The brief version of the World Health Organization Quality of Life instrument (WHOQOL-BREF) is a self-administered questionnaire that assesses qual-ity of life (QOL).1It has been culturally adapted into the Taiwan Chinese version,2–4which is used extensively in Taiwan.5–9 However, the Taiwan Chinese version cannot be applied to more than half of the elderly Taiwanese aged > 65 years, who use only a spoken language, Taiwanese,10because of significant differences between Taiwan Chinese and Taiwanese, mainly the pronunciation and characters used. Moreover, because these elderly
Taiwanese have received only a rudimentary edu-cation in the early part of the last century, they do not understand written/spoken Taiwan Chinese. To meet the needs of these elderly individuals, an alternative Taiwanese version of the WHOQOL-BREF has been developed to assess the QOL of Taiwanese-speaking elderly people by using an interview technique,11and its validity and relia-bility have been reported.12
The WHOQOL-BREF Taiwanese and Taiwan Chinese versions have been developed in compli-ance with the WHOQOL guidelines,13including
Agreement Between the WHOQOL-BREF
Chinese and Taiwanese Versions in the
Elderly
Chi-Wen Chien,1Jung-Der Wang,2,3Grace Yao,4I-Ping Hsueh,5,6Ching-Lin Hsieh5,6*
The brief version of the World Health Organization Quality of Life instrument (WHOQOL-BREF), including four domains, has been culturally adapted into Taiwan Chinese and Taiwanese versions for different tar-geted populations of elderly people. However, there is no evidence to suggest whether the results obtained from these two language versions can be directly compared or combined. This prevents the use of both versions together, which leads to missing data and a gap in the interpretation of results. The present study therefore examined agreement between the Taiwan Chinese and Taiwanese version of the WHOQOL-BREF in a group of 53 Taiwanese-speaking elderly people who can read Chinese. Each participant was evaluated using both versions in a random order within a 2-week period. There was acceptable agreement in 17 of the 28 items between both versions. Three of the four domain scores demonstrated moderate to high levels of agreement (0.65≤ intraclass correlation coefficient ≤ 0.81), with the exception being the social relationships domain. The results indicate that these three domain scores in the Taiwan Chinese and Taiwanese versions of the WHOQOL-BREF appear to be substantially equivalent, which allows direct comparison/combination of the results. [J Formos Med Assoc 2009;108(2):164–169]
forward and backward translation, as well as expert review, to minimize translation discrepancies and ensure the quality of mutual data comparison.2,11 However, no empirical evidence has been presented to establish that the QOL results obtained from the Taiwanese version can be directly compared or combined with those obtained from the Taiwan Chinese version. It is thus unlikely that researchers, clinicians or health policy makers will be able to make informative interpretations and decisions across Taiwanese- and Taiwan-Chinese-speaking elderly people. The present study aimed to exam-ine the level of agreement between the Taiwanese and Taiwan Chinese versions of WHOQOL-BREF in a group of Taiwanese-speaking elderly people who can read written Chinese, to provide evidence to support comparison/combination between the QOL results obtained from each version.
Methods
Participants
Participants were recruited from 15 long-term care institutions throughout eastern, southern, central and northern Taiwan by convenience sampling. Participants who met the following criteria were included: (1) spoke Taiwanese as their everyday language; (2) could read written Chinese inde-pendently, i.e. those with an educational level of junior high school or higher; (3) scored > 20 on the Mini-Mental State Examination,14which was administered using colloquial Taiwanese, to indi-cate no cognitive impairment; and (4) gave oral consent to participation.
Procedures
The Taiwan Chinese or Taiwanese version of the WHOQOL-BREF was randomly chosen to be ad-ministered to each participant at the first evalua-tion. After 2 weeks, the participants, whose QOL was determined to have remained stable, based on two additional questions about their self-reported QOL/health status, were assessed again with the other version. On each evaluation, the participants completed the Taiwan Chinese version
independently or received face-to-face interviews using the Taiwanese version administered by three trained interviewers. To ensure the quality of the interviewers, each interviewer received 2 hours of training from the first author, as well as a mini-mum of three interview practice sessions under supervision.
Instruments
The Taiwan Chinese and Taiwanese versions of the WHOQOL-BREF included 28 items, consisting of 26 standard items from the original WHOQOL-BREF and two Taiwanese national items.2,3 The 26-item standard WHOQOL-BREF contained two generic items (overall QOL and general health), and the remaining 24 items were further classi-fied into four domains: physical (7 items), psy-chological (6 items), social relationships (3 items), and environment (8 items). The two Taiwanese national items were “Do you feel respected by others?”, which was included in the social rela-tionships domain, and “Are you usually able to get the things you like to eat?” in the environment domain.2,3Responses from the two generic items (overall QOL and general health) were calculated as a single score with a range of 1–5. Domain scores were calculated by multiplying the mean of all item scores included in each domain by a factor of 4, and accordingly, potential scores for each domain ranged from 4 to 20. Higher scores indicate better QOL as reflected by the items/domains.
In terms of administration, the Taiwan Chinese version was a self-administered questionnaire. However, the Taiwanese version was administered face-to-face to each participant by interviewers with the assistance of an audio player. The con-tents of the Taiwanese version were prerecorded with a female voice to reduce variability in in-terviewers’ administration of the questionnaire. During the interview with the Taiwanese version, interviewers played/stopped the audio player as appropriate and recorded participants’ responses to each item. Replaying of the questions and their scale descriptors was allowed to ensure that partic-ipants understood the questions and descriptors. Standardized administration procedures of the
Taiwanese and Taiwan Chinese versions can be found in their respective manuals.2,11
Data analysis
Agreement between each individual item of the Taiwan Chinese and Taiwanese versions was ex-amined using weighted κ values. A weighted κ value > 0.4 indicated acceptable agreement.15At the domain level, the agreement between the two versions was conducted using a random effects model intraclass correlation coefficient (ICC). ICC values ≥ 0.8 indicated strong agreement, val-ues of 0.6–0.8 represented moderate agreement, and values ≤ 0.6 indicated weak agreement.16In addition, the Bland and Altman method,17which involved plotting the scores of the difference be-tween the two versions against those of the average between the two versions, was used to examine the limits of agreement between the domain scores of the two versions.
Results
At the first evaluation, 61 participants completed either the Taiwan Chinese or Taiwanese version of the WHOQOL-BREF. Eight participants were not eligible for the second evaluation, as their self-reported QOL/health status had been altered by physical or emotional disorders. For the re-maining 53 participants, mean age was 76.9± 6.2 years. Thirty-four participants (64%) were male. Seven (13%) lived together with their spouses, while the remainder were single (25%), divorced (13%) or widowed (45%). More than half of the participants (55%) lived in northern Taiwan.
of the social relationships domain. Agreement of the social relationships domain was poor (ICC= 0.48). The poor agreement of the social rela-tionships domain was further confirmed by the Bland–Altman plot, for which the limit of agree-ment was relatively larger than those of the other domains (Figure).
Discussion
To the best of our knowledge, the Taiwanese ver-sion of WHOQOL-BREF is the first QOL instru-ment specifically designed for elderly people who speak only Taiwanese. Taiwanese-speaking eld-erly people can benefit from this version, which assesses their QOL as well as addresses the limita-tions of the Taiwan Chinese version of WHOQOL-BREF. To facilitate direct comparison/combination between the two versions, it is essential to exam-ine the data equivalence between the Taiwanese and Taiwan Chinese versions for a group of par-ticipants. The present study found that more than half of the individual item scores obtained from the Taiwanese and Taiwan Chinese versions could be compared. Moreover, only the social relation-ships domain was found to have unacceptable agreement between the two versions. It is thus indicated that, except for the social relationships domain, the QOL of Taiwanese-speaking elderly can be compared or combined directly with that of Taiwan-Chinese-speaking elderly people, through use of the Taiwanese and Taiwan Chinese ver-sions of the WHOQOL-BREF. Direct comparison/ combination between the results obtained from both language versions can further enable
re-for the Taiwanese version was implemented re-for a group of elderly Taiwanese, to reduce the con-ceptual inequivalence and misleading wording.11,12 However, there are some inevitable discrepancies between spoken Taiwanese and written Chinese, such as idioms, thus leading to the possibility of the items showing poor agreement between ver-sions. Second, the effect of different administra-tion modes between the two versions might have resulted in a bias in the participants’ responses.
Previous studies have found that patients in an interviewer-administrated group report higher QOL scores than do those in a self-administered group.18–20The results of our study (not reported) were similar to these reports. The most likely cause is the interviewer effect, which may lead to posi-tive answers to potentially embarrassing questions, as a result of participants’ desires to present them-selves in a positive manner.18 In particular, the social relationships domain, which is related to Table. Agreement between the Taiwanese and Taiwan Chinese versions of the WHOQOL-BREF (n= 53)
Domain Item Statistics*
Generic 1 Overall QOL 0.48
2 General health 0.61
Physical 3 Pain and discomfort 0.12†
4 Dependence on medical substances and medical aids 0.28†
10 Energy and fatigue 0.45
15 Mobility 0.73
16 Sleep and rest 0.53
17 Activities of daily living 0.51
18 Work capacity 0.58
Psychological 5 Positive feelings 0.45
6 Spirituality/religion/personal beliefs 0.35†
7 Thinking, learning, memory and concentration 0.26†
11 Body image and appearance 0.54
19 Self-esteem 0.72
26 Negative feelings 0.49
Social relationships 20 Personal relationships 0.37†
21 Sexual activity 0.12†
22 Practical social support 0.35†
27‡ Being respected/accepted 0.18†
Environment 8 Freedom, physical safety and security 0.36†
9 Physical environment (pollution/noise/traffic/climate) 0.40
12 Financial resources 0.34†
13 Opportunities for acquiring new information and skills 0.09† 14 Participation in and opportunities for recreation/leisure activities 0.40
23 Home environment 0.68
24 Health and social care: accessibility and quality 0.49
25 Transport 0.50
28‡ Eating/food 0.49
Physical domain 0.81 (0.69 to 0.89)
Psychological domain 0.65 (0.47 to 0.79)
Social relationships domain 0.48 (0.24 to 0.66)
Environment domain 0.71 (0.54 to 0.82)
*Weighted k for the item level and intraclass correlation coefficient with 95% confidence interval for the domain level; †items demonstrating poor agreement; ‡Taiwanese national items.
participants’ perceived social support or sexual ac-tivity, might be more vulnerable to the interviewer effect. We thus speculate that the social relation-ships domain scores exhibited poor agreement between the two versions of the WHOQOL-BREF because of the different administration modes. However, future studies are needed to determine
also read Chinese live in long-term care institu-tions. Future studies that recruit more participants from the community or elsewhere are warranted to confirm our findings. In addition, elderly peo-ple who spoke exclusively Taiwanese or Chinese were excluded from this study. The reason for this exclusion was that recruiting elderly people
Difference between the two versions
20 18 16 14 12 10 8 6
Average score of the two versions
Social relationships Environment
Difference between the two versions
20 18 16 14 12 10 8 6
Average score of the two versions 10 8 6 4 2 0 −2 −4 −6 −8 −10
Difference between the two versions
20 18 16 14 12 10 8 6 10 8 6 4 2 0 −2 −4 −6 −8 −10 Average score of the two versions
Physical Psychological
Difference between the two versions
20 18 16 14 12 10 8 6
Average score of the two versions 10 8 6 4 2 0 −2 −4 −6 −8 −10 10 8 6 4 2 0 −2 −4 −6 −8 −10
Figure.Bland-Altman method for plotting difference of scores against mean scores of the Taiwanese and Taiwan Chinese versions in four domains. The two bold dashed lines define the limits of agreement (mean difference± 2 standard deviations).
social relationships domains of the WHOQOL-BREF Taiwanese and Taiwan Chinese versions. The social relationships domain of both WHOQOL-BREF versions may be individually used and ana-lyzed for the targeted population only.
Acknowledgments
This study was supported by research grants from the National Science Council (NSC-92-2314-B-002-137 and NSC-93-2314-B-002-032) and the National Health Research Institute (NHRI-EX95-9504PP). The authors would like to thank Ms Vanessa Carson for her assistance in editing this manuscript prior to submission.
References
1. The WHOQOL Group. Development of the World Health Organization WHOQOL-BREF quality of life assessment.
Psychol Med 1998;28:551–8.
2. The WHOQOL-Taiwan Group. The User’s Manual of
the Development of the WHOQOL-BREF Taiwan Version.
Taipei: Taiwan WHOQOL Group, 2001.
3. Yao G, Chung CW, Yu CF, et al. Development and verifica-tion of validity and reliability of the WHOQOL-BREF Taiwan version. J Formos Med Assoc 2002;101:342–51. 4. Yao G, Wang JD, Chung CW. Cultural adaptation of the
WHOQOL questionnaire for Taiwan. J Formos Med Assoc 2007;106:592–7.
5. Ay-Woan P, Sarah CP, Lyinn C, et al. Quality of life in depres-sion: predictive models. Qual Life Res 2006;15:39–48. 6. Lai KL, Tzeng RJ, Wang BL, et al. Health-related quality of
life and health utility for the institutional elderly in Taiwan.
Qual Life Res 2005;14:1169–80.
7. Jang Y, Hsieh CL, Wang YH, et al. A validity study of the WHOQOL-BREF assessment in persons with traumatic spinal cord injury. Arch Phys Med Rehabil 2004;85:1890–5.
8. Fang CT, Hsiung PC, Yu CF, et al. Validation of the World Health Organization quality of life instrument in patients with HIV infection. Qual Life Res 2002;11:753–62. 9. Yang SC, Kuo PW, Wang JD, et al. Development and
psycho-metric properties of the dialysis module of the WHOQOL-BREF Taiwan version. J Formos Med Assoc 2006;105: 299–309.
10. Tsai S. Language usage and occupational stratification in Taiwan: comparing ethnic differences among men.
Taiwanese Sociol (Taipei) 2001;1:65–111.
11. The WHOQOL-Taiwanese Group. Development and
Instruction of the WHOQOL-BREF Taiwanese Interview Version. Taipei: Taiwanese WHOQOL Group, 2007.
12. Chien CW, Wang JD, Yao G, et al. Development and validation of a WHOQOL-BREF Taiwanese audio player-assisted interview version for the elderly who use a spoken dialect. Qual Life Res 2007;16:1375–81.
13. World Health Organization. WHOQOL Protocol for New
Centers. WHO/MNH/PSF/94.4, Geneva: WHO, 1994.
14. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189–98. 15. McCluggage WG, Bharucha H, Caughley LM, et al.
Interobserver variation in the reporting of cervical colpo-scopic biopsy specimens: comparison of grading systems.
J Clin Pathol 1996;49:833–5.
16. Richman J, Makrides L, Prince B. Research methodology and applied statistics. Part 3: measurement procedures in research. Physiother Can 1980;32:253–4.
17. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement.
Lancet 1986;1:307–10.
18. Unruh M, Yan G, Radeva M, et al. Bias in assessment of health-related quality of life in a hemodialysis popula-tion: a comparison of self-administered and interviewer-administered surveys in the HEMO study. J Am Soc
Nephrol 2003;14:2132–41.
19. McHorney CA, Kosinski M, Ware JE. Comparisons of the costs and quality of norms for the SF-36 health survey collected by mail versus telephone interview: results from a national survey. Med Care 1994;32:551–67.
20. Leung KF, Wong WW, Tay MS, et al. Development and validation of the interview version of the Hong Kong Chinese WHOQOL-BREF. Qual Life Res 2005;14:1413–9.
