June 2010, Vol 32, No. 2 |
Original Articles
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Population norm of Chinese (HK) SF-12 health survey_version 2 of Chinese adults in Hong KongCindy LK Lam林露娟, Carlos KH Wong黃競浩, Elegance TP Lam林定珮, Yvonne YC Lo盧宛聰, Wen-wei Huang黃文偉 HK Pract 2010;32:77-86 Summary Objective: To establish the normative values of the SF-12 Health Survey_Version 2 (SF-12v2) for the Chinese adult population in Hong Kong (HK) and to determine factors associated with variations in scores. Design: Two cross-sectional telephone surveys with the Chinese (HK) SF-12 Health survey and a structured questionnaire in 2008 and 2009. Subjects: 2533 Chinese adults aged ≥18 years randomly selected from the general population in Hong Kong. Main outcome measures: Mean SF-12v2 domain scale and summary scores were calculated and population age-sex adjusted SF-12v2 norms were established. The effect of sociodemographic and morbidity factors on SF-12v2 scores, and the effect of SF-12v2 scores on service utilization rates were determined by multivariate regressions. Results: Mean SF-12v2 scores of subjects were significantly lower than the 1998 norm in all but the general health and vitality domains. Being female and the presence of chronic or acute illness were associated with worse SF-12v2 scores, but higher education level or being married were associated with better SF-12v2 scores. SF-12v2 scores were independent determinants of outpatient consultations and Accident and Emergency visits. Conclusion: The Chinese (HK) SF-12v2 data can now be interpreted more meaningfully with reference to the general population norms. Health-related quality of life (HRQOL) should be taken into account in the estimation of outpatient health service needs. Keywords: Health-related quality of life, SF-12, Norm, Hong Kong, Chinese, service utilization 摘要 目的：制定SF-12健康調查的第二版本應用於香港華人人口的標準數值及測定與分數變化的聯繫因素。 設計：在2008和2009年，以SF-12健康調查的中文第二版本和預設的問卷進行橫截面式電話調查。 對象：從18歲或以上的華裔成年人中隨機挑選2533人。 主要測量內容：計算SF-12版本之領域範圍的平均級別和摘要分數。確定經調整人口年齡和性別後的SF-12V2標準。用變值回歸法去測定個人因數和疾病因素對SF-12V2分數的影響，及SF-12V2分數對醫療服務使用率的影響。 結果：除整體健康和活力範疇外，SF-12V2的平均分數都比1998年的標準低。女性，急性和慢性病都與較低的SF-12V2分數有聯繫。但較高教育程度或已婚則與較高的SF-12V2分數有關。SF-12V2分數是使用門診和急症室服務的獨立決定因素。 結論：現可根據確定了的人口標準，更有意義地解釋SF-12中文第二版本數據。在預計門診健康服務的需要時，應考慮和健康相關的生活質素。 主要詞彙：健康相關的生活質素，SF-12，標準，香港，華裔，服務使用 Introduction Health-related quality of life (HRQOL) is becoming a standard outcome measure of the impact of illness and effectiveness of treatment. This outcome measure is particularly relevant to primary care that focuses on the person’s well-being. The MOS 36-Item Short-Form Health Survey (SF-36) is a widely used HRQOL measure. It can give an indication on the status of each of eight HRQOL domains on physical functioning (PF), role physical (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), role emotional (RE) and mental health (MH), which can then be summarized into two component summary scores related to physical (PCS) and mental (MCS) health, respectively.1,2 The shorter and more user-friendly SF-12 Health Survey (SF-12) was originally developed as an alternative to the SF-36 Health Survey for the measurement of the PCS and MCS scores.3 Both the SF-36 and SF-12 Health Surveys were translated and shown to be valid for the general Chinese population in Hong Kong.4-6 The SF-36 and SF-12 Health Surveys were revised to a second version (SF-36v2 and SF-12v2) in 1996 to improve the layout, clarity of the wordings, and differentiation of the response options.7,8 Major changes were also made to the scoring method of the SF-12v2 Health Survey so that it can give a full profile of eight domain scores as well as the PCS and MCS scores.8 The SF-12v2 Health Survey is gradually replacing the SF-36v2 Health Survey in clinical studies because it can give similar data with less administration burden and cost. The ability to use a smaller number of items for this purpose is most useful when data other than HRQOL need to be collected in the same study, so that the total number of questions a subject needs to answer will not be excessive. The validity and reliability of the SF-12v2 and its standard scoring algorithm have been confirmed in the American and Hong Kong Chinese populations.9,10 The establishment of the population norm is important for meaningful interpretation of the SF-12v2 scores since HRQOL scores are relative rather than absolute. Hong Kong Population based norm reference of the SF-36 Health Survey was established a decade ago11,12 but normative data for the SF-12v2 Health Survey is not available. The aim of this study was to determine the population SF-12v2 norm for Chinese adults in Hong Kong, which can serve as a ‘normal’ reference. Factors that might influence the SF-12v2 scores and the effect of SF-12v2 scores on health service utilization were also investigated. The results would facilitate the application of the SF-12v2 Health Survey to measure HRQOL in clinical service and research. Methods Subjects All 1350 and 1183 Chinese adults aged ≥18 years who had answered the Chinese (Hong Kong) SF-12v2 Health survey in two cross-sectional general population telephone surveys on primary health services utilization from March to April 2008 and on health status in May, 2009, respectively, were included in this study. Telephone-owning households in Hong Kong were contacted by random digital dialing through a computer-assisted telephone interviewing (CATI) system that had a 95% household coverage rate. A member of the household was recruited randomly by the use of the last birthday rule. 4510 persons were contacted and 2763 (61%) completed the survey. 182 children were excluded from further data analysis in this study because the SF-12v2 Health Survey was not applicable to them. The mean age of subjects was 45.52 (SD: 16.94) years. The socio-demographic characteristics of the subjects are compared to those of the Hong Kong general population13,14 and that of the 1998 population study with the SF-36 Health Survey11 in Table 1. The age-sex distributions of our subjects were similar to those reported by Census except for a lower proportion of elderly aged 65 and above. The study sample had a higher proportion of professionals or associate professionals and outpatient consultation rate than the general population. Study instruments and outcome measures The study instruments consisted of the Chinese (HK) SF-12v2 Health Survey and a structured questionnaire on socio-demography and morbidity in both surveys, and a questionnaire on service utilization rate and pattern, and outcomes of consultations in the 2008 survey. The instruments were administered by trained interviewers in Cantonese or Putonghua in the HKUSSRC. The data on consultation pattern and outcomes were not included in the present analysis. The primary outcomes were the Chinese (HK) SF-12v2 domain scale scores, which were calculated by summation of the relevant item scores and transformation to a range from 0 to 100 according to the standard scoring algorithm described in the Manual.8 Higher scores mean better HRQOL. The Physical Component Summary (PCS) and Mental Component Summary (MCS) scores were obtained first by z score transformation and then weighted summation of the scale scores with the population mean of 50 and standard deviation of 10. The Primary Health Care Service Utilization Questionnaire used in the 2008 survey was adapted from one used in previous population surveys on service utlilization in 1998 and that used in the HK Household thematic survey in 2005.11,14 Service utilization was measured by self-reported outpatient consultations with Western or Chinese medicine practitioners, Accident and Emergency Department (A&E) visits, and hospital admissions in the past four weeks. Chronic morbidity was measured by the presence of any self-reported chronic disease that was diagnosed by a registered doctor, and acute illness was measured by self-reported illness in the past four weeks. Data Analysis All statistical analyses were performed using SPSS Windows 16.0 program (SPSS, Chicago IL, USA). P-values <0.05 were considered statistically significant. Mean and standard deviations of SF-12v2 domain scores and component summary scores were calculated for all subjects and by age (18-40, 41-64, 65 or above years old) and sex groups. Population weighting factors stratified by 12 groups of age and sex combination derived from the mid 2008 Hong Kong population data15 was applied to estimate the age-sex adjusted SF-12v2 population norm (Appendix). The results were compared to the SF-12v2 data extracted from the 1998 population survey with the SF-36v2 Health Survey,12 to determine whether there was any difference in the population norm after 10 years. The significance of the difference was tested by two-sample t tests. Multivariate linear regressions on the data from the 2008 survey were carried out to determine the effect of sociodemographics, chronic morbidity or acute illness on the SF-12v2 PCS and MCS scores, and to determine whether the SF-12v2 PCS and MCS had an effect on the rates of consultation, hospitalization or A&E visit, controlling for confounding variables. Results Table 2 shows that the mean and distribution of the SF-12v2 domain scale and summary scores of our subjects, and the age-sex adjusted population normative values, in comparison with the mean SF-12v2 scores extracted from the 1998 population study. There were significant differences between the 2008-2009 and 1998 population mean scores in all domains and the effect size differences were greater than 0.5 for the BP score. The 2008-2009 norms were generally lower than those found in 1998 except the GH and VT scores. It was noted that the SF-12v2 had less ceiling effects, especially in the RP, BP, SF and RE domains, in 2008 than 1998. Table 3 shows the population mean Chinese (HK) SF-12v2 scores by age and sex groups. There was a trend for decreasing scores with increasing age, except for RE, MH and MCS scores. Subjects who had any chronic disease, acute illness, doctor consultations or A&E visits had significantly lower (worse) SF-12v2 scores than those who had not (Table 4). The scores of people who had been hospitalized were lower than those who had not but the difference did not reach statistical significance. Multivariate regression analyses found that the presence of an acute illness or chronic disease was independently associated with lower (worse) PCS and MCS scores, after controlling for other confounding variables (Table 5). Women had significantly lower PCS score than men, and higher educational level was associated with higher PCS scores. Older age or being married was associated with higher MCS scores. Table 6 shows the results of multivariate logistic regressions of service utilization rates on SF-12v2 scores, morbidity and sociodemographic variables. Every one point increase in PCS and MCS score was associated with about 5% and 3%, respectively, reduction in the odds of consultations, and 7% and 4%, respectively, reduction in the odds of A&E visits, after controlling for morbidity and sociodemographic variables. The adjusted effect of PCS and MCS on the odds of hospitalization remained insignificant. As expected, acute illness was associated with a higher utilization of all types of health services. Table 1: Characteristics of Study Subjects
N/A = Not available. Table 2: Descriptive Statistics of Chinese (HK) SF-12v2 Scores of 2008-2009 and 1998 Studies
PF = Physical Functioning; RP = Role Physical; BP = Bodily Pain; GH = General Health; VT = Vitality; SF = Social Functioning; RE = Role Emotional; MH = Mental Health; Table 3: HK 2008-2009 Population Mean and SD of Chinese SF-12v2 Scores by Age-sex Groups
PF = Physical Functioning; RP = Role Physical; BP = Bodily Pain; GH = General Health; VT = Vitality; SF = Social Functioning; RE = Role Emotional; MH = Mental Health.; PCS-12 = Physical Component Summary score; MCS-12 = Mental Component Summary score Table 4: Mean Chinese (HK) SF-12v2 scores by Chronic Morbidity and Service Utilization Groups
PF = Physical Functioning; RP = Role Physical; BP = Bodily Pain; GH = General Health; VT = Vitality; SF = Social Functioning; RE = Role Emotional; MH = Mental Health; PCS-12 = Physical Component Summary score; MCS-12 = Mental Component Summary score; A&E = Accident and Emergency Department. Table 5: Factors Associated with Chinese (HK) PCS-12 and MCS-12 Scores
PCS-12 = Physical Component Summary score; MCS-12 = Mental Component Summary score. Notes: Table 6: Factors associated with service utilization rates in the past 4 weeks
PCS-12 = Physical Component Summary score; MCS-12 = Mental Component Summary score; A&E = Accident and Emergency Department. Notes: All sociodemographic variables including age, sex, educational level, marital status and occupation were entered but the effect was not significant. Discussion This study has established the current population norm reference of the SF-12v2 scores for the Chinese population in Hong Kong. The effect size differences in the population mean SF-12v2 scores between the present and 1998 studies were lower than the generally accepted minimal clinically important difference (MCID) standard of 0.5 except in the BP scale. The mean SF-12v2 scores were generally lower in 2008-2009 than 1998 suggesting possible deterioration in the health and quality of life of our population. The decrease in the two role functioning and the social functioning scores from 1998 to 2008-2009 could have resulted from increasing work demands on people with the change in the economic and employment environment in HK over the decade. A decrease in mean BP score meant more impairment in quality of life as a result of pain in 2008-2009 compared with 1998, which could be related to ageing of the population. Furthermore, social expectation and roles could have changed over the decade leading to response shifts. On the other hand, General Health (GH) and Vitality (VT) scores improved since 1998, these domains are mainly subjective feeling of well-being that are less dependent on environmental and socioeconomic factors. The difference could also be the result of non-equivalence in the measurement methods in that the 1998 SF-12v2 scores were extracted from data collected by the SF-36 Health Survey instead of being collected by a stand-alone SF-12v2 Health Survey. There is no published data on the changes in SF-12 scores from any other population. Studies in the United States (US) found a decreasing trend in the SF-36 norms from 1990 to 1998 but the effect size differences were all smaller than 0.5.2,16 Further longitudinal studies in Hong Kong and other populations are required to determine when population SF-36 norms need to be updated. The Chinese (HK) SF-12v2 PCS and MCS scores were sensitive in differentiating between groups with different health status and health service needs, supporting its discriminative validity. The results replicated the age and sex associations with HRQOL scores observed in earlier studies.2,11,17,18 It is a universal observation that females rate their health and quality of life lower than males.2,11,17 An important finding was that age was not associated with poorer HRQOL after adjusting for the presence of acute or chronic illness (Table 5) although the unadjusted scores in a few specific domains seemed lower (Table 3). As a matter of fact, older adults had better mental-health related quality of life (higher MCS score) than younger people, also found in previous local and overseas studies.2,11 Age was associated with lower scores in the predominantly physical HRQOL domains (PF, BP and RP) but the effect was not significant for the PCS score that represents an overall summary indicator of all domains. Outpatient consultations and A&E visits were dependent on subjective HRQOL (measured by the SF-12v2 scores) because they are usually patient-initiated based on the subjective perception of need.17,19,20 On the other hand, HRQOL has no significant effect on hospitalization because the latter is mostly dependent on the doctor’s decision. It should be pointed out that HK population SF-12v2 norm was very different from that of the US population, with lower scores in all but the PF and VT domains.8 Such differences should not be interpreted as the HRQOL of the HK population being better or worse than that of the US population because HRQOL is relative rather than absolute. An HRQOL score is more meaningful if it can be interpreted with reference to what is regarded as ‘normal’ for the relevant culture. Therefore norm-based scoring of the SF-12v2 domain scale scores by Z score transformation is recommended16:- [Observed scale score – population mean/ population SD]x10+50 This enables the evaluation of the HRQOL impact of a specific disease, treatment or factor in terms of deviation from the population ‘normal’. Norm-based scores also facilitate comparisons of HRQOL data between different cultural groups.5,6 Our previous study showed that the norm-based SF-36 scores of people with specific chronic diseases in HK and the US were almost the same although the unadjusted scores were markedly different.5 Limitations The sample did not include people who lived in institutions or not accessible by the sampling CATI system, for whom the SF-12v2 norm found in this study may not be applicable. The study sample tended to be more educated, which might result in higher mean scores in the physical health-related quality of life domains. The sample size was relatively small in the elderly (≥65) group, which could affect the precision of the estimation of their mean scores. Furthermore, cross-sectional data on the association between SF-12v2 scores and service utilization might not be causal, which will need to be established by prospective studies. Conclusion The population norm of the SF-12v2 scores has been established for Chinese adults in Hong Kong. The results will enable norm-based scoring of the SF-12v2 Health Survey to evaluate the impact of an illness or treatment on HRQOL meaningfully. There were significant age-sex differences in the SF-12v2 scores, the appropriate age-sex normative values should be used to improve the precision of data interpretation. HRQOL should be included as a standard health outcome measure in clinical service and research. Acknowledgement The authors would like to thank the Social Science and Research Centre of the University of Hong Kong for their help in data collection. Thanks also go to Mr Richard Au Yeung in coordinating the project and Ms Mandy Tai in her help in data analyses. Notes: Users’ license of the Chinese (HK) SF-12v2 can be obtained from Quality Metric at www. at http://www.qualitymetric.com. A scoring sheet and syntax for scoring the SF-12v2 is available from the authors on request. Appendix B - General Population Weighting factors
The overall 2008 and 2009 population survey response rate was 61%. As no information was available from the non-responders, adjustment factor for non-response was assigned a value of one for all strata. Key messages
Cindy LK Lam, MBBS, MD(HK), FRCGP, FHKAM (Family Medicine) Carlos KH Wong, BSc, MPhil Elegance TP Lam, BSc, MMedSc Wen-wei Huang, BCM, Master of Chinese Medicine, Yvonne YC Lo, MBChB, FRACGP, FHKCFP, FHKAM (Family Medicine) Correspondence to: Professor Cindy L K Lam, Family Medicine Unit, HKU, 3rd Floor, Ap Lei Chau Clinic, 161 Main Street, Ap Lei Chau, Hong Kong SAR. Email: clklam@hku.hk References
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||