05PM457.qxd

9/8/01

1:55 pm

Page 387

Palliative Medicine 2001; 15: 387–397

Cross-cultural validation of the McGill Quality of Life questionnaire in Hong Kong Chinese Raymond SK Lo Senior Medical Officer, Palliative Care Unit, Department of Medicine and Geriatrics, Shatin Hospital, Hong Kong, Jean Woo Chief of Service, Department of Medicine and Geriatrics, Shatin Hospital, Hong Kong, Karen CH Zhoc Research Assistant, Palliative Care Unit, Shatin Hospital, Hong Kong, Charlotte YP Li, Research Assistant, Palliative Care Unit, Shatin Hospital, Hong Kong, Winnie Yeo Associate Professor, Department of Clinical Oncology, Prince of Wales Hospital, Hong Kong, Philip Johnson Professor and Chairman, Department of Clinical Oncology, Prince of Wales Hospital, Hong Kong, Yvonne Mak Medical Officer, Bradbury Hospice, Hong Kong and Joseph Lee Hospital Chief Executive, Bradbury Hospice, Hong Kong Abstract: The main focus of palliative care services is to improve patients’ quality of life (QOL). The potential value of assessment of QOL in palliative care is being increasingly recognized. The McGill Quality of Life questionnaire (MQOL) is designed specifically for palliative care patients, but its cross-cultural validity needs to be determined before it can be applied in populations of different cultures and ethnic groups. The cross-cultural validity of MQOL was investigated using a translated and modified version in Chinese – the MQOL-HK – in 462 palliative care patients in Hong Kong. Results show that the MQOL-HK is acceptable, valid and reliable. There is good acceptability, construct validity, convergent and divergent validity, test–retest and inter-rater reliability. Our study confirms that QOL does have cross-culturally robust constructs. Principal components analysis shows that the domains of physical, psychological, existential and support are all relevant and applicable in Chinese culture. Multiple regression analysis reveals that existential domain is the most important domain in predicting overall QOL. ‘Face’, eating and sex are additional facets of QOL that also need to be considered. The worst physical symptom on admission is the item of QOL with the lowest score, which need more care and attention by palliative care workers. A cross-culturally validated QOL instrument cannot just help ensure an accurate evaluation of profile, determinants, and changes of QOL, but is also a valuable asset for future comparison and evaluation of palliative care services and interventions across the world. Key words: quality of life; palliative care; hospice; culture; validation; end of life; Hong Kong Resumé: L’objectif principal des soins palliatifs est d’améliorer la qualité de vie des patients (QOL). L’intérêt de l’évaluation de QOL en soins palliatifs est de plus en plus reconnue. Le questionnaire McGill sur la qualité de vie (MQOL) est conçu spécialement pour les patients de soins palliatifs, mais sa pertinence transculturelle doit être évaluée avant de pouvoir l’utiliser auprès de populations de cultures et d’ethnies différentes. La pertinence transculturelle du MQOL a été évaluée en utilisant une version traduite en chinois et modifiée – le MQOL-HK – auprès de 462 patients montre que le MQOL-HK est utilisable, valable et fiable. Le test démontre qu’il est bien accepté, de conception valable, avec une validité convergente et divergente, une bonne reproductibilité chez

Address for correspondence: Dr Raymond SK Lo, Department of Medicine and Geriatrics, Shatin Hospital, 33 A Kung Kok Street, Ma On Shan, New Territories, Hong Kong. E-mail: [email protected] © Arnold 2001

0269–2163(01)PM457OA

05PM457.qxd

9/8/01

388

1:55 pm

Page 388

RSK Lo et al. le même patient et entre patients différents. Notre étude confirme que le MQOL a vraiment un profil de test transculturel. L’analyse des principaux items montre que les états physiques, psychologiques et existentiels ainsi que le soutien extérieur sont évalués de façon pertinente et que le test est applicable à la population chinoise. L’analyse de régression multiple révèle que l’état existentiel est le facteur d’information le plus important pour prédire la QOL globale. « L’apparence physique », la nourriture et l’activité sexuelle sont des éléments supplémentaires du QOL qui doivent être considérés. Le symptôme physique le plus mal au moment de l’admission est l’item du MQOL qui ont les scores les plus faibles; ils requièrent de ce fait un supplément d’attention de la part des soignants en soins palliatifs. Un instrument validé d’évaluation transculturelle de la QOL n’est pas destiné à fournir seulement une évaluation correcte du niveau, des déterminants et des changements de la QOL; c’est aussi un moyen valable pour les futures comparaisons et évaluations des soins palliatifs à travers le monde. Mots-clés: qualité de vie; soins palliatifs; services de soins palliatifs; culture; validation; fin de vie; Hong Kong

Introduction In 1987, the Royal College of Physicians in the UK recognized palliative medicine as a specialty, defining it as ‘the study and management of patients with far advanced disease for whom the prognosis is limited and the focus of care is quality of life.’1 In 1990, the World Health Organization also emphasized that ‘the ultimate goal of palliative care is the achievement of the best quality of life for patients and their families.’2 Quality of life (QOL) serves as the final common pathway for palliative care services,3 and should be considered as the main outcome measure.4 Measuring patients’ QOL is essential, so that aspects requiring intervention and counselling can be identified, and benefit of care and treatment can be assessed.5 QOL is however difficult to define and measure.6–9 It is multi-dimensional in that it involves various domains such as physical, functional independence, psychological, social, spiritual, and existential.10,11 It is subjective in that it should reflect the impact of a disease from a patient’s perception rather than what professionals or even carers perceive.12 It is dynamic in that it may change with time or disease progression.6 Notwithstanding the challenges and difficulties, progress has been made in recent years, in understanding and measuring QOL in palliative care.13 Various QOL measurement tools have since been designed for cancer patients,14–17 but they may not be ideal for use in the palliative phase.18,19 QOL measurement tools for palliative care patients need to be specially designed, in order to assess the areas

for which palliative care is most effective, such as psychosocial and spiritual problems.20 With the increase in researches in QOL in palliative care, it is especially important that the measurement tools used should also be valid and reliable. The McGill Quality of Life questionnaire (MQOL) was developed and validated in Canada, especially for palliative care patients.21–24 The MQOL is a multi-dimensional questionnaire measuring QOL in four domains: physical, psychological, existential and support. It is comprehensive and relevant for the palliative care population, and can be self-completed. It consists of 16 items and also a single item rating the overall QOL. This single item score (SIS) is useful in indicating the patient’s perception of his/her QOL taken as a whole. All the response categories are based on a numerical scale from 0 to 10, with verbal anchors at the ends of the scale. The questionnaire items were derived from informal patient interviews, literature reviews, and existing instruments. The MQOL has been demonstrated to have good acceptability, practicability and validity. The test–retest reliability and responsiveness to change have also been tested (SR Cohen, personal communication). The favourable psychometric properties of MQOL hence make it particularly suitable for QOL assessment in palliative care.25 Other scales are also available in palliative care, but various limitations have been identified which constrain their uses.6 The MQOL is also preferred because the existential domain can be measured; the physical domain is

05PM457.qxd

9/8/01

1:55 pm

Page 389

Cross-cultural validation of MQOL in Hong Kong Chinese 389 important but not predominant; and positive contributions are also taken into account. The existential domain explores the perception of purpose, meaning in life, and the capacity for personal growth and transcendence. A pilot study testing the applicability of the MQOL in Chinese palliative care patients had been performed in Hong Kong in 1997.26 The study was carried out, during a 2-month period, on all new admissions to a paired palliative care ward in Shatin Hospital. Results from this pilot study demonstrate that MQOL is also practical and acceptable for local Chinese. The translated version of the MQOL can be understood and completed by patients. Many patients appreciated the questionnaire as an opportunity to discuss issues which they found difficult to raise initially. Although some recruited patients could not complete the questionnaire due to poor physical condition, this should not preclude those who could complete it from doing so. However one potential limitation of the further wide use of MQOL, as with any other QOL measurement, is its validity in different ethnic groups and cultures. For example, the meaning and interpretation of QOL may not be the same for Chinese as for the western population. Some aspects like the contextual meaning of pain and emotional expression may be different for Chinese people.27 Different social environments and health care systems will also influence the population preferences for the desirability of different disability and handicap states. Perceptions and valuations of the quality of life domains may or may not vary from culture to culture. This is an important research issue, as it questions the extent to which a QOL scale developed in one country or ethnic group, can be uncritically applied in another. Given this preceding caution, there is early evidence, however, to suggest that the QOL concept in palliative care does have cross-culturally robust constructs.28 Physical, psychological, social functions, noxious symptoms and sensations and family hardship may appear broadly relevant across different cultures.13 In Hong Kong, the physical, psychological, social, and spiritual domains of QOL were confirmed to be relevant, from extensive field test involving healthy subjects and patients with different disease conditions, in the development of the World Health Organisation Quality of Life questionnaire.29 But still, as QOL emerges as a distinctly

Western notion, interpretations of its concepts in other cultures and religions should be cautious. Cross-cultural validation of the MQOL is needed before it can be formally applied in a population with a different culture. We investigate the cross-cultural validity and reliability of the MQOL in Hong Kong Chinese palliative care patients. The study will contribute to the understanding in the different cultural perspectives of QOL in palliative care, as well as further identifying palliative care patients’ respective needs. A cross-culturally validated QOL instrument will be a valuable asset, for future comparison and evaluation of palliative care services and interventions across the world.

Methods We evaluated the cross-cultural validity and reliability of MQOL in Chinese palliative care patients. This is part of an ongoing multi-centre prospective study designed to evaluate the profile, determinants, and longitudinal changes of QOL of palliative care patients in Hong Kong. Participating centres included Shatin Hospital Palliative Care Unit, Prince of Wales Hospital Department of Radiotherapy and Oncology, and Bradbury Hospice, which together serve a geographical cluster of Hong Kong with a population of approximately one million. Through this multi-centre collaboration, the representativeness of the study sample was enhanced. All consecutive new admissions of advanced incurable cancer patients to the above participating centres within a period of 18 months from August 1998 to January 2000 were recruited. Their QOL was assessed using a translated and slightly modified version of MQOL for the Hong Kong Chinese (MQOL-HK). Validity and reliability of the MQOL-HK were assessed using the QOL scores on admission of all the recruited cases. The QOL of recruited subjects was evaluated within 72 h of admission. These patients are also being followed up in an ongoing longitudinal study, with reassessment of QOL at regular intervals during hospitalization until death, or in the community post-discharge. The MQOL-HK was first developed from translation of the original MQOL into Chinese, by the first and second author of the pilot study during

05PM457.qxd

9/8/01

390

1:55 pm

Page 390

RSK Lo et al.

1997. Since then the translation had been further improved. The final version was then independently reviewed and successfully back-translated by the third author of this study, without previously seeing the original MQOL. The back-translated version has a close concordance with the original MQOL, as verified independently by a volunteer who is an university graduate with an interest in linguistics, with no previous background in medicine or health. All participants are fluent in both English and Chinese. The participation of non-health professionals in the back translation and verification process helped to ensure that the MQOL-HK could be easily understood by patients. In the MQOL-HK, the item of ‘feeling close to people’, which was modified in the later MQOL studies, was retained because it could be understood by the Hong Kong patients during the pilot study in 1997. Three experimental questions were also added in the MQOL-HK: ‘face’, eating, and sex. The first two questions are added as they are often conceived to be important components of Chinese culture, which may influence Chinese patients’ QOL.29 ‘Face’ refers to being respected or feeling being respected. Eating refers to the ability to eat, as well as enjoyment from food. A question on sex is also included, as its potential influence on palliative care patients’ QOL is uncertain, and needs to be investigated. All items are scored on a 0–10 categorical scale, with descriptive anchors at either end, like in the original MQOL. Some items are negatively phrased to avoid implicit response bias. For the final statistical analysis all scores are transposed on a 0–10 scale, with 0 representing the least and 10 the most desirable situation. All the recruited patients were interviewed by the research assistant, who is a university psychology graduate. During the interview the purpose of the study was explained and patients’ informed voluntary consent was obtained. The MQOL-HK was read by the patient unaided if possible, or otherwise read out aloud by the research assistant. Any queries on the questionnaire items were explained. All the item scores were rated by patients themselves. Patients with impaired mental status precluding the completion of the questionnaire were excluded. Validity and reliability of the MQOL-HK were assessed on the QOL scores obtained from the first interview of all recruited cases. First, factor

analysis using the principal components methods with orthogonal varimax rotation was performed for all the MQOL-HK item scores, to identify the different potential components with eigenvalues greater than 1 in the MQOL-HK. Sub-scales for the MQOL-HK were derived by grouping together those items with highest loading on each resulting factor. The internal consistency of the MQOL-HK and its sub-scales were calculated by Cronbach’s alpha. The construct validity of the MQOL-HK was then assessed by measuring the accuracy of the MQOL-HK sub-scale and total scores, in predicting the overall QOL of the patient. The MQOLsub-scale score is calculated as the mean of the scores on the individual items of that sub-scale, whereas the total score is calculated as the mean of all the sub-scales scores. As there is no established gold standard of QOL for Hong Kong palliative care patients as yet, the SIS measuring the overall QOL, rated by the patients as part of the MQOL-HK, was used for comparison. The correlation between the MQOL sub-scale scores and total score with the SIS was assessed by Spearman correlation coefficient. Multi-regression techniques were used to measure the degree to which the MQOL sub-scales and total score could predict the overall QOL. Convergent and divergent validity are also important components of construct validity. They reflect the extent to which the MQOL-HK correlates with similar and different constructs of another established measure of QOL. This was measured in the first 50 recruited cases by comparing patterns of correlation between the MQOL-HK and the Spitzer Quality of Life Index (SQLI), using Spearman correlation coefficient. The SQLI is a five-item questionnaire measuring activity, activities of daily living, general health, outlook and support in cancer patients.30 Although the SQLI was not designed for palliative care patients, it was used in the validity study of the original MQOL,21,24 and also chosen in our study for similar reasons. It has an item of outlook, a domain addressed by the MQOL-HK existential domain but not by most other questionnaires. It is also brief to administer without adding too much burden to advanced cancer patients. Only the first 50 consecutive patients were recruited for this evaluation, so that not too many patients were troubled with the task of completing two QOL questionnaires at one time.

05PM457.qxd

9/8/01

1:55 pm

Page 391

Cross-cultural validation of MQOL in Hong Kong Chinese 391 The test–retest reliability of the MQOL-HK was assessed on a separate group of 20 consecutively recruited patients. These patients were interviewed twice by the research assistant, first on admission and 2 days later. The MQOL-HK item, sub-scale and total scores obtained on these two occasions were compared using intra-class correlation coefficients. A time interval of 2 days was deliberately chosen to reduce the burden on the patient and to minimize recall bias, while accepting that there is a possibility that patients’ QOL may change during this interval. Inter-rater reliability was also assessed on another 20 consecutively recruited cases to evaluate the agreement between two different interviewers. The cases were interviewed twice, first by the research assistant on admission, and then by a palliative care nurse two days later. The MQOL-HK scores obtained on these two separate occasions were compared. The number of patients recruited for each of the test–retest and inter-rater reliability study was limited to 20, so that not too many patients were burdened with completing the QOL questionnaire twice within a short time. It is accepted that the test–retest and inter-rater reliability result will be limited due to this small sample. A larger scale evaluation would require another separate study. This cross-cultural validation and reliability study was part of an ongoing multi-centre longitudinal follow-up study of palliative care patients’ QOL. The research protocol received approval from the ethics committee of the Chinese University of Hong Kong, and Bradbury Hospice ethics committee.

Results A total of 462 cases were successfully recruited for the validity and reliability study. These constituted 45% of the total admissions during the recruitment period. Those who could not be recruited were comatose/semi-conscious, cognitively impaired with dementia/confusion, physically too ill with extremely poor general condition, or there was a language barrier. The mean age of recruited patients was 61.5 years (standard deviation was 14.5, age range between 16 and 89). Fifty-three per cent were male. The five most common diagnoses were lung, colorectal, liver, breast and nasopharyngeal cancers. A half of the recruited subjects had one or more

coexisting illnesses. Seventy-five per cent of the subjects had received primary education or above. Forty-seven per cent of the subjects were retired; 14% were housewives; and 15% of the subjects had not worked for at least a year due to the current illness. Forty-three per cent of subjects believed in traditional Chinese ancestral worship; 24% believed in either Christianity or Buddhism; and 32% were atheists. On the whole, the average time required to complete a questionnaire was 30 min, ranging from 15 to 45 min. Some subjects could complete the MQOL-HK on their own, whereas others needed some explanation by the research assistant. Many subjects seemed to appreciate the opportunity to talk about issues which they initially found difficult to voice. Some subjects became aware for the first time of their own inner conflicts which had been causing them distress. One patient was able to discuss with the research assistant private psychosocial problems, when health care professionals had had difficulty in establishing rapport. Research assistants reported any identified problems to the palliative care team for further follow-up, provided that the subjects gave consent. The mean and standard deviations of the MQOLHK individual item scores are listed in Table 1. The average QOL of this sample of 462 palliative care patients, as represented by the mean SIS score was Table 1 The mean and standard deviations of the MQOLHK individual items score MQOL-HK items

Sample mean score (total out of 10) and SD in brackets

Physical symptom 1 Physical symptom 2 Physical symptom 3 Physical well-being Depressed Anxious Sad Fear of future Personal existence Achieving life goals Life … worthwhile Life control Feel good about myself Closeness to people Every day ‘seems a gift’ World is caring Face Eating Sex Single item score

4.59 5.90 7.47 5.44 7.25 7.12 7.72 8.46 6.25 6.55 6.58 4.91 6.64 7.43 6.70 6.58 8.87 5.34 8.51 6.35

(3.01) (3.37) (3.37) (2.21) (3.56) (3.78) (3.39) (3.01) (3.37) (3.37) (2.21) (3.56) (3.78) (3.39) (3.07) (3.26) (3.04) (3.06) (3.70) (2.45)

05PM457.qxd

9/8/01

392

1:55 pm

Page 392

RSK Lo et al.

6.35 (median 7, SD 2.53). The individual item with the lowest mean score was the item on the first physical symptom, with a mean score of 4.59. The individual item with the highest mean score was the item on ‘face’, which had a mean score of 8.87. Factor analysis revealed five components of QOL: (1) physical; (2) psychological; (3) existential well-being; (4) support; (5) sex. All have eigenvalues greater than 1.0. The physical domain covers physical symptoms, physical well-being, and eating. The psychological domain covers aspects such as feeling depressed, nervous, worried, sad, fear and ‘face’. The existential well-being domain relates to aspects of meaning of existence, life goals, feeling life to be worthwhile, self-content, and seeing life as a burden. The support domain is about the aspects of feeling supported, and the world as caring. The fifth domain of sex consists of the item on sex alone. Together the five factors account for 59% of total variance. The factor loadings forming the five sub-scales are shown in Table 2. The grouping of items is the same between the MQOL-HK and MQOL apart from two items. The item on physical well-being does not load clearly with any factor in the MQOL, but group under physical in MQOL-HK; whereas the item on life control does not load clearly on any factors in the MQOL-HK. The internal consistency of the MQOL-HK is good with Cronbach’s alpha value of 0.83. Inclusion of the experimental questions of ‘face’, eating, and

sex did not affect the internal consistency. The psychological, existential and support sub-scales also show good internal consistency with Cronbach’s alpha above 0.7 (Table 3). Table 4 shows the Spearman correlation coefficients between the MQOL-HK sub-scale scores, total scores and the SIS. There is good construct validity of the MQOL-HK, as shown by significant correlation between the MQOL-HK total scores and the SIS, which represents the patient’s selfrated overall QOL (r  0.5, P  0.001). As for the sub-scales, the existential and psychological have the highest correlation with the SIS. There is good convergent validity of the MQOL-HK with the SQLI, with significant correlation between the MQOL-HK total score and the SQLI total score (r  0.40, P  0.004). There is good correlation between MQOL-HK sub-scales and the SQLI items which measure similar construct, e.g. MQOL-HK physical domain and SQLI health item. There is Table 3 Internal consistency of the MQOL-HK and its sub-scales Cronbach’s alpha Complete MQOL-HK with items of ‘face’, eating, and sex Physical sub-scale Psychological sub-scale Existential sub-scale Support sub-scale

0.83 0.68 0.85 0.82 0.70

Table 2 Factors loading for the MQOL-HK items. Sub-scales were formed from the mean of the items typed in bold in each column MQOL-HK items

Physical

Psychological

Existential well-being

Support

Sex

Physical symptom 1 Physical symptom 2 Physical symptom 3 Physical well-being Depressed Anxious Sad Fear of future Personal existence Achieving life goals Life … worthwhile Life control Feel good about myself Closeness to people Every day seems a gift World is caring Face Eating Sex

0.62 0.77 0.72 0.65 0.17 0.16 0.11 0.045 0.045 0.028 0.015 0.22 0.062 0.016 0.22 0.024 0.11 0.51 0.01

0.20 0.10 0.11 0.10 0.80 0.81 0.83 0.69 0.098 0.14 0.072 0.24 0.12 0.12 0.62 0.047 0.50 0.13 0.16

0.12 0.015 0.049 0.14 0.12 0.040 0.098 0.13 0.77 0.79 0.82 0.41 0.72 0.26 0.27 0.13 0.0088 0.15 0.10

0.062 0.0031 0.067 0.17 0.16 0.017 0.10 0.14 0.16 0.0047 0.18 0.097 0.11 0.80 0.032 0.87 0.057 0.099 0.099

0.025 0.012 0.017 0.15 0.03 0.035 0.035 0.13 0.088 0.19 0.016 0.5 0.093 0.087 0.078 0.018 0.015 0.069 0.86

05PM457.qxd

9/8/01

1:55 pm

Page 393

Cross-cultural validation of MQOL in Hong Kong Chinese 393 poor correlation between the MQOL-HK subscales and the SQLI items which measure different construct, e.g. MQOL-HK psychological domain and the SQLI activity item, thus demonstrating divergent validity. The SQLI total score correlates slightly less than the MQOL-HK with SIS (r  0.44 and r  0.45 respectively). Correlation between the MQOL-HK and the SQLI scores are listed in Table 5. As seen in Table 4, all the MQOL-HK sub-scale scores have statistically significant correlation with the SIS. Amongst the five sub-scales, existential well-being has the highest correlation at r  0.47

(P  0.000). Further multiple regression analysis also confirmed that the existential well-being was the most important sub-scale in the regression equation for predicting SIS, with the highest standardized  coefficient of 0.35 (P  0.000). The standardized  coefficients of the sub-scales are listed in Table 6. All the MQOL-HK sub-scales together are able to predict 31% of the variance in SIS (adjusted R2  0.31, P  0.000), which is better than the MQOL-HK total score on its own (adjusted R2 0.25, P  0.000). There is good inter-rater reliability of the MQOL-HK, with high correlation between the

Table 4 Spearman’s correlation coefficients between MQOL-HK total score, sub-scale scores, and SIS score MQOL-HK

SIS

Total

Physical

Psychological

Existential well-being

Support

Sex

r  0.50 P  0.000

r  0.26 P  0.000

r  0.39 P  0.000

r  0.47 P  0.000

r  0.24 P  0.000

r  0.12 P  0.014

Table 5 Spearman’s correlation coefficients between MQOL- HK and SQLI MQOL-HK Total

SIS

Physical

Psychological

Existential well-being

Support

Sex

SQLI total

r  0.40 P  0.004

r  0.44 P  0.001

r  0.42 P  0.002

r  0.35 P  0.010

r  0.29 P = 0.040

r  0.14 P  0.327

r  0.11 P  0.444

SQLI activity

r  0.13 P  0.374

r  0.31 P  0.024

r  0.27 P  0.060

r  0.06 P  0.651

r  0.22 P  0.114

r  0.02 P  0.892

r  0.008 P  0.956

SQLI daily living

r  0.11 P  0.446

r  0.23 P  0.095

r  0.13 P  0.361

r  0.13 P  0.371

r  0.14 P  0.321

r  0.23 P  0.095

r  0.11 P  0.424

SQLI health

r  0.29 P  0.040

r  0.12 P  0.410

r  0.49 P  0.000

r  0.36 P  0.010

r  0.02 P  0.870

r  0.13 P  0.342

r  0.13 P  0.362

SQLI support

r  0.29 P  0.037

r  0.34 P  0.013

r  0.03 P  0.827

r  0.15 P  0.296

r  0.17 P  0.228

r  0.26 P  0.069

r  0.18 P  0.196

SQLI outlook

r  0.40 P  0.003

r  0.36 P  0.010

r  0.36 P  0.002

r  0.38 P  0.006

r  0.38 P  0.005

r  0.17 P  0.228

r  0.08 P  0.565

The figures in bold represent the Spearman’s correlation coefficients which are statistically significant at P  0.05.

Table 6 Standardized beta coefficients of the MQOL-HK sub-scales in multiple regression analysis with SIS as dependent variable MQOL-HK Physical

Psychological

Existential well-being

Support

Sex

Standardized beta coefficient

0.16

0.21

0.35

0.093

0.028

Statistical significance

P  0.000

P  0.000

P  0.000

P  0.028

P  0.484

05PM457.qxd

9/8/01

394

1:55 pm

Page 394

RSK Lo et al.

research assistant and the palliative care nurse on conducting the MQOL-HK for the 20 palliative care patients. The intra-class correlation for the MQOL-HK total score was 0.83, which is strongly statistically significant at P  0.0002, whereas the intra-class correlation for the SIS was also high at 0.9, statistically significant at P  0.000. There is also good test–retest reliability, with a high intraclass correlation coefficient of 0.85 (P  0.000), between the MQOL-HK total scores rated by 20 patients 2 days apart.

Discussion The importance of QOL in palliative care is being increasingly recognized, and the potential value of QOL as an outcome indicator is relevant for different palliative care services across the world. When evaluating palliative care patients’ QOL, it is essential, however, that the QOL instrument used should be valid and reliable. The MQOL was developed in Canada with good acceptability, validity and reliability, but its cross-cultural validity needs to be examined before it can be applied in patient populations of different countries and ethnic groups. A cross-culturally validated QOL instrument will not only contribute towards providing valuable scientific information concerning the needs of palliative care patients, but also has an additional advantage in allowing international comparisons in QOL. Our study is able to demonstrate the cross-cultural validity of the MQOL in Hong Kong Chinese. The successful completion of the questionnaire by our series of 462 subjects confirms findings from our previous pilot study, that measurement of QOL using the MQOL is acceptable to palliative care patients in Hong Kong. Results from our study support the notion that quality of life in palliative care does have cross-culturally robust constructs. In our study, the QOL domains represented in the MQOL are applicable, and can be understood by Hong Kong Chinese. Factor analysis using the principal components methods groups most of the items under their respective domains as in the MQOL studies. The domains of physical, psychological, existential and support seem to apply across cultures in palliative care. Although the significance of existential well-being and spirituality as a separate domain of QOL has been debated,31,32 our study has

further provided confirmatory evidence of its importance and relevance in patients with advanced incurable illness. It is illustrated from our study that the existential well-being is more than just a Western concept, and that its influence on QOL is present in Hong Kong Chinese. Furthermore, regression results prove that existential well-being is the most important sub-scale in predicting the overall QOL for our patients. This interesting similarity between Hong Kong Chinese and the West in perception of QOL, may be partly due to the fact that there is much blending of the two cultures, owing to the historical connection of Hong Kong with the UK. Comparable findings are also found in the evaluation of handicap, where Hong Kong Chinese have similar valuations to UK subjects on different handicap scenarios.33 Handicap and QOL may truly have cross-culturally robust constructs. Nonetheless, there may still be other possible cultural factors, unique in influencing Chinese patients’ QOL. We investigated three potential factors: face, eating and sex. ‘Face’ is a commonly used word and a well-known component in Chinese culture. Its importance in Chinese palliative care patients, however, has not been extensively studied. Our results confirm that ‘face’ is relevant, and factor analysis identifies it as a factor belonging to the psychological domain. This makes clinical sense, as ‘having face’, i.e. being respected or feeling being respected, may help improve the psychological well-being of a patient. ‘Face’ refers to respect not just from family and friends, but also respect from health care staff. The item on ‘face’ actually has the highest mean score out of all the individual items in our study, probably reflecting the degree of its importance and the common traditional practice of ‘giving face’ in Chinese culture. ‘Face’ in QOL in palliative care deserves further study, and exploration of its relevance in the Western culture is worthwhile. The original MQOL does not contain an item measuring a similar aspect. The experimental question on eating also addresses another aspect pertinent to Chinese culture. The question asks our patients whether illness has affected their enjoyment from food. This item received an average score of 5.34 in our study. Factor analysis identifies it as belonging to the physical domain. Enjoyment from food will depend on the physical ability to eat, as well as appetite, which is influenced by presence and severity of

05PM457.qxd

9/8/01

1:55 pm

Page 395

Cross-cultural validation of MQOL in Hong Kong Chinese 395 other physical symptoms. Although eating and enjoyment from food may be related to other domains like psychological (the joy and satisfaction from a good meal) or support (cherishing a meal with family and friends), results from our sample of patients indicate that it is most relevant to the physical domain. Similar to ‘face’, the aspect of eating and enjoyment from food may not be confined just to Chinese. Its relevance in other culture needs to be explored. The third experimental question in MQOL asks about the effect of illness on patients’ normal sex lives. Sexuality in palliative care patients refers not just to physical intercourse. Sexual satisfaction can also be derived from kissing, embracing, caress or massage. It relates not only to just the physical aspect, but also to the psychological aspect in satisfaction, existential aspect in self-esteem, or the love and support aspect from spouse/partner. It crosses different domains. Previous research documents that this need of palliative care patients is often ignored.34 Its significance in different cultures also needs to be fully determined. Factor analysis from our results show that sex forms a separate domain on its own, though it has the lowest correlation with the SIS and is not useful in predicting overall QOL. In our study, the average score on the item of sex is high at 8.51. This may genuinely be that Hong Kong Chinese patients derive good sexual satisfaction in their advanced stages of illness. Alternatively, our patients may be embarrassed by the item, and simply rate it indifferently with an overall high score. In any case, the sex domain seems to be the domain with the least influence on well-being of our patients. The sex domain has the least correlation with the SIS. It also explains the least amount of variance amongst all domains. More studies are needed. Some differences were found between our study and the previous MQOL studies. The item on life control did not fit into the existential domain as in the MQOL. The impression from our interviewer was that, out of all items, life control is the item which our patients had the most difficulty in understanding. To be in control of one’s life may perhaps be a Western concept, not equally perceived in a traditional Oriental culture of collectivity and inter-dependency. This may represent one potential genuine difference between the two cultures. Another difference was that the physical well being

item can be grouped under the physical domain in our study, whereas it did not load clearly with any components in the MQOL. This may be simply related to the difference in wording between the two questionnaires. As expected, the physical subscale also has a low internal consistency in the MQOL-HK as in the MQOL, as it asks about the top three troublesome symptoms which are often unrelated. Our study is able to identify the aspects of QOL which will need better attention in future for our local Chinese palliative care patients. The item with the lowest score on admission is the first physical symptom ranked by patients. The mean score for this item is low at 4.59 out of 10. This illustrates the important fact that patients on transfer or admission to the hospice do still have difficult physical symptoms, which have not yet been adequately relieved. It reminds us again of the paramount importance of prompt symptom control, which is always the absolute fundamental and essential requirement in palliative care services. Furthermore, the effect of psychosocial or spiritual counselling can only be maximized when the patient is relatively free from physical distress. Another area highlighted from our study which requires more attention is the eating aspect. Paying more attention to diet, anorexia, and dysphagia problems will improve not just the nutrition, but also the satisfaction and enjoyment from eating, and ultimately the QOL. More care and effort is need in this facet of palliative care. Our study demonstrates that the translated and modified version of the MQOL, the MQOL-HK, is acceptable, valid and reliable for further use in Hong Kong. So far it has been successfully completed in over 400 patients. It has good construct validity as demonstrated by its correlation with the single item score representing the overall score of QOL. It also has good convergent and divergent validity as shown by its pattern of correlation with the SQLI. It has good inter-rater and test–retest reliability. It will therefore be a most useful tool for evaluating QOL in Chinese palliative care patients. Our ongoing longitudinal study using the MQOLHK will be able to contribute further information on this important aspect. Needs of patients and effectiveness of services can hence be better ascertained. One limitation with the MQOL-HK however, as with any other QOL scale, is that it cannot be applied in debilitated patients who could not

05PM457.qxd

9/8/01

396

1:55 pm

Page 396

RSK Lo et al.

complete the questionnaire. Further research is needed in addressing the QOL issue in these difficult situations, to ensure that this vital aspect in holistic care is not neglected for these patients in need. Finally, more qualitative or quantitative research will be needed, in further confirmation and evaluation of the QOL domains which are important to the population.

5 6 7 8

Conclusions 9

Our study demonstrates the cross-cultural validity of the MQOL in Hong Kong Chinese. The translated and modified version, the MQOL-HK, is acceptable, valid, and reliable for use in palliative care patients in Hong Kong. Our study results confirm that quality of life does have cross-culturally robust constructs. The conventional physical, psychological, social and existential domains are relevant in Chinese culture, with existential well-being as the most important domain. ‘Face’, eating and sex are also confirmed to be relevant facets in palliative care. Prompt physical symptom control on admission and aspects of eating, are the needs of patients identified from our study which require more focus and attention. Acknowledgements This study is supported by a grant from the Society of Promotion of Hospice Care Ltd, Hong Kong. We would like to thank Dr Robin Cohen and her team for agreeing to our translation and modification of the MQOL for our study, and for her valuable comments. We would like to thank all the staff in participating centres for their help. Above all, we would like to thank all the patients who have taken part in the research. Without their contribution, this study would not have been possible.

10 11 12 13

14

15

16

17

References 18 1 Hillier R. Palliative medicine – a new specialty. BMJ 1988; 297: 874. 2 World Health Organisation. Cancer pain relief and palliative care: report of a WHO expert committee. Geneva, Switzerland: World Health Organisation, 1990. 3 Schipper H. Quality of life: the final common pathway. J Palliat Care 1992; 8(3): 5–7. 4 Richards MA, Ramirez AJ. Quality of life: the main

19 20 21

outcome measure of palliative care. Palliat Med 1997; 11: 89–92. Finlay I, Dunlop R. Quality of life assessment in palliative care. Ann Oncol 1994; 5: 13–18. Donnelly S, Walsh D. Quality of life assessment in advanced cancer. Palliat Med 1996; 10: 275–83. Loew F, Rapin Ch. The paradoxes of quality of life and its phenomenal approach. J Palliat Care 1994; 10(1): 37–41. Hopgood P. Progress, problems and priorities in quality of life research. Eur J Cancer 1992; 28A: 1748–52. Schipper H, Levitt M. Measuring quality of life: risks and benefits. Cancer Treat Rep 1985; 69: 1115–23. Cella DF. Quality of life: concepts and definitions. J Pain Symptom Manage 1994; 3: 186–92. Cella DF, Tulsky DS. Measuring quality of life today: methodological aspects. Oncology 1990; 4: 29–38. Slevin ML, Plant H, Lynch D et al. Who should measure quality of life, the doctor or the patient? Br J Cancer 1988; 57: 109. Clinch JJ, Dudgeon D, Schipper H. Quality of life assessment in palliative care. In: Doyle D, Hanks G, MacDonald N eds. Oxford textbook of palliative medicine, 2nd edn. Oxford. Oxford University Press, 1998: 83–94. De Haes JCJM, van Knippenberg FCE, Nejit JP. Measuring psychological and physical distress in cancer patients: structure and application of the Rotterdam Symptom Checklist. Br J Cancer 1990; 62: 1034–38. Cella DF, Tulsky DS, Gray G et al. The Functional Assessment of Cancer therapy (FACT) Scale: Development and validation of the general measure. J Clin Onc 1993; 11: 570–79. Schipper H, Clinch D, McMurray A, Levitt M. Measuring the quality of life of cancer patients: the Functional Living Index – Cancer: development and validation. J Clin Onc 1984; 2: 472–83. Aaronson NK, Ahmedzai S, Bergman B et al. The European organisation for research and treatment of Cancer QLQ-C30. A quality of life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993; 85: 364–76. Cohen SR, Mount BM. Quality of life in terminal illness: defining and measuring subjective well-being in the dying. J Palliat Care 1992; 8(3): 40–45. Mount BM, Scott JF. Whither hospice evaluation. J Chron Dis 1983; 36: 731–36. Ahmedzai S. Measuring quality of life in hospice care. Oncology (Huntington) 1990; 4: 115–19. Cohen SR, Mount BM, Strobel MG, Bui F. The McGill quality of life questionnaire: a measure of quality of life appropriate for people with advanced

05PM457.qxd

9/8/01

1:55 pm

Page 397

Cross-cultural validation of MQOL in Hong Kong Chinese 397

22

23

24

25

26

27

disease. A preliminary study of validity and acceptability. Palliat Med 1995; 9: 207–19. Cohen SR, Mount BM, Thomas JJN, Mount LF. Existential well-being is an important determinant of quality of life. Evidence from the McGill quality of life questionnaire. Cancer 1996; 77: 576–86. Cohen SR, Hassan SA, Lapointe MJ, Mount BM. Quality of life in HIV disease as measured by the McGill quality of life questionnaire. AIDS 1996; 10: 1421–27. Cohen SR, Mount BM, Bruera E, Provost M, Rowe J, Tong K. Validity of the McGill quality of life questionnaire in the palliative care setting: a multi-centre Canadian study demonstrating the importance of the existential domain. Palliat Med 1997; 11: 3–23. Pratheepawanit N, Salek MS, Finlay IG. The applicability of quality of life assessment in palliative care: comparing two quality of life measures. Palliat Med 1999; 13: 325–34. Lo RSK, Ding A, Chung TK, Sze KH, Woo J. Pilot programme of clinical assessment of quality of life in palliative care unit in Shatin Hospital. Asia Pacific Hospice Conference 1997; Hong Kong (abstract). Chan KS, Lam ZCL,Chun RPK, Dai DLK, Leung CTL. Chinese patients with terminal cancer. In: Doyle D, Hanks G, MacDonald N eds. Oxford

28

29

30

31 32 33

34

textbook of palliative medicine, 2nd edn. Oxford: Oxford University Press, 1998: 83–94. Goh CR, Lee KS, Tan TC et al. Measuring quality of life in different cultures: translation of the Functional Living Index for Cancer (FLIC) into Chinese and Malay in Singapore. Ann Acad Med Singapore 1996; 25: 323–24. Leung KF, Tay M, Cheng SSW, Lin F. Hong Kong Chinese version World Health Organisation quality of life measure-abbreviated version. Hong Kong Hospital Authority, December 1997. Spitzer WO, Dobson AJ, Hall J et al. Measuring the quality of life of cancer patients: a concise QL-index for use by physicians. J Chronic Dis 1981; 34: 585–97. Cella D. Quality of life: the concept. J Palliat Care 1992; 8(3): 8–13. Donovan K, Sanson-Fisher RW, Redman S. Measuring quality of life in cancer patients. J Clin Oncol 1989; 7: 959–68. Lo RSK, Harwood RH, Woo J, Yeung F, Ebrahim S. Cross-cultural validation of London Handicap Scale in Hong Kong Chinese. Clin Rehab 2000; 15: 131–39. MacElveen-Hoehn P, McCorkle R. Understanding sexuality in progressive cancer. Sem Oncol Nur 1985; 1: 56–62.