feelings of mastery and of self-esteem. Most people manage to re-establish these over time. According to Taylor, this re-establishment is based on so-called positive illusions. People develop unrealistic beliefs that make it possible to regain control over the event and over one's life and to regain self-esteem [
19]. Although positive illusions may create unrealistic and maybe 'false' ideas, these illusions have been found to be important resources [
20].
Previous studies have shown that patients who score high on indicators of CAT have better psychological functioning [
21‐
24], they are less anxious and depressed, report more vitality and have a better mental functioning [
22,
25,
26]. Moreover, patients with a higher score on indicators of CAT reported better physical functioning [
22,
23], they showed fewer new coronary events or hospital admissions [
21,
26] and lived longer [
27]. It thus appears that patients who have higher self-esteem, mastery, and optimism, and who find a meaning in the experience have better abilities to adapt.
No standard method is available for investigating the ability to adapt based on CAT. Studies have used different indicators and methods for their analyses. For instance, studies have included indicators measuring optimism, mastery and self-esteem, but often exclude finding meaning. To our knowledge, only in two studies the effect of finding meaning was included [
27,
28]. The rationale to exclude benefit finding was described by Major et al. [
29] and Chan et al. [
23]. Both research groups suggest that mastery, self-esteem, and optimism are stable personality traits representing a persons' resilience, whereas finding meaning might be seen as a process facilitated by these personality traits.
Apart from this variety of indicators of CAT included to measure adaptive abilities, studies have also used different ways to measure these indicators. Some studies have analysed the effects of the different indicators separately [
25,
30], some have created a scale taking the indicators together [
26‐
28,
31], and again others have investigated each indicator separately as well as an aggregate scale of the indicators together [
21‐
23]. The latter studies revealed that besides the effect of the aggregate scale, often only one of the indicators had an effect on the outcome measurement. Since the overall results of these studies show different 'single' indicators to reveal an effect, indicators of persons' abilities to adapt cannot be simplified to one single indicator. Exploring the results of these studies further, it seems that significant effects have mostly been seen in studies using an aggregate scale. Therefore, in the present study persons' ability to adapt is constructed with an aggregate scale based on mastery, self-esteem and optimism.
The first aim of this study was to investigate if above HRQL, persons' adaptive abilities explain health state utilities. That is:
Do adaptive abilities account for the unexplained variance in health state utilities above the variance explained by HRQL?
Another possibility is that adaptation, in this study measured through persons' ability to adapt, has an indirect effect on utilities, through HRQL. As described above, adaptive abilities does affect psychological and physical functioning [
26]. This would fit the hierarchical model of Spilker and Revicki [
32], in which three levels of quality of life are distinguished that have mutual impact on each other. The hierarchy of this model ranged from a global level such as a health state utility, to HRQL domains, and to specific determinants of domains such as personality characteristics, [
32] which may include adaptive abilities. Thus, the second aim of this study was to investigate if adaptive abilities affect health state utilities via HRQL domains.
Is the relation between adaptive abilities and health state utilities mediated by HRQL domains [
33]?
Since we investigated psychological adaptive abilities we assume from a theoretical point of view that only mental health can mediate this relation.
Discussion
In discussion sections of papers and in theoretical manuscripts, the difference in health state utilities between people with a chronic illness and the public is often explained by adaptation [
1,
14]. The results of this study show that adaptive abilities are indeed related to utilities, but that this effect is fully mediated by mental health for the TTO, and partly mediated for the VAS. It seems that in people with a chronic illness a stronger ability to adapt may lead to better mental health, which in turn leads to higher health state utilities. The suggested relation between adaptation and health state utilities given by people with a chronic illness does not occur directly, but appears to be mediated by mental health. Admittedly, this conclusion has to be made with caution since not adaptation but adaptive abilities are studied here.
Adaptive abilities explained 46% of the variance in mental health, which in turn explained between 11 - 18% of the variance in health state utilities after correction for physical health. Arnold et al. [
52], already suggested such a mediation effect. They found that people with a chronic illness do not differ from healthy people in global quality of life and that global quality of life is mostly explained by mental functioning. Based on these findings they argued that people with a chronic illness psychologically adapt, causing a recovery of their mental health, which leads to recovery of global quality of life.
The cross-sectional design of this study limits the points described above. From this study no conclusions can be drawn about the causal relationship between persons' ability to adapt, HRQL, and health state utilities. Nevertheless, causal relations between persons' ability to adapt and HRQL have been described previously [
24,
47]. Future longitudinal research is necessary to further investigate this causal relationship.
The index based on CAT to measure persons' ability to adapt, has been used in several studies but has not yet been validated. Given the number of studies using such a scale based on the CAT, validation is pressingly needed. Further, this index has been suggested to reflect stable personality traits, which might not change over time [
29]. If adaptive abilities are indeed stable over time, then health state utilities of members of the public might be influenced in a similar way. Yet since members of the public find it difficult to anticipate on their ability to adapt [
11] we still would expect a less substantial effect of adaptive abilities on HRQL and health state utilities in this population.
HRQL predicted 20% of the variance in the TTO, and 49% of the variance in the VAS. These results are comparable with previous findings concerning the relationship between HRQL and health state utilities [
53]. The smaller amount of variance explained in the TTO compared to the VAS might be caused by the trading process. In this trading process, a series of information processing activities and construction of subjective values for dimensions are developed, making the variance in TTO-scores difficult to explain. Another explanation may lie in the cognitive nature of the TTO. Campbell [
54] suggested that quality of life can be assessed with cognitive or with affective measurements. Cognitive measurements depend on a more intellectual process while affective measurements depend on subjective feelings. The TTO might be seen as a more cognitive measurement, the VAS as a more affective measurement. After a life event, the affective component of well-being appears to be more impaired than the cognitive component, which means that this component is sensitive to change and the cognitive component is more stable [
55]. Finally, a more methodological explanation for the smaller amount of variance explained in the TTO might be that the TTO was skewed. When a dependent variable is skewed a smaller effect size might be anticipated [
56].
This study included patients with RA who had been diagnosed on average 13 years before. First, it can be questioned if patients still need to adapt to their illness so many years after diagnosis. It seems evident that adaptation takes place in the initial phase of the illness. However, the disabling, often progressive and uncontrollable characteristics of RA might result in adaptive processes, even after so many years. The results of this study indicate that adaptive abilities indirectly explain health state utilities, so this result might become more distinct when examining patients in the initial phase of their illness. Secondly, RA is a chronic illness characterized by pain and deformity of the joints, leading to physical limitations. There is evidence suggesting that people do not adapt to unpredictable stressors such as pain [
57]. On the other hand, patients with RA might be able to adapt to other aspects of their illness such as the physical limitations by learning new ways to perform activities and they might learn to accept their pain [
58]. More research is necessary to investigate the effect of adaptive abilities on health state valuations in other patient groups.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
YP and TPMV were involved in acquisition of data. YP performed the analysis and YP, AVR and AMS took part in interpretation of the data. The first draft of the paper written by YP was revised by AVR and AMS. All authors gave final approval of the version published.