Profile of EORTC QLQ – C30 and BR-23 scores
We assessed 348 predominantly middle aged Arab women who were mostly receiving chemotherapy. The mean scores for QLQ – C30 and BR-23 indicated that the patients had poor to average functioning and intense symptom experience. We shall now examine the significance of these findings in the context of the international data.
The literature on EORTC QLQ has emphasized that comparison of data should go beyond the usual presentation of mean scores and significant differences [
4,
35,
39,
40]. Towards this end, Table
2 shows that, although the mean scores for QLQ – C30 indicated that the patients had poor to average scale scores, only 5.8% to 11.2% had scores that met the ≤ 33% criterion for problematic functioning, while 12.0% to 40.0% met the >66% criterion for more severe symptoms. In other words, despite the reality and popular perception of cancer as a killer disease, many patients undergoing effective modern treatment can expect to have tolerable levels of physical, psychological and social functioning and subjective wellbeing [
41‐
43]. This was particularly the case for the functional scales of the BR-23, where majority of subjects had scores indicating good functioning. It is interesting that, despite their condition, about one-third were hopeful about the future. These findings constitute an evidence base for the country's cancer control program, to boost national health education about prognosis in cancer [
8].
With regard to the pattern of functional scale scores of the QLQ – C30, the pattern of responses of our patients was similar to the international data for moderately extensive/severe disease, because the lowest scores were noted for physical functioning and role functioning, while the highest scores were for cognitive and social functioning [
44]. This is in line with the reality of a severely debilitating physical illness, where the much available family social support and institutional support can help to boost psychosocial wellbeing [
28].
Factors associated with HRQOL
Of the factors investigated, the noteworthy associations with HRQOL in univariate analyses were age, stage of cancer, receipt of radiotherapy and fatigue. Although the correlations of age with scale scores were rather low, the results are in line with the international data in showing that younger women with cancer tend to have poorer functioning and more intense symptoms, especially if they are on chemotherapy [
12‐
16]. It has been suggested that younger patients are more likely to suffer adverse effects because of induction of early menopause and possible infertility [
12]. While the finding of association of disease stage with HRQOL scale scores indicates that the questionnaire has known groups validity, the significant associations were only for three scales (role functioning, diarrhea and future perspective). In the studies that found no significant association between QOL and cancer stage at diagnosis, it has been suggested that the diagnosis of breast cancer is so stressful that it may result in a pattern of psychological morbidity for women in early stages that is similar to that experienced by women with more advanced disease [
15]. Coupled with the indication that future perspectives and body image perceptions were better for those who had last treatment over a month ago, [
12,
45], the above findings should be an evidence base for the contents of a cancer counseling clinic for our patients.
Predictors of HRQOL
The highlight of our regression analyses is that, contrary to the third hypothesis[
30], the functional scale scores were more important in predicting functional scales, than the scores of the symptom scales, while social functioning accounted for the highest proportion of variance in GQOL. This is in line with a Canadian report, which indicated that social functioning was strongly correlated with global QOL, while physical and role functioning were highly correlated [
46]. This result has been explained from the perspective of the phenomenon of response -shift [
45], which occurs if the patient's perception of the severity of the symptoms changes over time. It is suggested that many patients gradually adapt to the situation following diagnosis and treatment, becoming better at coping with symptoms, such that the patients no longer perceive the burden of the symptoms to be as great as the first time [
45]. Hence, for such patients, the physical, emotional and social limitations of the illness tend to become more important determinants of QOL rather than the physical symptoms. However, in a German study that assessed the role of the symptom scales in predicting HRQOL, it was found that fatigue was the most important predictor of QOL [
47]. While pain was the only symptom scale that entered our regression model for GQOL, we note that, in univariate analysis, fatigue had the highest correlation with GQOL, of the symptom scales of the QLQ-C30; and fatigue was a significant predictor for social functioning (2.3% variance, P < 0.006). Hence, clinicians should pay particular attention to how to alleviate fatigue, in addition to the usual emphasis on pain [
47].
The predictive power of financial difficulty shows that, despite institutional supports, cancer is associated with significant family financial burden. It appears that, although treatment was provided free -of -charge, families still made significant out -of -pocket expenses.
Comparison with the international data (additional file 1)
The relative youth of our subjects, in comparison with the international data, is noteworthy. The life expectancy at birth for Kuwaiti women is 78.95 years (2009 estimate:
https://www.cia.gov/library/publications/the-world-factbook/geos/KU.html). The following pattern emerged in comparing our scale scores with those of the international EORTC QLQ data (additional file
1)[
48,
49], as well as previous Kuwaiti QOL data from a general population study [
50] and women with multiple sclerosis [
32], using the global QOL scale of the 26-item WHOQOL-Bref. First, for the functional scales of the QLQ – C30, the Kuwaiti global QOL score was less than those of other countries by at least 10%. In particular, the GQOL score from the neighboring UAE was 20% higher, while that for three Islamic countries was 14% higher. The pattern was the same for physical and role functioning, except role functioning data from China [
24]. For emotional functioning, Kuwaiti scores were at least 10% less than the data from Korea, Sweden, and other Western nations. The tendency for the emotional functioning data from the UAE and three Islamic countries to be higher than that from Kuwait, did not reach significance (Effect size: 0.22, 95% C.I. = -0.02 – 0.46). For cognitive and social functioning, Kuwaiti scores were at least 10% less than the data from the UAE, the three Islamic countries and the Western nations.
Second, for the symptom scales of the QLQ – C30, the Kuwaiti data indicated more intense symptoms for dyspnoea, appetite and diarrhea and more financial difficulty (i.e., at least 10% difference).
Third, for the functional scales of the BR – 23, our data was at the middle of the range of the international data for body image, being much similar to the data from Iran and China, indicating better functioning than the Korean data (over 20% difference), but worse functioning than most data from the Western world. For sexual functioning, the Kuwaiti data was at the top of the range of the international data, being similar to the Iranian and UAE data, but indicating better functioning for most data from the Western world. Fourth, of the symptom scales of the BR-23, the mean scores indicated that Kuwaiti women had more intense symptoms than women from all the comparison countries for systemic side effects, breast symptoms and arm symptoms. Regarding systemic side effects, Kuwaiti women had at least 10% difference with women from the Western world, Korea, and Iran, but similar scores with women from the UAE (Effect size, 95% C.I: 0.09, -0.15 -0.33). The pattern was the same for breast symptoms, except that the Kuwaiti women had clinically significantly more intense symptoms than the women from the UAE.
Fifth, using the 0–100% scale of the WHOQOL-Bref[
32,
50], Kuwaiti women in the general population, women in the general population who rated themselves as being sick, and women attending clinic for multiple sclerosis, all had significantly higher global QOL scores than the women with breast cancer (Effect size, 95% C.I: 1.22, 1.10 – 1.35; and 0.83, 0.60 – 1.05, respectively. Comparison Kuwaiti data were adjusted for socio-demographic variables).
In summary, the profile of scale mean scores for QLQ – C30 indicated that Kuwaiti women had clinically significantly poorer global QOL and functional scale scores, and more intense symptom experience for most scales, in comparison with the international data, including data from neighboring Islamic countries. For the breast specific scale (BR-23), Kuwaiti women seemed to have clinically significantly better sexual functioning, but more severe treatment side effect symptom experience, especially systemic side effects and breast symptoms. The poor global QOL score of Kuwaiti women with cancer was confirmed by the fact that they had significantly lower scores than their compatriots in the general population and those with multiple sclerosis. It appears, therefore, that the experience of clinically more severe symptoms among Kuwaiti women, as a group, was a significant contributor to their low global QOL and comparatively poorer physical, emotional and social functioning [
30]. In view of this, clinicians treating Kuwaiti women for cancer should take particular interest in preparing the patients for the acute toxicities of breast cancer treatment [
12]. Other factors that could have contributed to their comparatively poor HRQOL are, their relative youth, the fact that they were on chemotherapy [
12‐
16], and the aggressive histopathologic picture of breast cancer in Kuwait [
27]. In other words, the biological and treatment side effect factors seemed to be more important than family and institutional supports in the cross-sectional HRQOL outcome of our patients. A longitudinal study is needed to confirm this trend. However, we note that the laudable institutional supports did not include any program on psycho-oncology. Our general experience is that women with breast cancer do not receive any formal psycho-education on the nature of the illness, the implications of treatment and how to cope with cancer. Our findings, therefore, call for a formal program of psycho-oncology at the KCCC, to counsel patients on the nature of the disease and treatment side effects, address their psychosocial concerns and explore methods of coping [
51].
Limitations and strengths of the study
Our findings cannot be generalized to the general population of women with breast cancer because the study was cross-sectional, the participants were a selected clinic population, and there is no general population reference data for the EORTC QLQ – C30 in Kuwait [
52]. In addition, we did not sort the patients into categories of disease progression and disease free survivors [
28], and the fact that the responses were based on interview could have introduced a bias for patients who may have wished to respond privately. However, we have used an internationally well validated questionnaire to describe the HRQOL of a fairly large number of breast cancer survivors, and we have presented our results in such a way as to make them comparable with data from other countries. Also, there is published data on general population norms for a popular QOL instrument (WHOQOL -BREF) that was presented using the 0 – 100% scale (or the percentage scale maximum method – % SM) [
50]. The value of the %SM measure is that it can be used for making comparison with the scale scores of other questionnaires [
53].
Of the Arabic samples reported in the literature [
9,
10], ours is the largest. Furthermore, we categorized scale scores into those with/without significant problems [
20,
28,
37], and applied the principle of clinically significant difference in scale scores between groups [
35] to compare our results with the international data. Hence, we met the need to provide clinically-based benchmarks in interpreting QOL data, so as to make them meaningful to clinicians[
4,
5,
44]. Finally, it appears that our patients had similar demographic characteristics with those of breast cancer clinic populations in Kuwait [
27,
38].