Regional differences in health status in China: Population health-related quality of life results from the National Health Services Survey 2008

Health & Place 17 (2011) 671–680

Contents lists available at ScienceDirect

Health & Place journal homepage: www.elsevier.com/locate/healthplace

Regional differences in health status in China: Population health-related quality of life results from the National Health Services Survey 2008 Sun Sun a,b,n, Jiaying Chen c,nn, Magnus Johannesson d, Paul Kind e, Ling Xu f, ¨ a,b Yaoguang Zhang f, Kristina Burstrom a

Department of Learning, Informatics, Management and Ethics, Medical Management Centre, Karolinska Institutet, Nobels v¨ ag 15a, SE-171 77 Stockholm, Sweden Department of Public Health Sciences, Karolinska Institutet, Sweden c School of Health Policy and Management, Nanjing Medical University, Hanzhong Road 140, 210 029 Nanjing, P.R. China d Department of Economics, Stockholm School of Economics, Sweden e Centre for Health Economics, University of York, UK f Center for Health Statistics and Information, Ministry of Health, P.R. China b

a r t i c l e i n f o

abstract

Article history: Received 21 June 2010 Received in revised form 18 January 2011 Accepted 20 January 2011 Available online 28 January 2011

Purpose: To measure, describe and analyse regional differences in health-related quality of life measured by EQ-5D in China. Data were obtained via face-to-face interviews on a national representative sample (n¼120,703, 15–103 years). The EQ-5D instrument was used to measure health status. Results: Rural areas had worse health status than urban areas. Health status was worst in western areas and best in eastern areas, and such disparities were profounder in rural areas. In urban areas, health status was best in middle-sized cities. In rural areas, health status increased with the economic development level of a county. Conclusion: Our study enhances understanding of the urban–rural differences and east–middle–west differences in health and sheds light on inequalities in health status between different city categories in the urban areas and county categories in the rural areas. & 2011 Elsevier Ltd. All rights reserved.

Keywords: China EQ-5D Regional health inequalities Rural population Urban population

The EQ-5D is a copyrighted instrument and all requests for using it should be sent to the EuroQol Executive Office in Rotterdam, the Netherlands ([email protected]).

1. Introduction Disparities in population health exists not only between ¨ countries (Konig et al., 2009; Szende and Williams, 2004) but also within a country. Compared to developed countries ¨ (Bellanger and Jourdain, 2004; Burstrom et al., 2010; LopezCasasnovas et al., 2005; Salmela, 1993) such disparities are more serious in developing countries (World Health Organisation [WHO], 2001). Ranking by equality in health distribution from high to low, China came 101st out of 191 member states of WHO. Many factors contribute to regional differences, e.g. natural conditions, socio-economic development, health resource alloca¨ tion, etc. (Bellanger and Jourdain, 2004; Burstrom et al., 2010; n Corresponding author at: Department of Learning, Informatics, Management ¨ 15a, and Ethics, Medical Management Centre, Karolinska Institutet, Nobels vag SE-171 77 Stockholm, Sweden. Tel.: + 46 8 524 847 79; fax: +46 8 30 05 92. nn Corresponding author. Tel.: + 86 25 868 629 50; fax: + 86 25 868 629 53. E-mail addresses: [email protected] (S. Sun), [email protected] (J. Chen).

1353-8292/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.healthplace.2011.01.007

Fang et al., 2009; Lopez-Casasnovas et al., 2005). China is a heterogeneous country, with 1.3 billion inhabitants distributed over 9.6 million km2 of territory (Central People’s Government of the People’s Republic of China [CPGPRC], 2004). Regional differences in China present a complex picture, with different types and levels (United Nations Development Programme [UNDP], 2005; Pei and Rodriguez, 2006). There are three large geographic areas in China: the eastern, middle and western areas. Each of those areas contains about 10 provinces, in total 31 provinces in mainland China (CPGPRC, 2004). Under each province, there are cities in urban areas and counties in rural areas, on average 70 cities/counties per province. About 70% of the Chinese population live in rural areas and the majority live by agriculture. Regional disparities exist not only between urban and rural areas, but also between east, middle and west and between different cities (urban) and counties (rural). The regional disparities existing historically have been widening rapidly since the economic reform of 1980s (Office of WHO Representative in China and Social Development Department of China State Council Development Research Centre [OWHORC and SDDCSCDRC], 2006; UNDP, 2005). The rapid economic growth and dramatic social and political system transition have assumed different magnitudes in different regions, being more thorough and deeper in the east and urban areas (OWHORC and SDDCSCDRC, 2006; UNDP,

672

S. Sun et al. / Health & Place 17 (2011) 671–680

2005). This has impacted heavily on health; overall, the Chinese are living longer and are healthier (OWHORC and SDDCSCDRC, 2006), but at the same time, disparities in health between different regions are widening, e.g. large gaps in life expectancy between urban and rural (75.2 compared to 69.6 years), and eastern and western areas (71.9 compared to 68.4 years) (UNDP, 2005). The recent health care reform makes reduction of regional inequalities in health a priority field (National Development and Reform Commission [NDRC], 2009). Large studies of inequalities in health in China have focused on the urban–rural issues (Liu et al., 1999; Shi, 1993; Tang et al., 2008), but there is a shortage of studies regarding eastern–middle–western areas (UNDP, 2005), and differences between different types of cities and counties. A recent study revealed that there are distinct disparities in mortality and morbidity across different provinces in China (Fang et al., 2009). China has substantial regional differences in health care systems (Shi, 1993) and insurance schemes (Xu et al., 2007), which have attracted attention in recent years (Gao et al., 2002; Shi, 1993; Henderson et al., 1994). Regional differences in socio-economic developments and health resource allocation have been considered common factors leading to inequalities in health (Bellanger and Jourdain, 2004; Lopez-Casasnovas et al., 2005; Salmela, 1993), and have been investigated in China as well (Henderson et al., 1994; Fang et al., 2009; Pei and Rodriguez, 2006; UNDP, 2005). Studies of regional inequalities in health mostly used mortality and morbidity measures (Gao et al., 2002; Fang et al., 2009), as well as the usage of health service (Henderson et al., 1994); there are few studies using health-related quality of life (HRQoL) measures in China. HRQoL reflects multi-dimensions of health, such as physical, psychological, social, cognitive and role function, as well as general wellbeing (Spiker and Revicki, 1996). Outcome measures that neglect impacts on HRQoL might tend to underestimate the health disadvantage of deprived areas (Congdon, 2001). The EQ-5D instrument has been used worldwide to measure HRQoL (Rabin and de Charro, 2001). EQ-5D population health studies ¨ et al., 2001; Kind have been performed mostly in Europe (Burstrom ¨ et al., 1998; Konig et al., 2009; Sorensen et al., 2009; Szende and Williams, 2004), USA (Fryback et al., 2007; Lubetkin et al., 2005; Luo et al., 2005), Canada (Leung et al., 2007) and Asia (Kil et al., 2008; Hoi et al., 2010; Shafie et al., 2010; Ting et al., 2007; Tsuchiya et al., 2002), including China (Sun et al., 2010; Wang et al., 2005). Studies among other Chinese populations have also been performed in Singapore (Luo et al., 2003), USA (Lubetkin et al., 2005; Luo et al., 2005) and Canada (Leung et al., 2007). Results from an EQ-5D study performed in Beijing suggested that EQ-5D is valid for measuring health status in the Chinese population (Wang et al., 2005). The EQ-5D was included for the first time in the National Health Service Survey (NHSS) 2008 to measure population health status in all 31 provinces in mainland China (Sun et al., 2010). In the present study, we investigated regional inequalities in HRQoL on a national representative sample from four different levels: urban–rural differences, eastern–middle–western differences, inter-city differences and inter-county differences.

(Center for Health Statistics and Information of Ministry of Health of People’s Republic of China [CHSIMOHPRC], 2009). Face-to-face interviews were conducted by trained local interviewers (Center for Health Statistics and Information of Ministry of Health of People’s Republic of China [CHSIMOHPRC], 2008). The NHSS 2008 questionnaire includes more than 200 questions, on acute diseases and injuries, chronic and other diseases, hospitalisation, health-related behaviour, educational level, family income and employment status, social relations, safety and security, medical care fees, accessibility (distance and time) and satisfaction with health service, insurance coverage, vaccination and disease control and woman and child health services. The EQ-5D was placed in the beginning of the questionnaire, after questions on general information and diseases and injuries, but before questions on health-related behaviour. The whole interview for each individual took generally around 45 min. In NHSS 2008, 56,400 households were sampled using a multistage stratified cluster random sampling (CHSIMOHPRC, 2008). In the first sample stage, 2400 districts (urban area) and counties (rural area) were stratified based on socio-economic, health care and population structure to sample 94 districts and counties. In the second stage, 2350 sub-districts (Jiedao in Chinese) in urban area and townships (Xiang in Chinese) in rural area in the 94 counties were stratified based on population size and income per capita to sample 470 sub-districts and townships. In the third stage, 940 residential committees (Juweihui in Chinese) in urban area and villages (Cun in Chinese) in rural area were sampled using the same criteria as in the second stage. In each residential committee or village, 60 households were randomly selected, and all family members in a sampled household were interviewed individually. An oversampling of ten households in each residential committee (village) took place at the design stage, to make up for households, which could not be reached (less than 7%). In total, 177,501 respondents were included in NHSS 2008. Of these, about 18% aged below 15 years were excluded, since EQ-5D questions should only be administered to respondents aged 15 years and above. Respondents not answering the questions by themselves were excluded (13%). In total, less than 2% of the respondents had missing answers on age, sex, in at least one of the EQ-5D dimensions, on VAS or reported VAS higher than 100. After applying the previous exclusion criteria, 120,703 respondents were used for this study. Ethical permission for analyses of this study was granted by the Regional Ethics Committee, Stockholm, Sweden (Dnr: 2009/1892–31). 3.2. Interview procedure

3. Material and methods

The interviewers were recruited from local health workers. The supervisors for interviewers were trained at the national level (4 supervisors per county, recruited from local health authority staff and county interviewers). The supervisors then trained the interviewers in each county (30 interviewers per county). An instruction for performing face-to-face interviews on NHSS questions was provided by MoH (CHSIMOHPRC, 2008). As a quality control, the supervisors checked the completeness of the questionnaire at the end of each day. If the information was missing, the interviewer went back during the same day or next day to ask the missing question again. If all family members in one household were missing, the interviewer would try to contact that family three times during the following days. If no family members were reached, the interviewer would choose another sampled household so as to clock up 60 households in each residential committee (village).

3.1. Material

3.3. Health outcome measure

Data were derived from the NHSS 2008, which has been organised by the Ministry of Health (MoH) in China every fifth year since 1993

The EQ-5D instrument is a generic HRQoL outcome measure (Rabin and de Charro, 2001) which classifies respondents’ health

2. Aim To measure, describe and analyse regional differences in HRQoL measured by EQ-5D in China.

S. Sun et al. / Health & Place 17 (2011) 671–680

status in five dimensions (mobility, self-care, usual activities, pain/discomfort and anxiety/depression), with three severity levels (no problems, some problems and severe problems). The EQ-5D instrument in total defines 243 health states. A visual analogue scale (VAS) was used in the survey, with anchor points 0 (worst health state) and 100 (best health state). The scale consisted of a horizontal line where every 10th was marked and labelled 0, 10, 20, y, 100. The question was framed: ‘‘On the scale please point out which point best represents your own health state today.’’ The scale was harmonised to fit in with the NHSS questionnaire, and hence slightly differs from the EQ VAS. In the present study, the results are presented as VAS scores since they were not obtained with the original EQ VAS.

673

Government (Beijing, Tianjing, Shanghai and Chongqing) and province capital cities with a population of more than one million inhabitants. So-called prefecture cities, which are under the direct govern of the Provincial Governments, with a population of more than half million inhabitants and province capital cities with a population between half and one million inhabitants are in the middle-sized city category. Prefecture cities and provincial capital cities with a population less than half million inhabitants are in the small city category. The counties in the rural areas can be divided into four categories based on their socio-economic development level. Indicators such as economy, demography, literacy and health were used to define the socio-economic development level. The county categories are: Type I–IV (Type I is most developed and Type IV is least developed).

3.4. Data analyses 3.4.1. Variables for region categorisations Variables for region categorisations in our study followed the official categorisations (CPGPRC, 2004; CHSIMOHPRC, 2009). 3.4.1.1. Urban and rural areas. Each province contains both urban and rural areas, on average 70 urban cities and rural counties (including county-level cities) per province. The under-provincial administrative tiers are presented in Fig. 1. Below city (district) level in the urban areas, the administrative tiers from high to low are sub-districts (Jiedao in Chinese) and residential committees (Juweihui in Chinese); while below county level in rural areas, the administrative tiers are townships (Xiang in Chinese) and villages (Cun in Chinese). 3.4.1.2. Eastern, middle and western areas. The eastern area is alongside the sea coast, and is the most developed area. It includes 12 provinces: Liaoning, Beijing, Tianjin, Hebei, Shandong, Shanghai, Jiangsu, Zhejiang, Fujian, Guangdong and Hainan. Next to the eastern area is the middle area, including 9 provinces: Shanxi, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hubei and Hunan. The western area is the least developed area, including 10 provinces: Sichuan, Chongqing, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, Ningxia, Xinjiang, Guangxi and Inner-Mongolia. 3.4.1.3. City and county categories. In terms of administrative level and population size, the cities in urban areas can be divided into three categories: big, middle-sized and small cities. The big city category includes cities under the direct govern of the Central

Province Urban

City (district)

Rural

3.4.2. Variables for socio-economic status The highest accomplished educational level was classified into below primary school, primary school, junior middle school, senior middle school, college and above. An individual’s annual income was assessed by dividing household annual income by the numbers of persons living in the family within the last half-year, regardless of age and employment status. Respondents were then ranked from lowest to highest by their annual income and divided into five groups of equal size: the lowest income group had an income below 2500 RMB, the second group from 2500 to 3999 RMB, the third group from 4000 to 5999 RMB, the fourth group from 6000 to 99,999 RMB, the fifth group 10,000 RMB and above. Employment status was categorised into employed, unemployed, student and retired. 3.5. Statistical analyses All descriptive analyses by region were stratified by sex and age groups (15–44 years, 45–64 years, 65 +). Percentages of respondents reporting problems in each EQ-5D dimension and VAS score (mean) were analysed by region. We carried out multiple logistic regression analyses to estimate the likelihood of having no problems in each of the EQ-5D dimensions. Multiple linear regression analysis was used to estimate the variation in VAS scores. We created two sets of models; in one set we controlled for sex and age (5-year age groups were entered as dummy variables, except for the oldest, 85+ ), and in the other set, we additionally controlled for educational level, income group and employment status (dummies entered based on the categorisation of socio-economic variables). Separate regressions were carried out for each region (urban–rural areas, eastern–middle– western areas, big–middle–small city categories and Type I–IV county categories); dummy variables were entered based on this regional categorisation. All analyses were performed in SAS 9.1 (SAS Institute Inc., 2006). The significant level used in regression analyses was 5% level.

County

4. Results 4.1. Characteristics of respondents Sub-district (Jiedao in Chinese)

Township (Xiang in Chinese)

Residential committee (Juweihui in Chinese)

Village (Cun in Chinese)

Fig. 1. Under-provincial administrative tiers in China.

Characteristics of respondents by region are presented in Table 1. The proportion of women was higher in all regions, except in the rural Type IV county. In the rural area, there was a higher percentage of respondents with the lowest educational level (19%) and in the lowest income group (29%), compared to the urban area (7% and 6%, respectively). The proportion of retired was low in the rural areas (2%) compared to urban areas (32%), and the employment rate was higher in the rural areas (82%) than in the urban areas (41%).

674

S. Sun et al. / Health & Place 17 (2011) 671–680

Table 1 Characteristics of respondents (n¼ 120,703), by region, China 2008. Urban area

Rural area

Total n

Eastern Middle Western Big city 33,446 13,935 9931 9580 13,029

Middle-sized city 9472

Small city 10,765

Total

Eastern Middle Western Type Type Type I II III 87,257 28,369 23,244 35,644 21,209 27,549 27,927

Type IV 10,572

Sex Men Women

46.1 53.9

46.2 53.8

45.5 54.5

46.6 53.4

46.2 53.8

46.7 53.3

45.4 54.6

49.0 51.0

48.1 51.9

48.8 51.2

49.8 50.2

47.9 52.1

49.4 50.6

48.8 51.2

50.4 49.6

Age group (years) 15–44 45–64 65 +

42.0 38.2 19.8

39.6 39.2 21.2

43.1 38.6 18.3

44.5 36.2 19.3

36.3 39.5 24.2

44.4 36.8 18.8

47.1 37.7 15.2

50.3 37.4 12.3

47.3 38.3 14.4

49.2 39.9 10.9

53.4 35.1 11.5

46.5 39.2 14.3

50.8 38.0 11.2

48.7 38.3 13.0

60.8 30.0 9.2

6.8

5.2

7.6

8.4

4.2

5.2

13.4

19.0

17.2

17.1

21.6

17.5

15.5

19.0

30.9

14.6 30.4

12.5 30.5

15.0 32.2

17.2 28.3

11.2 27.5

11.3 31.9

25.7 20.4

32.9 37.7

30.4 40.1

33.7 39.8

34.5 34.4

31.7 37.5

30.7 42.8

33.2 37.8

40.3 24.5

Level of education Below primary school Primary school Junior middle school Senior middle school College and above Missing

29.7

30.8

30.8

26.8

31.0

32.5

30.0

9.2

10.7

8.7

8.3

11.5

9.8

8.9

3.8

18.5 0.0

21.0 0.0

14.3 0.1

19.2 0.1

26.1 0.0

19.0 0.1

10.5 0.0

1.1 0.1

1.5 0.1

0.6 0.1

1.1 0.1

1.6 0.2

1.1 0.1

1.0 0.1

0.5 0.0

Income group First group (low) Second group Third group Fourth group Fifth group (high)

5.7 9.0 13.4 25.5 46.4

1.7 3.9 7.9 24.5 62.0

11.1 15.2 20.5 26.6 26.6

5.7 10.1 14.0 25.9 44.3

1.9 4.5 8.2 23.2 62.2

5.0 7.9 12.4 26.5 48.2

10.8 15.6 20.6 27.5 25.5

29.3 26.4 21.0 14.8 8.5

17.0 23.0 23.4 21.2 15.4

23.7 30.6 24.5 15.3 5.9

43.0 26.2 16.8 9.3 4.7

20.1 20.7 20.1 20.8 18.3

24.0 27.9 24.3 16.9 6.9

30.5 30.6 21.7 11.8 5.2

59.2 22.5 12.2 4.7 1.4

Occupational status Employed 41.4 Retired 32.2 Student 4.3 Unemployed 22.1 Missing 0.0

42.1 36.9 4.9 16.1 0.0

41.4 27.2 3.4 28.0 0.0

40.3 30.4 4.5 24.7 0.1

38.0 41.8 4.5 15.7 0.0

44.0 34.2 3.5 18.2 0.1

43.2 18.6 4.7 33.4 0.1

81.7 1.8 4.5 11.7 0.3

78.2 2.2 3.6 15.8 0.2

82.9 1.1 4.1 11.5 0.4

83.8 1.9 5.3 8.6 0.4

77.3 2.6 4.1 16.0 0.0

81.4 1.3 5.1 11.9 0.3

83.0 1.9 4.3 10.4 0.4

88.2 0.9 3.8 6.2 0.9

Rural respondents reported more problems in all the EQ-5D dimensions and lower VAS scores than the urban respondents (Table 2). The difference in problems reported in EQ-5D dimensions between the rural and urban respondents increased with age.

women aged 45–64 years, respondents from big cities reported the lowest problems in all EQ-5D dimensions. For women aged 65+ , except for pain/discomfort and anxiety/depression dimensions, respondents from middle-sized cities reported the lowest percentage of problems in EQ-5D dimensions. For all the age groups, women from small cities reported the highest percentage of problems in most EQ-5D dimensions.

4.3. By eastern, middle and western areas

4.5. By county category in rural areas

Problems reported in EQ-5D dimensions increased and VAS scores decreased from eastern to western areas for both urban (Table 3) and rural respondents (Table 4). This gradient was steeper among rural than among urban respondents.

Overall, problems reported in EQ-5D dimensions increased and VAS scores decreased from Type I to Type IV county, from most developed to least developed county (Table 6). Exceptions were found for both men and women aged 15–44 years in the usual activities and pain/discomfort dimensions and men in mobility dimension, where Type II county respondents reported the lowest percentage of problems.

4.2. By urban and rural areas

4.4. By city category in urban areas For men aged 15–44 years, except for the anxiety/depression dimension, respondents from big city reported the lowest percentage of problems in all EQ-5D dimensions; for men aged 45 years and above, respondents from middle-sized cities reported the lowest problems in most EQ-5D dimensions (Table 5). For all age groups, men from small cities reported the highest percentage of problems in most EQ-5D dimensions. For women aged 15–44 years, except for the usual activities dimension, respondents from middle-sized cities reported the lowest percentage of problems in all EQ-5D dimensions; for

4.6. Regression analyses For each region, the likelihood of having no problems in a specific EQ-5D dimension and the variation of VAS score were analysed, controlling for sex and age in Table 7, and also for educational level, income group and employment status in Table 8. The likelihood of women having no problems was significantly lower than for men, and women had significantly lower mean VAS scores than men. Individuals with a higher

S. Sun et al. / Health & Place 17 (2011) 671–680

675

Table 2 Percentage of respondents reporting moderate or severe problems in each EQ-5D dimension, VAS score (mean), by age group, sex and urban–rural areas, total, China 2008. EQ-5D dimension

Urban and rural areas 15–44 years

45–64 years

65 +

Urban 6358 7707 %

Rural 20,900 22,993 %

Urban 6009 6757 %

Rural 16,356 16,286 %

Mobility Men Women

0.9 0.9

1.4 1.4

3.8 3.7

Self-care Men Women

0.5 0.5

0.8 0.9

Usual activities Men Women

0.6 0.8

Pain/discomfort Men Women Anxiety/depression Men Women

Men (n) Women (n)

VAS score (mean) Men Women

Urban 3052 3563 %

Rural 5488 5234 %

4.9 6.2

13.7 16.1

18.8 22.5

2.2 1.8

2.9 3.7

8.1 10.0

13.1 16.3

1.4 1.4

3.0 2.9

4.7 5.8

11.3 13.8

18.8 22.4

1.5 2.3

2.9 4.0

7.3 9.3

10.3 14.8

15.7 22.6

24.6 29.9

1.9 2.8

3.0 3.7

5.3 6.0

7.0 9.5

7.6 11.0

16.4 19.6

86.1 86.1

84.5 84.9

77.9 78.8

76.8 76.3

71.2 69.5

69.7 67.3

Table 3 Percentage of respondents reporting moderate or severe problems in each EQ-5D dimension, VAS score (mean), by age group, sex and eastern–middle–western area, urban area, China 2008. EQ-5D dimension

Urban area (eastern, middle, western) 15–44 years

Men (n) Women (n)

Eastern 2502 3017 %

45–64 years Middle 1912 2367 %

Western 1944 2323 %

Eastern 2597 1614 %

65 + Middle 1782 994 %

Western 1630 955 %

Eastern 1338 1614 %

Middle 824 994 %

Western 890 955 %

Mobility Men Women

0.5 0.5

0.9 1.2

1.3 1.0

3.2 3.2

4.8 4.0

3.6 4.1

12.9 16.9

14.0 16.0

14.7 15.1

Self-care Men Women

0.2 0.3

0.6 0.7

0.8 0.6

1.4 1.2

2.9 2.2

2.7 2.4

7.4 9.4

9.0 11.1

8.3 9.7

Usual activities Men Women

0.3 0.4

0.8 1.3

0.8 0.8

2.1 2.2

3.9 3.3

3.3 3.6

10.2 13.1

11.5 13.9

12.7 15.1

Pain/discomfort Men Women

1.2 2.0

1.6 2.9

1.6 2.2

6.8 9.6

7.6 8.6

7.8 9.5

15.8 23.3

15.3 22.6

15.8 21.3

Anxiety/depression Men Women

1.3 2.3

1.7 3.0

2.9 3.4

4.5 5.2

5.3 5.7

6.6 7.6

7.2 9.4

7.7 12.9

8.1 11.9

87.5 85.9

86.7 85.5

83.6 81.8

78.6 78.0

78.3 77.5

76.4 74.3

71.6 70.4

72.6 69.9

69.4 68.1

VAS score (mean) Men Women

educational level, in a higher income group and gainfully employed were more likely to have no problems in EQ-5D dimensions, and they had higher mean VAS scores (results not shown). In Table 7, when sex and age were controlled, rural respondents had significantly lower likelihood of having no problems in

all EQ-5D dimensions and had significantly lower VAS scores than the urban respondents. For urban areas, respondents in the eastern area were significantly more likely to have no problems in EQ-5D dimensions than those in the western area, with the exception of pain/discomfort. Difference in the anxiety/depression dimension was found between middle and

676

S. Sun et al. / Health & Place 17 (2011) 671–680

Table 4 Percentage of respondents reporting moderate or severe problems in each EQ-5D dimension, VAS score (mean), by age group, sex and eastern–middle–western area, rural area, China 2008. EQ-5D dimension

Rural area (eastern, middle, western) 15–44 years Eastern 6252 7161 %

Men (n) Women (n)

45–64 years Middle 5355 6079 %

Western 9293 9753 %

Eastern 5426 5448 %

65 + Middle 4620 4658 %

Western 6310 6180 %

Eastern 1978 2104 %

Middle 1363 1169 %

Western 2147 1961 %

Mobility Men Women

0.9 0.9

1.8 1.1

1.5 1.9

3.1 4.3

4.8 6.0

6.5 8.0

16.2 18.3

19.4 23.0

20.8 26.6

Self-care Men Women

0.5 0.6

1.1 0.7

0.8 1.1

1.9 2.5

3.0 3.5

3.8 4.9

11.3 12.9

14.0 17.3

14.2 19.3

Usual activities Men Women

1.0 0.9

1.7 1.2

1.4 1.9

3.1 4.0

4.7 5.7

6.0 7.6

17.0 17.5

18.6 23.1

20.4 27.1

Pain/discomfort Men Women

2.2 3.1

2.8 3.7

3.4 4.9

7.7 11.2

9.8 13.9

12.9 18.7

20.7 25.2

25.4 31.6

27.6 33.9

Anxiety/depression Men Women

1.6 1.7

2.1 2.2

4.4 6.0

4.1 6.3

4.7 7.1

11.1 14.2

12.6 15.4

14.2 17.2

21.2 25.5

88.6 87.6

85.8 84.4

84.7 83.2

81.5 79.1

78.1 75.5

77.0 74.5

71.9 69.5

67.7 65.2

68.5 66.1

VAS score (mean) Men Women

Table 5 Percentage of respondents reporting moderate or severe problems in each EQ-5D dimension, VAS score (mean), by age group, sex and big-middle-small city category, urban area, China 2008. EQ-5D dimension

City category (urban area) 15–44 years Big city

Men (n) Women (n)

2169 2623 %

45–64 years Middle-sized city 1934 2271 %

Small city

Big city

2255 2813 %

2468 2749 %

65 + Middle-sized city 1653 1835 %

Small city

Big city 1463 1737 %

Middle-sized city 841 938 %

1888 2173 %

Small city 748 888 %

Mobility Men Women

0.6 0.8

1.0 0.6

0.9 1.1

3.4 3.0

3.0 3.4

5.0 4.7

14.2 15.7

10.9 14.2

15.8 19.1

Self-care Men Women

0.3 0.4

0.6 0.4

0.6 0.6

1.8 1.1

1.9 1.5

2.9 3.0

7.9 9.5

7.1 7.7

9.5 13.3

Usual activities Men 0.4 Women 0.5

0.6 0.6

0.8 1.2

2.7 2.1

2.2 2.4

3.9 4.4

11.1 13.8

9.6 10.3

13.6 17.6

Pain/discomfort Men 1.0 Women 2.1

1.1 1.7

2.1 3.0

7.7 8.4

5.8 9.0

8.2 10.5

15.6 20.7

13.7 22.6

18.1 26.2

Anxiety/depression Men 1.9 Women 3.1

1.7 2.6

2.1 2.8

5.4 5.1

4.7 6.1

5.7 6.9

6.7 8.9

6.1 10.2

11.0 16.0

VAS score (mean) Men 84.7 Women 83.8

86.9 85.2

86.7 84.7

76.9 76.2

79.0 78.1

78.2 76.6

69.9 69.3

73.1 70.7

71.7 69.1

western area respondents. Both eastern and middle area respondents had significantly higher VAS scores than western respondents. For rural areas, an eastern–middle–western gradient in health was observed; there were significant differences for all dimensions

between all areas. In urban areas, overall, middle-sized city respondents had the best health, while the respondents in small cities had the worst. An exception was seen for VAS scores, where the lowest scores were found among big city respondents. In rural

S. Sun et al. / Health & Place 17 (2011) 671–680

677

Table 6 Percentage of respondents reporting moderate or severe problems in each EQ-5D dimension, VAS score (mean), by age group, sex and Type I–IV county category, rural area, China 2008. EQ-5D dimension

County category (rural area) 15–44 years Type I 4584 5282 %

Men (n) Women (n)

45–64 years

Type II 6741 7251 %

Type III 6381 7222 %

Type IV 3194 3238 %

Type I 4103 4203 %

65 +

Type II 5259 5211 %

Type III 5338 5357 %

Type IV 1656 1515 %

Type I 1479 1558 %

Type II 1611 1476 %

Type III 1916 1713 %

Type IV 482 487 %

Mobility Men Women

1.3 1.0

1.1 1.1

1.7 1.6

1.5 2.2

3.8 5.3

4.4 5.7

5.2 6.4

8.2 9.3

15.9 17.5

17.7 21.8

20.8 25.8

23.4 28.8

Self-care Men Women

0.6 0.5

0.7 0.6

0.9 1.1

0.9 1.4

2.1 2.9

2.8 3.7

3.2 3.7

5.0 5.5

10.4 12.1

13.2 17.1

14.2 17.9

16.8 21.4

Usual activities Men Women

1.3 1.1

1.2 1.0

1.5 1.6

1.5 2.7

3.3 4.7

4.2 5.4

5.1 6.1

8.0 9.5

16.2 17.9

17.9 22.0

20.0 24.7

24.1 29.4

Pain/discomfort Men Women

2.7 3.6

2.3 3.1

3.2 4.6

3.5 5.6

9.0 13.0

9.0 13.0

11.8 17.1

13.2 18.3

21.5 26.1

21.2 27.5

29.0 34.1

27.6 34.5

Anxiety/depression Men Women

1.9 2.1

2.0 2.2

3.4 4.5

5.6 7.7

5.0 7.1

6.1 7.8

7.7 11.6

12.1 15.0

13.3 15.2

14.3 17.9

19.0 23.2

22.2 25.9

87.2 86.3

86.6 85.4

86.2 84.7

83.6 81.8

80.1 77.6

79.1 76.5

78.4 75.9

76.1 73.8

70.8 68.7

69.5 67.4

68.8 66.1

68.4 66.6

VAS score (mean) Men Women

Table 7 Multiple logistic regression analyses on having no problems in EQ-5D dimensions and linear multiple regression analyses on VAS score, by region, controlled for sex and age group, China 2008. Mobility

Self-care

Usual activities

Pain/discomfort

Anxiety/depression

VAS

Odds ratio

95% CI

Odds ratio

95% CI

Odds ratio

95% CI

Odds ratio

95% CI

b-

Odds ratio

95% CI

0.64

(0.60–0.68)

0.56

(0.52–0.61)

0.52

(0.49–0.56)

0.61

(0.58–0.64)

0.59

(0.56–0.62)

 0.25

Urban area categoryb Eastern 1.17 Middle 0.95

(1.03–1.33) (0.82–1.08)

1.44 0.91

(1.22–1.70) (0.77–1.08)

1.45 1.00

(1.26–1.67) (0.86–1.15)

1.01 0.99

(0.92–1.12) (0.89–1.11)

1.44 1.19

(1.28–1.63) (1.05–1.35)

3.06 3.01

(1.25–1.60) (1.31–1.74)

1.66 1.65

(1.42–1.94) (1.38–1.98)

1.61 1.79

(1.41–1.84) (1.53–2.09)

1.31 1.36

(1.19–1.45) (1.21–1.51)

1.35 1.33

(1.20–1.52) (1.17–1.51)

 1.08 1.01

Rural area categoryd Eastern 1.84 Middle 1.27

(1.70–1.98) (1.18–1.37)

1.80 1.20

(1.64–1.97) (1.09–1.31)

1.83 1.26

(1.70–1.98) (1.16–1.36)

1.70 1.30

(1.61–1.80) (1.23–1.38)

2.57 2.19

(2.40–2.75) (2.04–2.34)

4.37 0.99

(3.75–4.99) (0.36–1.63)

Rural county categorye Type I 2.02 Type II 1.68 Type III 1.39

(1.81–2.25) (1.52–1.87) (1.26–1.53)

2.21 1.56 1.42

(1.93–2.53) (1.38–1.77) (1.26–1.61)

2.18 1.79 1.52

(1.96–2.43) (1.62–1.99) (1.37–1.68)

1.57 1.57 1.13

(1.45–1.71) (1.45–1.70) (1.04–1.22)

2.72 2.35 1.56

(2.48–2.97) (2.16–2.56) (1.44–1.69)

4.16 3.04 2.51

(3.40–4.91) (2.30–3.78) (1.77–3.25)

Rurala

Urban city categoryc Big city 1.41 Middle-sized 1.51 city

95% CI

estimate ( 0.81–0.31) (2.26–3.85) (2.18–3.83) ( 1.88- -0.29) (0.18–1.83)

a

Base line category: urban. Base line category: urban western area. c Base line category: small city. d Base line category: rural western area. e Base line category: Type IV county. b

areas, a gradient in the likelihood to have no problems was seen for all EQ-5D dimensions and in the VAS score over Type I–IV county, i.e. respondents in Type IV county reported worst health. In Table 8, after controlling also for socio-economic status, overall, the differences between regions were reduced and some

differences became insignificant, but the trend was similar. This may suggest that the regional differences in health are partly due to differences in socio-economic composition between regions. Some exceptions were found for the rural eastern–middle–western areas, where the effects of regional variables increased in the

678

S. Sun et al. / Health & Place 17 (2011) 671–680

Table 8 Multiple logistic regression analyses on having no problems in EQ-5D dimensions and linear multiple regression analyses on VAS score, by region, controlled for sex, age group, educational level, income group and employment status, China 2008. Mobility

Rurala

Self-care

Usual activities

Pain/discomfort

Anxiety/depression

VAS

bestimate

Odds ratio

95% CI

Odds ratio

95% CI

Odds ratio

95% CI

Odds ratio

95% CI

Odds ratio

95% CI

95% CI

0.78

(0.72–0.86)

0.71

(0.63–0.79)

0.66

(0.60–0.72)

0.75

(0.70–0.80)

0.91

(0.84–0.98)

1.52

(0.87–2.17)

Urban area categoryb Eastern 1.03 Middle 1.07

(0.91–1.18) (0.93–1.23)

1.28 1.01

(1.08–1.52) (0.85–1.20)

1.28 1.12

(1.11–1.48) (0.97–1.31)

0.89 1.11

(0.80–0.99) (0.99–1.24)

1.21 1.37

(1.07–1.36) (1.20–1.55)

2.34 3.51

(1.54–3.15) (2.68–4.34)

Urban city categoryc Big city 1.08 Middle-sized city 1.26

(0.94–1.24) (1.08–1.46)

1.22 1.32

(1.02–1.45) (1.09–1.59)

1.17 1.42

(1.01–1.36) (1.20–1.67)

1.02 1.13

(0.92–1.14) (1.01–1.27)

0.93 1.02

(0.82–1.06) (0.89–1.17)

 2.71  0.14

(  3.53–1.88) (  0.98–0.69)

Rural area categoryd Eastern 1.83 Middle 1.25

(1.69–1.99) (1.15–1.35)

1.89 1.20

(1.72–2.09) (1.10–1.32)

1.87 1.24

(1.72–2.02) (1.15–1.35)

1.64 1.27

(1.55–1.75) (1.20–1.35)

2.77 2.28

(2.59–2.96) (2.12–2.44)

3.42 0.45

(2.79–4.05) (  0.19–1.09)

Rural county categorye Type I 1.89 Type II 1.55 Type III 1.28

(1.69–2.12) (1.39–1.72) (1.16–1.42)

2.29 1.54 1.39

(1.99–2.64) (1.35–1.75) (1.22–1.57)

2.12 1.68 1.42

(1.89–2.38) (1.51–1.87) (1.28–1.57)

1.38 1.38 1.01

(1.27–1.51) (1.27–1.50) (0.93–1.09)

2.16 1.92 1.31

(1.97–2.38) (1.76–2.10) (1.21–1.42)

2.05 1.29 1.05

(1.27–2.83) (0.53–2.05) (0.30–1.81)

a

Base line category: urban. Base line category: urban western area. c Base line category: small city. d Base line category: rural western area. e Base line category: Type IV county. b

dimensions self-care, usual activities and anxiety/depression, after controlling for socio-economic status. The VAS score coefficient for the rural area became positive, which is the opposite of what can be seen in Table 7.

5. Discussion We investigated regional inequalities in HRQoL measured by the EQ-5D instrument among the Chinese population stratified by sex and age from four different levels. Disparities in health between urban and rural areas have been demonstrated by previous studies (Gao et al., 2002; Henderson et al., 1994; Liu et al., 1999), and our study reported similar findings with the EQ-5D descriptive system, though with smaller differences. Respondents’ health status decreased from eastern to western areas, which is in line with previous findings (Fang et al., 2009; Pei and Rodriguez, 2006; UNDP, 2005). We also found such differences to be profounder in rural than in urban areas. In the urban areas, respondents from middlesized cities reported best health; in rural areas, respondents’ health status diminished from Type I to Type IV counties. To the best of our knowledge, this is the first time that regional inequalities in health have been investigated by applying the EQ-5D instrument to a nationally representative sample in China, and the inter-city and inter-county differences in health have not yet been reported in any other articles published internationally. Although the Chinese government has devoted much effort to reduce the huge disparities between urban and rural areas – e.g. by providing free education, building up the rural pension system, providing a new Rural Cooperative Medical System, etc. – the urban areas are still much more developed than the rural areas– e.g. the illiterate population is found mainly in rural areas (illiteracy rate: 4.6% in urban areas, 11.6% in rural areas) (UNDP, 2005), the number of persons with university education is higher in urban areas and urban residents’ per person income is higher. These differences will be profounder if all types of subsidies to urban respondents are taken into consideration (UNDP, 2005). China’s urban–rural income inequalities might be the highest in

the world (UNDP, 2005). The majority of rural respondents are employed in agriculture, and as peasants most of them do not have unemployment benefit or pension, and their social security network relies mainly on the family. We expected a large difference in health status between the urban and rural respondents, but as the results showed, although urban respondents reported better health than rural respondents, compared with studies based on life expectancy, mortality or morbidity data, the magnitude of the differences was smaller. In the regression analyses, when socio-economic status was controlled, the opposite trend was found for VAS score. This might be linked with the way in which people judge their own health status. Studies suggested that respondents with lower expectation of health might rate their own health status higher (Sen, 2002; Symon et al., 2006). The rural respondents might have a lower expectation of health than the urban respondents, in which case, given the same health conditions, the rural respondents might perceive their health status higher than the urban respondents. This issue needs further investigation. The eastern–middle–western disparities in developments have existed historically. Although in recent years China’s development strategy has more focused on western areas – e.g. low taxation and better transport conditions, national level fiscal transferring to the middle and western areas, encouraging people to migrate to western areas – there is still a huge gap between those areas. For example, the school drop-out rate is highest in the western area (UNDP, 2005), and average household income decreased from east to west. From eastern to western area, health status went from high to low, and disparities in health are profounder in rural than in urban areas. This might be due to that middle-sized cities being located mainly in the middle and west of China, and health is better in middle-sized cities. The possible reasons are given in the following paragraph, where we discuss about inter-city differences in health. In rural areas with the same socio-economic status, respondents’ health status declined from eastern to western area. Health status was best in middle-sized cities. The reason could possibly be that there is a lower stress level, better housing

S. Sun et al. / Health & Place 17 (2011) 671–680

situation and a better environment in middle-sized cities than in big and small cities, but this issue needs further investigation (Van de Poel et al., 2009). Big city respondents had fewer problems in EQ-5D dimensions than small city respondents. However, they had lower VAS scores, and this difference was enlarged after controlling for socio-economic status. Respondents in big cities might have higher expectations of health than small city respondents, and might rate their health lower for this reason (Sen, 2002; Symon et al., 2006). Our results showed that from Type I to Type IV counties, health status went from high to low, and this gradient persisted when socio-economic status were all controlled. Type III and IV counties are mainly located in the mountain and desert region in the western area. Basic living conditions are quite rough there, with respondents facing such problems as lack of clean drinking water and difficulties in communicating with the outside world, due to poor road conditions. When interpreting the results, the minimum important differences (MID) in health status need to be addressed. To our best knowledge, there is no authoritative recommendation for the MID for EQ-5D. Some studies have investigated this issue, mainly focused on utility scores (Luo et al., 2010; Sullivan et al., 2005; Walters and Brazier, 2005). Most studies on MID for visual analogue scales are related to a certain health problem, e.g. pain or sleep problems (Salaffi et al., 2004; Zisapel and Nir, 2003). Further studies are needed to explore this issue. Some limitations must be addressed here, apart from limitations regarding how to define socio-economic status, interviewer bias and ceiling effects, which have been discussed elsewhere (Sun et al., 2010). The sampling design was complex, using a multi-stage sampling with both stratification and clustering. The difficulties of taking the effects of the sampling design into consideration at all stages of the analyses might affect the precision of our estimates. Nevertheless, the NHSS sampling design was examined by the MoH for all waves of the surveys, and the representativeness of the sample was considered good, i.e. proportions of the population from different regions, age, sex and socio-economic structures are representative of the Chinese population (CHSIMOHPRC, 2008), and are similar to the census data, except for the unemployment rate, which might be due to different ways of defining unemployment (National Bureau of Statistics of China, 2009). Another limitation of our study is related to the situation concerning inner-immigrants, those rural respondents working in urban areas. This number has increased, and by 2002 it will reach 150 million (UNDP, 2005). However, they were not included in our study. Inner-immigrants are often associated with low pay, harsh working and living conditions, and difficulties in accessing health care and social security networks, due to their not being registered residents of urban areas (Zhan et al., 2002). The NHSS question applied a proxy to investigate these persons’ living situation, but information on their health status is lacking. Further studies are needed among the inner-immigrants. There are large regional differences in income, and for the same amount of income the relative values are different in different regions, so that the income group could differ from one region to another. But in order to make cross-regional comparisons, we used the average income group for the whole country. The limitation of doing so is due to the fact that the same amount of money is worth more in poor regions. Comparison of our results with population studies in other countries showed similar findings, with most problems reported in the pain/discomfort dimension, followed by the anxiety/ depression dimension, more problems with increased age and ¨ et al., 2001; women reporting worse health than men (Burstrom ¨ Kind et al., 1998; Konig et al., 2009; Leung et al., 2007; Lubetkin

679

et al., 2005; Sun et al., 2010; Sorensen et al., 2009; Szende and Williams, 2004;). However, it is difficult to make a direct comparison for the proportion of respondents reporting problems in EQ-5D dimensions, as this differs from country to country. This might be due to several reasons other than the difference in health status between populations. The age and sex structure might be different across countries, or people in different countries might refer to levels of health differently (Szende and Williams, 2004; Sorensen et al., 2009; Bharmal and Thomas, 2006), or again the mode of administration might vary from one survey to another (Bowling, 2005; Weinberger et al., 1996), or some countries include proxy respondents. Therefore, one needs to be cautious when making international comparisons of population health status. A large ceiling effect (a large proportion of the population tending to report good health, which here refers to reporting no problems in any EQ-5D dimensions) as found in our study (87%) has also been observed in other studies including Chinese populations, in China (Wang et al., 2005), as well as in US (Fu and Kattan, 2006; Lubetkin et al., 2005) and Canada (Leung et al., 2007). This was also the case in a study in Japan (Tsuchiya et al., 2002) and in Malaysia (Shafie et al., 2010). We have investigated different types of regional inequalities in health status, by using the EQ-5D instrument in a national representative sample. Given the caveats, the EQ-5D distinguished well between regions that were a priori supposed to differ in health status. Our study adds knowledge regarding regional disparities in different dimensions of health, apart from the known regional differences in health using mortality and morbidity data. This study enhances the understanding of the urban–rural differences and east–middle–west differences in China, and also sheds light on inequalities in health status between different city categories in the urban areas and county categories in the rural areas. The results can be used as EQ-5D norms data for different regions in China, and as a reference for further studies regarding regional health status.

Acknowledgements We would like to thank the Ministry of Health in China for data collection and their support of data analysis work. We acknowledge financial support from the Swedish Research Council (Swedish Research Links programme 348-2009-6538). We are also grateful for the helpful comments and suggestions received on earlier versions of this paper from the Equity and Health Policy Research Group, Department of Public Health Sciences, Karolinska Institutet, at the 26th EuroQol Plenary Meeting in Paris in 2009, at the 7th World Congress of Health Economics (iHEA) in Beijing in 2009, at the 16th Annual Conference of the International Society for Quality of Life Research (ISOQOL) in New Orleans in 2009, and from the Health Economics Research Group, Medical Management Centre, Karolinska Institutet. We would also like to thank two anonymous reviewers, for their constructive comments and valuable suggestions. References Bellanger, M.M., Jourdain, A., 2004. Tackling regional health inequalities in France by resource allocation: a case for complementary instrumental and processbased approaches? Applied Health Economics and Health Policy 3, 234–250. Bowling, A, 2005. Mode of questionnaire administration can have serious effects on data quality. Journal of Public Health (Oxf) 27, 281–291. Bharmal, M., Thomas, J., 2006. Comparing the EQ-5D and SF-6D descriptive systems to assess their ceiling effects in the US general population. Value in Health 9 (4), 262–271. ¨ Burstrom, K., Johannesson, M., Diderichsen, F., 2001. Swedish population healthrelated quality of life results using the EQ-5D. Quality of Life Research 10, 621–635.

680

S. Sun et al. / Health & Place 17 (2011) 671–680

¨ ¨ Burstrom, B., Hemmingsson, T., Burstrom, K., Corman, D., Farah, A., Gustavsson, A., ¨ et al., (updated 21 May 2010). Livsvillkor, levnadsvanor och halsa i Stockholms ¨ ¨ ¨ orelser ¨ lan: oppna jamf 2010. Living condition, living habits and health in Stockholm County: open comparison 2010 (in Swedish). Summary in English, School of Public Health, Karolinska Institutet, Stockholm, Sweden. /http:// www.folkhalsoguiden.se/upload/folkh%c3%a4lsoarbete/Livsvillkor,%20levnads vanor%20och%20h%c3%a4lsa%20i%20Stockholms%20l%c3%a4n.pdfS (accessed 06.08.2010). Center for Health Statistics and Information of Ministry of Health of People’s Republic of China, 2009. An analysis report of National Health Services Survey in China, 2008 (in Chinese). Centre for Health Statistics and Information of Ministry of Health of People’s Republic of China, Beijing, China. Center for Health Statistics and Information of Ministry of Health of People’s Republic of China, 2008. The Protocol and Guidelines for the Fourth National Health Service Survey (in Chinese). Centre for Health Statistics and Information of Ministry of Health of People’s Republic of China, Beijing, China. Central People’s Government of the People’s Republic of China, 2004. Zhongguo de she hui bao zhang zhuang kuang he zheng ce. Situation and policy regarding China’s social welfare (in Chinese). Central People’s Government of the People’s Republic of China, Beijing, China. Congdon, P., 2001. Health status and healthy life measures for population health need assessment: modelling variability and uncertainty. Health & Place 7, 13–25. Fang, P.Q., Dong, S.P., Xiao, J.J., Liu, C.J., 2009. Regional inequalities in health and its determinants: evidence from China. Health Policy. doi:10.1016/j.healthpol. 2009.08.002. Fryback, D.G., Dunham, N.C., Palta, M., Hanmer, J., Buechner, J., Cherepanov, D., 2007. US norms for six generic health-related quality of life indexes from the National Health Measurement Study. Medical Care 45 (12), 1162–1170. Fu, A.Z., Kattan, M.W., 2006. Racial and ethnic differences in preference-based health status measure. Current Medical Research Opinion 22 (12), 2348–2349. Gao, J., Qian, J.C., Tang, S.L., Eriksson, B., Blas, E., 2002. Health equity in transition from planned market economy in China. Health Policy Plan 17, 20–29. Henderson, G., Akin, J., Li, Z.M., Jin, S.G., Ma, H.J., Ge, K.Y., 1994. Equity and the utilization of health services: report of an eight-province survey in China. Social Science & Medicine 39 (5), 687–699. Hoi, L.V., Chuc, N.T.K., Lindholm, L., 2010. Health-related quality of life, and its determinants, among older people in rural Vietnam. BMC Public Health 10, 549. Kil, S.R., Lee, S.I., Yun, S.C., H.M., & Jo, M..W., 2008. The decline of health-related quality of life associated with some diseases in Korean adults (in Korean). Journal of Preventive Medical Public Health 41 (6), 434–441. Kind, P., Dolan, P., Gudex, C., Williams, A., 1998. Variations in population health status: results from a United Kingdom national questionnaire survey. British Medical Journal 316, 736–741. ¨ Konig, H.H., Bernert, S., Angermeyer, M.C., Mattschinger, H., Martinez, M., Vilagut, G., 2009. Comparison of population health status in six European countries. Medical Care 47 (2), 255–261. Leung, B., Luo, N., So, L., Quan, H., 2007. Comparing three measures of health status (perceived health with Likert-type scale, EQ-5D, and number of chronic condition) in Chinese and white Canadians. Medical Care 45 (7), 610–617. Liu, Y., Hsiao, W.C., Eggleston, K., 1999. Equity and health care: the Chinese experience. Social Science & Medicine 49, 1349–1356. Lopez-Casanovas, G., Costa-Font, J., Planas, I., 2005. Diversity and regional inequalities in the Spanish ‘system of health care services’. Health Economics 14, 221–235. Lubetkin, E.I., Jia, H.M., Franks, P., Gold, M.R., 2005. Relationship among sociodemographic factors, clinical conditions, and health-related quality of life: examining the EQ-5D in the US general population. Quality of Life Research 14, 2187–2196. Luo, N., Chew, L.H., Fong, K.Y., Koh, D.R., Ng, S.C., Yoon, K.H., 2003. Validity and reliability of the EQ-5D self-report questionnaire in Chinese-speaking patients with rheumatic disease in Singapore. Journal of Rheumatology 30 (10), 2268–2274. Luo, N., Johnson, J.A., Shaw, J.W., Feeny, D., Coons, S.J., 2005. Self-reported health status of the general adult US population as assessed by the EQ-5D and health utility index. Medical Care 43 (11), 1078–1086. Luo, N., Johnsson, J.A., Coons, S.J., 2010. Using instrument-defined health state transition to estimate minimally important differences for four preferencebased health-related quality of life instruments. Medical Care 48, 365–371. National Bureau of Statistics of China, (updated 2009). China Statistic Yearbook 2008. /http://www.stats.gov.cn/tjsj/ndsj/2009/indexeh.htmS (accessed 08.24.2010). National Development and Reform Commission, (updated 8 April 2009). Opinions of the China People’s Congress Central Committee and State Council on Deepening the Health Care System Reform. /http://www.china.org.cn/govern ment/scio-press-conferences/2009-04/09/content_17575378.htmS (accessed 06.08.2010).

Office of the WHO representative in China and Social Development Department of China State Council Development Research Centre, (updated June 2006). China: health, poverty and economic development. /http://www.wpro.who.int/NR/ rdonlyres/A1F18401-BE93-44EF-9F76-55DDA2C6E12D/0/hped_en.pdfS (accessed 06.08.2010). Pei, X., Rodriguez, E., 2006. Provincial income inequality and self-reported health status in China during 1991-7. Journal of Epidemiology Community Health 60, 1065–1069. Rabin, R., de Charro, F, 2001. EQ-5D: a measure of health status from the EuroQol Group. Annals of Medicine 33 (5), 337–343. Salaffi, F., Stancati, A., Silverstri, C.A., Ciapetti, A., Grassi, W., 2004. Minimal clinically important changes in chronic musculoskeletal pain intensity measured on a numerical rating scale. European Journal of Pain 8, 283–291. Salmela, R., 1993. Regional inequalities in health and health care in Finland and Norway. Health Policy 24, 83–94. SAS Institute Inc., 2006. second editionBase SASs 9.1.3 Procedures Guide, vols. 1, 2, 3, and 4. SAS Institute Inc., Cary, NC. Sen, A., 2002. Health: perception versus observation: self reported morbidity has severe limitations and can be extremely misleading. British Medical Journal 324, 860–861. Shafie, A.A., Hassali, M.A., Liau, S.Y., 2010. A cross-sectional validation study of EQ-5D among the Malaysian adult population. Quality of Life Research. doi:10.1007/s11136-010-9774-6. Shi, L.Y., 1993. Health care in China: a rural–urban comparison after the socioeconomic reforms. Bulletin of the World Health Organization 71 (6), 723–736. Sorensen, J., Davidsen, M., Gudex, C., Pedersen, K.M., Bronnum-Hansen, H., 2009. Danish EQ-5D population norms. Scandinavian Journal of Public Health 37, 467–474. Spiker, B., Revicki, D.A., 1996. Taxonomy of quality of life. In: Spiker, B. (Ed.), Quality of Life and Pharmacoeconomics in Clinical Trials. Lippingcott-Raven Publishers, Philadelphia, pp. 25–32. ¨ Sun, S., Chen, J.Y., Johannesson, M., Kind, P., Xu, L., Zhang, Y.G., Burstrom, K., 2010. Population health status in China: EQ-5D results, by age, sex and socioeconomic status, from the National Health Services Survey 2008. Quality of Life Research. doi:10.1007/s11136-010-9762-x. Sullivan, P.W., Lawrence, W.F., Ghushchyan, V., 2005. A national catalogue of preference-based scores for chronic conditions in the United States. Medical Care 43, 736–749. Symon, Z., Daignault, S., Symon, R., Dunn, R.L., Sanda, M.G., Sandler, H.M., 2006. Measuring patient’s expectations regarding health-related quality-of-life outcomes associated with prostate cancer surgery or radiotherapy. Adult Urology 68, 1224–1229. Szende, A, Williams, A (Eds.), 2004. Spring Med Publishing Ltd., Hungary. Tang, S., Meng, Q., Chen, L., Bekedam, H., Evans, T., Whitehead, M., 2008. Tackling the challenges to health equity in China. Lancet 372, 1493–1501. Ting, J.C., Tarn, Y.H., Ching, L.H., Liou, W.S., Shaw, J.W., Chiou, X.G., 2007. Taiwanese version of the EQ-5D: validation in a representative sample of the Taiwanese population. Journal Formosan Medical Association 106 (12), 1023–1031. Tsuchiya, A., Ikeda, S., Ikegami, N., Nishimura, S., Sakai, I., Fukuda, T., et al., 2002. Estimating an EQ-5D population value set: the case of Japan. Health Economics 11 (4), 341–353. United Nations Development Programme, (updated October 2005). China Human Development Report 2005: development with equity. /http://ch.undp.org.cn/ downloads/nhdr2005/c_NHDR2005_complete.pdfS (accessed 06.08.2010). Van de Poel, E., O’Donnell, O., Van Doorslaer, E., 2009. Urbanization and the spread of diseases of affluence in China. Economics and Human Biology 7, 200–216. Wang, H., David, A., Mullahy, K., Mullahy, J., 2005. Variation in Chinese population health related quality of life: results from a EuroQol study in Beijing, China. Quality of Life Research 14, 119–132. Walters, S.J., Brazier, J.E., 2005. Comparison of the minimally important difference for two health state utility measures: EQ-5D and SF-6D. Quality of Life Research 14, 1523–1532. Weinberger, M., Oddone, E.Z., Samsa, G.P., Landsman, P.B, 1996. Are health-related quality of life measures affected by the mode of administration? Journal of Clinical Epidemiology 49 (2), 135–140. World Health Organisation, (updated June 2001). The world health report 2000—health systems: improving performance. /http://www.who.int/whr/2000/ en/whr00_en.pdfS (accessed 06.08.2010). Xu, L., Wang, Y., Collins, D.C., Tang, S., 2007. Urban health insurance reform and coverage in China using data from National Health Services Surveys in 1998 and 2003. BMC Health Services Research, 7–37. doi:10.1186/1472-6963-7-37. Zhan, S.K., Sun, Z.W., Blas, E., 2002. Economic transition and maternal health care for internal migrants in Shanghai, China. Health Policy Plan 17, 47–55. Zisapel, N., Nir, T., 2003. Determination of the minimal clinically significant difference on a patient visual analogue sleep quality scale. European Sleep Research Society 12, 291–298.