Health Insurance In China: After Declining In The 1990s, Coverage Rates Rebounded To Near-Universal Levels By 2011

Health insurance coverage in China has experienced a series of dramatic transitions in the past few decades. In the late 1970s health insurance was available to more than 90 percent of the Chinese population. ¹ That share declined to around only 5 percent in the late 1990s, ^2,3 followed by a rapid increase to near-universal coverage again in 2010. ⁴ Besides the dramatic changes in coverage rates, the types of available health insurance have also undergone dramatic changes in parallel with economic reform and reconstruction in China.

Although China returned to near-universal coverage in a relatively short time, it is not clear whether the current system can deliver equitable access to care across different types of coverage. Monitoring the health insurance coverage transition and providing a comprehensive description of coverage patterns across socioeconomic groups would be of great importance for health economics and health policy in China. Only a few studies have evaluated the effects of health insurance on health management and disease control. In a comparison of insured and uninsured inpatients with schizophrenia in China, Michael Phillips and coauthors found that patients with schizophrenia who had health insurance were 2.8 times more likely to receive inpatient treatment, compared to similar patients without insurance. ⁵ A more recent study reported that people with health insurance had better awareness, treatment, and control of hypertension, compared to people with no insurance. ⁶ However, the study did not examine differences between insurance types. It is also not clear whether health insurance coverage was associated with diabetes management or health care utilization.

In this study we analyzed the health insurance transitions in China during the period 1991–2011—a time of significant economic change in the country—and we provide an update and extension of earlier work that examined the rate of insurance coverage during the period 1989–2000. ^1,7 We also evaluated changes in health insurance coverage rates by different types of insurance. To provide additional context, we compared the prevalence, awareness, treatment, and control of hypertension and diabetes among people with different types of insurance. Self-reported health conditions and the use of health care were also compared by type of insurance.

Study Data And Methods

Study Population

The China Health and Nutrition Survey ⁸ is an ongoing, prospective, household-based study that includes multiple age groups and cohorts across nine provinces. There were nine waves of the survey in the period 1989–2011, collectively including 26,000 individuals in 4,400 households. The survey was first conducted in 1989 and was repeated in 1991, 1993, 1997, 2000, 2004, 2006, 2009, and 2011. A stratified probability sampling method was applied for the participants’ enrollment, which has been described in detail elsewhere. ⁸ See online Appendix Exhibit A1 for a brief description of the sampling method. ⁹

We excluded data from the 1989 wave because in that year only adults ages 20–45 were surveyed. We also excluded data for Beijing, Shanghai, and Chongqing (three megacities, or independent cities that are not part of any province) from the comparison of time trends, as those data were available only for 2011 (when the cities were added to the survey). Additionally, we excluded participants who were pregnant or younger than age 20 at the time of the survey. For the time trend analysis, eight waves of the survey with a total of 78,519 observations were included. Because the data for plasma glucose for the diagnosis of diabetes were available only for the 2009 wave, we used data from 2009 to estimate the associations of health insurance with the use of health care and with prevalence, awareness, treatment, and control of diabetes and hypertension.

The study was approved by the Institutional Review Boards of the University of North Carolina at Chapel Hill; National Institute of Nutrition and Food Safety, Chinese Center for Disease Control and Prevention; and the China-Japan Friendship Hospital, Ministry of Health of China.

Data On Health Insurance

Detailed data on health insurance use and the types of services covered by insurance were collected from 1991 to 2011. These data were updated annually according to the potential available categories of health insurance in China for a given year. The Chinese government initiated several types of insurance to promote insurance coverage after 1991. ¹ For the current analysis, seven types of insurance were examined: public health insurance, which is free; urban employee basic medical insurance (UEBMI); collective insurance, including the rural newly cooperative basic health insurance launched in 2003; unified planning health insurance, which was not available after 2004; commercial health insurance, which was not available before 2004; urban resident basic medical insurance (URBMI), which was not available before 2007; and all other forms of health insurance. The China Health and Nutrition Survey also included a category for “dependent health insurance for women/children.” Few respondents reported having this insurance, so we combined it with “all other forms of health insurance.” People who responded “have health insurance” but did not indicate which type of insurance they had were also classified as having “all other forms of health insurance.”

Identifying Patients With Diabetes Or Hypertension

In the 2009 wave of the China Health and Nutrition Survey, fasting blood samples were collected in neighborhood clinics and respondents’ homes. We categorized respondents as having diabetes if they had a fasting blood glucose measurement of at least 7.0 mmol/L or had ever been diagnosed as having diabetes by a physician. ¹⁰ Systolic blood pressure was measured on the right arm of respondents in a seated position. ¹¹ We categorized respondents as having hypertension if they had a mean systolic blood pressure of at least 140 mmHg, had a mean diastolic blood pressure of at least 90 mmHg, or reported taking antihypertensive medication. ¹² Detailed information is in Appendix Exhibit A1. ⁹

Assessing Patients’ Socioeconomic Status And Health Care Use

Trained interviewers used questionnaires to collect information on sociodemographic and household variables, self-reported illness or injury in the past four weeks (including chronic conditions and acute illnesses), use of formal health care (that is, care provided by a hospital or health care center with certifications) in the past four weeks, use of preventive care in the past four weeks, and any visit in the past year to a folk doctor. Also known as “barefoot doctors,” folk doctors are typically rural residents with no more than middle or high school education who have received minimal basic medical and paramedical training.

Statistical Analysis

Below we present the coverage rate for each type of health insurance by China Health and Nutrition Survey wave. We also present the coverage rate for each type of insurance according to individual-level characteristics that were investigated in a previous study ¹ and that we believed a priori to be important contributors to the use of health care.

Multilevel logistic regression models were used to examine the effect of insurance coverage on the likelihood of health management and use of health care, after adjustment for the individual characteristics described above. We categorized people as being aware of a condition (hypertension) or disease (diabetes) if they reported having been diagnosed with it by a physician. People who were aware of a condition or disease and reported receiving medication for it were categorized as being treated for it. People who were being treated for hypertension or diabetes and whose blood pressure or fasting blood glucose was controlled were categorized as having their condition or disease under control. We defined hypertension as being controlled if after treatment the average systolic blood pressure was below 140 mmHg and diastolic blood pressure was below 90 mmHg among individuals ages fifty-nine and younger, or the average systolic blood pressure was below 150 mmHg and diastolic blood pressure was below 90 mmHg among those ages sixty and older. ¹² We defined diabetes as being controlled if after treatment the fasting blood glucose level was less than 7.0 mmol/L.

We also conducted a sensitivity analysis by applying the propensity score stratification method. The propensity score was the estimated probability of health insurance selection conditional on observed covariates. ^13,14 The probability was calculated with health insurance coverage as the dependent variable and all of the covariates listed above as independent variables. Quintiles of the propensity scores and health insurance coverage were included in the analysis. The effect of health insurance coverage on outcomes was estimated within each stratum of the propensity score. Stratum-specific effects were then pooled to obtain an overall effect. ^13,14

Data were analyzed using SAS, version 9.2. We considered a two-tailed p value of $<0.05$ to be significant.

Limitations

Several limitations need to be considered in interpreting our results. First, the cross-sectional design of the association analysis limited our ability to infer causality in the observed associations. Second, the pool of respondents to the China Health and Nutrition Survey was not designed to be representative of China but to be a random sample that would include a range of demographic conditions, levels of economic development, and health indicators. Data on the randomly selected households in nine provinces provide a broad-based indication of the trends in China. A third limitation is that survey respondents with hypertension might have better awareness, treatment, and control of hypertension than the general Chinese population with hypertension because some of the respondents had their blood pressure measured repeatedly across different waves of the survey. This could explain why we did not observe a significant association between health insurance coverage and awareness of hypertension.

No such potential limitation applies to diabetes awareness because blood samples for fasting glucose measurement were collected only once. The awareness and treatment rates among patients with diabetes and the control rate among treated patients with diabetes in our study were 38.3 percent, 35.6 percent, and 14.5 percent, respectively (data not shown), which are comparable to 23.66 percent, 20.33 percent, and 8.28 percent, respectively, among Chinese respondents in the InterASIA study, ¹⁵ and 45.8 percent, 42.5 percent, and 20.9 percent, respectively, among Chinese people as reported in an article on diabetes in China during 1979–2012. ¹⁶ However, the size of our sample of people with diabetes was too small for the propensity score stratification analysis of diabetes management and control. It is possible that the lack of association between insurance coverage and diabetes awareness and management might also be due to a lack of statistical power, which warrants future studies with larger sample sizes.

Study Results

The health insurance coverage rate in our sample fell slightly from 32.3 percent in 1991 to 21.9 percent in 2000 before rebounding to 49.7 percent in 2006, after which it rapidly increased to 90.1 percent in 2009 and 94.7 percent in 2011 (see Appendix Exhibit A2). ⁹ In 1991 insurance coverage was most widespread among people who lived in urban neighborhoods (80.1 percent); had more than a high school education (87.8 percent); were professors, technicians, managers, or government officers (81.2 percent); and worked in state agencies or organizations (75.9 percent). In contrast, insurance coverage rates were lowest among people who lived in rural villages (9.5 percent); had never gone to school (18.3 percent); and were farmers, fishermen, or hunters (5.5 percent). These differences disappeared in 2011, when the insurance coverage rates were above 90 percent for most of the subgroups (Appendix Exhibit A2). ⁹

When we examined the trends over time in coverage by different types of health insurance, we found that the rate of public (free) health insurance decreased from 16.8 percent in 1991 to 3.5 percent in 2011 ( Exhibit 1 ). The decline was especially apparent among people in urban neighborhoods, where it decreased from 45.9 percent in 1991 to 6.8 percent in 2011 (Appendix Exhibit A3). ⁹ Urban employee basic medical insurance coverage rates decreased from 10.1 percent in 1991 to 4.9 percent in 2000 and 2004 and then increased to 19.6 percent in 2011 ( Exhibit 1 ). A dramatic increase in collective health insurance coverage rates began in 2006—mainly among people in rural villages ( Exhibit 1 and Appendix Exhibit A3). ⁹

SOURCE Authors’ analysis of data for 1991–2011 from the China Health and Nutrition Survey. NOTES Collective insurance includes the rural newly cooperative basic health insurance. Unified planning health insurance was not available after 2004. Commercial health insurance was not available before 2004. Urban resident-based basic medical insurance (“urban resident basic”) was not available before 2007, and data on it were not collected before 2009. People with “other insurance” include the few members of the study population who reported having “dependent health insurance for women/children” in the survey, those who responded “have health insurance” but did not indicate the type of insurance, and those who reported having any other form of health insurance. “Urban employee basic” is the urban employee–based basic medical scheme.

We did not find significant differences between people with and without health insurance in terms of the prevalence or awareness of diabetes ( Exhibit 2 ) or hypertension ( Exhibit 3 ), or of being sick or injured in the past four weeks ( Exhibit 4 ). We found that among patients who were aware of their disease or condition, those with insurance had a higher odds ratio of treatment for diabetes (age- and sex-adjusted odds ratio: 3.86; 95% confidence interval: 1.24, 12.0) and a higher odds ratio of treatment for hypertension (age- and sex-adjusted OR: 1.55; 95% CI: 0.96, 2.50), compared to patients without insurance. After we further adjusted for socioeconomic and lifestyle factors, those with insurance had a 6.77 higher odds ratio of treatment for diabetes (95% CI: 1.44, 31.9; Exhibit 2 ) and a 1.72 higher odds ratio of treatment for hypertension (95% CI: 1.04, 2.84; Exhibit 3 ), compared to patients without insurance. People who were sick or injured were more likely to seek formal health care when they had public insurance (OR: 2.76; 95% CI: 1.39, 5.49), urban employee basic medical insurance (OR: 1.84; 95% CI: 1.18, 2.87), or urban resident basic medical insurance (OR: 1.87; 95% CI: 1.15, 3.04), compared to people with no insurance ( Exhibit 4 ). Insured people were more likely to seek preventive care than uninsured people (OR: 2.41; 95% CI: 1.51, 3.83), a difference that was significant for all types of coverage.

The results comparing covariate imbalance before and after propensity-score stratification are presented in Appendix Exhibits A4 and A5. ⁹ We found that almost all covariates that were significantly different between uninsured and insured populations were no longer significant after being adjusted for propensity score quintiles. After we applied the propensity score stratification, we found that the multivariate-adjusted odds ratio between uninsured and insured patients was no longer significant (OR: 1.46; 95% CI: 0.91, 2.36) for receiving antihypertensive medications when patients were aware of their condition (data not shown). We also found that insured people were 2.30 times more likely to seek preventive care than those without insurance (OR: 2.21; 95% CI: 1.40, 3.51).

Discussion

Using data from the unique large-scale longitudinal study of the China Health and Nutrition Survey, we observed that health insurance coverage rates dramatically increased in both urban and rural areas in China during the period 1991–2011. When we examined the cross-sectional data for 2009, we did not observe a significant impact of health insurance coverage on health conditions, although we did observe an association between coverage and receiving medical treatment when patients were aware of their disease or condition. People with insurance were more likely to seek preventive care, compared to those without insurance.

Possible reasons for the dramatic expansion of health insurance coverage in China include the introduction in 1998 and promotion of urban employee basic medical insurance, ^17,18 the introduction in 2003 and promotion of the New Cooperative Medical Scheme for rural residents, ^19,20 and the introduction in 2007 and promotion of the urban resident basic medical insurance. ²¹ As a result of the introduction of the New Cooperative Medical Scheme, health insurance coverage rates in rural villages jumped from less than 20 percent in the period 1991–2004 to 50.7 percent in 2006, 95.2 percent in 2009, and 97.8 percent in 2011 (Appendix Exhibit A3). ⁹ Our data indicated that this scheme was essentially the only health insurance available to residents of rural villages.

With rapid economic development and improvement of living conditions, cardiovascular diseases have replaced acute diseases as the leading cause of premature mortality in China. ²² Although cardiovascular diseases are among the most common and costly of all health conditions, they are also among the most preventable. Our study indicated that 3.1 million cardiovascular events in 2011 of China could be prevented if systolic blood pressure could be managed to the theoretical minimum level of 115 mmHg, ²³ while 0.9 million cardiovascular events could be prevented if blood glucose levels could be controlled to the optimal level of 4.9 mmol/L.

Universal health coverage may make it possible to greatly improve the management and control of cardiovascular diseases or related health conditions. We did not observe significant associations between universal health coverage in China and the prevalence of diabetes or hypertension, although we observed higher rates of hypertension and diabetes treatment among patients who were aware of their condition or disease and who had insurance, compared to those who were uninsured. The prevalence of hypertension in China (33.5 percent in 2010) ²⁴ was comparable to that in the United States (30.4 percent in the period 2003–10). ²⁵ However, rates of awareness of the condition were much lower in China (24 percent in 2002) ²⁶ than in the United States (61 percent in 2003–10). ²⁵ Among those with hypertension, an estimated 46 percent of US adults had their condition controlled in 2003–10, ²⁵ compared to only 25 percent in China in 2002. ²⁶ Lack of awareness of diabetes in China is also a major public health concern. In 2010, among patients with diabetes, 30.1 percent were aware of their disease, only 25.8 percent of those who were aware received treatment for diabetes, and only 39.7 percent of those who received treatment had adequate glycemic control. ²⁷ Both hypertension and diabetes require long periods of outpatient treatment, and the associated costs would be a huge financial burden for people without insurance.

The observed association between health insurance coverage and a higher rate of treatment when people were aware of their disease or condition might be due to various factors. For example, people with health insurance also had higher incomes, more education, better transportation, and better health care facilities, compared to the uninsured. Although we attempted to limit confounding by adjusting our estimates for certain potential risk factors, residual confounding from poorly measured or unmeasured factors could not be completely ruled out. We also applied the propensity score stratification method to adjust for differences between uninsured and insured groups, which could have eliminated 90 percent of the bias due to the measured confounders. ¹³ Although our multivariable analysis controlled for a wide range of risk factors for diabetes and hypertension, unmeasured confounding may still exist. ²⁸ However, only a very strong unmeasured risk factor for diabetes and hypertension together with a very large prevalence imbalance between uninsured and insured groups could explain our findings.

Furthermore, even though health insurance coverage did improve treatment rates, coverage did not guarantee improved health status, especially in poor rural areas as reported in studies of the period 2009–12. ^29,30 The quality of health care was still quite poor in undeveloped rural areas in China in that period. ^29,30 In some village clinics, “certified” doctors—that is, doctors with board certification who are qualified to work at any level of the health care system—were not available or not yet prepared to provide primary care. ^29,30 This might be one reason why we did not observe significant associations between health insurance coverage and prevalence of hypertension and diabetes. The lack of association between insurance coverage and prevalence, awareness, treatment, and control of diabetes that we observed is consistent with previous observations. ^31,32 An earlier study did not find a significant difference between insured and uninsured patients in terms of the quality and use of diabetes care in China. ³¹ In another study of inpatients with diabetes, medical insurance did not have any impact on hospitalization costs of antihyperglycemic treatment or other costs resulting from the treatment of complications. ³²

We found some differences in health management across different insurance types. For example, compared to the uninsured, urban employee basic health insurance was associated with higher treatment and control rates of hypertension, which might be due to the insurance having patients pay a smaller proportion of direct medical costs through out-of-pocket spending. ³³ Based on data from the National Health Service survey, in 2013 the reimbursement rates for hospital admission were 68.8 percent for patients with urban employee basic health insurance, 53.6 percent for those with urban resident basic health insurance, and 50.1 percent for those with rural newly cooperative basic health insurance. ³⁴ Higher reimbursement rates and lower out-of-pocket spending on health care might improve future disease control, because medical services were particularly unaffordable to Chinese rural patients. ³⁵ However, a recent study indicated that the rapid increase in medical expenditures for patients with hypertension and diabetes had offset the financial protection for patients provided by the expanded rural health insurance of the New Cooperative Medical Scheme. ³⁶

Our study has a number of strengths. First, it was the most comprehensive study of different types of health insurance coverage in China, covering a larger group of people and a longer period of time than other studies, and it provides a timely description of insurance coverage rates before, during, and after the introduction, promotion, and popularization of several new types of health insurance.

Second, in this study the association of health insurance coverage with the prevalence, awareness, treatment, and control of diabetes was examined for the first time, and we identified a new potential strategy for improving hypertension and diabetes management in China—medical insurance coverage. The strengths of the China Health and Nutrition Survey include its originality and high response rate. The survey’s overall response rates for people who participated in previous waves and the most recent one were around 88 percent at the individual level and 90 percent at the household level. ³⁷

Our findings imply that although health insurance coverage is almost universal in China, its potential benefit for disease prevention and management had not been realized during the time period covered by this study. More efforts may be needed to improve the effective control of diabetes and awareness of disease status.

Conclusion

We observed dramatically increased rates of health insurance coverage in both urban and rural areas in China in the period 1991–2011. People with health insurance were more likely to receive treatment for diabetes or hypertension, when aware of their disease status, and were more likely to seek preventive care, compared to those without health insurance.

ACKNOWLEDGMENTS

This study was supported by the Swiss Re Foundation. The authors thank the National Institute of Nutrition and Food Safety, Chinese Center for Disease Control and Prevention, Carolina Population Center (Grant No. 5 R24 HD050924), the University of North Carolina at Chapel Hill, the National Institutes of Health (NIH) (Grant Nos. R01-HD30880, DK056350, R24 HD050924, and R01-HD38700), and the John E. Fogarty International Center of the NIH for financial support for the China Health and Nutrition Survey (CHNS) data collection and analysis files from 1989 to 2011 surveys, and the China-Japan Friendship Hospital, Ministry of Health, for support for CHNS 2009.

NOTES

• 1 Akin JS , Dow WH , Lance PM . Did the distribution of health insurance in China continue to grow less equitable in the nineties? Results from a longitudinal survey . Soc Sci Med . 2004 ; 58 ( 2 ): 293 – 304 . Crossref, Medline, Google Scholar
• 2 Liu X , Cao H . China’s Cooperative Medical System: its historical transformations and the trend of development . J Public Health Policy . 1992 ; 13 ( 4 ): 501 – 11 . Crossref, Medline, Google Scholar
• 3 Lei X , Lin W . The New Cooperative Medical Scheme in rural China: does more coverage mean more service and better health? Health Econ . 2009 ; 18 ( Suppl 2 ): S25 – 46 . Crossref, Medline, Google Scholar
• 4 Zhang D , Shi L , Tian F , Zhang L . Care utilization with China’s New Rural Cooperative Medical Scheme: updated evidence from the China Health and Retirement Longitudinal Study 2011–2012 . Int J Behav Med . 2016 ; 23 ( 6 ): 655 – 63 . Crossref, Medline, Google Scholar
• 5 Phillips MR , Lu SH , Wang RW . Economic reforms and the acute inpatient care of patients with schizophrenia: the Chinese experience . Am J Psychiatry . 1997 ; 154 ( 9 ): 1228 – 34 . Crossref, Medline, Google Scholar
• 6 Guo J , Zhu YC , Chen YP , Hu Y , Tang XW , Zhang B . The dynamics of hypertension prevalence, awareness, treatment, control and associated factors in Chinese adults: results from CHNS 1991–2011 . J Hypertens . 2015 ; 33 ( 8 ): 1688 – 96 . Crossref, Medline, Google Scholar
• 7 Du J . Economic reforms and health insurance in China . Soc Sci Med . 2009 ; 69 ( 3 ): 387 – 95 . Crossref, Medline, Google Scholar
• 8 Zhang B , Zhai FY , Du SF , Popkin BM . The China Health and Nutrition Survey, 1989–2011 . Obes Rev . 2014 ; 15 ( Suppl 1 ): 2 – 7 . Crossref, Medline, Google Scholar
• 9 To access the Appendix, click on the Appendix link in the box to the right of the article online.
• 10 World Health Organization . Definition and diagnosis of diabetes mellitus and intermediate hyperglycaemia [Internet]. Geneva : WHO ; c 2006 [cited 2017 Mar 10 ]. Available from: http://www.who.int/diabetes/publications/Definition%20and%20diagnosis%20of%20diabetes_new.pdf Google Scholar
• 11 Du S , Batis C , Wang H , Zhang B , Zhang J , Popkin BM . Understanding the patterns and trends of sodium intake, potassium intake, and sodium to potassium ratio and their effect on hypertension in China . Am J Clin Nutr . 2014 ; 99 ( 2 ): 334 – 43 . Crossref, Medline, Google Scholar
• 12 James PA , Oparil S , Carter BL , Cushman WC , Dennison-Himmelfarb C , Handler J , et al. 2014 evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8) . JAMA . 2014 ; 311 ( 5 ): 507 – 20 . Crossref, Medline, Google Scholar
• 13 Rosenbaum PR , Rubin DB . The central role of the propensity score in observational studies for causal effects . Biometrika . 1983 ; 70 ( 1 ): 41 – 55 . Crossref, Google Scholar
• 14 Austin PC . The performance of different propensity-score methods for estimating differences in proportions (risk differences or absolute risk reductions) in observational studies . Stat Med . 2010 ; 29 ( 20 ): 2137 – 48 . Crossref, Medline, Google Scholar
• 15 Hu D , Fu P , Xie J , Chen CS , Yu D , Whelton PK , et al. Increasing prevalence and low awareness, treatment and control of diabetes mellitus among Chinese adults: the InterASIA study . Diabetes Res Clin Pract . 2008 ; 81 ( 2 ): 250 – 7 . Crossref, Medline, Google Scholar
• 16 Li M , Su L , Liang B , Tan J , Chen Q , Long J , et al. Trends in prevalence, awareness, treatment, and control of diabetes mellitus in Mainland China from 1979 to 2012 . Int J Endocrinol . 2013 ; 2013 : 753150 . Crossref, Medline, Google Scholar
• 17 Liu GG , Zhao Z , Cai R , Yamada T , Yamada T . Equity in health care access to: assessing the urban health insurance reform in China . Soc Sci Med . 2002 ; 55 ( 10 ): 1779 – 94 . Crossref, Medline, Google Scholar
• 18 Xu L , Wang Y , Collins CD , Tang S . Urban health insurance reform and coverage in China using data from National Health Services Surveys in 1998 and 2003 . BMC Health Serv Res . 2007 ; 7 : 37 . Crossref, Medline, Google Scholar
• 19 Ministry of Health, Ministry of Finance, Ministry of Agriculture . [ The advices for establishment of New Cooperative Medical Scheme ]. Beijing : Ministry of Health ; 2003 . Chinese. Google Scholar
• 20 Liu J , Xu L , Cao X , Wang X , Sun H , Tang C , et al. Analysis of satisfaction about New Cooperative Medical Scheme and its influencing factors in Weihai, China . Health Policy . 2008 ; 86 ( 2–3 ): 239 – 44 . Crossref, Medline, Google Scholar
• 21 Lin W , Liu GG , Chen G . The Urban Resident Basic Medical Insurance: a landmark reform towards universal coverage in China . Health Econ . 2009 ; 18 ( Suppl 2 ): S83 – 96 . Crossref, Medline, Google Scholar
• 22 Zhou M , Wang H , Zhu J , Chen W , Wang L , Liu S , et al. Cause-specific mortality for 240 causes in China during 1990–2013: a systematic subnational analysis for the Global Burden of Disease Study 2013 . Lancet . 2016 ; 387 ( 10015 ): 251 – 72 . Crossref, Medline, Google Scholar
• 23 Li Y , Wang DD , Ley SH , Howard AG , He Y , Lu Y , et al. Potential impact of time trend of life-style factors on cardiovascular disease burden in China . J Am Coll Cardiol . 2016 ; 68 ( 8 ): 818 – 33 . Crossref, Medline, Google Scholar
• 24 Li YC , Wang LM , Jiang Y , Li XY , Zhang M , Hu N . [ Prevalence of hypertension among Chinese adults in 2010 ]. Zhonghua Yu Fang Yi Xue Za Zhi . 2012 ; 46 ( 5 ): 409 – 13 . Chinese. Medline, Google Scholar
• 25 Centers for Disease Control and Prevention . Vital signs: awareness and treatment of uncontrolled hypertension among adults–United States, 2003–2010 . MMWR Morb Mortal Wkly Rep . 2012 ; 61 : 703 – 9 . Medline, Google Scholar
• 26 Wu Y , Huxley R , Li L , Anna V , Xie G , Yao C , et al. Prevalence, awareness, treatment, and control of hypertension in China: data from the China National Nutrition and Health Survey 2002 . Circulation . 2008 ; 118 ( 25 ): 2679 – 86 . Crossref, Medline, Google Scholar
• 27 Xu Y , Wang L , He J , Bi Y , Li M , Wang T , et al. Prevalence and control of diabetes in Chinese adults . JAMA . 2013 ; 310 ( 9 ): 948 – 59 . Crossref, Medline, Google Scholar
• 28 Schneeweiss S . Sensitivity analysis and external adjustment for unmeasured confounders in epidemiologic database studies of therapeutics . Pharmacoepidemiol Drug Saf . 2006 ; 15 ( 5 ): 291 – 303 . Crossref, Medline, Google Scholar
• 29 Sylvia S , Shi Y , Xue H , Tian X , Wang H , Liu Q , et al. Survey using incognito standardized patients shows poor quality care in China’s rural clinics . Health Policy Plan . 2015 ; 30 ( 3 ): 322 – 33 . Crossref, Medline, Google Scholar
• 30 Ke L , Zhang Y , Wang X , Li S , Yang W , Tong N . Assessment of diabetes care and the healthcare system in economically and transport underdeveloped rural mountain areas of western China: a cross-sectional survey . J Diabetes . 2016 Jun 1 . [Epub ahead of print]. Medline, Google Scholar
• 31 Wang Z . Diabetes care in China: impacts of traditional Chinese medicine (TCM) and insurance on quality and utilization [Internet]. Santa Monica (CA) : RAND Corporation ; 2011 [cited 2017 Mar 10 ]. Available for download from: http://www.rand.org/pubs/rgs_dissertations/RGSD286.html Google Scholar
• 32 He M , Ma J , Wang D , Yu X . Costs for hospitalized patients with diabetes mellitus in China . Asia Pac J Public Health . 2015 ; 27 ( 2 ): NP925 – 35 . Crossref, Medline, Google Scholar
• 33 Wang W , Fu C , Zhuo H , Luo J , Xu B . Factors affecting costs and utilization of type 2 diabetes healthcare: a cross-sectional survey among 15 hospitals in urban China . BMC Health Serv Res . 2010 ; 10 : 244 . Crossref, Medline, Google Scholar
• 34 Center for Health Statistics and Information . [ An analysis report of fifth National Health Service Survey in China, 2013 ]. Beijing : Peking Union Medical College Press ; 2015 . Chinese. Google Scholar
• 35 Hu S , Tang S , Liu Y , Zhao Y , Escobar ML , de Ferranti D . Reform of how health care is paid for in China: challenges and opportunities . Lancet . 2008 ; 372 ( 9652 ): 1846 – 53 . Crossref, Medline, Google Scholar
• 36 Liu X , Sun X , Zhao Y , Meng Q . Financial protection of rural health insurance for patients with hypertension and diabetes: repeated cross-sectional surveys in rural China . BMC Health Serv Res . 2016 ; 16 : 481 . Crossref, Medline, Google Scholar
• 37 Popkin BM , Du S , Zhai F , Zhang B . Cohort profile: the China Health and Nutrition Survey—monitoring and understanding socio-economic and health change in China, 1989–2011 . Int J Epidemiol . 2010 ; 39 ( 6 ): 1435 – 40 . Crossref, Medline, Google Scholar