New Evidence on the Impact of China s New Cooperative Medical Scheme

Transcription

1 New Evidence on the Impact of China s New Cooperative Medical Scheme and Its Implications for Rural Primary Health Care Kim Singer Babiarz, a Grant Miller, Ph.D., M.P.P., b Hongmei Yi, Ph.D., c Linxiu Zhang, Ph.D., c and Scott Rozelle, Ph.D. d a Department of Agricultural and Resource Economics, University of California, Davis, CA USA b Stanford Medical School and National Bureau of Economic Research, Stanford, CA USA c Center for Chinese Agricultural Policy, Institute of Geographical Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing China d Freeman Spogli Institute for International Studies, Stanford University, Stanford, CA USA Abstract: Context: China s New Rural Cooperative Medical Scheme (NCMS) aims to provide health insurance to 800 million rural citizens and to correct distortions in rural Chinese health care. However, little is known about the effects of its individual policy features or its impact on village clinics (the primary care backbone of rural China). Objective: The objective of this study is to determine whether the NCMS program and it s individual policy features have impacted village clinic operations and clinic user rates. Design: We collected detailed data from 160 village clinics and 8,339 individuals in 100 villages across 5 Chinese provinces in 2004 and Using multivariate linear regressions that control for a set of clinic and individual attributes as well as village and year effects, we estimate the effect of individual NCMS policy features on clinic and individual outcomes. Main Outcome Measures: Clinic outcomes: log weekly patient flow, log annual income, revenue from medicine sales; Patient outcomes: patient probability of seeking care, patient log annual health expenditure, patient probability of catastrophic medical spending Results: For village clinics, we find that NCMS programs are associated with a 26% increase in patient caseload and a 29% increase in gross revenue but no change in net revenue or drug revenue. For individuals, NCMS participation is associated with a 6% increase in village clinic use but no change in overall medical care use. Out-of-pocket medical spending falls by 19% and catastrophic spending declines by 36% with participation as well. Across heterogeneous county programs, these changes occur even with minimal benefit packages. Conclusions: Although the NCMS provides financial risk protection and has partly corrected distortions in rural Chinese health care (reducing specialty services and drug sale rates), it may have also shifted uncompensated burdens to village clinics. Given renewed interest among Chinese policymakers in emphasizing primary care, this phenomenon deserves greater attention. New Evidence on the Impact of China s New Cooperative Medical Scheme

2 1. Introduction and Its Implications for Rural Primary Health Care During the post-mao era, China s market-based reforms produced unprecedented social and economic gains. However, an unintended consequence of decollectivization was the near collapse of its rural health system. 1,2 The dismantling of communes and the cooperative medical schemes (coupled with fiscal decentralization and sharp reductions in tax revenue) undercut the rural health system s public finance base1 and left 640 million people in rural China uninsured. 3 5 Consequently, household protection against unpredictable medical spending evaporated. 6 Without public subsidies, barefoot doctors from former communes also opened private clinics en masse and public sector health facilities raised prices. 1 Government regulatory responses made specialty medical care and drug prescription more lucrative and allowed doctors to over provide them The result was an escalation of medical spending and a re-doubling of household exposure to financial risk. 11,12 Recognizing these problems, in 2003 China s government established one of the largest public sector health insurance programs in the world: the New Rural Cooperative Medical Scheme (NCMS). 1 3 By the end of 2008, this program covered more than 800 million people in rural China. 13,14 NCMS allows households to purchase health insurance for modest premiums of 10 to 20 yuan (roughly $1.50 to $3.00) per person, and local and central governments contribute subsidies of 20 to 40 yuan each. County governments effectively operate their own versions of NCMS, defining benefit packages, designating participating providers, pooling risk across local subscribers and experimenting with policy innovations. Although all policies cover a portion of inpatient expenses at designated hospitals, county NCMS programs are otherwise heterogeneous. Some reimburse varying degrees of outpatient care at hospitals and/or village clinics. Others include compulsory household medical savings accounts (hereafter household accounts) that can be used for outpatient care and rolled-over year-to-year. Village clinic reimbursement is also frequently paired with more stringent licensing requirements, additional training for clinicians, and in some isolated cases, provider payment reforms to selectively reduce the supply of drugs and specialty services. 13 This paper uses provider- and individual-level data collected in 2004 (shortly after the introduction of NCMS in selected regions) and 2007 (after the dramatic expansion of NCMS across most of rural China) from 100 villages in 25 rural counties of 5 provinces. Given substantial heterogeneity in policy features across counties which we find to be a significant predictor of important outcomes, a particularly attractive aspect of our data is its detailed information about the nature of each county s NCMS program. This detail allows us to pinpoint precise policy features associated with changes in outcomes among village clinics and rural households. Although a small number of studies have analyzed changes in provider-level outcomes under the NCMS, 17 our nationwide collection of village clinic data is unique. Village

3 clinics constitute the primary care and public health service backbone of rural China. Clinics serve many more patients than local township health centers, and they often are the only reasonable source of medical care in China s most remote villages. Given their central role in rural health care delivery and reform efforts to promote primary over specialty care, it is essential to understand how NCMS is affecting their operations and viability. 2. Methods 2.1 Data We use two waves of a dataset collected by the Chinese Academy of Sciences Center for Chinese Agriculture Policy in collaboration with Stanford University. The dataset contains three major components: a village clinic module, a household module, and a county NCMS program module. The first wave was conducted in April 2005, collecting data from 2004 in a random sample of 100 villages from 25 rural counties in Jiangsu, Sichuan, Shaanxi, Jilin, and Hebei. Sample counties are illustrated in Figure 1. The resulting sample includes 156 village clinics and 3,257 individuals. The second wave was repeated in the same 100 villages in April 2008, collecting 2007 data from 160 clinics and 8,339 individuals. In the first wave wave, 8 households in each village were randomly drawn and surveyed; in the second wave, the sample was increased to 20 households per village (the original 8 plus additional randomly selected households). The second-round survey instrument was also an expanded version of the first130 round instrument. We use variables collected in both waves. The village clinic module collected information about clinic utilization, finances and operations. Specific variables include average weekly patient flow, average monthly gross income, total net annual income, and the share of income attributable to drug sales. Clinics also provided information about NCMS participation and program rules, including whether or not patient expenditures were reimbursable. Clinicians were also asked questions about NCMS-related work performed throughout the year and how they felt that the program had affected their business. The household module gathered detailed information about medical care spending and use of health services in the past year for each family member. Specific variables include whether or not a household participated in NCMS, if each individual was ill during the past year, if each person who was sick consulted a doctor, and the type of facility from which medical care was received. Respondents were also asked how much was spent out-of-pocket and how medical care was otherwise financed (for example, how much was reimbursed by the NCMS or financed by borrowing or selling assets). For out-of-pocket spending, respondents were instructed to report only expenses paid for clinical services and drugs (not spending on transportation and meals, for example). Finally, the county NCMS program module (executed in each of the 25 counties 5 counties per province) collected data on the features of each county s program from local NCMS officials and township health centers. These program attributes include the types of services that qualify for reimbursement (inpatient only; inpatient and outpatient) and

4 whether programs include household accounts. Descriptive statistics for these data are in Table Statistical Analysis Our analyses of how village clinic operations and individual welfare have changed under NCMS take three basic forms. For simplicity, we call these basic, full, and combined models. We perform each type of analysis with both clinic- and individual-level data using ordinary least squares (OLS) regression. For dichotomous dependent variables we use linear probability models given the well-known incidental parameters problem with probit and logit models when controlling for fixed effects. Our four key clinic-level outcome variables are weekly patient flow, monthly gross income, annual net income, and share of monthly income from drug sales. Our main individual-level outcomes are if an individual sought medical care when sick, if medical care was sought at a township health center or hospital, total annual out-of-pocket health spending, and exposure to financial risk (if out-of-pocket spending exceeded the 90th percentile of spending among the uninsured and if medical care was financed by borrowing or selling assets). The basic model estimates how clinic and individual outcomes are associated with a single NCMS measure (controlling for a detailed set of clinic and individual characteristics, respectively). For clinics, this variable is eligibility for NCMS reimbursement; for individuals it is participation in NCMS. At the clinic level, we control for the number of doctors, doctor training, doctor age, size of service area, size of medicine stock and average village income per capita. In the individual-level models, we control for gender, age, the square of age, education, urban/rural resident status, and mean per capita village income. Finally, to adjust for unobserved differences across villages and to account for common changes over time, we also include dummy variables for villages and years (in our case, for the year 2007). Given substantial cross-county heterogeneity in NCMS policies, the full model then estimates how village and individual outcomes are associated with distinct policy attributes of county NCMS programs (hereafter policy attributes) in addition to the single NCMS measure. Results from the full model provide a more nuanced picture of specific behavioral mechanisms embedded in NCMS that might explain its impact. The three specific policy attributes that we analyze for clinics are NCMS availability to village residents, coverage of outpatient services at local township health centers, and household account requirements. The policy attributes that we study for individuals are the availability of NCMS reimbursement for village clinic services, NCMS reimbursement of outpatient services at township hospitals, and whether or not an individual s family has a household account. All control variables included in the basic models are also incorporated into the full models. Finally, to evaluate the combined impact of common county-level bundles of policy attributes, we specify three frequently-observed bundles of policy attributes (hereafter policy bundles): a.) reimbursement for inpatient services at township health centers and hospitals (from the common fund), b.) reimbursement for inpatient services at township

5 health centers and hospitals from the common fund plus reimbursement for village clinics services from household accounts, and c.) reimbursement for inpatient and outpatient services at township health centers and hospitals from the common fund plus reimbursement for village clinics services from household accounts. (In other words, (b) = (a) plus household accounts, and (c) = (b) plus outpatient reimbursement.) To calculate the impact of each policy bundle, we use the full model to conduct joint significance tests of each bundle s components. When justified by significant F-statistics, we then sum across policy components to obtain the implied net impact of each of the three combinations. 3. Results We present results from our village clinic and individual analyses in single tables for each type of model (basic, full and combined). Estimates from specifications with natural log transformed dependent variables can roughly be interpreted as percent changes. Estimates from linear probability models (with dichotomous dependent variables) can be interpreted as percentage point changes. Table 2 Panel A reports results from the basic model for village clinics. Clinic designation for NCMS reimbursement is associated with a 26% increase in weekly patient flow and a 29% increase in monthly gross income. However, annual net income and share of income from drug sales remain unchanged. Taken together, the gross and net income results suggest possible reductions in the provision of high profit margin specialty services as well as possible increases in unfunded clinic responsibilities under the NCMS (such as enrolling participants, managing patient reimbursement applications, complying with required facility renovations, and financing mandated clinician training all of which may have raised costs but not net earnings). Table 2 Panel B shows basic model estimates for individuals. While NCMS participation is not associated with a change in the probability of using medical care if sick, it is associated with a change in type of facility used when ill. NCMS enrollees are 5% less likely to use township health centers or larger hospitals (or 5% more likely to use village clinics). The increase in patient flow found at village clinics therefore appears to be partially due to patients switching from township health centers to village clinics (as well as some unmeasured increase in the frequency of low profit margin service use). NCMS participation also is associated with a 19% reduction in total annual out-of-pocket medical spending and some reduction in exposure to financial risk. In particular, NCMS participation is associated with a 2 percentage point reduction (a 29% decline) in the likelihood of having net out-of-pocket health expenditures above the 90th percentile of spending among the uninsured and a 2 percentage point (47%) reduction in the likelihood of financing medical care through asset sales or borrowing. Table 3 reports results from the full model. Panel A shows that clinic eligibility is positively associated with weekly patient flow, but no longer significantly so. Instead, simply being in a village with a NCMS program is associated with a near-doubling of weekly patients (a 55% increase implying that NCMS programs generate large

6 spillovers for all local clinics that dominate the effect of eligibility for reimbursement). One possible explanation is that NCMS lowers the cost of medical care at township health centers, so patients are more likely to seek follow-up services and drug refills at more accessible village clinics. Estimates for clinic income follow the basic model s general pattern and also match the full model patient flow results. Clinics in NCMS counties experience large increases in gross income (71%) regardless of clinic eligibility for reimbursement but no change in net income or share of income from drug sales. Consistent with studies of medical savings accounts (in combination with greater regulation and oversight), 18,19 household account requirements are associated with substantial reductions in weekly patient flow (39%) and nearly significant reductions in gross clinic income (as individuals spend account funds more like their own money because they can be saved for catastrophic needs or spent in subsequent years). Outpatient reimbursement of township health center services is not associated with any village clinic outcome. Table 3 Panel B presents full model estimates for individual outcomes. As with the basic model, no NCMS program attribute is associated with statistically significant changes in the probability of medical care use when sick or with the choice of where to use services. Outpatient reimbursement is associated with an increase in the probability of using medical care, and clinic reimbursement is associated with a decrease in the probability of using a township health center or hospital (or an increase in village clinic use) although both fall just shy of conventional significance levels. The negative relationship between NCMS participation and out-of-pocket expenditures is no longer significant, but outpatient reimbursement of township health center services is associated with a 19% decrease in out-of-pocket medical spending. Household account requirements are also associated with a reduction (26%) in out-of-pocket spending (which together with the other clinic and individual full model results implies a reduction in intensity of service use if services are used). Finally, NCMS reimbursement of village clinic services is associated with a 3 percentage point (63%) reduction in the probability of financing medical care through borrowing or asset sales, suggesting meaningful financial risk protection. Table 4 reports combined model results, presenting cumulative effects for common NCMS policy bundles. Estimates and levels of statistical significance for policy bundles a, b, and c are sums across full model estimates for individual policy attributes and corresponding F tests of their joint significance. Panel A s clinic results for policy bundle a generally correspond to those of the basic model. The results for benefit policy bundle b are qualitatively similar, suggesting that reductions in patient flow and gross income associated with household accounts are dominated by increases associated with inpatient coverage. The most comprehensive of the three policy bundles is associated with larger patient flow and gross income gains (again perhaps due to spillover effects of coverage at township health centers). Policy bundles b and c are also associated with reductions drug sales as a share of clinic revenue (24% and 22%, respectively). No bundle, however, is associated with a significant change in village clinic annual net revenue. Table 4 Panel B shows combined model results for individual-level outcomes. No policy bundle is significantly associated with use of medical care when sick or changes in types

7 of facilities used. However, policy bundle c is associated with a 40% reduction in out-of-pocket medical care spending. Moreover, policy bundles b and c are associated with a statistically significant decline in the probability of borrowing or selling assets to pay for medical care. 4. Discussion Our study provides new evidence on the impact of the NCMS on village clinics and rural households. In doing so, it is one of the first to analyze individual policy attributes of heterogeneous county-level programs as well as the first to provide an assessment of how the NCMS is affecting village clinics (a small number of studies have studied the impact of smaller scale insurance schemes on village clinics ). For individuals, the NCMS is not clearly related to the use of medical care, but it may have re-directed patients away from specialized facilities and towards village clinics. Importantly, while other studies have found mixed NCMS results for financial risk protection 2,23, we find evidence that it has reduced outlier out-of-pocket spending and the need to borrow or sell assets to pay for medical care. Perhaps more importantly, the results of our clinic-level analyses provide new insight into NCMS s impact on primary health care in rural villages. Overall, we find that NCMS has increased clinic patient flow and gross income but not net revenue. Increases in patient flow and gross but not net clinic revenue may reflect desirable reductions in more specialized high profit margin services and drug sales. However, they may also raise important questions about uncompensated burdens that the NCMS has shifted to village clinics. Specific new responsibilities for which clinics are not paid include enrolling participants, managing patient reimbursement applications, complying with required facility renovations and financing additional mandated clinician training. In fact, many village clinicians participating in our survey reported that NCMS has hurt their overall business. Given renewed interest among China s policy-makers in emphasizing primary care, future work is needed to more closely examine the impact of the NCMS on rural primary care facilities. An important limitation of our study is the non-random placement of NCMS programs and non random household participation decisions.14 Our results control for a large number of potentially confounding factors, however, and we have also used other statistical approaches to assess their robustness. Repeating all of the paper s analyses using propensity score weights (which finds the best control group match for each treatment group observation using all observed characteristics simultaneously) yields the same pattern of results with slightly larger magnitudes (see Appendix 1).

8 Acknowledgements We acknowledge the funding of the following groups: Stanford s Presidential Fund for Interdisciplinary International Studies; MIT; Chinese Academy of Sciences (100 Talents) and KSCX2-YW-N-039), and Social Protection in Asia (SPA) policy-research and network-building programme which is funded by the Ford Foundation and IDRC and managed by the IHD New Delhi, India, and IDS Brighton, UK. References 1. Liu X,Yi Y. The Health Sector in China: Policy and Institutional Review. A Background Paper for the World Bank China Rural Health Study. Washington DC: The World Bank; Wagstaff A, Lindelow M, Wang S, Zhang S. China s Health Challenges at the Start of the New Millenium. In: Reforming China s Rural Health System. Washington DC: The World Bank; Insuring Rural China: Rising to the Challenge. Washington DC: The World Bank; Hsiao WC. The Transformation of Health Care In Rural China. N Engl J Med. 1984;141: Liu Y. Development of the Rural Health Insurance System in China. Health Policy Plan. 2004;19: Liu Y, Rao K, Hsiao WC. Medical expenditure and rural impoverishment in China. J Health Popul Nutr. 2003;21: Hsiao WC. Chinese Health Care System: Lessons for Other Nations. Soc Sci Med. 1995;41: Blumenthal D, Hsiao WC. Privatization and Its Discontents: The Evolving Chinese Health Care System. N Engl J Med. 2005;353: Lindelow M, Wagstaff A. China s Health Sector: Why Reform is Needed. Washington DC: The World Bank; Rural Health in China: Briefing Note Series. 10. Yip W, Eggleston K. Addressing Government & Market Failures with Payment Incentives: Hospital Reimbursement Reform in Hainan China. Soc Sci Med. 2004;58: Yip W, Hsiao WC. The Chinese Health System At A Crossroads. Health Aff. 2008;27: Hu S, Tang S, Liu Y, Escobar ML, de Ferranti D. Reform of How Health Care is Paid

9 for in China: Challenges and Opportunities. Lancet. 2008;372: New Rural Cooperative Medical Scheme in China. Beijing: Ministry of Health Department of Rural Health Management, Implementing the New Cooperative Medical Schemes in Rapidly Changing China: Issues and Options. Beijing: Office of the World Health Organization Representative in China, Brown P, debrauw A, Du Y. Understanding Variation in the Design of China s New Cooperative Medical System. China Q. 2009;198: Brown P, Theoharides C. Health-Seeking Behavior and Hospital Choice in China s New Cooperative Medical System. Health Econ. 2009;18:S47-S Wagstaff A, Lindelow M, Gao J, Xu L, Qian J. Extending health insurance to the rural population: An impact evaluation of China s new cooperative medical scheme. J Health Econ. 2009;28: Yip W, Hsiao WC. Medical Savings Accounts: Lessons from China. Health Aff. 1997;16: Hsiao WC. Behind the Ideology and Theory: What is the Empirical Evidence for Medical Savings Accounts? J Health Polit Policy Law. 2001;26: Yip W, Wang H, Liu Y. Determinants of Patient Choice of Medical Provider: a Case Study in Rural China. Health Policy Plan. 1998;13: Zhou Z, Gao J, Xue Q, Yang X, Yan J. Effects of Rural Mutual Health Care on Outpatient Service Utilization in Chinese Village Medical Institutions: Evidence from Panel Data. Health Econ. 2009;18:S129-S Yip W, Wang H, Hsiao WC. The Impact of Rural Mutual Health Care on Access to Care: Evaluation of a Social Experiment in Rural China. Working Paper, Wagstaff A, Lindelow M. Can Insurance Increase Financial Risk? The Curious Case of Health Insurance in China. Washington DC: The World Bank, Policy Research Working Paper Series no

10 Figure 1. Sample province location.

11

12

13

14