Quality of life in urologic cancer patients: importance of and satisfaction with specific quality of life domains

Prostate cancer is the most frequently diagnosed cancer among men, with about 1.3 million new cases being diagnosed annually worldwide [1]. The 5-year survival rate for prostate cancer is relatively high, about 83% in Europe [2]. Quality of life (QoL) has become an important quality and decision criterion for cancer patients, medical practitioners, and the healthcare system. Multiple studies have been performed to investigate QoL [3‐6] and mental health [7‐9] in urologic cancer patients and survivors. Most studies found that, compared with other cancer types, QoL among prostate cancer patients was relatively good [10, 11]. When compared with controls from the general population, prostate cancer survivors often rate global QoL dimensions similarly [12‐14]. However, when specific symptoms are considered such as urinary incontinence, bowel or rectal function, or sexual function, prostate cancer patients report significant detriments [15, 16].

Studies on QoL among cancer patients have historically focused mainly on health-related QoL [3, 17]. However, QoL is a broader concept that also includes factors such as professional life, family life, leisure activities, and finances. Factors like these need to be considered as well when evaluating cancer patients’ QoL and developing supportive services.

While QoL assessment instruments generally cover several domains of well-being, they do not consider the subjective importance of these domains to the individual respondents. It is implicitly assumed that each dimension has roughly the same meaning for all participants. This, however, is not necessarily the case. It is known that subjectively important QoL domains correlate more strongly with overall QoL than unimportant ones do. Therefore, attempts have been made to weight the QoL components with the corresponding importance ratings [18], but weighted results were generally similar to the results of the unweighted calculations [18‐20]. However, differing subjective importance of QoL domains is also interesting beyond the aspect of weighting. Which areas of life become more important and which become less important after a cancer diagnosis? Does the importance of health increase?, and does the importance of the other dimensions decrease in the perception of cancer patients? The effects of a cancer diagnosis or treatment on the subjective importance of QoL dimensions are largely unknown.

One instrument that allows to shed light on these issues is the Fragebogen zur Lebenszufriedenheit (Questionnaire on Life Satisfaction) FLZ-M [21] since this questionnaire also addresses the subjective importance of QoL dimensions. A normative study was performed that allows to compare the cancer patients’ importance and satisfaction assessments with those of the general population. Since importance and satisfaction can change over time in cancer patients, we performed a study with two measurement points spaced 3 months apart to investigate changes in these variables.

While one way to assess the importance of a life domain is to ask the participants directly to assess the importance they attribute to it, a second approach is to calculate the associations between the specific life domains and a global assessment of QoL. Direct importance ratings of QoL domains can differ from the associations between the satisfaction with these domains and global QoL. While the domains health and family received the highest mean importance ratings in a large general population study [22], the contribution of the domain finances to the variance explanation of global life satisfaction was higher than the contributions of health and friends. We intend to test whether such relationships between direct importance ratings and regression coefficients for the prognosis of global QoL can also be found in cancer patients.

The aims of this study were (a) to analyze satisfaction with and the subjective importance of QoL components in urologic cancer patients in comparison with the general population, (b) to investigate changes in QoL in the patient sample, and (c) to investigate the degree to which the QoL domains predict overall QoL using regression analyses.

While the impact of urologic cancer on health-related QoL has been examined in multiple studies, the aim of the work presented here was to test whether other QoL domains are also affected by the disease and whether the subjective importance ratings of various QoL dimensions differ between urologic cancer patients and the general population. Health is one of the several QoL dimensions included in our analyses; this allows us to investigate the relevance of health in relation to other areas of QoL.

The most relevant QoL dimensions were health and family life, each of which had mean importance scores above 4 on a scale of 1–5. However, the general population also considers health to be the most important dimension; there were no significant differences in the health importance assessments between the patients and the general population. Other general population studies have also reported health receiving the highest importance ratings [19, 32]. While the patients’ mean importance ratings were higher than those of the general population in four of the seven other dimensions, the other three dimensions showed an opposite trend. This means that the non-health domains do not become less relevant for people after they have been diagnosed with cancer. As such, it is important to consider problems concerning finances, work, and social relationships when studying patient QoL, as these less physical aspects of life appear to be highly relevant for patients as well [33].

Concerning satisfaction, it is not surprising that the most relevant difference between the patients and the general population was found for the health domain. Nevertheless, the patients’ mean satisfaction rating was 3.08 which is nearly exactly the middle of the 1–5 scale, rather than in the lower half of the scale as one might expect. The patients’ satisfaction ratings were higher than those of the general populations in all of the other domains and in five of the seven cases even with statistically significant differences. This could be a result of a judgment effect: when there are severe detriments in one area (health in this case), the problems in other areas seem to become less relevant. To gain a better understanding of a person’s satisfaction with their health state, it might be useful to consider not only their satisfaction with their health alone but also with their health in relation to their general satisfaction with other areas as well.

During the 3-month period between t1 and t2, satisfaction scores slightly improved in the health domain (effect size d = 0.17), but became worse in the partnership domain, with a large effect size of d = − 0.41. The item includes both partnership and sexuality. While in most cases the combination of partnership and sexuality in one dimension makes sense, for prostate cancer patients, these sub-domains can be experienced quite differently. Several patients reported for example that they were highly satisfied with their partnership but very dissatisfied with their sexuality. Since urologic cancer patients often experience urinary and sexual symptoms that do not disappear within the first months after surgery [12, 15, 34, 35], the loss in satisfaction with the combined partnership/sexuality dimension is understandable. Partnership and sexuality are areas of life that deserve special attention in the treatment of urologic cancer patients and survivorship care plans [15, 36‐39]. A US-American study showed poorer quality of sexual communication and more sexual dissatisfaction after treatment in patients than in the general population [40]. Moreover, patients´ relationship satisfaction, quality of communication about sexuality, and sexual satisfaction were strongly associated with their partner’s satisfaction with the overall treatment outcome [15] and partners´ level of depression and sexual activity [40].

When considering the changes in QoL scores from t1 to t2, one must take into account that they might have been affected by response shift processes, whereby the respondents’ frames of reference changed due to adaptation processes [41‐43]. A study with prostate cancer patients [42] tried to quantify this effect and to estimate “true” changes. So-called thentests [44] could be used to further explain such effects and to better understand the real changes.

How do the eight QoL dimensions contribute to global QoL scores at t2? The results of the regression analyses (Model 1) show that all of the dimensions positively contribute to this global score and that the only non-significant dimension is partnership. The highest contributions came from the dimensions health (β = 0.418) and income (β = 0.386). Even after including the baseline value in the regression analyses (model 2), the domains with the highest β values were health (β = 0.323) and income (β = 0.311). A general population study [22] found that income was the strongest predictor of general life satisfaction (r = 0.59), while health was a weaker predictor (r = 0.46) and comparable with the dimensions friends (r = 0.45) and job (r = 0.47). It would be interesting to compare the associations between health satisfaction and general life satisfaction between patients and the general population in a more systematic way. Our analyses were controlled for age, education, and tumor stage. Therefore, these factors cannot be considered confounders for the effects. The relevance of the domain income seems to contradict the low importance ratings of this domain. While the patients declare that income is not so relevant for them, those patients who are satisfied with their income report a higher overall QoL than those who are less satisfied with their income. There is no linear relationship between the direct, explicit importance ratings of the dimensions and the indirect assessments based on associations with global QoL. While both analytical approaches reveal the health dimension to be highly relevant, the income dimension shows that the results of these two approaches may differ considerably. A similar phenomenon was observed in a general population study [22] where the domains with highest mean importance ratings were not necessarily those with the highest capability for predicting global life satisfaction. This shows that direct assessments of subjective importance must be considered with caution.

Some limitations of this study should be mentioned. While multiple studies have investigated health-related QoL in urologic cancer patients, assessments of QoL areas beyond health are rare, and considering the subjective importance of other life domains is a relatively new pursuit. Until now, there are only few applications of the FLZ-M in oncological research, and we could not compare our main findings with results obtained in the scientific literature. The dimension partnership/sexuality included two components which, in the case of urological cancer patients, do not form a consistent scale. We showed that direct importance assessments and indirect assessments in terms of β coefficients can result in different outcomes. While health was relevant in both approaches, the income dimension showed contradictory results. We cannot derive conclusions about the best way to infer the subjective relevance; a more stringent comparison between these direct and indirect methods would be a task for future research. Though the response rate of this study was relatively good, it is possible that the proportion of patients with severe problems is underrepresented since the t2 sample included only those study participants who had survived until at least 3 months after t1 and who were willing and able to take part in the t2 assessment. The time interval of 3 months is not sufficient for conclusions about long-term changes in QoL; however, in five of the eight dimensions of QoL, significant changes in the satisfaction ratings could be established. Though we tried to select a control group with a similar distribution of age and education, there may be differences with regard to other aspects such as income we could not control for. We addressed several research questions in this paper, but the data set can also be used for testing other relationships, e.g., the correlations between the importance and the satisfaction ratings, or testing the “domain-importance-as-a-leveler-hypothesis” [32] that postulates a moderating effect of the domain importance on the associations between domain satisfaction and global QoL, or the associations between changes in importance (from t1 to t2) and changes in satisfaction. In our study, we compared the QoL dimension health with other dimensions of QoL and did not perform a detailed analysis of health-related QoL in urologic cancer patients. For those purposes, special instruments such as the prostate-specific module EORTC QLQ-PR25 [45] are available, and from our study, we cannot derive conclusions for dealing with QoL problems that are specific for urologic cancer patients. Studies with patients suffering from other cancer entities should be performed to further validate the instrument in oncologic settings and to evaluate the generalizability of the findings of this examination.

In summary, the results of this study underline that health is a relevant dimension of QoL but not solely so. The importance of the domain income/finances shows that this aspect is also meaningful for understanding cancer patients’ life situation, even if they do not explicitly state that to be the case. The domain partnership/sexuality is especially sensitive for urologic cancer patients and should be taken into account in the cancer care setting. Domain importance is meaningful. Even if importance ratings are not necessary for qualifying a weighted global QoL score, they are useful tools for better understanding what is truly relevant for patients [46].