With a quasi-experimental design, this study aims to assess whether the Zero-markup Policy for Essential Drugs (ZPED) reduces the medical expense for patients at county hospitals, the major healthcare provider in rural China.
Data from Ningshan county hospital and Zhenping county hospital, China, include 2014 outpatient records and 9239 inpatient records. Quantitative methods are employed to evaluate ZPED. Both hospital-data difference-in-differences and individual-data regressions are applied to analyze the data from inpatient and outpatient departments.
In absolute terms, the total expense per visit reduced by 19.02 CNY (3.12 USD) for outpatient services and 399.6 CNY (65.60 USD) for inpatient services. In relative terms, the expense per visit was reduced by 11% for both outpatient and inpatient services. Due to the reduction of inpatient expense, the estimated reduction of outpatient visits is 2% among the general population and 3.39% among users of outpatient services. The drug expense per visit dropped by 27.20 CNY (4.47 USD) for outpatient services and 278.7 CNY (45.75 USD) for inpatient services. The proportion of drug expense out of total expense per visit dropped by 11.73 percentage points in outpatient visits and by 3.92 percentage points in inpatient visits.
Implementation of ZPED is a benefit for patients in both absolute and relative terms. The absolute monetary reduction of the per-visit inpatient expense is 20 times of that in outpatient care. According to cross-price elasticity, the substitution between inpatient and outpatient due to the change in inpatient price is small. Furthermore, given that the relative reductions are the same for outpatient and inpatient visits, according to relative thinking theory, the incentive to utilize outpatient or inpatient care attributed to ZPED is equivalent, regardless of the 20-times price difference in absolute terms.
Citation: Zhou Z, Su Y, Campbell B, Zhou Z, Gao J, Yu Q, et al. (2015) The Financial Impact of the ‘Zero-Markup Policy for Essential Drugs’ on Patients in County Hospitals in Western Rural China. PLoS ONE 10(3): e0121630. https://doi.org/10.1371/journal.pone.0121630
Academic Editor: Harry Zhang, Old Dominion University, UNITED STATES
Received: June 11, 2014; Accepted: February 11, 2015; Published: March 19, 2015
Copyright: © 2015 Zhou et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: The data from Zhenping are open to the public and the data from Ningshan are available upon request. We have received permission from the IRB of Xi’an Jiaotong University Health Science Center to release the anonymous individual medical records. Zhenping County Hospital has approved of public access to the data. To ensure compliance with local regulations at Ningshan, the data from this county can only be accessed in Ningshan with the approval and supervision of Ningshan County Hospital. For researchers interested in acquiring the raw data, the data of Zhenping can be downloaded from the Harvard Dataverse Network (DOI: 28588) and the data from Ningshan can be obtained by contacting Ningshan County Hospital (contact name: Jiuhao Chen, email: firstname.lastname@example.org).
Funding: This study was funded by National Natural Science Fund of China (Serial number: 71203177), Shaanxi Social Science Fund (Serial number: 12Q036) and China Medical Board (CMB). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Medicines are frequently selected, evaluated, used, and spoken about in low-resource settings in ways that are strongly context-dependent . Medicines are not only symbols of healing and hope but are also valuable and sometimes unaffordable commodities within a market dominated by global pharmaceutical companies [1–3]. People in low-resource settings make triage decisions about which medicines to seek, not necessarily on the basis of ‘best practices,’ but rather on cost, as well as aesthetics and palatability . While aesthetic and palatability concerns are more locally engrained, the issue of cost can be addressed effectively by policy making . For instance, decreasing user fees for services and drugs has been shown to increase uptake of care in some low and middle-income countries . The extent to which medicines become available, accessible and affordable to people is deeply impacted by a country’s policy on essential drugs, which ensures the basic access of drugs at low cost . Over 150 countries, most of which are developing countries, have established a national essential medicines list . While it is believed that lowering drug costs increases access, reviews of pharmaceutical policies in developing countries have found virtually no evidence of this . China, which is undergoing rapid health reform, provides a ripe context to investigate the impact of lowering essential drug cost on access to drugs and health-seeking behavior.
China has undergone transformational health reform since 2009. At the heart of this reform have been the reconstruction of a national primary healthcare system and the enhancement of insurance programs targeting low-income citizens . The Chinese Central Government proposed the development of a national essential medicines system, and made it a top priority in health reform . This effort was bolstered by the National Essential Medicines Policy (NEMP), which aims to increase the availability of cost-effective medicines . These efforts represent a major attempt to “broaden” access to essential drugs in the country and ultimately improve health equity 
A recent study examining health equity in rural China shows that utilization of healthcare—both inpatient and outpatient—had been largely pro-rich between 1993 and 2008 . This inequity existed partly because drug costs from both inpatient and outpatient care pose a significant burden to low-income people. There is increasing empirical evidence showing that drug costs are primary contributors to medical impoverishment [15–17]. Patients with chronic conditions frequently visit hospitals and can incur high medication costs that are not covered under the catastrophic coverage model of the New Cooperative Medical Schemes (NCMS). Further, Yip and Hsiao et al. found that 11.6% of the rural poor households became impoverished due to outpatient expenses related to chronic disease . To address the financial burden experienced by patients in rural settings, China hopes to match its great success in “broadening” health coverage with a major “deepening” of health care . A critical part of deepening care is making essential drugs more affordable for patients .
In an effort to make essential medicines affordable to more people, the Chinese government passed the Zero-markup Policy for Essential Drugs (ZPED). The goals of the policy are to contain the costs of medicines and reduce the financial burden to the public. Since 2009, ZPED has been implemented in primary healthcare institutions, including township health institutions and villages in most provinces of China. Previous research investigating the effect of ZPED at township and village levels has been very promising. For example, Sun et al. showed that per-visit outpatient and inpatient expenses were reduced in six township health centers in Anhui Province . Research by Wang et al. investigated township health institutions in eastern, central, and western China and echoed the results by Sun et al. [19, 20]. Chen et al. conducted another study in Anhui Province with a larger sample size and found a reduction of cost per prescription in 88 primary health institutions . Further, Yang et al. found a reduction in cost per prescription in Hubei province . Song et al. and Shi et al. reached the same conclusions in Shandong province and Beijing, respectively [23, 24]. Lastly, a 2014 study by Song et al. found that the new policy led to price reductions in essential medicines in 149 primary healthcare centers across four Chinese provinces .
With promising results at the township and village levels, it is critical to now look at public hospitals at the county level and higher levels, where ZPED has been in a pilot phase for most of the provinces to date.
County public hospitals, which undertake the task of curing patients with common medical conditions, are the main provider of healthcare in rural China. The fourth National Health Service Survey shows that more than 48% of patients were hospitalized in county hospitals . Moreover, statistics from the Ministry of Health in China show that the medical expense in county hospitals consists of 40% of the total medical expenses among all medical institutions . Therefore, in order to control the rapid growth of medical expenses in rural China, it is strategic to hurdle drug expenses in county hospitals. Many studies [19, 20, 28, 29] suggest that ZPED could eliminate the reliance on drug revenue to compensate for the deficit in medical revenue in primary healthcare institutions.
In deepening health system reform in China, it is critical to implement ZPED in county hospitals. At present, ZPED has been piloted for some county hospitals in several provinces. For example, in September 2011, Fuyang County and another 28 counties in Zhejiang implemented ZPED. In June 2011, Shaanxi province made the policy, “Plan of Centralized Purchase of Medicine in Public Hospitals,” and piloted ZPED in county hospitals. However, evidence from a rigorous program evaluation of ZPED at the county hospital level is lacking. While some studies have attempted to assess ZPED at the county level about the effect on the top five most frequent conditions of inpatient services  as well as on county hospital revenue and government subsidy levels , very little has been done to assess the policy’s financial impact on general patients. To our knowledge, Shen and Wang are the researchers who have investigated ZPED’s financial impact. Shen (2013) conducted an evaluation in Zhejiang province  and Wang et al. did research in Henan province . The results from these two studies are contradictory to each other. Shen (2013) showed reduction in per-visit expense in outpatient and inpatient services. However, Wang et al. showed an increased trend instead .
Beyond these inconsistent findings in evaluating ZPED at the county level, there are important limitations in the previous research: 1) nearly all the research focused on the community-level medical institutions and few studies evaluated ZPED in county-level hospitals; 2) most research designs only collected data to conduct pre-post comparisons, which lacks scientific rigor in evaluating the policy; 3) only descriptive statistics methods were used in most of the research and confounding variables like age, gender, marital status and severity of medical condition could not be controlled. In order to overcome these limitations in the previous research and to fill an important research gap, this study aims to assess the impact of ZPED on medical expenses, including outpatient and inpatient expenses, in county hospitals, and to explore how the cost of essential drugs impacts individual behavior and pharmaceutical uptake.
Ningshan county hospital in Ankang city, Shaanxi province, was selected to be the study object. According to the documents provided by Ningshan county hospital, ZPED was implemented since December 1st, 2010 in Ningshan county hospital. Among essential drugs, 44% of them were selected to be subject to the zero-markup policy. From December 1st, 2010 to November 11th, 2011, in Ningshan county hospital, the cost of the drugs purchased was 7.88 million CNY and the revenue of the drugs sold was 6.96 million, in which the cost and revenue of zero-markup drugs were 1.27 million and 1.1 million, respectively, accounting for 16.12% and 15.8% of the total cost and revenue. In 2011, patients benefited from 194 thousand CNY due to ZPED, according to hospital financial records. On average, the markup rate decreased by 2.46% in 2011.
In this quasi-experimental study, Ningshan county hospital was the experimental group, compared to Zhenping county hospital, the control group. Zhenping county hospital in Shaanxi province was selected as the control group because of the following reasons. First, comparability between Ningshan county and Zhenping county was indicated in administrative district, geography, income per capita, and health resources, etc. (Table 1). Second, comparability between Ningshan county hospital and Zhenping county hospital was indicated in the size of hospital, clinical departments, and medical personnel, etc. (Table 2). Third, ZPED had not yet been implemented in Zhenping county hospital in 2011, making that hospital and year an excellent control for the study.
Our research team members have no direct contact with human subjects. All the patients’ information collected from medical records was anonymized and de-identified prior to analysis. The study protocol was reviewed and approved by the Ethics Committee of Xi’an Jiaotong University Health Science Center.
Summary statistics of the hospitals’ basic information include the hospital certificate, department, staff, and number of beds of Ningshan county hospital and Zhenping county hospital. The operation reports include the two hospitals’ outpatient expense per visit, inpatient expense per visit, and drug expense per visit.
In random sampling, 1,007 outpatient records in 2010 and 1,007 outpatient records in 2011 were selected from Ningshan county hospital, which include the outpatients’ demographic information, outpatient expense and its composition. As the information system of outpatient service was not established in Zhenping county hospital, the outpatient records from Zhenping county hospital were not collected in this study. All the inpatient records from 2010 to 2011 were collected in Ningshan and Zhenping county hospitals. In Ningshan county hospital, 2,728 and 3,078 records were collected for years 2010 and 2011, respectively. In Zhenping county hospital, as the information system of inpatient service has not been integrated, the personal information, disease diagnosis, and the inpatient expense were in separate systems. The data from the three systems were merged by admission number. Finally, 1,711 and 1,722 records in 2010 and 2011, respectively, were used in data analysis, which is about 39% and 38% of the total inpatient records in 2010 and 2011, respectively. The medical records include the inpatient’s basic information, disease diagnosis, inpatient expense, and its composition.
Indicators which were employed to measure the financial impact of ZPED include: per-visit outpatient expense, drug expense per outpatient visit, the proportion of drug expense out of outpatient expense, per-visit inpatient expense, drug expense per inpatient visit, and the proportion of drug expense out of inpatient expense. We hypothesize that the per-visit outpatient and inpatient expense, drug expense and the proportion of drug expense will be reduced after implementing ZPED. A theoretical framework on how ZPED is proposed to influence the outcome indicators is shown in Fig. 1.
The method of difference—in—difference (DID)  was employed to analyze the effects of ZPED. Because both the hospital-level data and individual-level data are available, we conduct analysis in two approaches, accordingly. We denote the DID using hospital-level data as ‘Hospital-data DID’ and the one using individual-level data as ‘Individual-data DID.’
The implementation effect of ZPED is: (1) In Equation (1), for each first subscript, “1” represents treatment-group hospital and “0” represents control-group hospital; for each second subscript, “1” represents the year of 2011 and “0” represents the year of 2010. Accordingly, y1,0 and y0,0 are evaluation indicators for treatment group and control group, respectively, in 2010; y1,1 and y0,1 are evaluation indicators for the treatment group and control group, respectively, in 2011. The effect of zero-markup policy is captured byβ.
In order to control both the observable confounding factors, linear regression models were employed to analyze the effects of ZPED .
- Dependent variables: inpatient expense, drug expense in inpatient service, the proportion of drug expense in inpatient expense.
- Independent variables: policy (policy = 1 means treatment-group hospital,policy = 0 means control-group hospital), year (year = 1 means the year of 2011, year = 0 means the year of 2010).
- Control variables: age, gender, marital status, occupation, health insurance, severity of illness, surgery, treatment outcome.
The regression model is as follow: (2) In Equation (2), y is the dependent variable, policy and year are independent variables,policy Year is the interaction term, xm is the control variable, μ is a residual term. β3, the coefficient of the interaction term, is the effect of ZPED. If no control variables are included in the regression model, β3 is the estimate of DID: (3) The notations in this equation are the same as those in Equation (1). When the control variables are included in Equation (2), the estimate of β3 would become more sophisticated.
Testing the assumption of DID
DID is based on the assumption that, without the intervention, the trend of the outcome variables from treatment group and control group should be similar. Of note, Chen et al. did apply the DID model in their study in China; however, the assumptions of DID were left unexamined . In this study, the assumptions are satisfied with the following evidence.
First, Ningshan county and Zhenping county are similar in terms of the administrative district, geography, income per capita, and health resources. Moreover, there is similarity between Ningshan county hospital and Zhenping county hospital in terms of the certificate of hospital, clinical departments, and medical resources.
Second, in order to demonstrate that the trends between two hospitals are similar for all outcome variables, we collected data from 2007 to 2011. Fig. 2 and Fig. 3 show that, for both per-visit outpatient expense and drug expense per outpatient visit, from 2007 to 2010, the trends in Ningshan county hospital are similar to those in Zhenping county hospital, which is consistent with the hypothesis of DID.
In addition, Fig. 4 and Fig. 5 show that the trends of per-visit inpatient expense and drug expense per inpatient visit are similar in two county hospitals from 2007 to 2010 as well. Therefore, DID can be applied to evaluate the effects of ZPED on per-visit medical expense in the county hospital.
Results from hospital-data DID
From the results of second-order difference, as shown in Table 3, with the implementation of ZPED in Ningshan county hospital, the per-visit outpatient expense, the drug expense per outpatient visit, and the proportion of drug expense in outpatient expense dropped by 19.02 CNY (3.12 USD), 27.20 CNY (4.46 USD) and 11.73 percentage points, respectively. The per-visit inpatient expense, the drug expense per inpatient visit, and the proportion of drug expense in inpatient expense declined by 389.11 CNY (63.87 USD), 278.42 CNY (45.70 USD) and 3.9 percentage points, respectively.
Results from individual-data DID
As the information of individual outpatient records from Zhenping county hospital was not collected, only the data of individual outpatient records from Ningshan county hospital in the years of 2010 and 2011 is utilized to conduct regressions. Three regressions were employed to analyze the effects of ZPED, in which the dependent variables are per-visit outpatient expense, drug expense per outpatient visit, and the proportion of drug expense in outpatient expense, the independent variable is year, and the control variables are gender, age, surgery, etc. The descriptions of the dependent variables are shown in Table 4. From the year of 2010 to 2011, the per-visit outpatient expense, per-visit drug expense, and the proportion of drug expense in outpatient expense declined by 16.53 CNY (2.71 USD), 17.37 CNY (2.85 USD), and 5.01 percentage points, respectively. The descriptions of independent variables are shown in Table 5.
The results of regressions are shown in Table 6. After controlling for the variables of gender, age and surgery, the per-visit outpatient expense, drug expense per outpatient visit, and the proportion of drug expense in outpatient expense declined in a statistically significant manner, by 15.32 CNY (2.51 USD), 17.14 CNY (2.81 USD) and 7.35 percentage points, respectively.
Using the data of inpatient records from Ningshan and Zhenping county hospitals in years 2010 and 2011, DID with control variables in regressions was employed to analyze the effects of ZPED. From Table 7, in Ningshan County hospital, compared with 2010, the per-visit inpatient expense, drug expense per inpatient visit, and proportion of drug expense in inpatient expense declined in 2011. In Zhenping county hospital, the per-visit inpatient expense and the drug expense per inpatient visit increased from 2010 to 2011. The descriptions of independent variables are shown in Table 8.
Table 9 shows that the coefficients of the interaction of 2011 and Ningshan in the models with per-visit inpatient expense, drug expense per inpatient visit, and proportion of drug expense in inpatient expense as dependent variables are statistically significant. After controlling for the confounding factors, the per-visit inpatient expense, drug expense per inpatient visit, and the proportion of drug expense out of inpatient expense dropped by 399.6 CNY (65.59 USD), 278.7 CNY (45.75 USD) and 3.92 percentage points, respectively.
Table 10 shows the reduction of expense in relative terms. For outpatient visits, the expense reduced by 11% according to hospital-data DID and by 9% according to pre-post comparison in regression. For inpatient visits, the expense reduced by 11% from both hospital-data and individual-data DID. Assuming the results for outpatient expense remains the same from hospital-data and individual-data DID, the estimate of relative change in outpatient expense from hypothetical individual-data DID is 11%.
In sum, in absolute terms, the total expense per visit reduced by 19.02 CNY (3.12 USD) for outpatient services and 399.6 CNY (65.60 USD) for inpatient services. In relative terms, the reduction of expense per visit is 11% for both outpatient and inpatient services. The drug expense per visit dropped by 27.20 CNY (4.47 USD) for outpatient services and 278.7 CNY (45.75 USD) for inpatient services. The proportion of drug expense out of total expense per visit dropped by 11.73 percentage points in outpatient visits and by 3.92 percentage points in inpatient visits.
Methodologically, the DID method could be only used to control the time-related unobservable confounding variables but could do nothing for the observable confounding variables like gender, age, severity of illness, treatment effect, etc. In order to control both of the observable confounding variables and time-related unobservable confounding variables, DID with control variables in regressions is employed to further analyze the effects of ZPED on inpatient expense. The results are similar with or without controlling for observable variables (see Table 11). It suggests that unobservable confounders rather than demographic characteristics contribute more to the point estimates of outcomes impacted by ZPED.
To analyze the outpatient data, both hospital-data DID and regressions are employed. Without access to the outpatient records in Zhenping county hospital, only data from Ningshan hospital are utilized in regressions for pre-post comparison. The observable confounding variables such as gender, age, and surgery are controlled in regressions. Table 6 shows that control variables such as age and surgery have an impact on the estimates. However, the unobservable confounders remain unaddressed. In the DID, the unobservable and time-related confounding factors are controlled. The results from the DID are different from those from regressions because unobservable confounders are controlled. It is assumed that unobservable confounders rather than demographic characteristics contribute more to the point estimates of ZPED on outpatient expense. From Table 3, the per-visit outpatient expense and the per-visit drug expense drop by 19.02 CNY (3.12 USD) and 27.20 CNY (4.46 USD), respectively. Importantly, this implies that the non-drug expense increased by 8.18 CNY (1.34 USD).
Because ZPED reduces healthcare prices for patients, it has the potential to change the structure of inpatient and outpatient utilization. First, the probability of an inpatient visit is independent from the change of outpatient price . This means that patients’ decision making on whether to utilize inpatient services are determined by factors other than outpatient price. Second, the impact of inpatient price on the probability of outpatient visit is significant and positive. The point estimate was 0.171 from pooled data from nationally representative surveys in 2003 and 2008 . This means that the outpatient visits by the general population will decrease by 2% if the inpatient price is reduced by 11%. A very interesting point made in the Zhou et al. (2011) study is that users of outpatient services are more sensitive than the general population to the change in inpatient price. For users of outpatient services, outpatient visits will decrease by 3.39% if the inpatient price is reduced by 11%. Third, it is necessary to emphasize that in the study conducted by Zhou et al. , among nine estimated cross-price elasticity values, only two are significant at the 95% confidence level. In most of the cases, outpatient services and inpatient services are un-exchangeable. Even though in the naïve case that outpatient services are substitutes for inpatient services, the magnitudes of 2% and 3.39% are small. However, these estimates in traditional economics may not hold from the perspective of behavioral economics. Substantial evidence from behavioral economics research has shown that people oftentimes make decisions not based on absolute, but on relative changes in price . This phenomenon is at the heart of relative thinking theory, which states that people are influenced more by relative changes than by absolute changes in a given base line.
There are some important limitations in this study. First, the individual outpatient records were not accessible in Zhenping County, which limits the analysis for outpatient services. Second, due to the segmented health record system for inpatient service in Zhenping county hospital, a part of the inpatient records was lost in the merge of the three datasets, which may introduce selection bias. However, it is reasonable to assume that the lost data are random rather than systematic and selection bias is kept at a minimum, if any. Third, the analysis of drug expense was on all drugs rather than on zero-markup drugs because it is impossible to distinguish zero-markup drugs in patient datasets.
Quantitative analyses indicate that implementation of ZPED is a benefit for patients in both absolute and relative terms. The reduced per-visit expense holds for both outpatient and inpatient services. This could have implications for the demand and utilization of these services in the future. Implementing ZPED will reduce the per-visit medical expense for patients in both absolute and relative terms. The absolute monetary reduction of per-visit inpatient expense is 20 times of that in outpatient care. According to cross-price elasticity, the substitution between inpatient and outpatient is small due to changes in expense. Furthermore, given that the relative reductions are the same for outpatient and inpatient visits, according to relative thinking theory, the incentive to utilize outpatient or inpatient care attributed to ZPED is equivalent, regardless of the 20-times price difference in absolute terms. As a result, it is more likely that comparative utilization of these two services will not change significantly following implementation of ZPED in county hospitals.
Conceived and designed the experiments: ZLZ YS JG. Performed the experiments: ZLZ ZYZ JC YP. Analyzed the data: ZLZ ZYZ QY. Wrote the paper: ZLZ YS BC. Critical revision of the manuscript: YS JG BC. Approval of the final version of the manuscript: ZLZ YS BC ZYZ JG QY JC YP.
- 1. Petryna A, Lakoff A, Kleinman A (2006) Global Pharmaceuticals: Ethics, Markets, Practices. Duke University Press Books.
- 2. Geest S, Whyte S, Hardon A (1996) The Anthropology of Pharmaceuticals: A Biographical Approach. Annual Review of Anthropology 25: 153–178.
- 3. Whyte S, Geest S, Hardon A (2003) Social Lives of Medicine. Cambridge University Press.
- 4. Nichter M (2008) Global Health: Why Cultural Perceptions, Social Representations, and Biopolitics Matter. University of Arizona Press.
- 5. Craig S, Adams L, Spielberg S, Campbell B (2009) Pediatric therapeutics and medicine administration in resource-poor settings: a review of barriers and an agenda for interdisciplinary approaches to improving outcomes. Soc Sci Med 69: 1681–1690. pmid:19796859
- 6. Lagarde M, Palmer N (2008) The Impact of User Fees on Health Service Utilization in Low- and Middle-Income Countries: How Strong Is the Evidence?”. Bull World Health Organ 86: 839–848. pmid:19030689
- 7. WHO (2000) WHO Medicines Strategy: Framework for Action in Essential Drugs and Medicines Policy 2000–2003.Geneva, Switzerland.
- 8. Hogerzeil H. (2004) The Concept of Essential Medicines. BMJ, 329: 1169–1172. pmid:15539676
- 9. Ratanawijitrasin S, Soumerai S, Weerasuriya K (2001) Do National Medicinal Drug Policies and Essential Drug Programs Improve Drug Use?. Soc Sci Med 53: 831–844. pmid:11522132
- 10. Süssmuth-Dyckerhoff C, Wang J (2010) China’s health care reforms. Health Int 12: 55–67.
- 11. The State Council of China (2009) Implementation Plan for the Recent Priorities of the Health Care System Reform (2009–2011). Beijing, China: The State Council of China.
- 12. WHO (2001) How to Develop and Implement a National Drug Policy. 2nd edition. Malta: World Health Organization.
- 13. Eggleston K (2012) Health Care for 1.3 Billion: China’s Remarkable Work in Progress. The Milken Institute Review. Second Quarter.
- 14. Zhou ZL, Su Y, Gao J, Campbell B, Zhou ZY, Xu L, et al. (2013) Assessing equity of healthcare utilization in rural China: results from nationally representative surveys from 1993 to 2008. International Journal for Equity in Health 12: 34. pmid:23688260
- 15. Hu J, Rao K, Qiang J, Wu J (2007) The Study of Economic Burden of Chronic Non-communicable Diseases in China. China Journal of Prevention Control Chronic Non Communicable Disease 15: 189–193.
- 16. Wagstaff A, van Doorslaer E (2003) Catastrophe and impoverishment in paying for health care: with applications to Vietnam 1993–1998. Health Econ 12: 921–934. pmid:14601155
- 17. Yip W, Hsiao WC (2009) Non-evidence-based policy: how effective is China’s new cooperative medical scheme in reducing medical impoverishment? Social Science & Medicine 68: 201–209.
- 18. Gulliford M, Figueroa-Munoz J, Morgan M, Hughes D, Gibson B, Beech R, et al. (2002) What does ‘access to health care’ mean? J Health Serv Res Policy 7: 186–188. pmid:12171751
- 19. Sun Q, Zuo G, Li K, Meng Q, He P (2012) Whether the essential medicine policy decreases the medical cost of rural population: experience from three counties in Anhui Province. Chinese Health Economics 31: 65–67. pmid:21880401
- 20. Wang H, Tang Y, Liu Y, Yang L, Zhang X (2012) Effect evaluation of national essential medicine system in China: based on survey data from Shandong, Hubei and Sichuan. Chinese Journal of Health Policy 5: 30–34.
- 21. Chen Y, Bai B, Dai T (2013).Effect on implementing essential medicine system at primary healthcare institutions in Anhui Province. Chinese Journal of Health Policy 6: 34–35.
- 22. Yang L, Liu C, Ferrier JA, Zhou W, Zhang X (2013) The impact of the National Essential Medicines Policy on prescribing behaviours in primary care facilities in Hubei province of China, Health Policy and Planning 28: 750–760. pmid:23161585
- 23. Song Y, Bian Y (2012) Impact evaluation of essential medicine system on prescription charges in township hospitals of Shandong province. Chinese Health Service Management 8: 586–587.
- 24. Shi Y, Li W (2013) Effect evaluation and consideration of essential drug system in Beijing Xicheng District around 6 years before and after the implementation of the System. China Pharmacy 24: 300–302.
- 25. Song Y, Bian Y, Petzold M, Li L, Yin A (2014) Effects of the National Essential Medicine System in reducing drug prices: an empirical study in four Chinese provinces. J Pharm Policy Pract 7: 12. pmid:25317336
- 26. Center for Health Statistics and Information MOH (2009) An Analysis Report of National Health Services Survey in China, 2008. Beijing: Xiehe Medical University Press.
- 27. Ministry of Health (2011) Statistical bulletin of China’s health service development of 2010. www.moh.gov.cn.
- 28. Li L, Liu GE, Chen Y, Yao L (2008) Evaluation and thoughts on the effects of implementing the policy “without added profit” to sale drug in community health service institutions of Beijing. Chinese Health Economics 27: 41–43.
- 29. He P, Liu B, Sun Q (2011) Comparative analysis on the drug price of township health center before and after the essential medicines system reform: Based on the sample survey of three counties in Anhui province. Chinese Journal of Health Policy 4: 11–16.
- 30. Zhou ZL, Su Y, Zhou ZY, Gao J, Yu Q, Chen J, et al. (2013) The Impact of Zero-markup Policy on Inpatient Expenditure: Evidence from County Public Hospital in Shanxi Province. Chinese Journal of Health Policy 6: 25–32.
- 31. Zhou ZL, Su Y, Campbell B, Zhou ZY, Gao J, Yu Q, et al. (2015): The impact of China’s Zero-Markup Drug Policy on county hospital revenue and government subsidy levels, Journal of Asian Public Policy.
- 32. Shen R (2013) Impact of drug zero-markup sale reform on drug use in public hospital. Chinese hospital 17: 62–63.
- 33. Wang J, Zhen J, Fan W, Zhang Q, Zhang G (2013) Investigation and analysis of the effects of national essential drug system on medical expenses in different levels of medical Institutions. China pharmacy 23: 2982–2984.
- 34. Angrist J, Pischke J (2008) Mostly harmless econometrics: An empiricist's companion. Princeton: Princeton University Press.
- 35. Wooldridge J (2005) Introductory Econometrics: A Modern Approach. Boston: South-Western College Publishing.
- 36. Chen M, Mao W, Chen W, Jiang H, Zhang L (2013) Impact of the essential medicines policies on operation and expenditure burden at primary health care institutions. Chinese Health Resource 16: 93–94.
- 37. Zhou Z, Su Y, Gao J, Xu L, Zhang Y (2011) New Estimates of Elasticity of Demand for Healthcare in Rural China. Health Policy 103: 255–265. pmid:22014842
- 38. Azar O (2007) Relative thinking theory. The Journal of Socio-Economics 36: 1–14.