If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address correspondence to: Eliza Lai-yi Wong, PhD, The Chinese University of Hong Kong, 4/F, School of Public Health Building, Prince of Wales Hospital, Shatin, New Territories, Hong Kong, China.
To establish a normative profile of health-related quality of life (HRQoL) for Hong Kong (HK) Chinese residents aged 18 years and above and to examine the relationship between socioeconomic characteristics and health conditions and the preference-based health index.
Methods
We recruited 1014 representative Cantonese-speaking residents across 18 geographical districts. The normative profiles of HRQoL were derived using established HK value sets. Mean values were computed by sex, age group, and educational attainment to obtain the EQ-5D HK normative profile for the general HK population. To explore the relationships among potential covariates (socioeconomic characteristics and health conditions) and the HK health index, a multivariable homoscedastic Tobit regression model was employed for the analysis.
Results
The mean index value was 0.919 using the EQ-5D-5L HK value set. Younger ages reported greater problems with anxiety or depression than did older ages, whereas older ages reported greater problems with pain or discomfort than did younger ages. Persons with higher educational attainment and those who reported higher life satisfaction reported significantly higher health index scores (P < .05). On the contrary, receiving government allowance and having experienced a serious illness were significantly associated (P < .05) with a lower health index.
Conclusions
The norm values fully represent the societal preferences of the HK population, and knowledge of societal preferences can enable policy makers to allocate resources and prioritize service planning. The study was conducted with the EuroQol International EQ-5D-5L Valuation Protocol and therefore enabled us to compare the EQ-5D-5L values with other countries to facilitate understanding of societal preferences in different jurisdictions.
To date, there are many different instruments for measuring health-related quality of life (HRQoL), including EuroQol 5-dimension (EQ-5D), Health Utilities Index (HUI),
The EQ-5D is a generic multi-attribute utility instrument that provides a preference-based measure of health status. It was developed by the EuroQol Group
and is globally used to measure health-related quality of life (HRQoL) for clinical and economic assessment. Among the different types of HRQoL measures, a profile-based measure (eg, SF-36 and SF-12) is one of the many where an individual health profile is calculated with weighted sums of the scaled scores.
A preference-based measure (eg, EQ-5D, SF-6D, and HUI-3) enables each subject’s self-classified information to be converted into a single index (utility value) and is a preferred method as it can be easily computed by applying scoring rules that reflect the relative importance of each dimension of health states, enabling comparisons between different populations and predictions of health outcomes.
A systematic review, psychometric analysis and qualitative assessment of generic preference-based measures of health in mental health populations and the estimation of mapping functions from widely used specific measures.
Health Technol Assess.2014; 18 (vii-viii, xiii-xxv): 1-188
Thus, the reference values for HRQoL in specific jurisdictions are very important for the interpretation of health status because they reflect the norms of one population in the health system and provide a meaningful anchor to another group of individuals for comparison, such as comparing the health status of a patient group with that of the general population. Moreover, these reference values not only determine whether a group or individual scores are below or above the average for that specific population
the EQ-5D is a recommended tool for estimating quality-adjusted life years (QALYs) using utility values for evaluating cost-effectiveness analyses of the intervention programs.
The descriptive system of the EQ-5D comprises 5 dimensions: mobility, self-care, usual activities, pain or discomfort, and anxiety or depression. In the past decade, value sets for the 3-level version (EQ-5D-3L) have been established and widely used in many countries. Nevertheless, several studies reported ceiling effects and low discriminatory power of the 3L version and criticized its feasibility with missing values.
The 5-level EQ-5D version (EQ-5D-5L) is an updated version of the original EQ-5D-3L with 5 levels (no problems, slight problems, moderate problems, severe problems, and extreme problems) for each dimension. It defines 3125 possible health states to improve the instrument properties, such as increasing discriminative capacity and reducing ceiling effects,
compared with the EQ-5D-3L, which only describes 243 health states. The EQ-5D-5L is available in more than 130 languages and in various modes of administration.
HK is a legacy of the British Colony and was returned to China as a special administrative region of China in 1997. HK has a different health system with different cultural, demographic, and socioeconomic characteristics compared with other provinces of China.
Total hip arthroplasty through the mini-incision (Micro-hip) approach versus the standard transgluteal (Bauer) approach: a prospective, randomised study.
Effectiveness and tolerability of transdermal buprenorphine patches: a multicenter, prospective, open-label study in Asian patients with moderate to severe chronic musculoskeletal pain.
The Hospital Authority (HA) is a statutory body managing all 42 public hospitals in HK, and it embarked on the first benchmark patient experience survey on inpatient service in 2010, for which the EQ-5D was included as a measure of health outcomes.
To gather a proactive collection of feedback from patients with different disease and health needs, the patient experience survey on specialist outpatient and accident and emergency services was followed in 2014
Examining the health-related quality of life using EQ-5D-5L in patients with four kinds of chronic diseases from specialist outpatient clinics in Hong Kong SAR, China.
There are no existing population norms using the EQ-5D from the HK general population. This study, therefore, aimed to establish a normative profile of HRQoL for HK Chinese residents aged 18 years and above using the preference-based measure, EQ-5D-5L HK, stratified by sex, age, and educational levels. It also explored the relationships between HRQoL and other socioeconomic factors, such as long-term conditions and disability, mental illness, and chronic diseases in the HK population.
Methods
Study Design
The current study was conducted using the data derived from the valuation study of the preference-based health index using the EQ-5D-5L in the HK.
These preferences (utilities) were measured on a scale from 0 to 1, where full health is anchored at 1 and death at 0. A negative value was obtained when the health state was worse than death. It was a cross-sectional, population-based, face-to-face survey using the locally validated EQ-5D-5L HK and was conducted between June 2014 and October 2015 in the HK general population. Sampling was performed using the stratified quota method in which the sample quota was assigned according to the HK population structure (age, sex, and education level).
Data Collection
A representative sample in terms of age, sex, and highest educational attainment was recruited across 18 geographical districts in HK. The survey included Cantonese-speaking HK residents aged 18 and above in both public and private housing estates. Face-to-face interviews were conducted by a team of 6 trained and experienced interviewers with the aid of computer-based valuation software (The EuroQol Valuation Technology, EQ-VT).
In addition, respondents were asked to report their own health status using the EQ-5D-5L descriptive system, the EQ-VAS, and their socioeconomic information, including age, sex, marital status, educational level, and experience with serious illness, including whether it was a longstanding health condition such as a disability, mental illness, or chronic disease.
All participants were informed of their rights, any possible risks or benefits were explained, and information about the purpose of the study and details of the research procedures were given before starting the interview. Written informed consent was also obtained from each of the participants at this stage. The participants were allowed to withdraw from the study at any point. All data were kept confidential and anonymous.
Measurements
EQ-5D-5L HK
The culturally validated HK Chinese version of the EQ-5D-5L instrument (EQ-5D-5L HK) consists of self-reported health on a 5-dimension descriptive system and self-reported overall health using the EuroQol Visual Analogue Scale (EQ-VAS).
The descriptive system comprises the dimensions of mobility (MO), self-care (SC), usual activities (UA), pain/discomfort (PD), and anxiety/depression (AD), and each dimension has 5 response levels describing the level of severity: (1) no problem, (2) slight problems, (3) moderate problems, (4) severe problems, and (5) extreme problems. There are 3125 possible health states defined by combining each level on each dimension, ranging from the best possible states “11111” as “full health” to “55555” as “the worst health.” The EQ-VAS is the self-reported overall health perception of the respondents. It records the respondent’s self-rated health on a vertical scale from 0 (the worst health) to 100 (the best health) scale where the respondents picture their health status on the interview day.
Socioeconomic characteristics
The socioeconomic characteristics of the respondents were collected in the interviews. The items included age, sex, educational attainment, marital status, employment, housing, having children, and receiving any allowance from the government (ie, Comprehensive Social Security Assistance [CSSA], old age allowance, disability allowance, etc). The participants were also asked to report whether themselves, their family members, or someone they knew had experienced serious illnesses and whether the subjects themselves had any longstanding health conditions such as a disability, mental illness, or chronic disease. In the traditional Chinese culture, it is expected that family members take care of each other, and caring for the sick and elderly is an unavoidable responsibility. Therefore, a person who has experienced serious illnesses either himself/herself or their family members, would have an impact on their underlying health status in HK, which provides a distinct normative evaluation of health in HK compared with other jurisdictions. In addition, they were asked to rate how satisfied they were with their life on a scale of 0 to 10, where 0 is not satisfied at all and 10 is completely satisfied.
Statistical Analysis
The utility score of EQ-5D was derived using the established HK value set
by weighting each respondent’s self-report health status to a single preference-based health index (utility scores). Data were analyzed using IBM SPSS Statistics version 24 and Stata version13.0. Descriptive summary statistics were estimated for the percentage of people reporting any problem on each EQ-5D-5L dimension or the EQ-VAS scores, the top 10 most frequently reported EQ-5D-5L health statuses, and self-reported life satisfaction. Nevertheless, owing to rounding, the total percentage of distribution may not be possible to add up to 100. The normative profiles of the HK health index were described (mean, standard deviation, and 25th, 50th, and 75th percentile) according to the stratified respondents’ characteristics, including age groups, sex, and education level. Mean-difference tests were conducted to calculate the differences between age groups. To explore the relationships among potential covariates (socioeconomic characteristics and health conditions) and the HK health index, a multivariable homoscedastic Tobit regression model was employed for the analysis. EQ-5D utility scores usually show a severe ceiling effect, with most participants rating themselves in full health with a utility score of 1.0, and therefore the data could be interpreted as being bounded or censored at 1.0. This nature of the EQ-5D score distribution could produce biased and inconsistent estimation when using OLS regression.
Statistical significance was considered if the P value was <.05.
Results
Background Characteristics
A total of 1033 HK residents aged 18 and above responded to the study recruitment, and 19 respondents either dropped out from the interview or gave incomplete responses, leaving 1014 respondents who participated in this study. Table 1 shows the background characteristics of the respondents: demographics (sex and age), socioeconomic characteristics (highest education attainment, marital status, employment status, living status, had any children, receiving any government allowance, and self-reported living satisfaction), and health conditions (experience with serious illness and self-report longstanding health conditions) and a comparison with the structure of general HK population.
Of the study participants, 59% were women, with 16% aged 18 to 24 years, 17% aged 25 to 34, 17% aged 35 to 44, 12% aged 45 to 54, 22% aged 55 to 64, and 16% aged 65 years and above. Regarding socioeconomic characteristics, the majority of the respondents (81%) attended secondary school/subdegree or above, and around half of them were married (58%) and were in paid employment (43%). The majority of them (93%) lived with family, and 63% of them had children. Approximately 14% of the respondents received at least one allowance from the local government. In addition, approximately 77% of the respondents rated their life satisfaction 7 or above (out of a score of 0-10). In all of health conditions, 70% of the respondents had experienced severe illness, to the self (26%), relatives (39%), or caring for others (49%). For the self-reported health status, around one-third of the participants (30%) reported that they had at least one longstanding health condition, and approximately 90% of them were found to have the condition more than 6 months before participation in the study. The most common self-reported longstanding health conditions were hypertension (19%, 192 respondents) and diabetes (8%, 80 respondents). The study sample was overall reasonably representative of the local population in terms of sex, educational attainment, and marital status. Nevertheless, the sample had fewer respondents from the 45 to 54 age group and those in paid employment than the general HK population.
Table 1Background characteristics of study samples and comparison against HK general population
Table 2 shows the distribution of self-reported descriptive EQ-5D-5L and overall health VAS of the sample by different age groups. Around half of the participants (46%) reported no problems in all domains. The highest proportion of participants reported problems in Pain/Discomfort (41%), followed by Anxiety/Depression (26%), Mobility (12%), and Usual Activities (9%), whereas the lowest percentage reported problems in Self-care (1%). The mean (standard deviation [SD]) of the VAS was 82.7 (11.8) for the overall sample. Regarding the 5 descriptive dimensions, the population seemed to experience greater problems in the dimension of pain or discomfort, especially in 55 to 64 age group. Interestingly, younger age groups (age 18-24 and age 25-54) experienced greater problems with anxiety or depression.
Table 2Self-reported health using the EQ-5D-5L descriptive system and EQ Visual Analogue Scale by age group
The mean health index value, using the EQ-5D-5L HK value set, for the general population in HK was 0.92 (SD 0.12), with values ranging from 0.02 to 1.00, which was left-skewed to “Full Health” status (ie, utility index at 1.00) (Fig. 1). The EQ-VAS score (Fig. 2) was ranged from 25 to 100, which was also left-skewed to “the best health you can imagine” (ie, EQ-VAS score of 100). The top 10 most frequently reported EQ-5D-5L health states out of the 3125 possible health states mainly settled around level 1 (no problem) and 2 (slight problem). The 10 health states included “11111” (46%), “11121” (18%), “11112” (9%), “11122” (8%), “21121” (3%), “21111” (2%), “11221” (2%), “21122” (2%), “21222” (1%), and “11212” (1%). These health states represented the majority (91%) of the participants.
Figure 1Distribution of Hong Kong Chinese EQ-5D-5L health index.
Table 3 shows the mean values for the EQ-5D-5L index scores by age group, sex, and educational level. Generally, elder people had lower index values in males and females than the young groups, but they were not significant at P < .05. In addition, educational level seemed to be another factor affecting the mean values where persons with lower educational level had roughly lower index values, particularly in the 25 to 34 age group.
Table 3Hong Kong Chinese Population norms using the EQ-5D-5L HK value set
Health Index by Socioeconomic Characteristics and Health Conditions
Table 4 shows the means and standard errors (SE) of the EQ-5D-5L health index and different background characteristics (ie, demographics, socioeconomics characteristics, and health conditions). After adjusting for the demographics, the results revealed that some of the socioeconomics characteristics and health condition were significantly (P < .05) correlated with the health index. For the socioeconomic characteristics, persons with higher educational attainment and those who reported higher life satisfaction had significantly higher health index (P < .05). Nevertheless, persons who received government allowance had significantly (P < .05) lower health index than those not receiving the allowance. In addition, persons who experienced serious illness also had significantly lower health indexes (P < .05).
Table 4The correlation of EQ-5D-5L health index and socioeconomic characteristics and health conditions
Health index, mean (SD)
Coef.
95% CI
P
Sex
Male
0.920 (0.193)
Ref
Female
0.918 (1.607)
−0.005
(−0.021; 0.011)
.378
Age group
18-24
0.938 (0.161)
Ref
25-34
0.935 (0.225)
−0.009
(−0.042; 0.025)
.617
35-44
0.942 (0.193)
0.004
(−0.035; 0.042)
.845
45-54
0.925 (0.257)
0.001
(−0.040; 0.042)
.969
55-64
0.894 (0.289)
−0.014
(−0.055; 0.027)
.501
65+
0.884 (0.482)
0.010
(−0.036; 0.056)
.672
Highest education attainment
Primary and below
0.868 (0.418)
Ref
Secondary/subdegree
0.928 (0.001)
0.034
(0.013; 0.056)
.002**
Postsecondary/degree
0.940 (0.193)
0.038
(0.010; 0.067)
.008**
Marital status
Single
0.932 (0.161)
Ref
Married
0.916 (0.161)
−0.002
(−0.035; 0.030)
.884
Divorced/separated
0.912 (0.514)
0.026
(−0.019; 0.071)
.267
Widow
0.875 (0.836)
0.008
(−0.039; 0.056)
.735
Employment status
Full-time student
0.934 (0.225)
Ref
Retired/homemaker/unemployed
0.899 (0.225)
0.003
(−0.037; 0.042)
.897
Employed/full-time/part-time
0.936 (0.129)
0.009
(−0.025; 0.043)
.614
Living status
Live alone
0.874 (0.672)
Ref
Live with family
0.922 (0.129)
0.021
(−0.011; 0.053)
.192
Had any children
With children
0.912 (0.161)
Ref
Without children
0.930 (0.129)
−0.009
(−0.037; 0.019)
.549
Receiving any government allowance
Received
0.840 (0.546)
Ref
Not received
0.931 (0.096)
0.065
(0.061; 0.141)
<.001***
Self-reported living satisfaction
Score 0 to 8
0.906 (0.005)
Ref
Score 9 or above
0.947 (0.005)
0.091
(0.064; 0.118)
<.001***
Experience with serious illness
Without experience
0.945 (0.129)
Ref
With experience
0.907 (0.161)
0.043
(0.024; 0.061)
<.001***
Caring family
Yes
0.911 (0.111)
Ref
No
0.921 (0.132)
0.007
(−0.008; 0.022)
.370
Caring others
Yes
0.914 (0.122)
Ref
No
0.926 (0.127)
0.005
(−0.009; 0.019)
.512
Self-report longstanding health conditions
At least one longstanding health conditions
0.873 (0.321)
Ref
Without any longstanding health conditions
0.938 (0.096)
0.019
(0.003; 0.039)
.050*
Note. SD indicates standard deviation; Coef., coefficient; CI, confident interval.
This study reports the first population norm of HRQoL for Chinese residents aged 18 years and above that was derived based on a representative sample (1014 respondents) in HK using the preference-based value set of EQ-5D-5L HK.
In the study, around half of the participants (46%) reported no problems in all health domains. The results are in line with the findings of the normative values estimated by EQ-5D-5L in other populations such as the USA (44%),
The proportion of “no problem” responses for each dimension (mobility, self-care, usual activities, pain or discomfort, and anxiety or depression) were 88.3% (mobility), 98.5% (self-care), 91.4% (usual activities), 59.5% (pain/discomfort), and 74.0% (anxiety/depression). The findings showed that our population experienced fewer problems in self-care and usual activities, but greater problems in pain or discomfort and anxiety or depression. The pattern was similar to the EQ-5D-5L population studies in other countries, such as China, Australia, Poland, and Germany.
Interestingly, our younger population reported more problems with anxiety or depression, but the proportion declined with age, which was similar to the population of China.
No previous normative study using the EQ-5D-3L in HK was performed to determine if there was improvement in the ceiling effect by increasing the levels of responses in the EQ-5D-5L, but reductions were shown in other countries.
Nevertheless, a direct comparison of the cross-country utility scores is not recommended because each jurisdiction has a different demographic structure, societal values, and health system, which all could have an impacted perceived utility in the 5 dimensions of the EQ-5D-5L.
Furthermore, the index derived from the EQ-5D-5L in the local population was higher than the norms derived from different versions of the short-form instrument.
Undoubtedly, utility valuations from different instruments cannot be directly compared because of the differences between utility scale effects, variation in the descriptive systems, and the preferences of the people interviewed.
Why do multi-attribute utility instruments produce different utilities: the relative importance of the descriptive systems, scale and ‘micro-utility’ effects.
In previous studies, the normative profile by sex, age, and educational level were shown to provide important reference information. In this study, although only educational level showed significant effects on people’s HRQoL, which is inconsistent with other findings in Asia, we should not neglect that sex and age might play an important role in the evaluation of people’s HRQoL. There was a clear trend that EQ-5D utility score decreased with increasing age, which was consistent with studies reported in Singapore and Japan.
These normative values can be used to compare health profiles of patients with specific conditions of the same age or sex and to identify disease burden in particular patient populations.
In addition, the normative profile fully represents the societal preferences of the Chinese adult population in HK. Thus, the estimated EQ-5D-5L norms could be used as baseline to enable the comparisons of the treatment effectiveness of pre- and postintervention in healthcare evaluation. It also could be used to evaluate the healthcare burden on a given disease and compare it with different adjacent countries. Moreover, the population norms could be used to conduct the economic evaluation to determine policy formulation and on resource allocation. In addition, additional local research is also required to assess the responsiveness over time for time trend analysis.
In this study, the mean health index decreased with increasing age, and women reported slightly worse health status than men. Nevertheless, no significant differences between different age groups or sexes were identified in the multivariate analysis after controlling for all socioeconomic characteristics and health conditions in the study. The results were different from prior studies in some countries, in which they found that elder groups and females were significantly more likely to have lower health indexes than the others in the multivariate analysis.
Our findings revealed that people with higher educational attainment and life satisfaction reported a significantly higher health index. Nevertheless, individuals who received government allowance and had experienced serious illness had a significantly lower health index than the HK general population, which was consistent with previous studies.
One of the strengths of this study is that it covered a large, representative sample of the adult population in HK through quota sampling by age, sex, and educational attainment over 3 geographical areas of HK: Hong Kong Island, Kowloon, and New Territories. Nevertheless, a limitation of our study is that only the land-based, noninstitutional population in HK was covered. Those who lived in institutions (ie, hospitals, old age homes, or nursing care homes) and persons living on board vessels were not involved in the study. These individuals may have significantly different health indexes. This limitation might have introduced some bias in our estimations and implies that people with more or severe health problems might be underestimated in the study. Nevertheless, this group makes up only 1% of the HK population,
and further studies are suggested to confirm differences and supplement the current findings in the local population.
Conclusions
This is the first study to provide normative profiles of HRQoL for HK Chinese residents aged 18 years and above that are derived using the EQ-5D-5L HK measure and its preference-based weights stratified by sex, age, and education levels. Because a health index is different from profile-based measures and is shaped by societal context and health systems, the HK normative data can be used as a reference that enables health evaluation and comparisons of different healthcare interventions with similar socioeconomic characteristics in the local population. Thus, it enables local policy makers to formulate healthcare policy and planning in HK to maximize efficiency in allocating limited health resources.
Acknowledgments
We gratefully acknowledge Dr Slaap Bernhard (executive director) and the scientific team of the EuroQol Research Foundation. We thank the district officers from 18 districts over Hong Kong for their help with recruitment. We are also grateful to Mr Dicken Chan, Ms Nicole Huang, Ms Yeung Yeung Pan, Mr Ringo Sze, and Mr Sky Chan for conducting the survey. Finally, we are grateful to all HK survey respondents for their participation in the study. Without their participation and engagement, the study would not have succeeded.
Source of Financial Support
This study was supported by the Health & Medical Research Fund from the Food and Health Bureau of Hong Kong, Hong Kong Special Administrative Region (HKSAR) (Grant No. HMRF11120491) and EuroQol Research Foundation.
References
Torrance G.W.
Feeny D.H.
Furlong W.J.
Barr R.D.
Zhang Y.
Wang Q.
Multiattribute utility function for a comprehensive health status classification system: Health Utilities Index Mark 2.
A systematic review, psychometric analysis and qualitative assessment of generic preference-based measures of health in mental health populations and the estimation of mapping functions from widely used specific measures.
Health Technol Assess.2014; 18 (vii-viii, xiii-xxv): 1-188
(2011 Population Census – Thematic Report)Census and Statistics Department, The Government of the Hong Kong Special Administrative Region. Ethnic Minorities,
2012
Total hip arthroplasty through the mini-incision (Micro-hip) approach versus the standard transgluteal (Bauer) approach: a prospective, randomised study.
Effectiveness and tolerability of transdermal buprenorphine patches: a multicenter, prospective, open-label study in Asian patients with moderate to severe chronic musculoskeletal pain.
Examining the health-related quality of life using EQ-5D-5L in patients with four kinds of chronic diseases from specialist outpatient clinics in Hong Kong SAR, China.
(2011 Population Census – Main Report)Census and Statistics Department, The Government of the Hong Kong Special Administrative Region. Census and Statistics Department,
2012
Why do multi-attribute utility instruments produce different utilities: the relative importance of the descriptive systems, scale and ‘micro-utility’ effects.
30182-2&id=gr1.jpg){kind=link}
30182-2&id=gr2.jpg){kind=link}