ReportNo. 15278-CHA : Issuesand Options in Health Public Disclosure Authorized Financing

August12, 1996

Chlin, in Mn(N'loli(i Dep,irtmenelt I Ii111,1n I)evelOpmrnntDep,ilrillent Public Disclosure Authorized Public Disclosure Authorized

Public Disclosure Authorized Documentof the World Bank Exchange Rate CuLrrencyUnit = Yuan (Y) 1993 US$1.00 = Y 8.0 (commercial rate) January, 1996 US$ 1.00= Y 8.3 (official rate)

Note: Ulnlessotherwise specified, all yuan valueshave been adjustedto 1993 constantyuan.

Acronyms

CEM Country Economic Memorandumii(World Bank) CMS Rural Cooperative Medical System DALY Disability-adjusted Life Year DRG Diagnosis Related Grouips EPI Expanded Program on EPS Epidemic PreventionlService FP Family Plannilig and Reproductive HealthiProgram GBD Global Burden of GIS Government Employees' Health Insurance System [also retirees, military, university studenits] HEI Health Economics Institute; sometimes called National Healthi Economics Researcih Institute HFA Health-for-All in the Year 2000 (Goal of the WHIO) IMR InfanitMortality Rate LIS Labor Health Insurance System (covers employees of state-owned enterprises) MCH Maternal and Child Health Program MOA Ministry of Agriculture MOF Ministry of Finance MOH Ministry of Health NCD Nonconimunicable disease OECD Organization for Economic Cooperation and Development PRC People's Republic of Chinia

SOE State-owned enterprise SPC State Planning Commission SSB State Statistical Bureau TCM Traditional Chinese TFR Total fertility rate THC Township health centers (middle level of the rural three-tier health system, between village clinic and county ) U5MR Under-five mortality rate UNICEF United Nations Children's Fund WHO World Health Organization

Glossary This glossary provides definitions of some of the key terms used in this report. Adverse Selection - The tendency of individuals expecting high health expenditures to purchase insurance or purchase a more generous insurance package than people expecting low levels of expenditures. Basic Benefits Package - A minimum set, or core, of health services. Capitation Payment - A fixed payment to a provider for each listed or enrolled person served per period of time. Payments will vary accordingto the number, age, and sex of patients enrolled but not witlhthe number of services rendered per patient. Community Financing - A community-managedscheme whereby a local community group collects and manages funds from households, government, and local industries and organizes the delivery of a package of health benefits for communitymembers. The health risks of community membersare pooled, and the benefits include some catastrophic as well as basic care. Cost Containment - A set of steps to control or reduce inefficiencies in the consumption, allocation, or production of services whicihcontribute to higher than necessary costs. Inefficiencies in consumption can occur when health services are inappropriately utilized; inefficiencies in allocation exist when a different mix of services could produce greater health benefits; and inefficiencies in production exist wlhenithe cost of producing health services could be reduced by using a different combination of resources. Cost-sharing - A provision of health insuranceor third-party payment that requires the individual who is covered to pay part of the cost of medical care received. This is distinct from the payment of a health insurance premium, contribution, or tax, which is paid whether medical care is receivedor not. Ccst-sharing may be in the form of deductibles, co-insurance, or co-payments. Disability-adjusted Life Year (DALY) - A unit used for measuring both the global burden of disease and the effectiveness of health interventions, as indicated by reductions in the disease burden. It is calculated as the present value of the future years of disability-free life that are lost as the result of the premature deaths or cases of disability occurring in a particular year. Diagnosis Related Groups (DRGs) - Groupings of diagnostic categories drawn from the InternrationalClassification of and modified by the presence of a surgical procedure, patient age, presence or absence of significant comorbidities or complications, and other relevant criteria.

.. DRGs are the case-mix measure used in Medicare's prospective payment system in the United States. Modifications of DRGs have been adopted in several other countries as a tool for hospital management and/or reimbursement. Externalities - Costs or benefits arising from production or consumption that fall on individuals and groups not directly involved in the production or consumption concerned; and which are not compensated for by exchange. For example, immunization of an individual against an infectious disease can block the transmission of the disease to other individuals who are not directly involved and who pay nothing for the protection they receive. This creates an external benefit. And one person's inappropriate use of an antibiotic will frequently hasten spread of resistance to it, creating external costs borne by others. Fee-for-Service - Payments to a provider for each item or service rendered. Global Budget - An aggregate cash sum, fixed in advance, intended to cover the total cost of a service, usually for one year ahead. Health Insurance - Financial protection against the medical care costs arising from disease or accidental bodily injury. Such insurance usually covers all or part of the medical costs of treating the disease or injury. Insurance may be obtained on either an individual or group basis. Health Maintenance Organization (HMO) - An organization that accepts responsibility for organizing and providing a defined set of health services for its enrolled population, in exchange for a predetermined fixed, periodic payment for each person or family unit enrolled. The payment is fixed without regard to the amount of actual services provided to an individual enrollee. Health Planning - Planning concerned with improving health, whether undertaken comprehensively for a whole community or for a particular population, type of health service, institution, or health program. The components of health planning include: data assembly and analysis, goal determination, action recommendation, and implementation strategy. Managed Competition - Government regulation of health insurance and health care markets using competition as the means to achieve efficiency objectives within a framework of government intervention. Moral Hazard - In health insurance, this refers to the fact that people who are insured tend to act in ways that increase health expenditures. They may take fewer precautions against avoidable events, or, more importantly, patients and providers increase the use of services because insurance subsidizes prices at the point of service. Out-of-Pocket Payments - Payments borne directly by a patient without reimbursement by any insurance. They include cost-sharing. Prepayment - Usually refers to any payment to an organization for anticipated services (such as an expectant mother paying in advance for maternity care). Prepayment is distinguished from insurance because it involves payment to organizations which, unlike an insurance company, take responsibility for arranging and providing needed services as well as paying for them (such as health maintenance organizations, prepaid group practices, and medical foundations). Public Good - A good or service whose benefits may be provided to a group at no more cost than that required to provide it for one person. The benefits of the good are indivisible and individuals cannot be excluded. For example, a measure that eradicates smallpox protects all, not just those paying for the .

iv Public Health - The science dealing with the protection and improvement of the population'shealth by organized community effort. Public health activities are generally those that are less amenable to being undertaken by individuals or that are less effective when undertaken on an individual basis. They do not typically include direct personal health services. Public health activities include: ; sanitation; preventive medicine, quarantine, and other disease control activities; occupational health and safety programs; assurance of the healthfulness of air, water, and food; health education; epidemiology, and others. Risk Pooling - The process of distributing the probability of financial loss across multiple parties. Health insurance, for example, is a way of distributing the uneven burden of high medical expenditures across a subgroup of the population. Salary Payment - Remuneration that is fixed per period of time and does not vary either with the number of individuals served or with the number of services rendered, although rate of salary change can depend on performance. Selection Bias - The tendency for multiple health plans or providers to attract,intentionally or unintentionally, an uneven (biased) distribution of health risks. Also known as risk selection. When done intentionally by insurance companies to discourage participants likely to need substantial care, the practice is sometimes known as "cherry-picking". Social Health Insurance - A term mainly used to denote compulsory, or public, health insurance, usually part of a social security system, which is funded from specific (mainly payroll) contributions and is managed by a government agency or autonomous organization such as a sickness fund, mutual aid society, or private insurer. Supplier Induced Demand - The ability of providers to use their authority to boost demand for their services, for the services of colleagues, or for the sale of drugs or tests in which they have a financial stake, above the level that would be demanded if patients had adequate and accurateinformation on medical benefits and costs. Third Party Payer - Any organization, public or private, that pays or insures health care expenses for beneficiaries at the time they are patients. The first party is the patient and the second party is the provider. Third parties may be private insurers; quasi-public bodies such as sickness funds; and government bodies themselves. Voluntary Health Insurance - Health insurance that is taken up and paid for at the discretionof individuals, or employers on behalf of individuals. Voluntary insurance can be offered by a private, public, or quasi-public body. Yellow Book - Generic term in China for hospital and health service price lists for the thousands of products and services sold.

References OECD: The Reform of Health Care, A Comparative Analysis of Seven OECD Countries, pp. 9-11, Health Policy Studies No. 2, Paris, 1994. Glossary of Terms as Commonly Used in Health Care, Alpha Center,Washington DC.

v

Contents

Page

Executive Summary xiii

1. Introduction: Health and Health Policy in China 1 2. Health Services And Their Financing 13 3. Strengthening Public Health Programs 25 4. Meeting the Needs of the Poor 33 5. Implementing Reforms in Pricing and Planning 41 6. Options for Efficient Risk Pooling in Rural Areas 49 7. Options for Efficient Risk Pooling in Urban Areas 59 8. Recommendations and Implications for Public Finance 67

Annexes

1. China's Health Expenditure and Health System Data 75 2. Outside Influences on Health Status 83 3. Deaths and Disease Burden in China 89 4. Child Mortality Trends in China 115 5. Health Indicators from Disease Surveillance Points System 121

Background Papers and Consultant Reports Prepared for this Study 127

Bibliography 129

vii Tables in Text

Table 1 Summary of Recommended Health Policy Actions, 1996-2001 Table 1.1 Trends in Under-Five Mortality Rates--China and Other Asian Countries, 1960-90 Table 1.2 Recommended Health Finance Policies: Impact on the Health Sector Table 2.1 Sources and Uses of Health Financing, 1993 Table 3.1 Government Finance of EPS and as a Share of GDP, Selected Years Table 3.2 Immunization Coverage in China, 1993 Table 3.3 Characteristics of Smokers in China, 1984 Table 3.4 Estimates of Tobacco Price Elasticity of Demand, Various Countries Table 4.1 Incidence of Absolute , 1978-90 Table 4.2 Health Status by Income, Rural China, 1993 Table 4.3 China's 592 Poor Counties: Comparison of Average Income, 1992-93 Table 4.4 Poverty-Related Disease Conditions, China, 1990 Table 5.1 Comparison of Production Costs and Fees for Selected Procedures in Shanghai , 1989 Table 5.2 Comparison of Production Costs and Fees for Body CT Scan, 1988 Table 6.1 Prevalence and Benefits of Community Financing in Thirty Poor Counties, 1993 Table 6.2 Management of Community Financing, 1993 Table 6.3 Percentage of Community Financing by Source Table 6.4 Prevalence and Benefits of Community Financing in Five Provinces, 1991 Table 8.1 Summary of Major Health Care Issues in China Table 8.2 Public Finance Implications of Proposed Health Initiatives: Illustrative Costs Table 8.3 China's Government Expenditures in International Perspective Table 8.4 Revenue Implications of Recommended Tax Measures in the Year 2000 Table Al.l China: GDP, Nominal Exchange Rates and Price Deflators, 1978-1994 Table Al.2 GDP, Health Spending and Population Growth, 1978-93 Table A1.3 China: National Health Expenditures, 1978, 1986, and 1993 Table A1.4 China: Government Spending on Health, Selected Years, 1978-93 Table A1.5 China: Components of Recurrent Health Budget, Selected Years, 1978-93 Table Al.6 China: Health Insurance Coverage and Spending, Rural Population, 1993 Table Al.7 China: Revenues and Government Subsidies of Health Institutions Providing Hospital Services, 1993 Table Al.8 China: Average Number of Hospital Beds and Health Care Personnel, 1993 and 1994 Table A 1.9 China: Utilization of Hospital Beds at County Level and Above, Selected Years, 1985-94 Table A2.1 Income and , 1978-94 Table A2.2 Illiterates and Semi-literates, Selected Provinces, 1982 and 1990 Table A2.3 Access to Safe Drinking Water and Sanitation, China and Comparators, 1990 Table A2.4 Per Capita Nutrient Availability in China, 1950-95 Table A2.5 Key Demographic Indicators for China: 1970, 1995, and Projections to 2020 Table A3.1 Number of Deaths in China by Age, Sex, and Cause, 1990 Table A3.2 Number of Deaths in China by Age, Sex, and Cause, 1990, by Percent Table A3.3 Burden of Disease in China by Age, Sex, and Cause, 1990 (in DALYs)

viii Table A3.4 Burdenof Disease in China by Age, Sex, and Cause, 1990(in DALYs) by Percent Table A3.5 DALYs by Cause,as a Percentageof Totals, China and other Regions, 1990 Table A3.6 DALYs by Cause, as a Percentageof Totals, China and other Regions,2020 Table A3.7 Deaths and Burden of DiseaseAttributable to TobaccoUse, China, 1990 and Projectionsto 2020 Table A4.1 EstimatedRates of Changein Under-FiveMortality, China, 1960-91 Table A4.2 Estimatesof China's Under-FiveMortality, 1960-90 Table A4.3 Income Growth Rates, China and OtherCountries, 1960-90 Table A5.1 Age-SpecificMortality Statistics, Urban and Rural Areas, 1990-94 Table A5.2 Male Age-SpecificMortality Statistics, Urban and Rural Areas, 1990-94 Table A5.3 FemaleAge-Specific Mortality Statistics, Urban and Rural Areas, 1990-94 Table A5.4 Male Age-SpecificMortality Statistics, Rural Areas, by Income Quartile, 1990-94 Table A5.5 Female Age-SpecificMortality Statistics, Rural Areas, by Income Quartile, 1990-94

Boxes in Text

Box 2.1 Health Researchand Development:A NeglectedComponent of the Chinese Health System? Box 3.1 The Challengeof STDsand HIV in China Box 5.1 The LeveragingEffect Box 5.2 The High TechnologyEquipment Race Box 5.3 Problemsof Rural Health Delivery: A Case Studyof Maternal and Child Health Programs Box 5.4 RegionalPlanning for Health Box 6.1 Estimatingthe Costs and Contentof a Basic BenefitPackage Box 7.1 Wage Tax Flows from Contributorsto Individualand CommonAccounts Box 7.2 The Three Tiers of FinancingHealth Servicesin the Jiujiangand Zhenjiang Experiments Box 8.1 Lessonsfrom Reformof OECD Health Systemsand Singapore

Figures in Text

Figure I Estimatesof Under-FiveMortality Rates in China, 1960-91and Projectionsto 1995 Figure 2 InsuranceStatus and Sourcesof Health Finance, 1993 Figure 1. I Deviationof China's Annual Rate of Declinein Under-FiveMortality from Rates Predictedby the GrowthRates in Income Figure 1.2 Distributionof DiseaseBurden by BroadGroups of Causefor China and Comparators,Estimates for 1990and Projectionsto 2020 Figure 1.3 InsuranceStatus of China's Population,1981 and 1993 Figure 1.4 Growth in Health Spendingas a Shareof GDP, SelectedCountries, 1960-93 Figure 2.1 Growth in Health Expendituresin China by Source, 1978-93 Figure 2.2 Per CapitaAnnual Health Spendingon PopulationGroups, 1993 Figure A2.1 EvolvingPatterns of Age Distributionand Mortalityin China, 1970, 1995, and2020 Figure A4.1 Trends in China's Under-FiveMortality Rate, 1960-91and Projectionsto 1995

ix Figure A4.2 Relation Between Change in Child Mortality and Change in Per Capita Income, China and Other Countries, 1960-89 Figure A4.3 Deviation of China's Annual Rate of Decline in Under-Five Mortality Rates (from Rates Predicted by Growth Rates in Income)

x Acknowledgments

This study was prepared by the World Bank at the request of the government of the People's Republic of China, which is considering policies to improve the equity and efficiency of health services. It is the third World Bank policy study of China's health sector. The series began with a general study entitled China. The Health Sector (1984), followed by China: Long Term Issues and Options in the Health Transition (1992). This report, China. Issues and Options in Health Financing (1996) looks in greater detail at the impact of China's move toward a market economy on the financing and organization of health care, and recommends a number of steps China can take to reform the financing of health care.

Research for this report was carried out jointly by the Chinese government and World Bank staff. Work in China proceeded under the direction of Mr. Liu Peilong, Director of the Foreign Loan Office of the Ministry of Health. Two World Bank task managers in the Human Development Department managed the project: William P. McGreevey from June 1994 to February 1996 and Helen Saxenian from March to August 1996. The report was authored by Helen Saxenian, together with William Hsiao, Dean T. Jamison, William P. McGreevey, and Winnie Yip. Initial work at the World Bank was carried out under the direction of Mr. Vinay Bhargava, Chief, Human Resource Division, China and Mongolia Department, followed by Mr. Joseph Goldberg, Chief, Rural and Social Development Division. Mr. Jagadish Upadhyay, health group manager, and Dr. Janet Hohnen, public health specialist, managed the work on behalf of the Division. Mr. Nicholas Hope, Director, China and Mongolia Department, guided the preparation work and Mr. Richard Newfarmer, Lead Economist in the Department, and Mr. Michael Walton, Chief Economist, East Asia and Pacific Region, helped set the overall context for the report. William P. McGreevey and Helen Saxenian worked under the general direction of Dr. Richard Feachem, Senior Adviser, Human Development Department. World Bank staff in Beijing, including Mr. Pieter Bottelier, Resident Mission Director, Mr. Ramgopal Agarwala, Ms. Kathy Ogawa, and Mr. Zhao Hongwen, also provided assistance.

China's Minister of Health, Professor Chen Minzhang, guided the early report preparation in October 1994. Ministry of Health staff, including Mr. Liu Xinming, Deputy Director, Planning and Finance Department, Professor Cai Renhua, Director, Legal Affairs Department, Ms. Liu Yingli and Mr. Liu Junguo of the Foreign Loan Office, and Mr. Fei Zhao Hui, now of the Ministry of Finance, provided considerable help during field visits. Mme. Sun, Ministry of Finance, and Mr. Ying Li, State Council, reviewed many phases of the study's work and provided advice throughout.

The study draws extensively on 15 background papers prepared between November 1994 and September 1995 by leading specialists in China's health economics and finance. A list of the papers and their authors appears before the bibliography attached to this report.

A related study of China's national health accounts was partially funded by the World Bank and a special grant from the Canadian government. Contributors to that study include Professor Peter Berman of Harvard University, Mr. Gilles Fortin of the Canadian Institute for Health Information, Mr. Vernon Hicks of Health Economics Consulting Services in Halifax, Nova Scotia, and Mr. J. Brad Schwartz, consultant, of Chapel Hill, North Carolina.

An advisory group has been helpful throughout in defining the principal issues and approaches in this study. This group consists of Professor William Hsiao, the K. T. Li Professor of Health Economics at Harvard University; Professor Hu Shanlian of Shanghai Medical University and

xi Deputy Director, Health EconomicsInstitute, Beijing; Professor Dean T. Jamnison,Director, Center for Pacific Rim Studies,UCLA; and ProfessorWei Ying, BeijingMedical University and Director, Health EconomicsInstitute.

Peer reviewersin the World Bank includeMr. Willyde Geyndt,Mr. CharlesGriffin, Mr. Jeffrey Hammer, Mr. EmmanuelJimenez, and Mr. NicholasPrescott. Membersof the World Bank's Health Group in the HumanDevelopment Department provided valuable comments,including Mr. Howard Barnum,Dr. Denis Broun, Mr. Philip Musgrove,Dr. Mary Young, and Mr. George Schieber. Mr. Richard Bumgarner,Mr. Yuanli Liu, and Mr. RichardPeto also providedvaluable input.

Bank staff visited the provincesof Shanxiand Jiangsu in October 1994, and Sichuan,Hebei, Jiangxi, and Guizhou,in April 1995. Bank staff membersand consultantswho joined these missions and contributedto the report includeProfessor John S. Akin, Dr. Harry E. Cross, Mr. Jeffrey Hammer, ProfessorWinnie Yip, and Mr. Zhou Ji An. Ms. HeleneGenest and Mr. Paul Hutchinson providedvaluable assistanceto the missionwork. An initialdraft of this report was discussedwith the Chinese governmentin October, 1995. Follow-upwork on nationalhealth accounts was done during a December,1995 mission. The full report was discussedwith Chinese governmentofficials in a workshop from April 22 - 24, 1996, and this version incorporates their comments.

This report was editedby MadelynRoss. MyleneDomingo prepared the manuscriptwith the help of Susan Sebastian,Akosua Hudgens,Yvette Atkins,and Euna Osbourne. Jillian Cohen, Ellen Lukens,and other Health Group staff helped in checkingsources and manuscriptreview.

xii Executive Summary

BACKGROUND

1. Before 1949 China's populationwas among the least healthyin the world. This burden of disease was both a consequenceand a cause of the nation's poor economicperformance. Subsequent investmentsin improvinghealth have contributedto China's high rate of economicgrowth, while at the same time directly improvinghuman well-being, particularly among the poor. China's dramatic success in improvinghealth conditions--lifeexpectancy rose from less than 40 years in 1950to 69 years by 1982--wasaccompanied by two related but less frequentlynoted achievements:

* By 1975, insurancecoverage and the rural CooperativeMedical System (CMS) reached close to 90 percent of the population--almostall the urban populationand 85 percentof the rural. China's citizens thus had reasonableaccess to a broad range of cost-effective preventiveand curativehealth serviceswhile sharingthe risks of medically-caused financial misfortune.

* The systemfor financeand deliveryof health servicescontained costs. Health care costs were held by 1981to just over 3 percentof GDP, despite the remarkablegains in health status and insurancecoverage.

2. Beginningin 1978,the Chinesegovernment introduced radically new economicpolicies that moved China away from a centrallyplanned economyand towarda competitivemarket system. This change in economicpolicies was accompaniedby devolutionof power from the central governmentto provincialgovernments. Many of these changeshad profoundrepercussions for the health system. For example,the transitionfrom agriculturalcollectives to the agriculturalhousehold responsibilitysystem weakened the financialbase of the CooperativeMedical System in rural areas. Some changes were also introducedto the health sector:the governmentgave health programsand facilitiesa great deal of financial independenceand encouragedthem to rely on user fees to support their operations. However,the governmentcontinues to administermany input and output prices in the health sector, the majority set below cost, and to control staffing in publicfacilities.

3. China's health sector is facing deep problemstoday as measuredby four dimensionsof health sector performance: health status, access to health care, efficiency,and total costs. Some of these problems are commonto many countries. Others stem from, or are exacerbatedby, the Chinesegovernment's failureto reformulatehealth financeand to redefineits roles in health. China needs to act now to correctthese problems,before they deepen and becomemore entrenched. Correctiveaction will require high-leveland joint inter-ministerialaction. Health is a sector that cannot simply be left to market forces.

Problemsin RecentHealth SectorPerformance

4. Health status. While China's overallhealth status is excellentcompared to other countries of similar income level, recent trends in child mortalityare a cause for concern. After falling steadily for 40 years, China's mortalityand under-fivemortality rates appear to have leveled off in the mid-1980s. Figure I shows trends in under-fivemortality in China for the period 1960-91, with projectionsto 1995. The graph indicatesthat the steady declinein the mortalityrate from 1960 ended by 1985. Fromthis point on, the under-fivemortality rate apparentlyremained stagnant, or

xiii even rose slightly, to a level in 1991 of about 45 per 1000. (For comparison, the under-five mortality rate is about 22 in Sri Lanka, 36 in Thailand, 111 in Indonesia, and 130 in India.)

Figure 1. Estimatesof Under-FiveMortality Rates in China, 1960-1991and Projectionsto 1995 200

Under-5 150 Mortality Rate 100 per 1,000 50

1960 1965 1970 1975 1980 1985 1990 1995 Year Source: See Annex4. Data for 1992-95are projections.

5. Mortality rates are, of course, determined by many factors, including income and education. China performed as well or better than other developing countries in terms of the rate of decline of under-five mortality relative to the rate of income growth before 1985. This positive performance relative to other countries sharply reversed during the period 1985-90, despite continued strong increases in China's real income per capita of 6 percent per year. However, even as per capita incomes rose overall, the incidence of rural poverty--which fell from 33 percent to 12 percent between 1978 and 1985--stopped declining in the second half of the 1980s and subsequent declines have been slower than during the early years of reform. It is unclear how much of the stagnation in child health from 1985 onwards is due to the decline in the availability of basic health services in rural areas following the dismantling of the CMS, to the persistence of rural poverty, or to other factors. What is clear is that China's performance in child health improvement since 1985 has been disappointing.

6. Access. Access to health care in China is inequitable, with especially deep divisions between the urban and rural population. For China as a whole, health spending per capita (public and private) was estimated at 110 yuan per year, or US$13.50 in 1993. (Note that health spending in purchasing power parity terms would be 4.8 times higher because of international price level differentials.) But the average health spending of 235 yuan per capita in urban areas was almost four times the average of 60 yuan per capita in rural areas. And the poorest quartile of the rural population accounted for only about 4 percent of all health spending in 1993. Only 10 percent of the rural population is insured compared to 50 percent of the urban population. While the two urban insurance systems--the Government Employees Health Insurance System (GIS) and the Labor Health Insurance System (LIS)--cover only 15 percent of China's total population, they absorb two-thirds of all public spending on health, and 36 percent of total health spending (Figure 2).

7. Health insurance coverage has declined rapidly in rural areas since the late 1970s due in part to new agricultural policies making more autonomous economic agents and in part to government policies that discouraged reliance on the rural Cooperative Medical System (CMS). As

xiv communes gradually disappeared, so did the CMS that the communes had supported. Only about 10 percent of the rural population is now covered by some form of community-financed health care, down from a peak of about 85 percent in 1975. There is also much variation in coverage across provinces depending on interpretation of national policy at the provincial level. The decline of the CMS has led to the loss of prepayment or insurance coverage by perhaps 700 million rural Chinese, requiring them to pay out-of-pocket for virtually all health services. Without insurance, medical expenses can lead to deferral of care, untreated illness, financial catastrophe, and poverty.

8. Efficiency. Facility duplication and excess capacity from some vertical national programs result in inefficiency and waste. In urban areas, the overlap involves Ministry of Health facilities, state-owned enterprise facilities, and traditional Chinese medicine facilities. In rural areas, there is growing duplication and overlap of services between maternal-child health centers, family planning services, township health centers, and epidemic prevention stations.

9. Public spending on health is skewed towards hospitals, while priority public health programs are increasingly underfunded. And with fiscal decentralization, the poorest counties are least able to finance public health programs. As a result of funding difficulties, public health workers have been diverted from important public health work, such as immunizations and disease surveillance, to activities such as routine testing of water and food in urban areas, for which they can more easily charge fees. In many parts of the country, the Epidemic Prevention Service (EPS) is now even charging for immunizations and tuberculosis treatment. This has reduced coverage and--in the case of tuberculosis treatment--led to medically inappropriate but profitable patterns of care.

10. Prices of most health services and many inputs to the health sector are fixed well below costs under guidelines issued by the Price Commission. Health care providers inappropriately promote certain profitable items--especially pharmaceuticals and high-technology diagnostic tests--in order to generate profits and cross-subsidize under-priced products and services. This leads to misallocation of spending, medically inappropriate services, and upward pressure on overall health expenditures in both rural and urban areas. Given the incentive structure, it is not surprising to find that pharmaceuticals account for a remarkably high share of aggregate health spending--52 percent in 1993.

11. Total Costs. Total health expenditures per capita grew by 8 percent per year in real terms from 1978 to 1986, accelerating to 11 percent per year from 1986 to 1993. Health expenditures now account for about 3.8 percent of GDP. Over the same time period GDP per capita grew by 7.7 percent per year in real terms. Health spending will continue to grow in real terms as China's income grows, but this growth is likely to be accelerated by China's price distortions in the health sector and heavy reliance on fee-for-service provider payment methods--particularly fee-for-service under third party insurance systems like the GIS and LIS. Spending growth in the GIS and LIS systems is simply not sustainable.

12. The aging of the population will increase health care costs, because the elderly have higher health costs than the young. China's population aged 55 and over now makes up 13 percent of the population, and will reach 18 percent by the year 2010. While the overall aging of the population is inevitable, government policies can influence how efficiently the health care system is prepared to address the elderly's health care needs. Further, effective health promotion and disease prevention programs implemented now--particularly to control tobacco use--will improve the health outlook for China's elderly.

xv Figure 2. Insurance Status and Sources of Health Finance, 1993

InsuranceStatus by Population Subgroups, 1993 (as percent of total population)

URBANUNINSURED 15% GIS COVERED TOTAL (Government Insurance INSURED System) 2% 21%

LIS COVERED RURALUNINSURED (LaborInsurance 64% ~~~~~~~~~~~~System) 64% ~~~~~~~~~~~12%

TOTAL RURALINSURED UNINSURED 7% 79%

Health Expendituresby Sources of Finance,1993 (as percent of total spending)

OTHER TTLPBI (publicand private) TOTAL PUBLIC 6% GOVERNMENT 55% RURALCMS BUDGET (largelypublic) % 2%_ GOVERNMENT INSURANCESYSTEM RURALOUT-OF- (GIS) POCKET 9% (private) 26% _ V

TOTAL PRIVATE 45% LABORINSURANCE URBANOUT-OF- SYSTEM(LIS) POCKET 27% (private) 16%

Total1993 Health Expenditures = 132 billion yuan PerCapita 1993 Health Expenditure = 110 yuan (60 yuan in ruralareas and 235 yuan in urbanareas) Source:Wei 1995.

xvi 13. Another cost to the economy of the present health financing and delivery system is the economic distortions that result from the current urban insurance systems. Because health coverage is tied to the employer--the government or state-owned enterprise--workers cannot retain their social benefits if they move from one job to another. Reforms, therefore, are needed so that workers can transfer jobs without jeopardizing their health (and pension) benefits.

14. In short, despite the remarkable early successes of China's health policy, achievements in child mortality, access, efficiency, and cost containment appear to have been either stagnating or eroding for over a decade. The Chinese government has reached consensus that these important concerns must be addressed by strong policy initiatives. The rest of this summary presents recommendations for dealing with these issues, following the same order as the chapters in the report. Finally, the implications of the recommendations in this report for public sector expenditures are discussed.

REPORT RECOMMENDATIONS

1. Strengthen Public Health Programs (Chapter 3)

15. Beginning with the founding of the People's Republic in 1949, China has complemented the development of local health services with a series of strong national programs for high priority public health activities, including disease surveillance, environmental monitoring and improvement, mass immunization, and health education. The government also supported treatment of infectious diseases, implicitly recognizing the large social benefits from infectious disease control. This was an appropriate role for government, since most public health programs provide services for which there are large social benefits, but for which individuals are less willing to pay on their own.

16. Three interrelated problems increasingly limit the effectiveness, scope, and population coverage of China's national public health programs, however. The first is simply that budgetary pressures constrain the operation and efficiency of existing programs, even though resource requirements are modest in the context of overall health expenditures. Part of this is due to the fact that almost all spending on public health is from provincial and local governments, and the poorest areas--which experience the most severe public health problems--have the least capacity to finance these programs. The second problem, related to the first, is that cost recovery policies constrain the demand for certain public health services (e.g. immunization), particularly among the poor. Third, the general movement toward fee-for-service payment has diverted the energies of public health providers to those activities for which fees can most easily be charged, instead of the highest priority activities for public health.

17. China needs to return to policies that vigorously finance and support public health, recognizing that these services must be financed by the govemment if they are to be provided at socially optimal levels. Particular attention needs to be given to reaching the unregistered urban population with public health programs. This report recommends that the Epidemic Prevention Service's budget of 1.3 billion yuan in 1993 be increased to at least 6.5 billion by the year 2001 and that the EPS be prohibited from charging user fees for most of its services. The budgets of other agencies carrying out priority public health activities also need additional support. At the same time, the government must ensure that public health programs are implemented efficiently and that China's highly effective disease surveillance system is maintained and adapted to the changing pattern of disease burden.

xvii 18. Nearly one million Chinese die each year of smoking-related diseases. Given the looming problem of a massive burden of costly illness and premature death from tobacco-related diseases (tobacco-related deaths are projected to increase to over two million per year by the year 2020), another recommended public health action would be to increase tobacco taxes substantially, accompanied by other measures to reduce smoking. This would help reduce eventual illness and death from smoking-related diseases and, if the incremental revenue were applied to public health (as in Australia), would help to finance public health programs adequately.

2. Ensure Essential Health Services for the Poor (Chapter 4)

19. The second priority for government health spending should be to ensure that the country's neediest citizens have access to priority health services. China's absolute poor reside almost entirely in rural areas. They are more likely to suffer from ill health, and their health problems can contribute to keeping them in poverty. Among the poorest quarter of the rural population, the infant mortality rate is 3.5 times greater than among city dwellers. The urban poor, especially unregistered migrants, also face relatively high health risks and need to be reached more effectively. There is strong justification on poverty assistance (or equity) grounds for government subsidies aimed at improving the poor's access to important health services. The Chinese government's role at present is minimal in this area. Resources need to be redirected or expanded to assure key health services for the poor. At the same time, given scarce public resources, subsidies need to be carefully targeted.

20. There are several ways to target the poor:

- Geographical targeting to areas where the poor are concentrated. For example, poor administrative villages in China's 592 officially designated poor counties could be targeted for subsidized services. This population totals about 75 million.

e Individual or household targeting, by identifying and certifying the poor for subsidized services. (China may be one of the only developing countries in the world where this targeting method is actually feasible, given good government records, but it entails heavy administrative costs);

* Program targeting to health services that particularly benefit the poor in both rural and urban areas, such as deworming and management of acute in children. (This would require subsidies in addition to those for the public health programs discussed in the previous section).

This report recommends phasing in a blend of geographic targeting (probably most practical at the village level) and universal finance of a limited number of services that particularly benefit the poor. It also recommends monitoring these approaches to guide policy improvements over time.

3. Reform Prices and Provider Payment Mechanisms (Chapters 5, 6 and 7)

21. Price distortions and irrational allocation of existing health resources have diminished the quality and effectiveness of China's existing health services. Prices for most medical services are set by the government at levels well below cost. To offset losses on basic services, the government has allowed high prices to be charged for drugs and high technology equipment. This results in a distorted pattern of services with overprovision of some services, such as CAT scans and ultrasounds, and underprovision of other services, especially those with a high labor content.

xviii Providers have strong incentives to overprescribe drugs of ail kinds, especially expensive drugs, in order to bring in additional income. Fee-for-service provider payment methods also encourage overprovision of services. Outpatients, for example, are frequently treated with intravenous drip solutions of glucose, vitamins, antibiotics, and other drugs. In almost all cases, this does not constitute justified medical practice. Given the incentive structure, it is not surprising to find that spending on drugs accounts for over half of all health spending in China, compared with 5 to 20 percent in OECD countries and 15 to 40 percent in most developing countries. Government action is needed for a variety of efforts to promote price and provider payment reform. These efforts will help to contain costs, reduce waste, and improve overall efficiency.

22. Price reform is urgently needed to bring administered prices in the health sector in line with marginal costs. A major study of pricing is needed to lay the basis for reform. Price and related provider payment reform would likely need to be implemented gradually over several years to be politically acceptable. Price reforrn in the health sector is an integral part of addressing an economy- wide problem of price distortions.

23. China also needs to move away from its dependence on fee-for-service provider reimbursement. Unconstrained fee-for-service reimbursement promotes the excessive use of such services because consumers rely on providers to recommend needed services, while providers, in turn, have a financial incentive to increase the volume of services. Experience from other countries shows that case-based reimbursements and various partial and full capitation payments can help contain costs and improve quality. Alternatively, fee-for-service payment methods can be used under a global budget constraint, or a mix of provider payment approaches can be used to improve incentive structures.

24. Options for provider payment reform, of course, hinge fundamentally on what types of risk- pooling arrangements are developed in rural and urban areas. One approach to provider payment would be to move, initially under urban insurance and community-financing schemes, from fee-for- service toward more aggregated products, as has already been tested in Zhenjiang, and finally to prepayment for a complete package of services. Under such a system, the provider would assume more risk and would have a decreasing incentive to over-provide services.

4. Control Investments and Improve Regional Planning (Chapter 5)

25. Governments can play an important role in cost containment through exerting oversight and controls on major human and capital investments in the health system. China is moving away from a centrally planned economy, but some oversight needs to be retained over new investments (across the many delivery systems in China) in hospital beds (especially at the tertiary level), expensive medical equipment, and the mix and numbers of health personnel. Experience from other countries shows that, once excess supply is created in any of these areas, the problem is politically difficult to correct. This excess capacity also drives up spending due to supplier-induced demand.

26. Government supply-side controls could be complemented by efforts to improve regional planning in health. China's health system is plagued by poorly coordinated vertical delivery systems in both rural and urban areas. Regional planning efforts should involve all relevant actors, including GIS, LIS, traditional Chinese medicine facilities, medical schools, and the MOH. As urban insurance centers are expanded, they would also become major stakeholders in regional planning efforts. China can build on experiences from planning exercises already underway in Baoji in Shaanxi, Jiujiang in Jiangxi, and Jinhua in Zhejiang provinces.

xix 5. Promote Efficient Risk Pooling in Rural and Urban Areas (Chapters 6 and 7)

27. In the poorest countries of the world, more than half of all health spending comes from private sources, mainly in the form of out-of-pocket expenditures. Within these countries, the poor purchase most of their services themselves. As country incomes rise, there is a remarkably uniform trend around the world: the share of out-of-pocket health spending gradually falls as prepaid risk- pooling mechanisms develop. Without risk-pooling mechanisms in place, catastrophic health expenses are a major hardship for both the poor and nonpoor. Government support to encourage the development of risk-pooling mechanisms can make health services more accessible and efficient. Risk-pooling mechanisms can be financed by general tax revenue, social health insurance (mandated payroll taxes), private voluntary insurance, or community financing (a package of services funded from a combination of households, local communities, and the government, which are organized and managed by the local community).

28. By the 1970s, China had risk-pooling mechanisms in place for a remarkably high percentage of its population relative to its still very low income level. With the widespread dismantling of the CMS in the 1980s, however, China has become much more like the rest of the low-income world in this area: perhaps 700 million rural people in China have lost their access to prepaid care and are now paying out of pocket for almost all of their health care. As incomes rise, they are seeking to regain such coverage.

29. Urban populations still have relatively high health insurance coverage but the two urban insurance systems, the GIS and the LIS, are in urgent need of reform. The GIS now covers about 30 million people, including current government workers, government retirees, the military, and university students. It spends almost four times the average health spending in China--389 yuan per covered member, compared to the national average of 110 yuan. The LIS covers an estimated 140 million people who are employees and retirees of state-owned enterprises (SOEs). The LIS currently spends about 259 yuan per covered member.

30. Rapid cost escalation in the GIS and LIS has led to a fiscal crisis in both systems. Government spending on the GIS system grew by 15 percent a year in real terms from 1978 to 1993. In recent years, the government has had to allocate additional funds to cover deficits. In the LIS system, some enterprises have been unable to cover the health care bills of their employees and retirees. Since the systems are pay-as-you-go financed, state-owned enterprises (SOEs) and government units with large numbers of retirees have particularly high costs.

31. Both the GIS and LIS systems also have major inefficiencies. Except for dependents in the LIS system, enrollees do not make significant copayments and therefore have few financial constraints on their consumption of medical services. In addition, the care that is not provided directly is reimbursed on a fee-for-service basis, resulting in incentives for overprovision of services. Finally, and most importantly, health insurance coverage is tied to place of work and therefore impedes the labor mobility essential for a modem economy. For a Chinese worker the loss of a position in a state-owned enterprise means loss of health benefits, housing, and pension rights.

32. In both rural and urban areas, the govemment can play an important role by promoting an appropriate policy framework for the development of efficient and equitable risk-pooling, or health insurance, mechanisms. The government also needs to monitor and systematically evaluate new approaches to rural and urban insurance in order to adapt and improve the policy framework.

xx 33. Rural reform options. Because of public finance constraints, a health system for the rural population financed largely from general revenue does not appear feasible over the medium term in China. (Government subsidies at present cover only a small fraction of publicly provided services: about 85 percent of costs are recouped from fees.) Mandatory wage taxes are also not feasible given that the majority of the rural population are self-employed. Community financing appears to be the most promising way to promote universal, or near-universal, health coverage in local communities and efficient service delivery, without being a major drain on government funds. China has considerable experience with community approaches to rural health insurance, beginning with the commune- based CMS system, and including a number of ongoing community-financing schemes that cover about 10 percent of the rural population today. A recent study of thirty poor Chinese counties showed that villages with community financing (compared to those without) are associated with higher use of health services at lower level facilities, lower rates of morbidity, lower fees for primary care services, and a lower proportion of income from drug sales in township health centers and county hospitals (see China Network 1996, Jin 1995a). To the extent possible, coverage should be universal at the local level.

34. China's experience suggests that, with appropriate government commitment, community financing is likely to be both administratively and financially feasible in many rural areas. Community financing has many advantages over the private, voluntary insurance that conceivably might develop in the richest rural areas. Private, voluntary insurance would exclude both the poor and individuals with high health risks and, if based primarily on fee-for-service payments, would also lead to cost escalation. And community financing has many advantages over the present system, in which rural residents must pay out-of-pocket for services on a fee-for-service basis. The government would need to monitor community-financing approaches in different areas, however, in order to evaluate their impact on health spending, efficiency, equity, and consumer satisfaction to inform medium-to long-term policy choices for rural health. Some of the key design features to be evaluated include the content of the basic benefits package, the size of the risk pool, different management models, reimbursement methods for individual doctors, township health centers, and county hospitals, the level of copayments by type of service and facility, and alternative measures to rationalize pharmaceutical use.

35. Urban reform options. Short- and-medium-term measures are urgently needed to reform the GIS and LIS insurance systems.

* Management of insurance needs to be taken out of the SOEs and government units in order to achieve larger risk pools and economies of scale in administration.

* Health insurance needs to be portable so that workers can move from one job to another without losing social benefits.

* The system needs to move from a pay-as-you-go system to one in which contributions allow for expected expenditures in old age. (Or, alternatively, other mechanisms need to be developed to cover workers in old age).

* Benefit packages need to be redesigned to be financially sustainable.

* Provider payment needs to move from fee-for-service to mechanisms that help contain costs, such as capitated payment methods.

xxi 36. The State Council has sponsored experimental health insurance centers that are testing key reforms of the GIS and LIS. Started in December, 1994, experiments in the cities of Jiujiang and Zhenjiang provide for wage-based enterprise/employee contributions into individual and pooled accounts managed by the insurance centers. The State Council decided to expand these experiments to fifty more cities throughout China in 1996. All of these experimental health insurance programs need to be carefully monitored and evaluated in order to inform policy-making. Government financing is needed for technical assistance, monitoring and evaluation, and strengthening regulatory capacity. Many issues still remain to be addressed, however, including how to pay for the unfunded health obligations of GIS and LIS workers (for example, for retirees of SOEs that will close) and what to do with SOE health facilities.

37. Over the medium- to long-term, urban health insurance coverage will also need to be broadened to include the increasing proportion of workers outside the state-owned sector and their dependents, as well as the rapidly growing unregistered or "floating" urban population. Urban insurance centers would, over time, need to harmonize benefits and contributions and portability across different areas of China.

Implications for Public Finance (Chapter 8)

38. Because of the erosion in public health programs and the unmet demand for health services outside the insured population, and especially among the rural poor, the central government should consider a substantial increase in its health spending for the purposes discussed above. In terms of broad direction, this report argues that spending on strengthened national public health activities should increase as a share of GDP by 2001. Public spending is also needed for new programs to provide essential services for the poor and to promote risk pooling in rural and urban areas.

39. China can afford these priority programs. Moreover, anticipated improvements in China's tax revenue performance will make them even more affordable. Today, China's budgetary expenditures as a share of GDP are far below those of other countries. Government budgetary expenditures were 14.1 percent of GDP in 1994, and extra-budgetary government expenditures were an additional 3.8 percent. Central government expenditures as a share of all government expenditures are also unusually low--only 40 percent compared to an average of 78 percent in other developing countries. According to World Bank projections, China's GDP will double between 1993 and 2001, from about 3,450 billion yuan, to 7,500 billion yuan, in 1993 prices. The World Bank also projects that government revenues, as a share of GDP, will rise significantly during the Ninth Five-Year Plan (1996-2000) if China implements suggested changes in tax administration and tax structure. With these changes, tax revenues should increase by 6 percent of GDP. The priority health programs recommended in this report might cost on the order of 13 billion yuan by the year 2001, which would be less than one percent of projected government revenue. Most of this spending--90 percent--would be for public health activities and subsidized services for the poor.

40. One related recommendation involves increasing the tax on tobacco by 20 percent or more. A 20 percent increase in the tobacco tax is projected to generate 10 billion yuan in incremental revenue annually. By decreasing smoking, this tax would produce important health benefits as well as revenues. These incremental revenues might also be used to help finance the public spending increases recommended in this report.

41. Experience from the high-income countries suggests that China may now be at a critical juncture for redirecting its health policies. If China adopts the package of policy reforms

xxii recommended in this report, it could expect, within ten to fifteen years, to achieve much higher levels of prepaid health coverage, to eliminate most of the excessive disease burden among the poor, to have maintained general improvements in life expectancy, and to have stabilized health expenditures at 5 to 7 percent of GDP (just below the range for OECD countries). Failure to adopt these policies would risk leaving a large proportion of the population without health insurance, jeopardize health indicator improvements, and encourage growth in health expenditures to 10 percent of GDP or more (as in Argentina, France, and the U.S.A.).

42. The choices that China makes in its health financing policy in the coming years will not rest only on financial and economic analyses. These choices hinge fundamentally on judgments about what kind of society China wishes to be and what value it places on social cohesion, poverty alleviation, equity, consumer choice, and quality of care. This report argues that achieving these broader social goals can be combined, in the right policy environment, with measures that improve economic efficiency in the health sector.

xxiii Table 8.5. Summary of Recommended Health Policy Actions, 1996-2001

Objectivesand Short-Term(1 to 2 Years) Medium-Term(3 to 5 Years) Programs

1. Strengthenpublic Providefull fundingfor the currentEPS. Increasefunding for public health to health programs ProhibitEPS from collectinguser fees for about 6.5 billion yuan by 2001. (Chapter3). publichealth services. Increasethe tobaccotax by 20 percentor Strengthenanti-tobacco programs, such more to reduce smoking. Earmarkthe as health education,prohibition of tax revenueto fundpublic health smoking in publicplaces, and regulations activitiesand health servicesfor the poor. on levelsof tar and nicotinein tobacco products. Upgradethe skillsof staff in EPS and other agenciesinvolved in public health Developa strategicplan for public health, to carry out their new mandate. given China's changingpattern of disease burden and risk patterns,to guidethe Ensurethat prioritypublic health work of EPS and other agenciescarrying programsreach the poor in urban areas, out public health functions. particularlythe unregisteredurban poor.

2. Ensure essential Phase in a programof geographically- Continueto subsidizeservices for the health services for the targeted subsidiesfor health servicesfor poor and expand coverage. poor (Chapter4). the populationin poor administrative villagesin China's 592 officially- Monitorimpact on the poor to guide designatedpoor counties. If the policy improvementsover time. administrativevillage has a viable community-financingscheme, channel Over the long term, when more the subsidiesthrough the scheme; comprehensiveurban insurancesystems otherwisedirect them to health providers. are set up, the governmentmight consider subsidizingthe poor's contributionto Phase in program-targetedsubsidies for a such schemes. limitednumber of health programsthat particularlybenefit the poor.

3. Reformprices and Carryout a major study on price reform, Graduallybring prices more in line with providerpayment focusingboth on prices of individual marginalcosts, with periodicupdates to mechanisms(Chapters health servicesand ways to bundle accountfor inflation. 5, 6 and 7). servicesinto packaged fees (this relates to reformsof the GIS and LIS, below).

Move awayfrom the relianceon fee-for- Assist rural communityfinancing plans in serviceprovider payment methods. establishingprovider payment These paymentmethods encourage mechanismsthat encourageefficiency overprovisionand drive up costs. (such as salarieswith performance bonusesfor village doctors,and Experimentwith and evaluatealternative capitationpayments to county hospitals). providerpayment methods that encourageefficiency in urban insurance experiments.

xxiv Objectives and Short-Term (I to 2 Years) Medium-Term (3 to 5 Years) Programs

4. Control investments Improve and expand regional planning Institute nationwide regional planning and improve regional techniques already developed under guidelines. planning (Chapter 5). World Bank Health Loan III to better integrate various hospital systems. Develop government oversight mechanisms to avoid creating Change government's role in manpower oversupplies of physicians, particularly from central management of staffing of specialist physicians, tertiary hospital individual health facilities to more macro beds, and expensive diagnostic manpower planning nation-wide and equipment. province-wide to support the broad health reforms proposed here.

5. Promote efficient Rural areas: risk pooling in rural and urban areas Develop national guidelines for the Provide training, technical assistance, and (Chapters 6 and 7). operation of community financing that modest central government subsidies to address the size of the risk-pooling base, rural communities that establish management structure, and community community health financing. control mechanisms. Monitor impact of alternative forms of Organize provincial technical assistance community financing on health status, teams to assist community programs. access, organization and delivery of services, pharmaceutical use, health Implement pilot programs. spending, etc.

Urban areas:

Extend urban insurance experiments to Gradually extend social health insurance other provinces. Systematically evaluate to cover workers employed by joint these additional experiments. Some of ventures, smaller collective industries, the start-up costs of new schemes (such and private enterprises. Mandate open as information systems, capital access on a voluntary basis to all urban equipment, training) might be financed residents not covered by their employers. with government assistance. The government could eventually finance participation by the indigent in insurance.

Experiment with risk pooling in larger areas such as the provincial level.

Reform provider payment methods. Insurance Fund Centers could move toward negotiating capitation contracts with providers covering all levels of services, under risk-adjusted capitation rates. Alternatively, they could move towards DRG-type payments for hospital services, based on reasonable cost, with periodic adjustments as needed.

xxv

1. INTRODUCTION: HEALTH AND HEALTH POLICY IN CHINA

1.1 A nation's healthpolicies affect directlyboth the health of the populationand the operationof the health care system. A growingbody of evidenceand analysis from around the world allows individualcountries to learn from the policiesof others. This chapter sets the analytical frameworkfor evaluatinghealth financepolicies in China by assessingthe performanceof China's health sector and examininghealth policy issues raised by this performance.

1.2 To assess China's health sector, four broadmeasures of performanceare used in this chapter. The first measure is the health status of China'spopulation. Overall China's health status has improvedenormously since 1949. A good measure of this is life expectancy: A typical country of China's incomelevel in 1990achieved a life expectancyof about 64 years, whereasChina achieved69 years. As will be discussedin detail below,however, other importanthealth status indicatorsin China appear to have stagnatedsince the early 1980s. The next three performancemeasures relate to the financingand provisionof health services: access to health services, efficiency,and total costs. Ensuringbroad access to health services helps ensurethat health care is deliveredequitably. This is a desirablegoal best achievedthrough risk-sharing (insurance) mechanisms that provide prepaid coveragefor a reasonablerange of services. The percentageof a populationcovered by such risk-sharing mechanismsis thereforean importantindicator of a health system's performance. Health policiesalso affect the efficiencyof care--inother words, whetherservices are producedat the least possiblecost and whether spendingis efficientlyallocated (producing value for money). Finally the total cost of the health resourcesused by a nation is another important policy outcome. Experiencearound the world suggeststhat spendingmore on health is not required for improvinghealth outcomes,since some high spendingcountries (such as the United States) get relativelylittle return for their resourcecommitments. Anotherdimension of costs is indirecteconomic losses that result from the health care system. For example, labor mobilitymay be impededby insurancesystems that tie the possibilityof health insurance coveragejust to certain employers. This report deals with policy issues and alternativesin all four of these areas.

1.3 It shouldfirst be noted, however,that health policy choices are not the only influenceson a nation's collectivehealth. Outsidefactors also have a major impacton a nation's health status and the operationof its health system. Income and educationlevels, for example,form the foundationupon whicha nation's health policiesare built. As income increases, so does the ability to acquirethe necessitiesfor good health--adequatefood intake, clean water and sanitation,satisfactory shelter, and accessto health services. Similarly,as educationlevels rise, so does the capacity of the populationto make informedchoices concerninghealth, dispositionof income,and personalbehavior.

1.4 Demographicchanges are another importantoutside influenceon the health sector. China's demographicpatterns have changed dramaticallyover the past four decades. Rapid declinesin fertility and mortality,the aging of the population,and the potential for explosive urban growth affect both health conditionsand planning for the evolutionof the health

I system and its finance. Annex 2 examines these and other outside influences on health status in greater detail.

Health Status: Progress and Problems

1.5 In 1984 the World Bank's first health sector report on China (World Bank 1984) called for completing the first Chinese health care revolution: extending successful programs for improving child health and control of endemic infections into poor rural areas; consolidating and deepening the health gains achieved in most of rural China by reversing the ongoing breakdown in cooperative medical services; and seeking new means to finance public health programs that were being neglected by local providers embracing fee-for- service practices. The report's other concern was to encourage a second revolution: preventing and managing the emerging epidemic of noncommunicable disease (NCD) at much lower cost than in the high-income countries, which preceded China in emergence of NCDs as dominant problems. The World Bank's second health sector report (World Bank 1992a) dealt much more extensively with problems of NCDs and their risk factors. This section reviews the status of these two health issues in China today.

Child Health and Infectious Disease

1.6 Despite rapid income growth in the past decade, China's earlier progress in improving child health appears, in the aggregate, to have come to a stop. Analyses from inside and outside China suggest that the infant mortality rate (IMR) stopped declining in about 1982 (World Bank 1992a, pp 6-7). A more recent overview (Parker n.d.) reaches a similar conclusion, pointing to a decline in immunization coverage beginning about 1988 and to recent unexpected outbreaks of immunizable diseases. Parker also presents evidence from surveys in nine provinces that a key indicator of child malnutrition--the percentage of children with very low height for their age--increased in rural areas between 1987 and 1992. Urban malnutrition, by contrast, declined sharply in those provinces during this period. These overall findings are even stronger in poor areas. The Study of Thirty Poor Counties found that the median infant mortality rate increased in those counties from about 50 to 72 deaths per 1,000 live births from the late 1970s to the late 1980s (China Network 1996 and Liu Yuanli and others, 1996).

1.7 Recently available census and survey data now permit a careful reassessment of trends in child mortality. This report commissioned a complete analysis of available data on national trends in the under-five mortality rate (U5MR). While this assessment could also have been undertaken for the infant mortality rate, demographers have concluded that under- five mortality estimates are consistently more robust and reliable than those for the IMR. UNICEF regards U5MR as the best indicator of social development because it accounts for the mother's health and level of knowledge, immunization levels, use of appropriate health services, access to water supplies, sanitation conditions, and the overall safety of the child's environment (UNICEF 1989, p. 82). Table 1. I compares China's results to those of other Asian countries; Annex 4 provides more detail on methods and results. The analysis concludes that U5MR in China declined steadily until the early 1980s and then began a slight upward drift.

1.8 Experience from other countries suggests that the under-five mortality rate need not plateau as China's has. Sri Lanka's per capita income is slightly higher than China's and its

2 Table 1.1. Trends in Under-FiveMortality Rates-China and Other Asian Countries, 1960-90 Year Under-Five Mortality Rate China Hongkong India Indonesia Japan Sri Lanka Vietnam 1960 173 53 235 214 37 140 105 1965 144 1970 115 1975 85 17 195 151 11 69 68 1980 60 1985 44 1990 44.5 7 127 111 6 22 46 Sources: China-- Annex4; OtherCountries -- WorldBank, WorldDevelopment Report 1993. Note: The under-Smortality rate is defined as the numberof childrenwho die betweenbirth and their fifth birthday,expressed here as the numberper thousandlive births.

1975 under-five mortality rate was moderately lower, but by 1990 its under-five mortality rate had dropped to half that of China's. Japan's IMR in 1951 was about the same as China's was in 1976, but it then dropped by a third in six years. Indeed, until 1951, the decline in Japan's IMR was remarkably similar to that which occurred in China 25 years later; Japan's IMR, however, continued to decline while China's decline soon stopped (Parker, n.d.).

1.9 This suggests that China's performance has deteriorated not only in absolute terms but also relative to other countries. To address this question, data from China and more than 80 other countries were analyzed to assess the relationship between changes in under-five mortality and changes in income. The actual rate of decline in China's U5MR was then compared to that which would be predicted based on the experience of the sample of countries. Figure 1.1 plots the results for each of the time periods into which the data were divided. China's best relative performance was in the period 1975-79; the under-five mortality rate declined a full 3 percent per year more quickly in China than would have been predicted. Within a decade, in 1985-89, the position reversed and China's U5MR was changing 3 percent per year less favorably than predicted. (Indeed China's U5MR was by then actually increasing slowly.)

1.10 What might explain this major and disturbing change? The large decline in China's performance in child health relative to other countries suggests that multiple factors are at work, including at least the following three possibilities:

* First, China's best performing period corresponds reasonably closely with China's period of rapid fertility decline, which stopped by the early 1980s. To the extent that the total fertility rate (TFR) decline reduced the number of closely spaced births (with their known risks for child health and survival), it would have contributed to China's good performance in the 1970s and early 1980s. The birth spacing effect, however, may be partially counterbalanced by the greater risk associated with first births.

3 Figure 1.1: Deviationof China'sAnnual Rate of Decline in Under-FiveMortality Rates from RatesPredicted by GrowthRates in Income

0.060 0.050 0.040 t 0.030 Deviation0.020 -1 from 0.010 predicted 1 I 0.0 1960-64 16-69 1980-84 1985-89 -0.020 197 7 -0.030-1907 _ -0.040 0.040 ~~~~~~1975-79

Source: Adaptedfrom Annex4. Note: Negativedeviations indicate that China's under-fivemortality rates were declining more rapidlythan would be predicted--inother words, China was doing well. Positive deviations(e.g. 1985-89)point to poor performance--thedecline was slowerthan would be expected.

* Second, China's success in the period 1978-85 in reducing the percentage of the population in absolute poverty (from 33 percent to about 12 percent) would have helped improve health conditions overall. The percentage in absolute poverty, however, decreased no further from 1985 to the early 1990s,. Although changes in the rate of poverty reduction almost certainly contributed, China's improvements in relative U5MR performance began earlier and started to decline earlier than the changes in poverty. Changes in poverty rates probably contributed to, rather than drove, China's relative performance in U5MR.

* Finally, China's child health performance over this entire period tracks closely with the rise and fall in the percentage of the rural population covered by the rural Cooperative Medical System (CMS). This system reached its peak coverage of about 85 percent in 1975, and declined rapidly after 1979 (see Chapter 6). The CMS did a good job of providing cost-effective interventions to address major causes of child mortality. Both statistical information (e.g. on immunization coverage) and anecdotal accounts suggest that China's public health function has suffered widely with the decline of the CMS, and its rise and fall probably account for some of the change in China's relative performance in reducing the U5MR.

4 1.11 The relative importanceof these various causeswill be clearer when analyses can be undertakenon a province-by-provincelevel. 1 The deteriorationof China's under-five mortalityrate relative to the rest of the world suggeststhe need for serious reconsiderationof how health policiescan be appropriatelyadapted to market-orientedreforms in the economy. This report explores optionsfor beginningthat process.

Noncommunicable Disease and Injury

1.12 The young and the old are afflicted by very differenthealth problems. Stroke, cancer, ischemicheart disease, and chronic lung disease accountfor most of mortalityin people of late-middleand older ages. Children,on the other hand, die from a relativelyshort list of infections,most of which are inexpensiveto treat or prevent. Figure 1.2 illustratesthe increasingimportance of noncommunicablediseases in China and other countries, with projectionsto 2020. As noncommunicablediseases increasetheir share of the disease burden in China, the contributionof injury is expectedto changevery little, while infectious diseasescontinue their steady declinein relative importance. Annex 3 providesmore detail.

1.13 Much of the projectedincrease in the relative importanceof noncommunicable disease results from unalterabledemographic changes. Part, however,comes from projected patterns of disease incidencethat are substantiallyhigher than they would be if patterns of risk were better controlled. Tobaccouse providesthe most importantexample. By 1990 tobacco use alreadyaccounted for about 800,000of China's 8.9 millionannual deaths; projectionsof the effects of past and future tobaccouse indicatethat over 2 million deaths related to tobacco use are likelyto occur in 2020. Tobacco-attributabledeaths will have almost tripled between 1990 and 2020 and their percentageof all deaths almost doubled unless tobacco use can be curtailed. (See Annex 3, TableA3.7).

1.14 It is possibleto devote enormousresources to interventionsfor managementof NCDs that are costly and of limitedefficacy. More cost-effectiveapproaches to the preventionand treatmentof NCDs may be neglectedunless the incentiveenvironment encouragesexperimentation with and adoptionof these approaches. This report pointsto options for improvingChina's approachto decisionsabout NCD preventionand management.

Performance of the Health Care System

1.15 A nation's health policies have a direct impact on the operation of the health care system as well as on the health of the population. This section highlights trends and policy issues associated with these non-health outcomes of health policy--access, efficiency, and costs of China's health care.

I China's DiseaseSurveillance Points data also provide informationon age-specificmortality rates. These data are presentedin Annex 5. The data, unfortunately,are only truly comparablefor the period 1990-94as the samplingbase was changingin earlieryears. Over this short period,no clear trend emergesin life expectancy,the probability of dyingbefore age one,the childmortality rate, or the adultmortality rate.

5 Figure 1.2. Distribution of Disease Burden by Broad Groups of Cause for China and Comparators, Estimates for 1990 and Projections to 2020

1990 2020

1 24g/ii

CHINA II 68%79

18%

2k 9 29% 24

INDIA 56%

, / / 15% s %

7% lil 5%

1:2% /~ R10% HIGH-INCOME COUNTRIES -i (Europe,US, 81% 85% Japan)

1: Communicable, Maternal, Perinatal, and Nutritional Conditions a II: Noncommunicable Diseases 0111:Injuries

Source: Murrayand Lopez(1996); See Annex3.

6 Issues of Access and Equity

1.16 The Chinese public has relatively good physical access to basic health care services. High population densities and a well developed health infrastructure make geographical barriers relatively modest in China, except for a significant minority living in mountainous or remote rural areas. Until recently, the government also made cost-effective public health services widely available, thus minimizing financial barriers to public health services. Finally, the cost of routine, basic outpatient health services is low enough that most nonpoor Chinese households are able to pay for such services out of their own current income or savings.

1.17 Catastrophic care poses more of a problem, however, because it involves services that are so expensive, relative to household income, that many households must either forego treatment or go deeply into debt to pay for them. As much as 70 percent of total health expenditures in many countries are for catastrophic care. An important performance indicator for a nation's health system, then, is how well it solves this problem of efficiently arranging for these services to be financed and provided. Given limited resources, what can be included under "catastrophic care" coverage will vary with a country's income level and health infrastructure.

1.18 Since only a fraction of the population needs catastrophic health care in any given year, pooling of risks provides the best mechanism for financing these services--either under a system financed by general tax revenue, social insurance, or private voluntary insurance. This is an issue both of equity--the poor will require subsidies--and of efficiency, in that all but the very wealthy (who can be self-insured) will generally benefit from risk pooling. Low-income countries typically lack the institutional and financial capacity to offer most citizens the advantages of risk-pooling arrangements. As development proceeds, however, risk pooling typically benefits an increasing percentage of the population--often, at least initially, through straightforward government or collective finance of clinics and hospitals open to everyone.

1.19 From the late 1960s through the early 1980s, China provided an exceptionally high percentage of its population relative to its income level with some form of risk pooling that guaranteed access to care. Figure 1.3 illustrates that only 29 percent of China's population in 1981 lacked access (i.e. were uninsured) but that by 1993 the uninsured proportion had grown to almost 80 percent. This change occurred largely as a result of fundamental economic changes in China's economy. The challenge is to reintroduce broad access to health care in the context of China's new economic environment.

Efficiency

1.20 The fact that total health expenditures have been growing rapidly in China, even as some key indicators of health status have stagnated or even begun to decline, suggests a mounting problem of inefficiency. The allocation of public spending favors less cost- effective hospital services over highly cost-effective public health activities. Distorted prices encourage the overuse of drugs and high technology tests. Fee-for-service payment encourages overprovision. Multiple vertical health delivery systems have led to excess capacity and waste. Reversal of these trends will require reallocation of resources, both

7 Figure 1.3. Insurance Status of China's Population, 1981 and 1993 1981

Government Employees' State-Owned Enterprise Insurance (GIS) Employees' Insurance TOTAL 2% (LIS) TOTAL ~~~~~~~~~12% UNINSURED 29% CollectiveIndustry Insurance Uninsured 5 29% Commune Industry Insurance 4%

TOTAL

Rural Insured INSURED 48% 71%

1993

Government Employees' Insurance (GIS) State-Owned Enterprise TOTAL 2% Employees'lnsurance INSURED Urban Uninsured (LIS) 21% 15% 12%

RuralInsured 7%

Rural Uninsured ;;1. t 64%

, 5 ! i,. if ' X' 2 )~~Mi. ~i X TOTAL UNINSURED 79%

Source: World Bank 1984 andWei 1996.

8 public and private. Risk-poolingand providercompensation arrangements need to be designedto contain costs, extend access, and promote attentionto value-for-moneyand quality of care.

Cost of Health Services

1.21 The cost to the economyof operatingthe health systemis anotherimportant performanceindicator. Demographicchange and economicgrowth virtually ensurethat health expendituresin China will grow as a percentageof GDP. Policy choices can, however,influence whether that expendituregrowth is excessive,and whetherit efficiently expandsaccess and improveshealth outcomes.

1.22 Figure 1.4 shows that somecountries have controlledhealth care costs much better than others. In 1960,Canada, Japan,the United Kingdom,and the United States wereall spending3-5 percentof national incomeon health,a share similarto what China spends today. But spendingrates divergedsharply over the next 30 years, with health spending reaching 14 percent of GDP in the United States by 1993--eventhough 15 percentof the populationstill is not insured. By contrast, Japan managedmuch more successfullyto contain spending.In 1993 Japan spent only 7.3 percentof GDP on health,with nearly universal coverageand the world's highest life expectancy.

Figure 1.4 Growth in Health Spendingas a Shareof GDP(in percent),Selected Countries, 1960-93

16 14- g' 12 -

2L910 v 0 g! 8 s 6 4 2 0 1960 1970 1980 1990 Year

Source:OECD Health Data 1995.

9 1.23 Perhaps the most important factor outside the health sector that leads to rising costs is the aging of the population and accompanying epidemiologic changes. The World Bank has concluded that the agingfactor alone would increase health expenditures from about 3.2 percent of GDP in 1992 to 5 percent in 2010 and 7 percent by 2030 (World Bank 1992a).

1.24 Factors within the health sector influence costs even more, however. China's health expenditures grew rapidly between 1978 and 1993--averaging 10.9 percent per year over the full 15-year period. Factors placing upward pressure on expenditures include a worldwide tendency for the utilization of health services to rise faster than income, and the transition from salaried compensation of providers to fee-for-service compensation. Distorted prices and high profitability from drugs and high technology diagnostic tests (discussed in detail in Chapter 5) also put upward pressure on health care costs in China.

1.25 During the same 15-year period, risk-pooling arrangements in China underwent significant changes as shown in Figure 1.3. Risk pooling in rural areas declined sharply after 1978, from 48 percent in 1981 to 7 percent in 1993. Under these conditions, the willingness and ability of patients to pay personally for services limits providers' capacity to overprescribe or even to provide genuinely necessary services. No such constraint operates under existing urban insurance programs such as the Government Health Insurance System, or GIS. Under the GIS, a "third party" (the government) reimburses providers for essentially all the patient procedures that GIS agrees with. Consequences of these different rural and urban incentive regimes were clear: Between 1981 and 1993 per capita expenditures in rural areas increased from 21 yuan per capita to 60 while GIS expenditures increased much more rapidly, from 96 yuan per capita to 389. As this report will discuss, there are many ways to achieve the efficiency gains of wide risk pooling without creating incentives that lead to excessive cost escalation. The GIS, however, embodies incentive arrangements of the type that have led to excessive growth in costs in other countries--such as Korea, Singapore, and the United States. And, as incomes grow in rural China, the demand for risk pooling and prepaid care will grow apace. Reliance on out-of-pocket financing to keep costs down is not only undesirable because of the efficiency losses but also impractical in the face of probable demand for prepaid arrangements.

1.26 Over the medium term, health-financing policies also need to minimize indirect costs to the economy (principally distortions in the labor market) by, for example, separating the provision of health services from .

Conclusion

1.27 China faces some disturbing trends and challenging policy issues in all four of the health sector performance areas reviewed in this chapter. There is stagnation (and perhaps even deterioration) in the child mortality rate. Insurance coverage and access to care have markedly declined. At the same time, real expenditures per capita have increased by a factor of more than 2.5 in rural areas and by as much as 4 times in urban areas since 1978. Incentives for preventive care have eroded while incentives (and opportunities) for providing excessive or inappropriate care have multiplied. And initiatives to undertake the collective action required for efficient risk pooling have not been widely or consistently implemented.

1.28 Clearly, major economic reforms begun in China in the late 1970s have brought rapid economic growth, but they have also had unintended and sometimes detrimental

10 consequencesfor the health sector. This report lays out options for adaptinghealth policy to China's new macroeconomicenvironment. The followingchapters recommendspecific policy measuresto deal with each of these challenges. Table 1.2 providesa previewof these measures,and shows the impactthat each would have on the health sector issues outlined above.

11 Table1.2 RecommendedHealth Finance Policies: Impact on the Health Sector

POTENTIALIMPACT ON HEALTH SECTOR PERFORMANCE

REPORTRECOMMENDATIONS Promoting Improving Accessand Improving Containing Health Equity Efficiency Costs Status

FullyFund Key Public Health ff f f t Programs SubsidizeEssential Health Services ft ft t forthe Poor ReformPrices and Provider Payment ft t Mechanisms ControlInvestments in Capital and ft ft Manpowerand Improve Regional Planning PromoteUniversal Risk Pooling ff

W2= nosignificant impact

E = moderateimpact

ft =strong impact

12 2. HEALTH SERVICES AND THEIR FINANCING

2.1 This chapter describesChina's current health deliverysystem: infrastructure,personnel, and the mechanismsused to finance services. It updatesearlier WorldBank health sector reports on China done in 1984and 1992 (World Bank 1984 and 1992a)and sets the stage for the analysis and recommendationsthat appear in subsequentchapters of this report.

Infrastructureand Personnel

2.2 To providehealth services to its populationof 1.2billion, China has some 200,000 health establishments,and a wide array of supportingresearch organizations. The country has 5.3 million health professionals,who make up about 0.8 percent of the total labor force. This includes 1.9 million doctors,or about 1.6 doctorsper thousandpeople, and I million nurses. Doctors are trained in one of three categories: junior doctors (19 percentof the total); senior doctors (62 percent);and doctorsoftraditional Chinese medicine(19 percent). Senior doctors are concentratedin medium-and large-sizedcities. Villagedoctors (also known as Barefoot Doctors) with much less training are excludedfrom these estimatesand categories.

2.3 China has some3 million hospitalbeds in operation,averaging 2.4 beds per thousand people. This is a higher per capita ratio than is found in other low-incomeregions of the world, includingAfrica and the rest of Asia, and nearly as high as in Latin Americaand the Caribbean region. Only the OECD and Eastern Europeancountries have higher bed-per-capitaratios. In 1994, China's bed-occupancyrate was 69 percent,and the averagelength of hospitalstay was 15 days (State StatisticalBureau 1994;World Bank 1993a;Ministry of Health 1993a.).

2.4 Three-quartersof China's healthspending goes to pay for inpatientor outpatienthospital care (Table 2.1). About 60 percentof hospitalspending is on pharmaceuticals.Most hospital facilitiesare part of the Ministryof Health (MOH) system, includingits provincialand county affiliates, or are operatedby state-ownedenterprises (SOEs). Others are run by village and townshipcollectives. China also has an estimated 161,000clinics operated by private practitioners in urban areas, although there is no officialestimate of the numbersof patientsthey serve.

2.5 More than half of China's health workersare employeesof the Ministryof Health or its provincial Health Bureaus. At the end of 1993,the Ministryof Health and its provincial affiliates employed 1.7 million hospitalworkers, 1.I million health workers in townshiphealth centers (THCs),250,000 workers in the Epidemic PreventionService (EPS), and a little under 100,000 in maternal and child health(MCH) programs(MOH 1995). MOH also finances the education of 220,000medical studentsenrolled at 120 medical schools. Thirty of these schools belong to the MOH;others belong to local or provincialgovernments.

2.6 China's SOEsemploy another 1.4 million healthworkers and operate 700,000hospital beds, both accountingfor roughlyone-quarter of the nationaltotal. Since the 1950s,SOEs have provided health services directlyto their employees,retirees, and their families. Most SOE health facilitiesare smallclinics and healthposts, but they includesome hospitals and large service centers as well. Some large SOE hospitalsand clinics servethird parties on a fee-for-service basis. The Taiyuan MachineryWorks in Shanxiprovince, for example,owns and operates a

13 Table 2.1. Sources and Uses of Health Financing, 1993 (millions of 1993 yuan)

Sources of Finance Government Eudget (excluding GIS) Insurance Society Finandng 2/ Out of ocket Uses of Health Traditional Finance Recurrent Chinese Community Govt. Other Total Expenditure 1/ Medicine Others GIS LIS Financing and NGOs Enterprises Village Private Other Urban Rural (Percerit) Hospitals Outpatient 1,912 349 1,266 3,737 737 1,276 5,828 9,459 24,564 Inpatient 2,571 445 2,569 7,583 139 0 1,096 1,170 15,573 Total 4,483 794 3,835 11,320 876 1,276 6,924 10,629 40,137 ______30.4% Pharmaceuticals Hospitals 7,639 22,552 1,367 10,812 17,844 60,214 Individual providers, 45.6% retailers 3,300 5,417 8,717 6.6% Testing 112 103 215 ______~~~~~~~~~~~~~~~~~~~~~~~~0.2%. Epidemic 1,305 137 195 1,604 3,241 Prevention Service 2.5% Maternal/Child 324 193 569 137 195 534 381 799 3,132 Health Program _ 2.4% Other Primary 1,344 78 610 2,032 Health Care 1.5% Family Planring 2.292 677 2,969 2.2% Medical Research 408 42 1,160 816 2,426 and Education 1.8% Construction 2,810 3 1,144 342 877 143 1,067 6,386 4.8% Other 2,579 2.579

Total 10,786 917 7,175 11,667 34,441 2,243 1,618 1,151 1,067 143 4,734 21,417 34,689 132,048 Percert 8.2% 0.7% 5.4% 8.8% 26.1% 1.7% 1.2% 0.9% 0.8% 0.1% 3.6% 16.2% 26.3% 100.0%

1/ Health recurrent budget includes Ministry of Health and local government public health departments. Most spending is at the local, not central, level. 2/ In the Chinese health accounts data, society financing can be either public or private finance. For example, under EPSand MCH, society financing from others' refers to user fees.

Source: Wei Ying, 1996. 300-bedhospital that receivesa quarter of its revenuefrom patients not affiliatedwith the enterprise. In 1993,China's SOE health facilitiesdelivered 18 percentof all outpatientand emergencycare in China, and 13 percent of all inpatienttreatment. The proposed reformsof SOEs would separatesuch services from the regular business of the enterprises. In some cases, however,the SOE may prefer to expand its medicalcare businessrather than to give it up. The TaiyuanMachinery Works plans to expand its hospitalto 500 beds, for example.

2.7 In addition to the health care infrastructureof the MOH and the SOE systems,thousands of health workers are employedby other governmentinstitutions such as the militaryand prison systems.In addition,there were some 150,000health workers in private practice in 1990,and an estimated 190,000in 1993 (excludingvillage doctors).

The Three-Tier Rural Delivery System

2.8 The rural three-quartersof the Chinesepopulation is served by a three-tiersystem of health services and referral. Normally, farmersand their families enter the system througha visit to one of China's approximately1.44 million villagedoctors. These health workersengage in both medical service and farming, and often earn as much from farming as medicine. Many of them received rudimentarytraining as Barefoot Doctorsin the 1960sand 1970sand continue working in the villagesof their birth. They work independentlyand are not countedamong the 5.2 million health personnel who work in China's medical institutions.

2.9 In this first tier of the system,village doctorsdiagnose and treat patients,prescribe pharmaceuticals,and refer patients to higher levels of service as warranted.Village doctors generallyoperate on a fee-for-servicebasis, but they also depend on the mark-upon pharmaceuticalprescriptions (typically about 15 percent) as a source of medical income.

2.10 The village doctor may also refer patients to the nearby townshiphealth center/hospital (THC). This second tier of rural health care consistsof some 52,000rural townshiphealth centers operating730,000 beds, or about one-quarterof all hospitalbeds in China. Township health centers are staffed by junior doctorsand other medical personnel,and can deliverbabies, treat infectionsand wounds,and providesome basic surgery such as appendectomies.The THCs depend substantiallyon patient fees to financetheir services,but local governmentsalso provide subsidiesthat cover part of their costs. These facilitiestend to have lowerbed-occupancy rates than the higher-levelhospitals. Some farmersreferred for hospitalattention bypass the township facility because they question its quality,and go directly to the county hospitals.

2.11 There are about 4,000 county hospitalsin China, which make up the third tier of the rural health care system. These hospitalsare usually the last point of referral for inpatienttreatment of rural residents, since few farmerscan afford to be treated at specializedbig-city hospitals. County hospitalshave on averageabout 300 beds. They typically have five departments: obstetrics and gynecology,, general surgery,internal medicine,and laboratories/X- rays, as well as emergencyroom facilities.

2.12 There are also three importantvertical public health servicesthat function independently of this three-tier systemin rural China. These are the EpidemicPrevention Service (EPS) and the Maternaland Child Health Program(MCH) under the MOH, and the FamilyPlanning and ReproductiveHealth Program(FP) under the FamilyPlanning Commission. All of these programsreceive budgetsfrom provincialand countygovernments and also collect fees from

15 their clientele. User fees supplement the public funding to a greater or lesser degree depending on economic conditions in the area where the services operate. In wealthier provinces, such as Jiangsu, these services are largely self-supporting and receive financing, if needed, through rural collective enterprises. In poorer provinces, such as Shanxi and Guizhou, these services are less able to obtain reimbursement from clients and enterprises and are more dependent on government finance.

Sources of Health Spending

2.13 China allocated about 3.8 percent of GDP to health in 1993, the most recent year for which comprehensive national health statistics are available.' Chinese society spent somewhat less on health care in earlier years--an estimated 2.9 percent in 1978 and 3.0 percent in 1986. Figure 2.1 shows the growth in China's health expenditures by sources and uses. The most dramatic change in health financing between 1978 and 1993 was the demise of the rural Cooperative Medical System (CMS) and its replacement mainly with private out-of-pocket spending for health care services in rural areas. The change from cooperative to personal financing may have hit the poor hardest, since they are among the least able to pay for health care out of savings. Other significant changes were as follows:

* The share of health spending provided by the government budget, exclusive of subsidized care for government workers provided through GIS, declined from 36 percent to 16 percent of national health expenditures between 1986 and 1993;

- The share of spending contributed by the CMS fell from 20 percent of health spending in 1978 to 2 percent by 1993;

- Out-of-pocket payments rose from 20 percent of revenue in 1978 to 26 percent in 1986 and to 42 percent in 1993, transforming the financing base of the health sector.

* In constant-price terms, government health spending (excluding GIS) increased threefold between 1978 and 1993. However, the rapid growth of the economy induced even faster growth in private health spending, which went up by a factor of 10 in constant-price terms. Spending by the Government Employees Health Insurance System (GIS) and the Labor Health Insurance System (LIS) rose only slightly as a share of national health expenditures, from 30 percent of all health spending in 1978 to 33 percent in 1986 and 36 percent in 1993.

Uses of Health Spending

2.14 The main uses of China's health sector funds--public, private, and insurance-based--are for hospital services and the purchase of drugs. Public-sector staff and institutions deliver most health services, collecting fees from patients or their employers.

2.15 Most hospitals manage to balance revenues and expenditures. Hospital managers are expected to cover 85 percent or more of their costs from patient revenues. To attract patients, hospitals sometimes borrow from suppliers or their own staff to purchase highly profitable high

' The Health EconomicsInstitute in Beijing(HEI), produceda revised estimateof nationalhealth expendituresin January 1996for spendingin 1993. The revisionraised the 1993 estimatefrom 3.6 percent to 3.8 percentof GDP. HEI has not revisedestimates for earlieryears.

16 Figure2.1. Growthin HealthExpenditures in Chinaby Source 1978-93 140

120

,f, 100 - L CD Out of Pocket

A 80/...

Govemment 60 ~~~~~~~~~~~~~~~~~~~InsuranceSystem (GIS)+ Labor Insurance X- System(LIS) c 40 : Govemment 20 Budget (excludingGIS)

oI RuralCooperative MedicalSystem 1978 1986 1993 technology diagnostic equipment. In 1993, 76 percent of all spending on health was used to finance hospital inpatient and outpatient services, including pharmaceuticals.

2.16 Chinese public hospitals are fairly autonomous compared to public hospitals in many other countries. Basic salaries are set by the government, but the hospital establishes bonuses, which can be two to three times the basic salary. Hospitals can move funds across budget categories, and can make their own capital investment decisions. They have little autonomy over personnel, however. Personnel are assigned to hospitals by the Ministry of Health and provincial health bureaus.

2.17 A 52 percent share of China's total health spending went to purchase pharmaceuticals in 1993. About 85 percent of pharmaceutical sales occurred in hospital inpatient or outpatient settings (Table 2.1). This 52 percent share of national health expenditures allocated to pharmaceuticals, whether in or out of hospitals, is higher than in most low-income countries. For OECD countries, spending on drugs averages 14 percent of health spending.

2.18 Overprescription and misallocation of resources towards drugs is one major efficiency issue in China's health system. Other problems are overuse of high technology diagnostics, and the long average length of hospital stays (15 days). These efficiency issues are discussed in more detail in Chapter 5.

Government Recurrent Budget

2.19 In China's national health accounts, spending under the government's "recurrent health budget" (spending by the Ministry of Health and related departments at the provincial and

17 county levels) amounted to about 8 percent of all health spending in 1993 (Table 2.1). It includes spending on public hospitals (40 percent of the total), EPS (12 percent), and maternal child health services (3 percent). Public spending on family planning is outside of MOH's recurrent health budget. It amounts to about 1.7percent of all health spending. Public spending on traditional Chinese medicine facilities (also separate from the recurrent health budget) is smaller yet, accounting for 0.7 percent of all health spending.

2.20 Public spending for the health care of government employees and related groups (8.8 percent of total health spending) is included as part of the government health spending, but is not part of the recurrent health budget. OECD countries generally keep this account separate as part of government non-wage employee compensation. GIS payments rose from 13 percent of government spending on health in 1978 to 34 percent in 1993. The bulk of increased government spending in the government health budget therefore went to pay the health care costs of government employees. There was also a big increase for the high priority family planning program. These changes have taken funds away from basic public health services.

2.21 China's EPS programs, which were entirely subsidized before 1985, are among those being pushed hardest to finance their work through user fees. Government regulations now require restaurant workers, for example, to be tested by the EPS, while their employers must pay for these tests. Shanxi's provincial EPS was entirely financed by the government until 1991. By late 1994, it received only 65 percent of its income from the provincial treasury, with the remainder coming from fees for testing services in cities and towns. There has been a significant decline in the frequency of field visits to villages, however, where costs cannot be recovered as easily. The EPS in Jiangsu province now receives just 60 percent of its income from the government treasury. In the relatively poor province of Guizhou, however, most EPS income still comes from the government because there is little capacity to sell testing services even in cities and towns. The EPS of Guizhou has cut its village field work even more than wealthier provinces like Jiangsu (Jin 1995a, pp. 25-26).

Capital Expenditures

2.22 Duplication of facilities is a serious issue in China today. In urban areas, the overlap involves the Ministry of Health, state-owned enterprises, and traditional Chinese medicine facilities. In rural areas, there is some duplication and overlap of services between maternal and child health centers, family planning services, township health centers, and epidemic prevention stations (see also Chapter 5, Box 5.3).

2.23 From 1985 to 1989, as much as 80 percent of health investments went into hospital construction and equipment and less than 10 percent supported public health and high priority basic clinical services. In the Eighth Five-Year Plan period (1991-1995) the central government established a special fund of 1.1 billion yuan to strengthen public health and basic health care, known as the Three Items Construction Program. Local governments had to provide complementary funds for EPS and MCH programs at the provincial and county levels (Hou and Zhou 1995). The idea was sound, but the poorest counties and townships have been unable to generate counterpart funds, and have thus been the least able to benefit from this program.

2.24 Public hospitals have the autonomy to make their own capital investment decisions. Hospitals' reputations and their ability to attract clients now depend on having high technology equipment because CT scanners, ultrasound, and other diagnostics have come to symbolize

18 satisfactory health care to Chinese consumers. At the end of 1993, China had 1,300 CT scanners, 200 MRIs, and 1,200 color Dopplers (Hu Haobo 1995). Many specialists believe these instruments are not the most cost-effective investment for a country at China's stage of development, or indeed even for much wealthier countries.

2.25 Investments in health research can contribute substantially to meeting China's remaining health sector problems, and government support of research is an essential element of public health policy (Box 2.1). The Chinese national health accounts data do not disaggregate investments in medical research. These investments are included in the estimate for medical education and research, which totals about 1.8 percent of all health spending (Table 2.1).

Government Fiscal Structure

2.26 Two fiscal problems complicate government efforts to finance health services and to promote redistribution between rich and poor areas of the country. First, revenues are decentralized, leaving the central government with few resources to transfer from rich to poor provinces. Only about 4 percent of the total recurrent health budget in 1993 fell under the direct control of the central government (Berman and others, 1995, p. 28). Province, prefecture, county, and township spending accounted for the remainder. This decentralization inhibits special assistance for the poor, since they generally live in provinces with less capacity to tax and redistribute benefits through publicly subsidized health services. Furthermore, regional income disparities are growing.

2.27 A second problem that constrains the government's ability to finance priority programs and promote redistribution has been the steady reduction in government revenue as a share of GDP. China's overall budgetary expenditures declined from 33.8 percent of GDP in 1978 to 13.8 percent in 1994 (World Bank 1996a).

2.28 These problems of decentralization and limited resource mobilization are addressed in a recent World Bank report that outlines the prospects for increasing China's central government revenues through more effective collection of the value-added tax from payroll taxes, and from a tax on pollutants. With these changes, government revenue would rise by 6 percent of GDP by the year 2000 (see Chapter 8, Table 8.4). The central government's share of revenues would increase to 60 percent, from about 40 percent in 1994, thus facilitating fiscal transfers to poorer provinces and counties. A key challenge for intergovernmental fiscal relations is to design and implement a grants scheme that will redistribute the central government's revenue surplus to the poorer provinces (World Bank 1996a). Proposals in later chapters of this report are consistent with those suggestions.

Who Gets Health Services?

2.29 Two major concerns of this report are that health insurance coverage in China is declining and that the rural poor have inadequate access to health services. Population sub- groups differ markedly, however, in income levels, health insurance coverage, and health spending (Figure 2.2). In 1993, risk-pooling mechanisms (GIS, LIS and rural community financing schemes) covered only 21 percent of the population, but accounted for 38 percent of health expenditures. This wide difference suggests that those who are covered have far better

19 Box 2.1. HealthResearch and Development:A NeglectedComponent of the ChineseHealth System?

The high-incomecountries of the OECD investa much more substantialfraction of their health expenditureson researchand development(R&D) than do low- and middle-incomecountries. In the OECD, 3 percentto 4 percent of health expendituresgoes to R&D; in developingcountries the amountlies between 1/2percent and I percent. China's R&D expendituresfor health appearto be in the range. In China,as elsewhere,elite universitiesand specializedinstitutions (such as the Chinese Academyof PreventiveMedicine) undertake most of the health research. These institutions,along with industry,also engage in new productdevelopment, including the developmentof new pharmaceuticals.In the OECDcountries, R&D expendituressplit approximatelyevenly between research (mostly public sector) and product development(mostly private sector). There appearsat presentto be relativelylittle new product developmentin the Chinesepharmaceutical industry as indicatedby the virtual absenceof newly registereddrugs from China (or, indeed,from any developingcountry).

The issue for health policy concernsthe extent to which Chinashould consider investing a larger proportionof health expendituresin R&D and, if so, how this might best be done. One line of argument is that, since the OECD countriesinvest so heavilyin R&D, the rest of the worldneed not do so: it can simply use the knowledgegenerated elsewhere even if, occasionally,such use requirespayment of licensingfees or high costs for patent-protectedproducts. A differentline of argumentis advancedin a recent report of the WorldHealth Organization(Ad Hoc Committee1996) which suggestsfour broad areas where China might productivelyinvest R&D resources:

(i) Operationalresearch focused on controlof the diseasesof extremepoverty that still affect perhaps 100 million Chinese; (ii) Biomedicalresearch and new productdevelopment to counter still evolvinginfections (such as AIDS or drug-resistanttuberculosis) with better vaccines,drugs, and diagnostictools; (iii) Epidemiological,preventive, and clinicalresearch to address rapidlyincreasing problems of noncommunicabledisease and injurywith interventionsthat are inexpensiveenough to be widely implementedand sustained;and (iv) Healthpolicy and systemsresearch to address questionsof cost containment,access, and qualityof service.

Thecase for expandingChina's efforts in these four areas appearsstrong. The 1996 WHOreport also stressed the importanceof competitiveallocation of R&Dresources and full engagementwith the internationalR&D communitiesto help China avoid reinventingwhat has alreadybeen learnedelsewhere.

Given the importanceof new knowledgefor improvinghealth outcomesand constrainingcost growth, China may wish to undertakean in-depthreview of this aspect of its system. One approachmight be to forn a committeelike the one that preparedthe WHOreport--on which a Chinesescientist participated--toundertake the task. Whilemost membersof the Committeewould of course be from China,it might also be useful to includethe perspectivesof a few eminentoutside scholars and industrial scientists.

capacity to meet their health care needs than the nearly 80 percent of the population who are not insured.

2.30 In terms of health service access, the 1993 data show that the poorest quartile of the rural population, i.e. the poorest fifth of all Chinese--virtually none of whom have prepayment or insurance arrangements to assure funding for health services--accounted for only about 4 percent

20 of total health service spending. This Figure2.2. Per Capita AnnualHealth Spending low share supports the view that the rural on PopulationCroups, 1993 poor receive an inequitably small share of the available health services, as discussed further in Chapter 4. 250 Urban 2.31 Urban dwellers, in contrast, who 200 - account for approximately 53 percent of c 150 earned income, received at least two- thirds of all health spending. The high >~ 100 Rural income elasticity of demand for health 50 services would explain only part of these o large disparities.

Health Coverage in Urban Areas Urban residents account for four times as much on health spending as do rural residents. Spending per 2.32 Official data indicate that 30 person is far lower among poor farmers, than among million persons, or some 2.5 percent of richer farmers, or groups covered by insurance. the Chinese population, are eligible beneficiaries of the Government Employees' Health Insurance System, or GIS. They receive free care in the government's clinics and hospitals or are reimbursed by their employing 400 - government agency. Beneficiaries include employees and retirees of 350- central, provincial, and local governments, disabled veterans, and 300 - university students. The 1993 health survey revealed an apparent discrepancy 250 between official coverage of GIS and 2 survey responses: 5.8 percent of those > 200 ainterviewed said they were covered by 60= 9_ ffi , GIS, which would be the equivalent of 70 million persons. The difference 100 3 between the 30 million people officially covered by GIS and these responses may 60i be due to dependents stating they are eligible even though the government may 0 * i _ not count them as eligible. About 10 percent of those who said they were covered by GIS in 1993 lived in rural Health spendingfor the lower-income quarter of areas. More analysis of actual GIS rural Chinese is one-tenth that of the government eligibility and coverage is needed, and SOE workers. especially in light of the very high per capita spending revealed by official Source: See AnnexTable Al.6 and Wei 1995. estimates of the costs of this insurance program (Zhao Yuxin 1995).

21 2.33 In terms of insurance coverage and eligibility, the next best-covered group in China is the 11.7 percent of the population employed by SOEs. In 1951, the PRC government adopted a policy that SOEs should provide or finance health services for employees and retirees, with dependents treated on a 50-50 cost-sharing basis. This is the LIS insurance system. The government requires SOEs to contribute 14 percent of wages to cover both health and--to a much smaller extent--welfare benefits such as child care. With the erosion of profitability in recent years, many SOEs now restrict employee eligibility for insurance-paid health care. Surveys in 1992 and 1993 show that many workers who were covered in principle did not receive any insurance-paid care (Hu Teh-Wei 1995; Henderson and others, 1995).

2.34 SOEs spent 34.4 billion yuan for health care in 1993 (Table 2.1)--over a quarter of aggregate health spending in China. One-sixth went to the SOEs' own services, and the rest to other parties, especially public hospitals. SOE spending on their own services is about half as large as that of the government recurrent health budget.

2.35 This comparison underscores the risks for the government in trying to replace the services now delivered directly by SOEs to their associated workers and retirees. The government strategy for reform of SOEs recommends the separation of SOE-owned and managed health services from the principal business of SOEs. The creation of health insurance centers as part of demonstration projects in Jiujiang and Zhenjiang in late 1994 may offer a model for such change. Chapter 7 discusses options for reform of SOE health insurance.

Health Coverage in Rural Areas

2.36 In the 1960s and 1970s, the rural Cooperative Medical System (CMS) reached the majority of rural Chinese. Village authorities used funds from agricultural collectives to pay for the salaries and training of Barefoot Doctors. These locally-recruited health workers focused on the villagers' basic health needs. The collectives helped to pay for farmers' health care, but most services required substantial copayments. By 1985, however, less than 10 percent of China's villages maintained CMS arrangements, most of them with village incomes falling within the upper quartile (Zhao Zhuyan and Lusheng Wang 1995). The decline of the CMS program is discussed in more detail in Chapter 6.

2.37 Some vestiges of the CMS system remain even in poor villages, however. Village doctors throughout China sell contract or prepayment insurance for selected EPS and MCH services. There are both one-year and four-year contracts available, depending on how long the series of basic shots and dosages will take to administer. A typical EPS contract provides four years of immunizations (the Expanded Program on Immunization (EPI) standard groups of DPT, measles, and polio) at a prepaid price of a few yuan. The MCH contract covers ante-natal and post-natal care for a child up to age seven and includes nutrition, growth monitoring, and referral, if necessary. These special insurance contracts reach as much as 40 percent or more of the targeted child population.

2.38 Community financing schemes similar to the former CMS system still operate in the rural areas of some wealthy provinces like Jiangsu. Community health funds pay for most medical fees and pharmaceuticals. The funds come from township enterprises, village tax collections, and voluntary contributions. A THC health committee supervises use of the funds by each village. Each village clinic has its own account to pay for drugs and other materials

22 consumed. Salary and subsidies of the village doctors do not depend upon pharmaceuticals sales, but rather on the number of patients served.

2.39 In the village clinics in Shanxi province, a different financing pattern prevails. Community financing still exists and requires a 4 yuan per person-year prepaid fee to cover four services. Pharmaceutical coverage is excluded. The villages provide the clinics with office space, and village doctors make their income from selling drugs: the more they sell, the higher their income. In October 1994, the majority of village clinics in rural Jinzhong prefecture, Shanxi, resembled drug stores. Those visited had stocks of more than 1,000 kinds of drugs (Jin 1995).

Trends in Paying for Health Care

2.40 Most Chinese--some 800 million in rural areas, and perhaps 100 million in urban areas-- pay directly for health services when they receive them. Of those surveyed in 1992-93 who had been referred to a hospital for care, 40.6 percent did not seek hospitalization on grounds of excessive cost and their inability to pay (Zhao Zhuyan and Lusheng Wang 1995). Even middle- income farmers would be unlikely to have enough savings to pay for a long hospital stay. The share of health spending that is out-of-pocket has risen steadily since the late 1970s. While many countries are moving towards a curative health system that is financed publicly but provided largely privately, China is moving in the opposite direction.

23 I 3. STRENGTHENING PUBLIC HEALTH PROGRAMS

3.1 Governmentshave a large role to play in assuringpublic health servicesand this role needs to be strengthenedin China. Public healthprograms are designedto address the health problemsof entire populationsor subgroupsin the population. Public health activitiesmay involvespecific health servicesfor the community(such as immunizations),the promotionof healthy behavior(reducing tobaccoconsumption, limiting salt intake,avoiding sexually transmitteddiseases), or the promotionof healthy environments(improvements in sanitation). Some clinical services--suchas the treatmentof tuberculosisand sexually-transmitteddiseases-- are considered public health activitiesbecause they yield large social benefits by preventingor curing these infectiousdiseases. Publichealth programscan also play an importantrole in providinghealth educationand in ensuringthat infectiousdiseases are detected and adequately treated at the clinical level. Diseasesurveillance is another public health activity. Most public health programsprovide servicesfor whichthere are large social benefits, but for which individualsare unwillingto pay the full costs. As such, public finance for priority public health programscan correct these "marketfailures."

3.2 China has a long history of well developedpublic health programs,but in recent years these programshave faced fundingdifficulties, as well as coordinationand other operational problems. In addition,China's disease patterns and health risks are changing,and public health programsneed to adapt to new challenges. This chapterreviews the statusof public health programsin China today. It is not intendedto be an exhaustivereview but instead uses tuberculosis control, immunizations,and tobaccocontrol to illustrateproblems in current public health programsand suggest broad directionsfor change. It concludesthat public health programsneed considerablestrengthening in China--interms of public finance,program strategies,and content.

Structureand Financeof Public Health Programs

3.3 The Ministryof Health (MOH),under the authorityof the State Council,provides technical leadershipand sets guidelineson public health activitiesas part of its overall leadership in the health sector. The EpidemicPrevention Service (EPS) is the backboneof public health programsin China. At the county level,the EPS is made up of the epidemicprevention stations and affiliatedstaff. Many other agenciesalso carry out some public health activities,or public health research, includingthe GeneralOffice of the NationalPatriotic Health Campaign Commissionand its local branches,the ChineseAcademy of PreventiveMedicine, the Center for Health Statisticsand Information,and the National Institutefor Health Education.

3.4 The EPS employsa quarter of a million workersand extendsdisease control programs throughoutrural China. EPS maintainsthe cold chain (refrigerationequipment) for immunizations,makes field visits to assure water quality, is responsiblefor the control of diarrhealdiseases, and runs specific endemicdisease control programsin many areas (such as those against malaria and schistosomiasis).For several decades EPS was fully funded,both in terms of staff and operatingcosts, from provincialbudget outlays that drew on general revenues from each province. Public health activitiesuntil the 1980swere delivered in a highly organized fashion under EPS supervision,drawing on the village doctor, townshiphealth center (THC), and county hospitalas needed.

25 3.5 China's fiscal decentralization in the early 1980s weakened both the financing and the coordinated operation of public health activities. The fiscal decentralization gave much more budgetary autonomy to local governments. Provincial health bureaus develop their own programs according to national guidelines. County hospitals, epidemic prevention stations, maternal and child health (MCH) centers, and THCs continue to receive some public subsidies for salaries, but they, too, are now required to generate substantial additional revenue from user fees. With the fiscal decentralization, poorer counties now have the least capacity to develop and maintain public health programs. Coordination has weakened between the epidemic prevention stations, THCs, and the village doctor. The epidemic prevention stations have difficulty supervising and influencing the activities of village doctors, who now operate as independent practitioners and generate income from fee-for-service medicine.

3.6 In absolute terms, public financing of EPS has Table 3.1. GovernmentFinance of EPS and as a Share of remained at about 1.5 billion GDP, Selected Years yuan since 1986. As a share of GDP, however, it has fallen PublicFinance from 0.11 percent in 1978 to 0.04 percent in 1993 (Table Year 3.1). The EPS faces funding EPS EPS EPS EPS difficulties due to rising input Budget" Budget Budget as Revenue costs and the fact that the (1993 per a share of fromUser personnel budget needed to billion capita GDP Fees cover retired health workers is yuan) (yuan) (percent) (1993 increasing. In order to cover the billion costs of its services, the EPS has yuan) had to rely increasingly on revenue generation. Over the 1978 0.94 1.0 0.11 N.A. past several years, ancillary 1982 1.13 1.1 0.10 N.A. services whose costs cannot be 1986 1.54 1.4 0.09 N.A. recouped through user fees have been cut. For example, in 1990 1.58 1.4 0.06 N.A. Shuoyang county, Shanxi, 80 1993 1.53 1.3 0.04 1.6 percent of the total EPS budget Source: Wei Ying 1995and 1996. went to staff salaries in 1993. N.A. = not available The number of days assigned to 1/ Thesenumbers differ from thosepresented in Table 2. 1, Chapter2, field work in villages in 1994 becausethe numbershere includeboth the recurrentand capitalbudget. was less than a quarter of what it had been five years earlier.

3.7 EPS generated an estimated 1.6 billion yuan in revenue from fees paid by individuals and institutions in 1993 over and above the 1.53 billion yuan it received from the government budget. These user fees have had two detrimental effects. First, EPS staff have reallocated their attention to services for which fees can most easily be charged, such as food inspections, although these services are not necessarily the highest priority activities. Second, fees have reduced demand, particularly among the poor, for such services as tuberculosis control and preventive health services.

26 Immunizations:Improve Coordinationand Funding

3.8 The Expanded Programon Immunization(EPI) is a key public health programin China. Since the late 1970s, China has provided political and financial support for immunization against tuberculosis, diphtheria, pertussis, tetanus, and polio, with measles coverage introduced later. Coverage of the four basic vaccines (BCG, DPT, polio, and measles) reached at least 80 percent in all provinces in 1988, and nearly 80 percent in all counties by 1990--with dramatic benefits. The incidence of pertussis decreased from 126 to I per 100,000 between 1978 and 1993, and of measles from 250 to 10 per 100,000. China recently expanded its immunization goals. The government hopes to eradicate polio from China soon and B was added to the EPI program, although it is not yet clear how widespread the coverage of newborn will actually be. Given the high levels of post-natal transmission of in China, the addition of immunization at birth against Hepatitis B provides a cost-effective way to prevent cancer and of the liver.

3.9 But the immunization program also faces major challenges. With the decline in funding for EPS, transport and refrigeration facilities are breaking down, and many local programs face shortages in supplies. Major system breakdowns occurred in 1993 and 1994 in many poor areas, which led to declines in immunization coverage. Village doctors operate now as independent practitioners, and no longer coordinate closely with EPS. Although village doctors are provided with vaccines from EPS and are paid a flat fee for carrying out immunizations, the fee is not adequate to cover their operational costs (for needles, syringes, sterilization, and operation of the cold chain). This can lead to inappropriate sterilization practices and loss of vaccine potency. EPS and village doctors, in turn, charge for immunizations in many areas, presenting a financial obstacle for poor households. To generate additional revenue, some EPS staff perform and charge for unnecessary antibody tests before providing immunizations. Table 3.2. Immunization Coverage in China, 1993 Finally, coverage differentials are still (percent) significant between provinces and BCG Polio DPT Measles between urban and rural areas (Table National 90 88 86 85 3.2). Measles coverage in rural Shanxi Urban 95 94 93 85 and Guizhou is as low as that of many Rural 85 86 84 73 Sub-Saharan African countries, while in Shanxi 65 66 62 49 the cities the increasing unregistered Guizhou 72 74 67 53 urban or "floating" populationlacks Source: WorldBank 1995e,p.3. Note: National,urban, and access to these public health services. rural figuresfor measlesare inconsistent.

3.10 Additional funding is needed to consolidate and expand China's gains in immunizations. This requires changes in other priority public health programs as well: providing for adequate salaries and supplies, appropriate training and supervision, performance-related incentives--both for EPS staff and village doctors who coordinate with them--and good working conditions to maintain the commitment of public health workers.

Tuberculosis Control: Expand and Subsidize the New Tuberculosis Program

3.11 The tuberculosis control situation in China illustrates both what can be achieved with a well-run public health program and what can go wrong. Although the death rates are decreasing, tuberculosis still remains a major health problem in China, accounting for an estimated 3 percent

27 of all deaths in 1990 (Annex 3, Table A3.1). Active tuberculosis is a highly infectious disease. Public health programs throughout the world try actively to identify and treat infectious persons early in the course of the disease to interrupt transmission to others. Without appropriate treatment, 60 percent of TB patients will die. Well-run programs can cure 80 to 90 percent of patients; poorly administered programs cure 30 percent or less. Most programs around the world try to avoid any financial barriers to tuberculosis treatment. Treatment is provided free of charge, and some programs even pay patients to comply with tuberculosis treatment. China made substantial progress in controlling tuberculosis during the 1960s and 1970s, using standard antibiotic treatment that was essentially free of charge.

3.12 Changes in China's health financing in the 1980s, however, had a detrimental impact on tuberculosis control programs and their effectiveness. As public subsidies were reduced, public facilities were encouraged to recoup their expenses from user fees. This resulted in many distortions. When doctors and hospitals expected to be reimbursed by GIS or LIS, they provided excessive diagnostic tests and examinations during treatment and dispensed higher-cost antibiotics that should have been reserved for more resistant cases. Daunted by the costs, many low income people infected with tuberculosis failed to enter treatment or dropped out early. There were no incentives to ensure that patients completed their treatment or were cured. As a result, many cases of tuberculosis remained infectious, spreading the disease to others. The spread of drug-resistant strains has also accelerated in China since the 1980s as a direct result of poor treatment practices.

3.13 Recognizing the problems that result from charging for tuberculosis therapy, China has launched a new tuberculosis control program that is already operational in certain areas of the country. The WHO-recommended program of directly observed short-course chemotherapy provides subsidies for treatment and appropriate incentives for providers of care. In the new tuberculosis program, patients with symptoms suggestive of tuberculosis are referred by health providers to the tuberculosis county dispensary under the EPS for physical examination and fluoroscopy. If a smear of the sputum tests positive for tuberculosis, drugs are to be provided free of charge. All smear-positive patients are to be treated with short-course chemotherapy and every dose is observed by the village doctor. The village doctor is paid for performance--an initial payment when the patient is enrolled, a further payment at two months, and a final payment after completion of treatment. Tuberculosis cases are to be closely monitored at the county level. Supervision is emphasized: at the beginning of each treatment, county tuberculosis dispensary staff and township disease control officers meet each patient and the village doctor managing the patient to review the treatment plan. Other aspects of the program are also supervised, including the laboratory protocol and county registry. In the first four years, almost 1.6 million patients with suspected tuberculosis were referred to the program. The cure rate is now 90 percent among new cases, compared to about 50 percent previously. The failure rate in previously treated cases, which is an indicator of drug resistance, fell from 18 percent in 1991 to 6 percent in early 1994 (China Tuberculosis Control Collaboration 1996). This experience shows that careful supervision, adequate funding, and appropriate financial incentives to providers can make a dramatic difference for this major public health problem. Unfortunately, many patients (those outside the project area, and those within the project area who are not referred to the tuberculosis program for treatment) are still being charged for treatment and their cure rates are low. The cost for drugs plus all other services, including unnecessary laboratory exams, x-rays, and traditional , may total more than 1,000 yuan.

28 Anti-TobaccoEfforts: A MultisectoralApproach

3.14 Smoking is a major health problem in China today. If current smoking patterns persist, about 50 million of the Chinese who are now aged 0-19 will eventually die as a result of smoking (Peto 1986). According to a 1984 nationwide survey, 61 percent of men and 7 percent of women in China smoked manufactured cigarettes (Table 3.3). Chinese men account for about 10 percent of the adults in the world, but consume 30 percent of the world's cigarettes. Smoking prevalence is negatively related to educational status. More recent surveys indicate that smoking is becoming even more widespread.

Table 3.3. Characteristicsof Smokersin China, 1984 Male Female

Regular Occasional Regular Occasional Smokers Smokers Smokers Smokers Ig.tal 55.5 5.5 6.4 0.7

Ag= <20 14.1 5.0 0.2 0.1 20 -60 63.3 6.0 5.8 0.6 >60 63.6 3.0 16.5 1.4 Selected

Worker 60.7 5.0 6.6 0.7 Peasant 58.1 5.7 6.0 0.6 Cadre 54.9 4.3 5.6 1.0 Teacher 44.8 5.3 2.6 0.3 Doctor 49.1 7.6 2.2 0.2 Education College 39.6 5.2 2.8 0.8 Middle School 47.5 6.3 1.7 0.2 Primary 61.2 5.3 5.5 0.6 Illiterate 63.9 4.0 10.6 1.0 Source: NationalSurvey on Smokingin China, 1984: See Teh-WeiHu 1995,Ministry of Health, 1991.

3.15 Over 800,000 deaths were attributable to in 1990, including deaths from coronary heart disease, chronic obstructive lung disease, and lung cancer (Annex 3, Table A3.7). Given present patterns of smoking, this number is projected to increase dramatically over the next thirty years to at least 2 million deaths per year. The economic loss in terms of health care expenditures related to smoking was estimated at 6.9 billion yuan in 1989 (Jin 1995c).

3.16 Effective public health programs to reduce tobacco use are based on two complementary strategies: tobacco taxation, to provide an economic deterrent to consumption, and measures such as health education, banning of smoking in public places, and banning of cigarette advertising and promotion, particularly by foreign tobacco companies. One important legislative option would be a complete ban on all direct or indirect advertising and promotion of any tobacco goods or trademarks. If advertising and promotion are not prohibited, then not only will the high prevalence of smoking among males be perpetuated, but also a substantial proportion of

29 females may be induced to smoke, further aggravating the growing epidemic of death from tobacco use in China. China already has cigarette taxes in place. In fact, cigarette taxes are a major source--31 billion yuan, or 9.5 percent--of government revenue (1992 figures; see Teh- Wei Hu 1995). The effective tax rate on cigarettes was about 38 percent in 1991. This compares to 60 percent in Japan and Korea and 85 percent in Denmark. Several Chinese cities have already taken steps to ban cigarette advertising and smoking in public places, but these restrictions are being circumvented by sport sponsorship and other forms of tobacco promotion, and more action is needed. The government is considering increasing the cigarette tax and using the revenue to finance anti-tobacco campaigns and other public health efforts.

3.17 The amount of revenue generated from an Table 3.4.aEstimatesDeof bacco increase in tobacco taxes depends upon the price ParicesElastricit ofDemand,elasticity of demand in China. In other countries, the Various Countries price elasticity of demand has been estimated at Country Elasticity between -.35 and -.74 (Table 3.4). Assuming a rough average elasticity of -.5 in China, a 10 percent USA -0.50 increase in tobacco prices would reduce demand by Ireland -0.38 an estimated 5 percent and would generate an United Kingdom -0.39 additional 5 billion yuan of revenue per year. A 20 percent incremental tobacco tax would generate an -0.50 additional 10 billion yuan in revenue. An ad valorem Austria -0.54 tax (a tax based on a fixed proportion of the retail price) would be preferable, so that the tax would Finland -0.35 adjust to changes in tobacco prices. It is also Canada -0.74 preferable to impose the tax on all tobacco products, Median -0.45 so as to reduce the possibility of substitution. Such a SoreZmiganeso95pptax would be more progressive, because high-income Source: Zimringand Nelson 1995, pp52 5 smokers who purchase more expensivebrands would be willing to pay more. A 10 or 20 percent incremental tax on tobacco products, accompanied by further efforts in health education, bans on tobacco advertising and sports sponsorships, limits on the tar and nicotine content of manufactured cigarettes, and media campaigns against tobacco, would generate important health benefits in China.

3.18 Given the rationale for raising the tobacco tax to reduce tobacco-related illness now and in the future, there is an argument for targeting all or part of the revenue from this tax for complementary anti-smoking activities, such as health education programs. China should consider using the additional revenue generated to support both anti-smoking and other priority public health programs.

Conclusions

3.19 China needs a strong publicly-funded public health service adaptable to the changing patterns of risk factors and disease burden in China. This may be achieved by a restructured and reformed EPS in coordination with other agencies. Central and provincial governments should fully finance high priority public health services since these services must be financed by the government if they are to be provided at socially optimal levels. Central and provincial funding is also necessary to ensure that the poorest counties, with some of the biggest public health

30 problems, can pay for their programs. Fundingfor activitiesthat are now the responsibilityof EPS, for example, needsto be considerablyincreased over the mediumterm from its present level of 1.3 billion yuan (in 1993)to at least 6.5 billionyuan in 2001, based on current cost estimates of fully fundingtuberculosis control, the expandedEPI (includinguniversal hepatitis B immunizationfor infants),endemic disease control, health education,and other important programs,as well as projectedcost increasesto the year 2001. (As income is projectedto double in real terms, EPS costs will also rise, particularlydue to wage increases). The new tuberculosis program needs to be expandedthroughout the country,with tuberculosistreatment provided free of charge. Special outreachmeasures are needed for the unregisteredurban population.

3.20 At the same time, the governmentmust ensurethat public health programsare implementedefficiently. This requires: (i) providingfor an adequate salary structure, appropriatetraining, performance-relatedincentives, and good working conditionsto maintain the commitmentof public health workers,(ii) financingand developingnew diseasecontrol programsthat address the changingpattern of disease in China,and (iii) ensuringthat China's highly effective health surveillancesystem is maintainedand adaptedto the changingpattern of disease burdenand risk factors. Inter-sectoralcooperation and collaborationwill increasinglybe needed since it is beyondthe scope of the health sector alone to deal effectivelywith many public health problems such as lead in air, indoorair pollution,drug abuse, traffic accidents, suicides, and HIV/AIDS(Box 3. 1).

3.21 Given China's loomingburden of costly illness and prematuredeath from tobacco- related diseases,a substantialincrease in tobaccotaxes is another recommendedpublic health action, accompaniedby strongeradditional measures to reduce smoking. This would help reduce eventual illness and death from smoking-relateddiseases and, if the incrementalrevenue were directed to public health (as in Australiaand California),would also help finance public health programs. The tobaccotax recommendationis discussedfurther in Chapter 8.

31 Box 3.1 The Chaflengeof STDs and HIV in China

The preventionand treatmentof sexually-transmitteddiseases (STDs) and the Human ImmunodeficiencyVirus (HIV),the that causesAIDS, poseparticular public health challengesin China today. WHO estimatesthat Chinahad 100,000cases of HIV infectionat the end of 1995. Despitethis relatively low number,.China faces the risk of rapid spread of HIV and an explosiveepidemic. Several factorscontribute to.this.risk.With economicmodernization, increasing mobility of the populationcan be a vehicle for spreading the disease.Rapidly growing migrant populations are another problem. Migrantstend to have littleaccess to medical services,little knowledge about HIV/AIDS,and migrantwomen from rural areas are more likelyto engage in commercialsex work to supplementtheir incomeor as their main source of income. More broadly, many vulnerablesubgroups in the populationhave almostno knowledgeabout the HIV virus and how to prevent its transmission.:Finally, the bloodsupply in Chinais poorly screenedand overlycommercialized, posing a threat to blood recipients.

The areas where HIV is most prevalentin Chinaand the modesof transmissionare changing.About 80 percent of the reportedcases of HIV infectionwere previously in the southwesternprovince of Yunnan,where drug traffickingand prostitutionare major problems. But the numberof cases in coastalareas and large municipalitiesis now rising. While about64 percentof currentHIV infectionsin Chinaresult from injection drug use (Ministryof Health 1995b),the govemmentbelieves that the major mode of HIV transmissionin the coming years will be sexualtransmission, as a result of increasingrates of high-risksexual behavior. Evidence for this comes, in part, from China's risingSTD rates. STDshave reemergedas a significantpublic health problem in China,after major efforts to eradicatethem in the 1950sand 1960s. ReportedSTD cases rose from 5,838 in 1985 to 237,573in 1993 (Ministryof Health 1995b).

Concertedprevention efforts now, whilethe prevalenceof HIV is still relativelylow in China, can be expected.tohave a huge pay-off. The cost effectivenessof interventionsdrops sharply,however, when infectionscross from high-riskgroups to the generalpopulation. Moreover,increased prevalence of HIV is likelyto generatemany additionalhealth problems. It would:greatly exacerbatethe alreadylarge TB problemin China,for example,since TB is one of the major opportunisticinfections of HIV. Preventionefforts should includedisseminating information on how to avoid ,promoting condom use, treatingother sexually- transmitteddiseases, and reducingblood-borne transmission. A growingbody of researchshows that STD preventionand treatment has a significantimpact on avertingHIV transmission.

To be effective,a multisectoralapproach to controllingthe spread of HIV and STDswill be needed. The government's NationalStrategic Plan for the Preventionof AIDSand STDs(Ministry of Health, 1995b) lays out a plan that would involve twenty-twogovernment agencies and groups in societyin prevention activitiesaimed at differentsubgroups. These subgroupsinclude drug users, prostitutesand clients,STD patients,migrant populations,transport workers, overseas laborers and businessmen,and men having sex with men. The Ministry of Railroads,for example,could help in condompromotion and in educatingmigrant groups about AIDS.

With the support of the WorldBank-financed Disease Prevention Project, the Chinesegovernment has recentlyembarked upon programsto help prevent and controlthe spread of STDsand HIV. Theseprograrns, which will strengthenthe planningand implementationcapacity at the center,are being implementedin seven major cities and in Yunnan Province. They focus on intersectoralcoordination, policy reform measures(for example,to promoteanonymity of treatment,improved availability and use of condoms,and sex educationin. schools),behavior risk factor surveillance,training of health workers,improved management of STD cases,.and other health promotionmeasures.

32 4. MEETING THE NEEDS OF THE POOR

4.1 Poverty is both an importantcause and consequenceof ill health. The poor are more likelyto suffer from ill health, and their health problemscan contributeto keeping them in poverty. There is a strongjustification on poverty assistance(or equity) groundsfor government subsidiesaimed at improvingthe poor's access to importanthealth services. The Chinese government's role at present is minimalin this area. Resourcesneed to be redirectedor expanded to assure key services for the poor. At the same time, given scarce public resources, subsidiesneed to be carefullytargeted to the very poor and to those services that provide the greatest impacton healthfor the moniesspent.

4.2 China has experiencedan Table4.1. Incidenceof AbsolutePoverty, 1978-90 enormousreduction in poverty since 1978 1985 1990 economicreforms began. But TotalPopulation 963 1059 1143 ~progress has been unsteady. (millions) Improvementsin the qualityof life (millions)______for the poor proved rapid in the early Urban 172 251 302 1980s, driven by fast rural growth. Rural 790 808 841 The pace then stalled as the locus of economicexpansion shifted to urban Incidenceof AbsolutePoverty (millions) and coastalregions. The majority of the poor in China are now Total Poor 270 97 98 concentrated in resource-deficient (% of population) (28.0%) (9.2%) (8.6%) rural areas, and comprise entire UrbanPoor 10 I 1 communitieslocated mostly in (% of population) (4.4%) (0.4%) (0.4%) upland sections of the interior provincesof northern,northwestern, RuralPoor 260 96 97 and southwesternChina. Their (% of population) (33.0%) (11.9%) (11.5%) health status is bleak. Infant and Source: World Bank,1992b. maternal mortalityrates in very poor Note:The urban populationfigures underestimatethose actually counties are at least 50 to 100 residingin urban areas. By convention,the Chinesestatistical system treats unregisteredurban residentsas rural residents. percent greaterthan the national average, and are much greater yet in the poorest townshipsand villages.

4.3 There are signs that progressin poverty reductionresumed in the early 1990s. While the trends are clear, the number of Chineseconsidered to be poor dependson the poverty line. The national definitionof povertyyields some 80-100million poor (Table4.1), whereasthe figure jumps to 350 million (in 1993)if internationalnorms of minimumacceptable living standards ($1 per person per day) are used (World Bank 1996c).

4.4 While absolutepoverty is largelya rural problem in China, large numbersof urban poor, particularlyunregistered migrant populations,also have little access to public health programs and clinical services.

4.5 A broad range of measuresis needed to reduce povertyin China, includingpolicies to improve labor mobilityand foster rural enterpriseand agriculturedevelopment in poor areas.

33 Targeted subsidies to improve the health status of the poor will also improve their welfare and economic productivity. This chapter discusses how targeted health assistance might best be achieved and the likely costs of such assistance. The bulk of the discussion is on the rural poor, since that is the locus of most absolute poverty in China.

Health Problems of the Poor

4.6 The poor face a greater disease burden than the nonpoor. According to the findings of the World Bank's World Development Report 1993 (World Bank 1993a), selected childhood diseases, tuberculosis, maternal problems, micronutrient deficiencies, sexually-transmitted diseases, and a few others, account for much of the incremental disease burden borne by the poor relative to the middle classes.

The Rural Poor

4.7 This pattern holds true in China Table 4.2. Health Status by Income, Rural China, as well. Much of China's infectious and 1993 parasitic disease burden, including Income I II I IV tuberculosis, diarrheal disease, and iodine Quartile deficiency disorders, is concentrated in Average Per 927 677 561 441 poor and remote areas. Roughly 50 Capita percent of children in households at or Income below the absolute poverty line are at (yuan/year) least mildly malnourished (stunted) while iron, vitamin A, and other micronutrient Infectious deficiencies remain common among the Disease/1000 poor. As many as 90 percent of poor children suffer chronic worm infections. Infant 29 34 44 72 Table 4.2 shows that the poorest quartile Mortality of the rural population reports an Rate/1000 infectious disease rate three times that of Life 71 69 68 64 the richest quartile of the rural population, Expectancy _ I and an infant mortality rate over twice Source: The 1993National Health Services Survey and the that of the richest quartile. 1990Population Census.

4.8 The rural poor not only have a larger disease burden, but they use health services less than higher income rural residents. One-third of low-income households sought no health care according to the Study of Thirty Poor Counties (China Network for Training and Research in Health Economics and Financing 1996), whereas only 16 percent of the high-income rural households sought no health care in the previous year. The number of visits per person for the lowest quartile income group was only 60 percent of the number of visits made by the highest income quartile; the number of inpatient days per person of the lowest income quartile was just 50 percent that of the highest quartile. Data from the Study of Thirty Poor Counties also show that the poor spend a high relative share of their cash income on medical care: 23 percent for those with household cash incomes under 250 yuan per month compared to 11 percent for those with incomes between 430 and 690 yuan per month (Luo Wujin as cited in Hammer 1996).

34 The Urban Poor

4.9 Much less is known about the health problems of the urban poor, notably the unregistered urban poor. The urban migrant population is much more likely to live in crowded, unsanitary conditions, and face higher risks of contracting communicable diseases, such as tuberculosis, than other urban residents. Migrant women are more likely to engage in commercial sex work on a full-time or part-time basis, with its concomitant serious health risks. China's cities have few community-based health service facilities, and the outreach of public health programs to unregistered urban populations is weak, at best.

The Role of Government Spending

4.10 Government spending on health has not effectively reached China's poorest residents. China's 1981 public finance reforms decentralized public finance to the provincial and county level, limiting the central government's ability to redistribute funds from richer to poorer areas of the country. An analysis of public expenditure over eleven years shows that the allocation of public expenditure is skewed towards richer regions and, within regions, to those provinces that are growing fastest (Hammer 1996). Within provinces, government spending is concentrated on government employee health insurance expenses and hospital inpatient and outpatient spending. Services that disproportionately benefit the poor, such as the Maternal and Child Health program and the Epidemic Prevention Service, have been constrained and increasingly forced to rely on revenue from user fees. Not surprisingly, traditional public health activities such as immunizations have the greatest coverage in the highest income provinces. In China's poorest provinces, less than half the children receive their third dose of DPT, compared to more than 90 percent in the richest provinces (Hammer 1996).

Bringing Key Health Services to the Poor

The Rural Poor

4.11 The State Council Leading Group Office of Poverty Alleviation and Development has identified 592 counties with the greatest incidence of rural poverty (Table 4.3). These counties, with a total population of about 210 million people, make up about one-quarter of all Chinese counties. In these counties, 58 percent of the population had incomes below 500 yuan and about 26 percent of the population had incomes below 300 yuan per year in 1992-93. Experts disagree as to exactly how many of China's absolute poor reside in the 592 poor counties, but it is generally agreed that at least half are in these counties. Thus, the 592 counties constitute a useful beginning reference point for subsidies to improve the poor's health.

4.12 The key issue here is targeting: how to reach the poor (and only the poor) with the appropriate transfers without incurring excessive administrative or political costs. There are several ways to target the poor:

* Geographical targeting to areas where the poor are concentrated. For example, poor administrative villages in China's 592 officially designated poor counties could be targeted for subsidized services.

35 * Individual or household targeting, by identifying and certifying the poor for subsidized services. (China may be one of the only developing countries in the world where this is actually feasible, given good government records, but it is administratively costly.)

* Program targeting to health services that particularly benefit the poor, such as deworming and management of acute infections in children. (This would be additional to the public health programs discussed in the previous section.)

4.13 This report recommends a blend of geographic targeting (probably most practical at the administrative village level) in China's poor counties and universal finance of a limited number of services that particularly benefit the poor. Program targeting could be for the entire population in China's poor counties (which would provide an element of geographic targeting as well) or for the nation. Costs would be higher if targeting is done at the national level, but the one-third of the absolute poor who do not live in officially designated poor counties could be reached in this way. The financing for subsidies would have to come largely from the central and provincial level, since poor counties have little means to finance such programs.

4.14 Under a program of geographic targeting, the government might direct subsidies at health care providers in poor administrative villages (supply-side subsidies). Providers would be responsible for offering services free or at low cost to village residents. The government might further define what services would be provided for free, including highly cost-effective services to ensure maximum health impact from the program. If the administrative village was part of a viable community-financing scheme, the government could instead channel the subsidies to the community-financing scheme for health care of the residents (demand-side subsidies). Community-financing schemes are defined and discussed in the Chapter 6.

4.15 Under a program of service targeting, the government might fully fund certain health programs that disproportionately benefit the poor. The major disease conditions in China that are almost entirely associated with poverty are listed in Table 4.4. All told, these conditions accounted for an estimated 23 percent of China's disease burden in 1990. These conditions can be largely addressed by cost-effective public health and clinical interventions, such as immunizations, short-course chemotherapy for tuberculosis, iodized salt treatments, deworming, and prenatal and delivery care. Since these interventions largely benefit the poor, one option for targeting public subsidies to the poor is to focus public finance on these services.

4.16 Per capita health spending in rural areas is currently about 60 yuan per year, including public and private spending. For the sake of illustration, the public subsidy for a package of services for poor villages in poor counties might be 25 yuan per capita in 1993 in order to finance a 30 yuan package (with the other 5 yuan from other sources). The services that could be provided under a 30 yuan-per-person package are described in Box 6.1 in Chapter 6. With income growth projected to double between now and the year 2001, the per capita costs could be assumed to double as well. If the population of approximately 75 million in poor villages in poor counties were targeted, the costs would be about 3.75 billion yuan in the year 2001. An additional 1 billion yuan or more might be directed at program subsidies, either to the entire population in poor counties (210 million) or nationwide. These calculations could be altered depending upon assumptions about the size of the population to be targeted and the size of the per capita subsidy by the year 2001.

36 4.17 These poverty subsidies should be accompanied by systematic monitoring and evaluation to determine whether these programs indeed help the poor as intended and to guide policy improvements over time.

The Urban Poor

4.18 While poverty is concentrated in rural areas, measures also need to be taken to improve access to essential health services for the urban poor, who are outside the coverage of the LIS and GIS urban insurance schemes. As a first step, fully funding priority public health programs with public finance would help the urban poor (Chapter 3). For clinical care, most urban health services are provided by hospitals. China could consider measures to revitalize community care, especially in poor neighborhoods, as a way of reaching the urban poor. Perhaps some of the current hospital subsidies could be redirected, over time, to subsidies for services for the urban poor, either at community clinics or hospitals. If some urban areas form a strong enough community, with sufficient stability and cohesiveness to enter into a social contract with each other, community-financing schemes might be a feasible channel for financing and organizing services for low-income urban residents (Hsiao 1995d). -Public subsidies for interventions that particularly benefit the poor could also help the urban poor.

37 Table 4.3. China's 592 Poor Counties: Comparison of Average Income, 1992-93

Total Population Population in designated poor counties of Poor with per capita incomes under: Counties 500 yuan 300 yuan 200 yuan

Total (millions) 210.65 121.4 55 18.4 Percent 58% 26% 9% By province (percent):

Yunnan 83 52 20 Guizhou 75 48 23 Zhejiang 74 26 Sichuan 74 29 7 Gansu 68 27 10 Shaanxi 66 29 6 Ningxia 64 34 9 Hebei 64 24 10 Shanxi 63 16 4 Hunan 60 29 11 Guangxi 58 35 11 Henan 58 22 6 Hubei 57 23 3 Qinghai 51 21 6 Xinjiang 42 19 8 Anhui 38 9 1 NeiMenggu 37 15 2 Jilin 33 20 12 Hainan 30 14 3 Guangdong 26 2 Jiangxi 23 5 1 Shandong ______22 4 2 Liaoning 17 4 Heilongjiang 15 8 3 Fujian 6 1 Xizang N.A. N.A. N.A.

* indicatesless than 0.5%.

N.A. indicatesnot available. Source: Chineseauthorities.

38 Table 4.4. Poverty-Related Disease Conditions, China, 1990 1/ Cost-effective Percent of Condition intervention available 2/ total disease burden, 1990 (DALYs) Infectious and parasitic diseases

Tuberculosis Short-course chemotherapy 2.0

Diarrheal diseases Monitored oral rehydration treatment 1.8

Measles, polio, pertussis diphtheria, tetanus Immunizations 1.1

Intestinal nematode infections and anemia School-based deworming 0.7

Other infectious and parasitic diseases Acute chemotherapy 1.2

Lower respiratory infections Early identification and antibiotic 5.7 treatment Maternal conditions Emergency obstetric services with 1.2 prenatal care Perinatal conditions Obstetric care from trained personnel 4.9

Nutritional deficiencies Protein-energy malnutrition Breastfeeding; prevention and 1.0 treatment of childhood infections

Vitamin A and iodine deficiency Supplementation/food fortification 0.4

Anemia Supplementation/food fortification 3.2

Total 23.1

1/ This list is intendedto highlightonly the main diseaseconditions that are almostentirely conditions of the poor. In addition to these conditions,the poor also sufferdisproportionately from many others,such as diabetes,chronic obstructivepulmonary disease, sexually-transmitted diseases, and HIV infection. (Althoughexisting projectionsof the STD/HIVcomplex's importance suggest little problemin China (Annex3, Table A3.6), increasingclinical and public health reports suggestthat these epidemicscould become serious threats.)

2/ Cost-effectiveinterventions refer to interventionscosting less than $150 per disability-adjustedlife year (DALY).

Source: Annex 3, TableA3.4; WorldBank 1993a.

39

5. IMPLEMENTING REFORMS IN PRICING AND PLANNING

5.1 Efficiencyis another basic dimensionof performancein the health sector. Government policies can have a major impact on whether healthspending is allocatedto the "best buys," whether services are producedat least cost, and whethercare is clinicallyappropriate. Well- designed incentivesare key to efficiency. The present pricing structure in China producesmajor distortionsand inefficienciesby providingincentives for excessiveand inappropriatecare. In addition, the parallel public deliverysystems produce considerable waste in tenns of excess capacity and idle resources. This chapterfocuses on pricing issues and on planning and coordinationof infrastructureand manpower. Major reforms in these areas could significantly improvethe overall efficiencyof the health sector in both rural and urban areas. Chapters 6 and 7 then turn to the issue of developingand strengtheninghealth insuranceor risk-pooling mechanismsin rural and urban areas, respectively.

Price Distortions in the Health Sector

5.2 Prices for health services in China are set under guidelinesestablished by the Price Commission,often at levels well belowcosts. Price Commissionofficials aim to make the prices high enoughto protect and develop the servicesprovided, yet low enoughto assure affordability to the users. During the 1960s and 1970s, the governmenttried to increaseaccess to health care by reducing the prices of medical visits and hospitaldays to levels that a poor could afford. Prices for most medical services in China are still below costs,especially for services with a large labor input. The prices of high technologydiagnostic tests, however,have been set far above costs to offset losses on other services. The system thus has two pricing extremes: most services are priced too low, leadingto under-the-tablepayments to physiciansand other problems. At the same time, high profit marginson drugs and diagnostictests encourage overprovisionin those areas. For example,now that village doctors' incomedepends on fees charged for drugs, injections,and diagnostictests, these doctorshave strong incentivesto overprescribedrugs and tests. This drives up health spendingwithout improvinghealth.

5.3 At the heart of the issue of pricing policy,as it affects goods and servicessold by the health sector to patients, is China's "YellowBook" price list--a detailed listing of thousands of medical procedures,services, and diagnostictests, which sets the price of each. Despite double- digit inflation in recentyears, these listedprices are rarely updated in many provinces,thus fixing most health serviceprices at extremelylow levels.

5.4 The difficulties caused by these pricing policiescan be clearly seen in the hospital setting. Most Chinese hospitalscharge patients(or the GIS and LIS insuranceprograms they belong to) for each item of service rendered and drug given; about 85 percentof revenues come from these charges, which are fixed and publishedin the YellowBook. Hospitalsalso receive a provincialgovernment subsidy for personnelwages (excludingsupplementary wages and bonuses),and some funds for capital investment.

41 5.5 Faced with this irrational price structure, hospital managers use profitable products to cross- Box 5.1. The LeveragingEffect .. i subsidize under-priced products. Unfortunately, hospitals have to over-sell a high volume of profitable Whenhospitals have to generate services to generate a small profit. This "leveraging profitsto cross-subsidizeservices: effect" is described in Box 5.1. Besides distorting the operatingat a loss, the amount dtht patterns of medical services, the leveraging effect can be generateddepends. on the. increases total health costs. profit marginof particularservices, For example,drugs have anallowed mark-upof 15 percent. hi ordert Pricing of Health Services. generate 100 yuan of: revenuesE above : cost, a hospitalhas to increasesales 5.6 Table 5.1 shows the average costs for fairly of its drug prescriptionsby 666 yuan. routine hospital procedures in Shanghai, using modern This "leveragingeffect" contribues cost accounting methods, and the official prices for to rapid inflationof health: i insured and uninsured patients. The actual costs for the expendituresi procedures are two to four times greater than the allowed fees for patients paying out-of-pocket, and two to three times the allowed fees for insured patients.

5.7 In contrast to these routine Table 5.1. Comparisonof ProductionCosts and procedures, the price of newly- Fees for SelectedProcedures in ShanghaiHospitals, introduced diagnostic tests has been 1989 set at a level high enough to (in 1993 yuan) encourage their rapid adoption. A AllowableFees comparison of the cost of a CT scan Procedure Average Insured Self-Pay with allowed fees in Shanghai and Costs Tianjin shows the large profit from Appendectomy 118 56 28 this test (Table5.2). Cataract removal 142 44 22 5.8 An unintended consequence Gastrectomy 326 292 97 of this pricing policy has been that urban hospitals, and even county and Cholecystectomy 199 111 56 township hospitals and health Exploratory 167 56 28 centers, have come to see high laparotomy technology equipment as their financial salvation (Box 5.2). These organizations now routinely Cesarean section 140 Ill 56 organize investor groups to buy such Source: The ChineseJournal of HospitalManagement equipment. They borrow from 1993;9:55 banks and often sell investment shares to staff members to purchase such equipment. Sometimes hospitals lease equipment from international equipment suppliers, with the lease payment set at a percentage of gross revenues generated from the use of the equipment. A hospital's reputation now is said to depend on possession of the latest equipment, be it computerized tomography, magnetic resonance imaging, fetal monitoring, upgraded intensive care units, burn units, or any of the wide variety of diagnostic techniques that manufacturers have successfully brought to market. China is in the midst of a diagnostic equipment race involving most of its hospitals and many health centers.

42 Box 5.2. The HighTechnology Equipment Race

In Jurong CountyHospital in Jiangsu province,the governmentsubsidy now accountsfor only a small part of hospital income:it declinedfrom 17 percentof revenuesin 1985to only 2.2 percentby 1993. As a result,the hospitalhas been forcedto collect fundsfrom othersources such as drug sales and tests using high technologyequipment. In 1993,the hospitalreceived capital investmentsof 0.2 millionyuan from the localgovernment, borrowed 2.2 million yuan from local banks, and raisedanother 0.6 million yuan from hospitalstaff to buy a CT scannerand other new equipment. About 1,000patients were diagnosedby CT in 1994,bringing the hospital 0.35 million yuan in revenuethat year. The incomefrom high technologydiagnostic testing (mainly CT) more than doubledfrom 1993to 1994. Althoughpatient volume grew only 2.1 percent,the numberusing high technologyequipment increased 50 percent. Totaloutpatient fees increased168 percent(Jin Shuigao 1995a,p. 21). Thesestories are beingduplicated throughout China--first in the somewhat richer coastalprovinces and, more recently,in the interior. Even Guizhouprovince, which has the lowest per capita incomeof all of China'sprovinces, has seen an increasein high technology medicine.

Table 5.2. Comparisonof ProductionCosts and Fees for BodyCT Pricing of Scan, 1988 (1993 yuan) Pharmaceuticals Cost AllowableFees 5.9 D prices at Fixed Variable Total Self-Pay Insured the retail level are Patient controlled by the State Shanghai 109 47 156 181 362 Pharmaceutical Agency. Tianjin 80 32 113 181 362 The price structure Tianjin32 80 113 181 362 ~~~allowsmark-ups of 15 Source:Chen Jie 1994,pp. 4-5 . allows ar-pof 15 percent at both the wholesale and retail level. Hospitals have an incentive not only to overprescribe drugs they purchase from manufacturers, especially expensive drugs, but also to manufacture their own products to maximize the permitted margin. Patients are frequently treated with intravenous drip solutions of glucose, vitamins, antibiotics, and other drugs. In almost all cases, this does not constitute justified medical practice, but is done to maximize profit. Village doctors earn almost all of their health-related income from drug sales and injections, with injections preferred over oral prescriptions in order to maximize revenue.

5.10 Given this incentive structure, it is not surprising to find that spending on drugs accounts for 52 percent of all health spending in China, compared to an average of 14 percent in OECD countries and 15-40 percent in most other developing countries.I This is not only inefficient, but the inappropriate use of drugs can generate high health costs (for example, from the development of antibiotic resistance).

The high share of spendingon pharmaceuticalsin China is in part due to the fact that the price of pharmaceuticalsis relativelyhigh comparedto many other health inputs,such as labor. Nonetheless, there is considerableevidence of overprescribing.

43 5.11 There are several ways to combatoveruse and misuseof pharmaceuticals:

* Take the profit out of prescribing. If hospitalpayments were case-based,they would be independentof the number of drugs and tests prescribed.

* Incorporatehigh copaymentlevels for outpatient drug prescriptionsin insurance benefit packages to reduce excessdemand.

* Educate providersand consumers. Some countriesgive providersprescribing handbooks and educate the public on the appropriateuse of drugs and the negativehealth impact from their overuse.

* Develop and enforce hospitalformularies/essential drug lists to guide cost-effective prescribing.

* Monitorprescribing patterns and providefeedback to physicians. Core indicatorsof appropriateprescribing patterns includethe averagenumber of drugs prescribedper patient encounter,the percentageof drugs prescribedby genericname, the percentageof patient encountersfor whichan antibioticis prescribed,the percentageof encounters in whichthe patient receivesone or more injections,and the percentageof drugs prescribedthat are not includedon the relevant essentialdrug list or local formularyfor that level of care.

5.12 On the issue of prices more generally,previous analyses emphasized the health sector problems caused by pricing policiesthat distort supplyand demand(World Bank 1992a). The standard advice in such situationsis to move towardmarginal-cost pricing, even if fees continue to be fixed. This step has been given extensiveconsideration by the authorities in China, but has not yet been adopted (see backgroundpapers prepared for this report, especiallythe papersof Zhao Yuxin, Cai Renhua,Chen Xiaoming,Meng Jianguo,Hu Haobo, Hu Shanlian,and Hou Yan and Zhou Heyu, all prepared in 1995).Senior policymakers in China seem to accept price reform in principle,but fear that jumps in price could set off extremelynegative reactions.

5.13 A major studyof costs and prices is neededto lay the basis for reform. Such a study shouldavoid attemptingto revise thousandsof prices in detail. Instead, it shouldfocus first on graduallybringing prices closer to costs. Someprices, especiallythose for services with high labor content, would be raised, while others would be lowered. Second,the studyshould analyze methods of defining and pricing broad packagesof care, in line with providerpayment reforms discussed in the followingchapters on risk pooling in rural and urban areas. Price reformsmust include a mechanism,perhaps on an annual basis, to respondto inflation.

Pricing Reforms and Reallocation of Government Spending

5.14 Chinese governmentspending on health is both inequitableand inefficient. Most governmentspending on health is directed at the hospitalsector throughMOH hospitals, includingtownship health centers (4.5 billion yuan in 1993)and traditionalChinese medicine facilities (.8 billion yuan). As discussedearlier in this report, relativelylittle is devoted to more cost-effectivepublic health (1.3 billionyuan) and maternal and child health services (0.3 billion yuan). The governmentinsurance system, in addition,spends almost all of its monies on the hospitalsector (11.5 billion in 1993). Governmentsubsidies to hospitalsare a small share--no more than 15 percent--oftotal hospitaloperating costs. The rest of hospitaloperating revenue

44 comes from out-of-pocketpayments by individualsor insurancepayments. The government currentlysubsidizes the basic salaries of MOH employees,which are supplementedby bonuses paid by the hospital.

5.15 The Ministry of Health has recentlyargued for more public subsidiesto the hospital sector. This would only exacerbatethe existingmisallocation of resources,however. Those who benefit from the present subsidiesare middle-incomeurban residentsand those insuredunder the GIS and LIS systems. The first priority for governmentresources should be public health and cost-effectivecare for the poor.

5.16 Over the medium term, reallocatingpublic spendingaway from hospitalstowards other prioritieswould be facilitated by the reformsproposed in this report. Price reform would better align allowablefees with marginal costs so that hospitalswould have less need to provide excessive diagnostictests to cover loss-makingactivities. Ongoingurban insuranceexperiments are already trying alternativeprovider paymentapproaches. Providerpayment methods need to encourageefficiency, and ensurethat efficienthospitals can recovertheir costs. This is discussed in more detail in Chapter 7. In addition,public hospitalsneed to gain greatercontrol over their personnelso they can operateefficiently. Public hospitals in China alreadyhave considerableautonomy over their budgets,investments, and fee collection,but not their personnel. Instead,the governmentassigns new medical graduatesto public hospitals without sufficientregard for staffing needs. As a result, some hospitalshave too many doctors and too few nurses, while others have too many personnelin total.

Planningand Coordinationof Infrastructure,Manpower, and Health Services

5.17 China has at least three separatevertical systemsinvolved in the planning,financing, and organization of urban hospitalfacilities--the public health system,the state-ownedenterprise system,and the traditionalChinese medicinesystem. Each vertical system protects its own institutionalinterests, and has little incentiveto coordinatewith others. Moreover,the public health hospitalsare owned and managedby differentlevels of government--national,provincial, and county. In a major city, three major hospitalsmight be located quite close to each other, one operated by the central government,one by the provincialgovernment, and one by an SOE. A few kilometersaway, across the city line, there could well be a county hospitaloffering similar services.

5.18 In rural areas, the collapseof the CMS weakenedthe referral and supervisionchains that had existed between the village, township,and county-levelhealth organizations. There is duplicationof supplybetween the Maternaland Child Health(MCH) and FamilyPlanning (FP) facilitiesas well as betweenWestern medicine and traditionalChinese medicine facilities. This duplicationstrains the government'shealth budgets and misusesscarce resources.This problem is likelyto worsenas the FamilyPlanning Commission extends its facilitiesto the townshiplevel, creatingmore duplicationand a greaterdrain on governmentfunds. Someof these problemsare illustratedin Box 5.3 usingthe exampleof maternaland child healthservices.

45 Box 5.3. Problemsof Rural Health Delivery:A CaseStudy of Maternaland ChildHealth Programs

A look at problemswith matemal and child health servicesin rural China provides insights into overall problems facing the rural health system.. Several health facility and householdsurveys, a cost- effectivenessstudy, and a programreview conductedby the Ministryof Health(MOM) have all documented the need to improvethe qualityof maternaland childhealth services.

Qualityand Cost-Effectivenessof MCH Services

Issues in the deliveryof maternaland child health servicesmirror those of the rural health system overall. Because Maternal and Child Health (MCH) centers at the county level, like many other health providersin China,are now forcedto generatea high shareof their incomefrom user fees, they have shifted the focus of their work from preventivepublic healthmeasures to more curativefunctions for which they can chargefees. Thiseffectively denies health care to a portionof the rural population,since most rural familiesno longerhave health insuranceand manycannot afford fee-for-service care. Thishas even,reduced the coverage of such.essential preventive programs as immunizations,while creatingduplication of servicesand quality problemsas well.

In order to generatefees, MCH centersare now performingfunctions that might best be referredto the county hospital. Many MCH centersare investingin equipmentto perform cesareaji deliveries,for instance,instead of referringpatients who need this;surgery to nearby county hospitals.Not only does this duplicatethe hospitals'services but such surgicalprocedures: should be concentratedin high volume centers, to the extent possible,where professional skills are moreeasily kept up to date.

ThroughoutChina, hospitals themselves have becomeincreasingly dependent on profitsfrom drugs and other services, particularly capital-intensiveinpatient care for which public subsidies are highest. Treatmnentof childhooddiarrhea--still a commonproblem in some areas of China--illustratesthis problem. Anti-diarrhealtreatment in Chinanow often involvesunnecessary hospital stays, routineintravenous fluids, and potentiallydangerous drug treatrnents.Most caseswould be betterhandled using basic oral rehydration, and appropriateantimicrobial treatments at outpatientfacilities when necessary,The costs of such inefficient service are borne by three groups: the patients'families, .who pay more than necessaryfor treatment;the public sectorbudget; and patientswho do notreceive treatrnent because their familiescannot afford it. .

Organizationand Coordination of MCHServices

Another basic problem is the lack of functionalcoordination among the many different groups providingmaternal and child health services. An organizationchart of the MOHshows a numberof parallel national programs at the central, provincial, prefectural, and county levels, converging at tha townshiphospital and village level.The Maternaland ChildHealth Departmentis generallyresponsible for maternaland child.. health policies, althoughthe MedicalAdministration. Department develops hospital care policies,inctuding A MCH referral servicesat the provinciallevel and below. EPS under the Departmrentof Disease Control. developspolicies on childhoodimmunizations and diarrhealdiseases as well as trainingsites and materialsffor the managementof childhooddiarrhea. The Medical Scienceand EducationDepart ent handles health workers'training, includingMCH training, although MCH in-service training is designedand implementedby the MCH Department.The existenceof these overlappingdepartnents and functionsat all levels seriously constrains the development of consistent policies and guidelines, efficient training,.and quality MCH preventiveand curativeservices.

46 5.19 China seems to have an adequatesupply of total hospitalbeds, based on international standards. There appears to be overcrowdingin tertiary (level III) hospitals,but the average length-of-stayof 15 days for patients in these facilitiesis excessivelylong by international standards. With policy changes, many of these patients mightbe adequatelymanaged at lower levels and the average length-of-staymight be reduced. Currently,secondary hospitals and township health centers have low occupancyrates. China needs a rational hospitalreferral systemto make appropriateuse of its hospitalresources.

5.20 Regionalplanning could improvecoordination between services. Regionalplanning can be used to develop approachesto disease control, to establishreferral systems, and to guide capital investmentsin order to avoid duplicativeand fragmentedsystems. Regionalplanning efforts should involve all relevantactors, includingGIS, LIS, traditionalChinese medicine facilities,medical schools,and the MOH. As urban insurancecenters are expanded,they would also be a major stakeholderin regionalplanning efforts. Plans shouldcover a large enough populationbase to support comprehensivehealth services. Box 5.4 describeshow regional health planning has worked in three prefectures.

5.21 Regionalplanning bodiesmight be complementedby central and provincialgovernment efforts to better control investmentsin facilities,expensive equipment, and the mix and number of health personnel. Once expensiveand sophisticatedfacilities are established,they have to generatesufficient revenues to fund their operations,which raiseshealth care costs in the long term. Many countries have realizedtoo late that they have too many hospitalbeds or too many physicians,especially specialists,which also puts upwardpressure on health costs. Other countries are strugglingto reform their health system withoutsufficient numbers of well-trained hospital managers.

5.22 Many countries have found that governmentsneed to play an active role in controlling the growth of health infrastructure,both capital and human, since investmentsin infrastructure have such long term and powerfulimpacts on health service deliveryand costs. In the Netherlands,for example,hospitals must apply and receive approval to purchasespecific types of equipmentand technologyor to providecertain specializedmedical services. This regulation has effectivelyprevented an oversupplyof medicaltechnology. Of coursethese policies also have drawbacks. Supply-sidecontrols reduce innovationand restrict marketentry of, for example, lower-costproviders of specializedmedical services. The Netherlands' success has been largely attributedto its severe sanctions: hospitalsmay be fined, the service may be closed down, and/or insurancecompanies may refuse to reimburse for the service. Many countrieshave limitedthe number of health specialists,either by limitingtraining opportunities or by restructuringphysician payment,to lessenthe financial incentivesto specialize. China might start first with efforts to control the proliferationof high technologydiagnostic equipment, both through controlson the amountof equipmentand throughprovider payment incentives.

47 Box 5.4. Regional Planning for Health

Regionalhealth planningbegan in Chinathrough the IntegratedRegional Health Development Project. Each of three prefectures--Baojiin Shaanxi,Jiujiang in Jiangxi,and Jinhua in Zhejiang provinces--agreedon a five-yeardevelopment plan. Each regionalhealth committeeincludes governmentrepresentatives involved in finance,planning, civil works,and health. These regional planningefforts have led to better diseasesurveillance and a new TB control strategy. An:anti- smokingcampaign in schools and improvedtraining and outreachfor MCH servicesare in place. Ambulanceservices now functionwith radio networksin the provincialand regionalcities. Managersaccepted centralization of the high technologydiagnostics in one hospitalto serveothers. Someresources moved down from hospitalfacilities to countyand townshiplevels. Patient: physicalrehabilitation and care for the mentallyill improved. There is better in-servicetraining for..... medical and health workers,and strongernetworks for assessmentand maintenanceof equipment.

The experimentsin these three regionshave attractedincreasing attention from the senior leaders. One of the three prefectures,Jiujiang, was also chosenas one of two sites for the demonstration. insuranceproject describedin Chapter7.

48 6. OPTIONS FOR EFFICIENT RISK POOLING IN RURAL AREAS

6.1 In additionto strengtheningfunding for public health(Chapter 3), improvingthe poor's access to health services(Chapter 4), and introducingsystem-wide measures to improve efficiency (Chapter 5), China needsto improve risk-poolingmechanisms in both rural and urban areas. Pooling the risks of medical expendituresis desirablebecause in China,as in other countries around the world, health care expendituresare highly skewed: about two-thirdsof medical costs each year are concentratedon the 10 percentof the populationwith the highest health expenses. In China's rural areas, the main insuranceissue is how to best provide some form of risk pooling for the perhaps 700 million rural residentswho have lost their access to prepaid health care since the collapseof the rural CooperativeMedical System (CMS) in the early 1980s. Nearly 90 percent of farminghouseholds now pay out-of-pocketfor almost all of their health services. Publicly-runhealth institutionsreceive minimalgovernment subsidies and so charge high user fees, even for emergencyservices. If China allowspresent trends to continue,the majority of the rural populationwill have to continueto finance health services, includingcatastrophic care, out-of-pocketand pay on a fee-for-servicebasis. These fee-for- service payment systemswill also tend to escalate health spending.

6.2 The governmentcan play an importantrole in the developmentof health insurance by providinga policy frameworkthat would, ideally,(1) encouragethe establishmentof risk- pooling mechanismsto meet the population'sdemand for insurance;(2) minimizethe well- known market failures that occur in insurancemarkets, and (3) promote forms of provider payment and health care organizationand deliverythat contain costs and promoteefficiency over the long term.

6.3 How can China best encouragerisk pooling for the general rural population? There are several possibleapproaches to the collectivefinance of a health benefitspackage that includes catastrophiccare.

* Generalrevenue.financing is an approachtaken in manycountries, from Sri Lanka to Sweden. Under this approach,the publicly-financedhealth system,organized either nationally (e.g. Britain)or subnationally(e.g. Canadianprovinces) finances almostall health services. In OECD countrieswith generalrevenue-financed systems, some of the population typically purchasessupplemental private insurancefor coverageof additionalbenefits such as private hospitalrooms. Whilethe Chinesecentral and provincialgovernments do provide some small subsidiesfor publicly-operatedhealth facilities,increasing these subsidiesto a level sufficientto effectivelyprovide insurancecoverage to the 800 million rural Chinese appears prohibitive,at least over the short to mediumterm.

* Mandatorysocial insurance,funded by a wage tax or premium,is anotherapproach that has been adoptedby many countries,and is used in urbanareas in China (the LIS system). But this has limitedpractical feasibility in rural areas in Chinatoday because social insurancerelies on employersto enrollbeneficiaries and collectcontributions, and peasantsare largelyself- employed. In addition,China lacksthe institutionaland organizationalcapacity to manage large social insuranceprograms that would coverhundreds of millionsof beneficiariesin rural areas.

49 * Privatevoluntary insurance is anotherapproach to risk-pooling,but it presentsseveral problems. Internationalexperience has demonstratedthat the privateinsurance market does not emergeto supplyhealth insurancein developingcountries, except for the mostaffluent urban populations(Musgrove 1996). Even if that were not the case, privateinsurance is not an equitableor efficientapproach to insuringbasic healthbenefits because of information asymmetryand selectionbias. Riskselection by insurancecompanies leaves the disabled, elderly,and less healthypopulation uninsured. Countries such as the United States are developingregulatory mechanisms to partially address these problems,but that regulation requires sophisticatedadministrative and institutionalcapacity.

* Community-financingschemes provide collectivehealth financingfor entire rural communities. In community-financingschemes, funds generallycome from three sources: households,government, and local industries. The local community--notthe government-- organizes and manages the financingand deliveryof services on behalf of the consumer (Hsiao 1995d). Analysis of ongoingcommunity-financing schemes in China shows that they can reduce inappropriateuse of drugs, improvequality of services,and reduce overall service costs--inpart by encouragingservice provisionat lower-levelhealth facilities wherever possible. (Studyof Thirty Poor Counties--seeChina Network 1996;Jin 1995a;Liu Yuanli and others, 1996). Becausecommunity financing can promote universal or near universal coverageat the local level and efficient service deliverywithout being a major drain on governmentfunds, communityfinancing appears to be a more promisingoption for risk pooling in rural China over the mediumterm than the othersoptions reviewedabove.

6.4 This chapter looksat the ongoingand past experienceof risk-poolingschemes in rural China, and examinesoptions for reestablishingsome type of risk-poolingarrangements under current economicconditions, given financial and institutionalcapacity in China. While China has much experiencewith communityfinancing, particularly under the CMS, it is unclear how well communityfinancing will work on a large scale today, given the reformsthat have changed the economicorganization of the countrysideand given the size and heterogeneityof China. Therefore,this approachwould need to be phased in and systematicallymonitored and evaluated. As noted in Chapter 4, such community-financingschemes, if they prove feasible, could also be used to channelsubsidies for health care to the poorestrural residents.

The Rural Cooperative Medical System in China: 1960 to 1983

6.5 China pioneered ruralcommunity-based health financing with the rural Cooperative Medical System (CMS), which operated under the agricultural commune system in the 1960s and 1970s. Under the communalsystem of agriculture,communes took in all farm revenuesand paid them out to individualsand householdson the basis of work pointsprovided to the commune. BarefootDoctors receivedwork pointsfor their medicalwork and, in that way, public health serviceswere financedby the townshipsand villageswith littleor no additionalsubsidy from higher levels of government.Barefoot Doctors delivered free preventiveand primarycare services at the village level. Patientstypically paid a coinsurancefee for drugs, secondaryservices, and hospitalservices, which was higherfor inpatientservices. SinceCMS was largelyfinanced by a village's communalwelfare fund, the benefitpackage varied depending on the wealth of each community.The poorercommunities often couldonly affordto coverprimary care servicesand did not cover inpatientservices at countyhospitals.

50 6.6 The CMS system has since collapsed in 90 percent of Chinese villages. The demise of China's CMS system can be explained by several factors. First, China shifted from a system of agricultural communes to an individual household responsibility system beginning in the late 1970s. The communal welfare fund disappeared, and with it went the major source of CMS financing. The source of support for the Barefoot Doctors was gone, as were CMS funds to pay for drugs and other health services.

6.7 As a result of the disbanding of China's communes, just 40-45 percent of China's villages were still covered by CMS by 1983. At about that time, an ideological shift prompted some high government officials to declare that the remaining CMS programs should be abolished. Thus, most communities that still had CMS in 1983 subsequently disbanded their programs by the mid-I 980s.

6.8 Even before its demise, patronage and corruption had weakened China's CMS. The CMS was controlled and managed by local officials, some of whom used their power for selfish gains. As a result, farmers lost confidence in the government-run CMS and refused to pay premiums once the system became voluntary after the late 1970s. This experience underscores the importance of effective organization and management of any new community-financing schemes, and the need for an adequate financial base.

6.9 The government did not replace the CMS with a new health-financing structure, but instead adopted a laissez-faire policy. In response, many communities voluntarily designed their own new funding mechanisms. Many villages fell back on a system of fee-for-service payments. In some villages, the poorest farmers, who could not afford to pay for services, could ask for support from the village welfare fund. That fund was maintained by a specific tax ofjust under 5 percent of the village's net output for farm production. One-fifth of this agricultural tax was designated for welfare assistance and to defray health care costs of those in need.

China's Recent Experience with Community Financing

6.10 China can also look to its more recent experiences with relatively small-scale community-financing schemes to identify sound policy options in this area. These include the completed Sichuan Rural Health Insurance Experiment (see Mao 1995; Sine 1994; Cretin, Williams, and Sine 1995) and the ongoing World Health Organization (WHO) Fourteen Counties Study of Community Financing (Yang 1995). In addition, the Study of Thirty Poor Counties produced a wealth of data on rural health finance frequently cited in this report (Luo 1995 and China Network 1996). Surveys done in preparation for the World Bank-supported Rural Health Workers Development Project also revealed important information on the prevalence and operation of community-financing schemes in China today. These experiments and studies are described below.

Sichuan Rural Health Insurance Experiment

6.11 The Sichuan Rural Health Insurance Experiment, which took place in 1989-90, assessed the potential impact of insurance and coinsurance on the demand for health care and on the likely costs of providing services under an insurance regime. This experiment involved a sample of twenty-six villages from two counties and 40,443 individuals (Sine 1994).

6.12 Three insurance benefit plans were implemented with varying reimbursement rates for inpatient and outpatient services. Premiums were set at 1.5 percent of average income. Insured

51 individualscould visit facilitiesat the villageand townshiplevel, but could onlyvisit county hospitalsin an emergencyor with the approvalof the townshiphealth center. The experiment showedthat householdswere willingto join such a scheme--over90 percentof householdsin the test areas voluntarilyjoined the programand 95 percentvoluntarily re-enrolled after the first year. It also showedthat administrativecosts could be kept low (8 percentof total reimbursements).In addition,the studyfound that:

* Coinsurance(the requirementthat the patientpay part of the cost of healthservices at the point of service)exerted a significantnegative effect on demandfor care acrossdifferent population subgroups.No interactionswere foundbetween the effectof coinsuranceand age, income,or healthstatus.

* Withthe exceptionof one village,users surveyedstated a high degreeof satisfactionwith the insurancearrangement (Mao 1995,p. 16);

* Serviceswere used less when there was no functioningvillage health station, underlining the importanceof an adequatesupply of basic services(Mao 1995;Sine 1994);and,

* As in other countries,a smallproportion of the populationaccounts for a large shareof total healthexpenditures, underscoring the need for catastrophicinsurance. Approximately11.5 percent of the coveredpopulation incurred 70 percentof the total healthexpenditures.

WHO Study of Fourteen Counties

6.13 An ongoingWHO studyis assessingChina's existingcommunity-financing schemes to learn how to improveforns of organization,financing, and servicedelivery. The fourteen participatingcounties are locatedin Beijing,Henan, Jiangsu, Zhejiang, Jiangxi, Hubei, and Ningxia. In each county,a researchteam interviewed540 householdsand surveyedhealth services. The study is in an intermediatephase.

6.14 The studyhas foundthat a typicalcommunity fund mightcollect 5 yuan per person from families, I yuan per person fromthe village'ssocial welfare fund, and I yuan per personfrom the township. Coveragetypically involves a deductible(for example,100 yuan)paid fully by the patient,with copaymenton expendituresabove the deductible. Coverageof drugs is limitedto 120 kinds of medicine,including traditional Chinese medicines, with a maximumreimbursement limit set for diagnostictests.

Study of Thirty Poor Counties

6.15 The Studyof Thirty Poor Countieswas conductedin 1993-1995by a networkof Chinese universitiesand HarvardUniversity. It foundthat 16.5percent of villagessurveyed still maintained some type of community-basedhealth finance schemes, covering 11.6 percentof the sampled population. About two-thirdsof the schemescovered only primarycare servicesat the village level and one-thirdcovered comprehensive services ranging from primarycare to inpatientservices. All of the benefitstructures incorporated coinsurance features and often set very high copaymentrates for inpatientservices. Table6.1 summarizesthe prevalenceand benefitsof communityfinancing in the counties surveyed.

52 Table 6.1. Prevalenceand Benefitsof CommunityFinancing in ThirtyPoor Counties,1993 Numberof Percentageof Percentageof Type of Benefit VillagesCovered VillagesCovered PopulationCovered

Comprehensive 29 5.1% 4.4% PrimaryCare Services Only 59 11.4% 7.2% Total 88 16.5% 11.6% Source: Studyof ThirtyPoor Counties(China Network 1996).

Note: Comprehensivebenefit refers to cooperativemedical schemes that reimburse30 to 100 percentof hospitalizationfees for townshipand county-levelhospitals, as wellas 50 to 100 percentof outpatientfees. Primarycare services refers to coverageof fees(or discountedprices) for most village-levelservices, with feesat the townshipand countylevels paid out-of-pocket by patients.

6.16 The study found that the most prevalent type of community fund management was by village committee or by the village and township jointly. Table 6.2 presents the existing management arrangements in the thirty counties surveyed.

Table 6.2. Managementof Community Financing, 1993 BenefitType Management Form Comprehensive Primary Care ServicesOnly Township Government 17.2% 3.4% Township Health Center 20.7% 6.8% Villageand Township Jointly 20.7% 10.2% VillageCommittee 34.5% 47.5% Villageand Township Doctors 6.9% 32.1% Source: Studyof Thirty Poor Counties (China Network 1996).

6.17 The study appears to provide strong support for the reestablishment of community- financed health care. As part of the study, 11,044 randomly selected households were asked about their preferences for such organized financing schemes. Seventy percent of the households who were not covered by community financing or child immunization and MCH prepayment schemes responded that they would like to see an improved CMS-type scheme reestablished, while 88 percent of those covered by community-financing schemes stated they would like them to continue. Among the 70 percent of the uncovered population that favored reestablishing community- financing schemes, about one-fourth preferred to see such a scheme managed by the village, one- fourth preferred it to be managed by the township, and the remainder preferred joint management by the township and village or by the health facility.

6.18 Another indication of management preferences was obtained in the course of preparing the World Bank-financed Rural Health Workers Development Project. More than 1,000 village cadres in five provinces were asked about their preference for the management of community-

53 financed health schemes. Close to 60 percent of the village cadres preferred management by a village committee and roughly one-fourth preferred joint management by township and village.

6.19 In most rural areas, particularly poor areas, it is not possible to derive adequate revenues for any organized financing scheme solely from households. Funding must come from multiple channels. Both the Study of Thirty Poor Counties and the preparation work for the Rural Health Workers Development Project obtained information on the source of financing for existing community-financed health plans. The two studies found that approximately one-half of the revenue cane from household contributions, about one-fourth came from village social welfare funds, and approximately 10 percent came from the government. The detailed results are shown in Table 6.3.

Table 6.3. Percentageof Community Financing by Source Village Sources of Financing Government Social Househol Other Welfare d Fund Fundssurveyed in 30 CountyStudy (1993) 16.1% 20.3% 48.1% 15.5% Fundssurveyed in 5 ProvinceSurvey 8.0% 30.3% 58.7/o 3.0%

(1991) I _ _ _ I _ I_ _ I_ I Source: Studyof Thirty Poor Counties(China Network 1996) and WorldBank Rural Health Workers Development ProjectPreparation.

6.20 Table 6.4 presents information on the prevalence of community-financed health schemes and their benefit structures in Hebei, Shanxi, Fujian, Guizhou, and Henan provinces. The relatively poor province of Shanxi has the highest percentage of coverage--close to two-thirds of the villages maintained some form of community financing. But another poor province, Guizhou, has very few villages with community financing (only 0.8 percent). In these poor provinces, the schemes were largely financed by household contributions; benefits only covered primary care services because of the relatively small contributions that poor households were able to make.

PolicyChanges in 1994

6.21 The Chinese governmentannounced a new policy direction for the financingof rural health care in a front-pagearticle in the People'sDaily, on July 2, 1994. This new policy appearedto draw in part on lessons from China's earlier experiencewith CMS and its other experiences in community financing. The government called for the development of community-basedschemes to fund and organizehealth care for the rural population,guided by the followingprinciples:

* The government'srole is to establishpolicy and provideleadership; * Each communityorganizes its own collectivefinancing for basic healthcare; * Fundingwill be derived frommultiple sources (government, collectives, and individuals); * Priorityshould be given to coveringpreventive services;

54 Table 6.4. Prevalence and Benefits of Community Financing in Five Provinces, 1991

Of which Province Number of Percentage of Comprehensive Primary Care Villages with Villages with Coverage Servkes Only Community Community Financing Financing Hebei 3992 13.1% 42% 58% Shanxi 4727 65.6% 15% 85% Fujian 512 6.3% 25% 75% Guizhou 160 0.8% 6% 94% Henan 1590 6.2% 7% 93% Total 10981 12.2% 24% 76% Source: WorldBank Rural Health Workers Development Project Preparation (World Bank 1993b) Note: Data from the Studyof ThirtyPoor Countiesindicates that almost80 percentof the "servicesand drugs coverage"category is comprehensivecoverage. Therefore, for the purposesof this table,the Five Province surveydata was recategorizedwith "servicesand drugs"plans counted as comprehensivecoverage and the remainingcategories (services only, drugs only, other) counted as coverageof primarycare services only.

* The schemesand benefitpackage should vaiy accordingto communityconditions and economiccapacity; and * The schemesshould be supervisedby, and accountableto, the people.

6.22 The 1994policy announcementemphasized that communityfinancing should be voluntaryrather than compulsoryat the individuallevel. The governmentalso suggestedthree managementmodels with varyingdegrees of communitycontrol.

6.23 Althoughthese health fundingprinciples are alreadysupported by the State Council, there is wide variation in the interpretationof the policy in differentprovinces. There is some confusion,for example, over the Ministryof Agriculture'sdecree that the tax burdenon farmers must be reduced--someprovinces view this decreeas constrainingany community-financed schemesthat involve householdcontributions. As a result, the government'ssupport of community-financedhealth care has had little impactto date.

6.24 China's villagesand townshipswould have strongerincentives to develop community- based collective financingif the central governmentclarified and elaboratedits priorities. The governmentcould also provide technicalassistance and perhapsa subsidyfor the establishment and operationof communityfinancing. There are strong economicarguments for such a subsidy, in order to encourageequitable and efficient risk-poolingmechanisms in rural areas. The basic elements needed to establish such equitableand efficient communityfinancing are set forth below.

55 Making Community Financing Work

6.25 An affordable and acceptable benefits package. The first major issue involves deternining what benefit package and coinsurance levels are acceptable and affordable to most community residents. While economic principles such as cost effectiveness should help guide the design of the benefit package, the end product must have popular support, and people must be willing to pay their share of the required premium. The Chinese Network of Health Economic Institutions developed several potential benefit packages to test consumer demand and costs. The prototype test packages cover the full range of services, but with higher coinsurance rates for drugs, demand-elastic services, and inpatient hospital care. In aggregate, the proposed coinsurance rate was about 30 percent. The estimated cost for these illustrative benefit packages was 30 yuan per person per year in 1993 prices, which amounts to 5 percent of the disposable income of "modest income" rural households (see Box 6. 1).

' ' 6'1' Estimating theaCost and Contentof

l- H-ii0-g00|tsiao and iLu0(1996) jused data from householdsurveys, financial records of health ii0 0 -0faciiies andresultsfof thieSichuan Rutal Health Experimentto model the likelyutilization iand costsi ;0 0---of.abasic p:ackg of servicesunder a: community-financing scheme. Sincethe majorityof Sthe:rural "opulation}scurently uninsured, he provisiontof a basic benefit packagewould increasethe 'Equanityoftservices!demanded. T o estimatethe likelyimpact of this, Hsiao and Liu used demand: elasticityassuniptions Xfrots estimatesfrom the SichuanRural HealthInsurance experiment-that a 10 percen inl:0crease intinsurance coverage would increaseoutpatient expenditures by 6 percentand ; l-0inpatiet-exeditures: by:4- percent. Thley:assumed that supplyis perfectlyEelastic becauseChina has:E l;an excess Ssupplyof personnel and beds at thievillage Xand township level. They also assumedthiat thei cormimuity-fiancinggschemewvould pay inpatientservices Eon a case-basedpayment, with payenelt - leveldetenined prospectively, ratherthan on a fee for servicebasis. This was assumedto generate 5percent15 savingseon inpatient costs.:

il;000ll00Thelebasic Xbenefitspackage modeled by Hsiaoand Liu wouldcover servicesat the level of ; -vWage,tQwnship health center, and countyihospital. This package includes catastrophic care, but thei :0 defini{tioniof catastrophiccare is limnitedby whiatthie county hospital can provide in termsof its 0;: ' 'eq''ipment,4rugs; and supplies,Xand personnel. (A countyhospital, for example,is not equippedand i0 : -staffedto catrtyoutopen- heart- surgery.) Two prototypepackages were developed,withi different E0 Joev ofcoiitsurance(mediumr and high). The mediiumlevel requireda 30 percentcopayment for " wnship healthcenter outpatient fees, a 40 ipercentcopaymnent for drugs, and a 35 percent - ioamn fforinpatient chargs at thiecounty hospitals.i Thlehigh level required a 30 percenti " cpaymetfotwnship: health center outpatientfees, 50 percentfosr drugs, and 45 percentfor: E: lllinpatitChares gat: the county hospitals.:istop-loss Ai level for coinsurancewas built in tat 500 yuan. 1 The packagewZith ahigh level of coinsurancewas estimatedto cost 28 yuan per person. The; i tpacag 4lwith a mediumilevel of coinsurancewas estimatedto cost 31 yuan per person. i; iii

' -Some ecoomists arguethiat health insurnce shouldonl covercatastrophic expenses. In praci0e:0one raely finds healthtinsurance Xarrangements that coveronly catastrophiccare, WhileIt E -i'jes' notvmake sensef to haveArisk pooling thiat excludes catastrophic expenses, there are sound0 i ; argumientsfor covering both cata;strophicand Xsonicnoncattrphic expenses in a risk-pooling 0 airl-;gems8ent.0 For ex'ampl1e,if only gcatastrophiccare gis covered under tinsurance, there are Xincentives; 0 ti providcare in a hospitalsetting (t4 beeligible for insurancereimbursement) when less costlybut --ote0onequally effectiveoutpatient care mightbe available.aIn addition,if only catastrophiccare is 0 l '0proided,,consumnersmay 0have Wdisincentives for seekcingealy treatment,although early treatmentibg0Is tpicalymore ost-effective. f 000Dj:00t;if03:;

56 The Sichuan Rural Health Insurance Experiment and household survey responses indicate that people want coverage for a wide range of products and services, from drugs to village doctors to county hospitals. Furthermore, rural residents seem to accept coinsurance payments. Field tests could reveal whether households are willing to make voluntary payments that will cover the expense of providing these services.

6.26 Choice ofproviders. To promote consumer satisfaction and competition on the supply side, beneficiaries should be given a choice of primary care provider wherever possible. If beneficiaries seek services from qualified providers outside of those included in the scheme, the collective financing scheme could reimburse them at a reduced rate.

6.27 Universal coverage and adequate size of risk-pools. If possible, enrollment should be mandated at the local level to minimize adverse selection. In designing schemes, consideration should be given to an adequate population size for risk pooling at the village and township level. The population of a village generally averages 1,000 people--a relatively small size for pooling hospital expenses. A township generally averages 12,000 people, which is adequate for pooling the risk of hospital expenses. The size of the optimal risk pool needs to be weighed against declining accountability as the size of the pool increases.

6.28 Referral and supervision. Effective referral systems need to be established between village doctors, township health centers, and county-level hospitals. Provisions need to be included for monitoring quality of care.

6.29 Provider payment. The method of paying health workers affects overall costs and quality of care. Strong consideration should be given to replacing fee-for-service payments with a salary plus performance bonus for village doctors. Village doctors would be responsible for providing public health and basic preventive and curative services. This payment method would minimize incentives to overprescribe tests and drugs. It would also reduce administrative expenses by eliminating costs for claims processing. Capitated payment contracts might be made for hospital services so that hospitals would have a stronger incentive to provide services efficiently to the covered population.

6.30 Organization and management. Misuse of funds, favoritism, nepotism, and, in the worst cases, corruption will destroy the solidarity necessary to make community financing work. Ideally, community-based financing schemes would be established as non-governmental entities, with directors elected by those enrolled in the schemes. There must be frequent and periodic financial and quality-of-service reports. Under such a system, residents of a community would have a strong sense of control over their community's financing scheme.

6.31 Adequate financing. Sources of financing for such schemes can include households, rural collective enterprises, village social welfare funds, and central, provincial, and local government subsidies. In the Sichuan Rural Health Insurance Experiment, farmers paid 1.5 percent of their incomes as premiums to buy health insurance, in addition to making copayments. Rural collective enterprises voluntarily fund community health in some richer villages. Currently, about one-fifth of the village social welfare fund is spent on health. Local government subsidies would vary with the local conditions and with the priority given to health care by the local government.

6.32 Government Support. To provide an incentive for the establishment and operation of community financing without resorting to politically unpopular mandates, the central government

57 should consider providing technical assistance in the establishment of community financing. It could also consider providing a matching grant to supplement the contributions of households, rural collective enterprises, village social welfare funds, and local governments. The subsidy might only go to townships in which collective-financing schemes meet certain basic government guidelines. There is an efficiency argument for such a government subsidy, in that the incentive would promote the development of efficient risk-pooling mechanisms. For illustrative purposes, the central government financing might initially be 5 yuan-per-person for a 30 yuan-per-person benefits package for nonpoor townships. (Poor villages might receive considerably more as discussed in the previous chapter.) This would grow over time with income growth. If, by the year 2001, 120 million people in rural areas received the matching subsidy, the costs to the government might be 1.2 billion yuan (120 million people times a 10 yuan per capita subsidy in the year 2001).

6.33 China's experience with community financing indicates that it may be a promising approach to reestablishing risk-pooling arrangements for catastrophic medical expenses in China's rural areas. Depending on the design of the scheme, community financing can also promote efficient service delivery. Nonetheless, China is a large and heterogeneous country. While community financing has many advantages in theory, it may prove difficult to administer on a widespread basis. It may, for instance, be difficult for the community-financing scheme to collect premiums, administer contracts, and remain solvent, especially in poor communities without much administrative capacity. This approach, therefore, needs to be phased in with technical assistance and systematically monitored and evaluated. While some general guidelines are important, as discussed earlier, local experimentation and adaptation should also be encouraged. Key design elements that could be tested include:

- Alternative benefit packages, including coinsurance levels, to test consumer demand and costs; -- Alternative provider payment arrangements (testing payment of primary care providers with a salary plus performance bonus and capitated arrangements for hospital stays and visits); e Alternative management models; and * Simple methods for monitoring quality and appropriateness of care, including pharmaceutical prescribing.

58 7. OPTIONS FOR EFFICIENT RISK POOLING IN URBAN AREAS

7.1 Health insurancecurrently covers about half of China's urbanpopulation. This is in sharp contrast to rural areas, wherethe vast majorityof the populationhas no access to any form of health insurance for catastrophiccare and must rely on informalrisk-pooling mechanisms such as the help of extendedfamilies. Whilecoverage is better in urbanareas, there are still urgent problems in the urban insurancesystems. The two formal insurancesystems, the government(GIS) and labor (LIS) insurancesystems, cover just 15 percentof China's total population,yet they accountfor 36 percent of China's health spending. There are many problems in the operationof the GIS and LIS, particularlyin terms of financial sustainability, labor market impact, and efficiency. The two urban insurancesystems are urgently in need of reform to avoid becomingan obstacleto economicmodernization and reform to the state-owned enterprisesystem. Change will be difficult,however, given the many powerfulgroups with vested interests in these insuranceplans. Graduallyexpanding insurancecoverage to uninsured groups in urban areas will be an additionalchallenge.

7.2 GIS beneficiaries,particularly retired veteransbut also civil servantsand university students,will oppose any reductionin their benefits. Enterprisesor governmentunits that have relativelyyoung workforcesor few retireesoppose pooling their risks with employersof primarily older workers. Governmenthospitals oppose paymentreforms that may reduce their revenuesor force their closure or downsizing. Various ministriesrepresent and defend conflictingpositions. The Ministryof Finance is mainlyconcerned with containinggovernment outlays. The State PharmaceuticalAdministration wants to maintainand increasedrug sales. The Ministryof Health wants to protect its hospitals' revenues. The Ministryof Labor defends the interests of China's state-ownedenterprise (SOEs), which themselves operate thousands of health facilities. Closinghospitals and firing staff is just as unwelcomeas closing a steel plant or a coal mine. It is not a viable policy option exceptat the margin.

7.3 These constraints require a balancingact. The governmentneeds to promote policies that control beneficiarydemand for health care, improveefficiency and reduce waste in hospitals to keep costs down, yet allow hospitalsto earn enoughto cover their costs. The urban health financeexperiments described below--in Shenzhen, Shanghai, and Zhenjiang/Jiujiang--are identifyingstrategies for nationalreforms of the GIS and LIS and, eventually,for broadening urban insurance coverage. These experimentsfocus on improvingincentives in the system, particularlyon provider paymentreforms and coinsurancerates. The most recentexperiments, in Zhenjiang and Jiujiang,appear to hold the greatestpromise as national models. As a result, in January 1996,the governmentdecided to replicatethe Zhenjiang/Jiujiangreforms in two additional prefectures,or cities, in each province,for a total of roughly50 sites.

7.4 This chapterreviews current experimentsin reformingthe GIS and LIS in selected Chinese cities. It then examinesthe issues involvedin expandingthese reforrnsand in extending insuranceto other urban residentsoutside the GIS and LIS systems.

59 Experiments in Reforming the GIS and LIS Systems

Mixed resultsfrom the Shenzhen Experiment

7.5 Shenzhen is a newly establishedcity of three millionpeople. One million are permanent residents and two million are temporaryresidents who tend to be young contract workers. In 1992, Shenzhenimplemented a new social health insurance program with the dual goals of extendinginsurance coverageand reducinghealth costs.

7.6 All employers were required to enroll their workers in the program. The insurance program provides a two-tieredbenefit structure. The first tier is an individualmedical savings account. The employerdeposits 6-10 percent of a worker's average annual wage (varyingby age) into this individualaccount, to be drawn upon only to pay for medical expenses. If expendituresexceed the amount availablein the individualsavings account, the patient is reimbursedfrom the second tier of financing--thesocial risk-poolingfund. For these expenses, the worker must pay a 10 percent coinsurancefee with a "stop-loss"ceiling. This ceiling for each worker was set at 8 percent of the wages that the worker earned in the previous year. Beyond this ceiling, the social risk-poolingfund pays 100 percent of the worker's medical bill. At the end of the year, 20 percent of any balance remaining in an individualaccount is carried over to the next year and 80 percent of the balance is transferred to the social risk-pooling fund.

7.7 Shenzhenfinances its social health insurance by an 8-10 percent payroll tax on employers, varying by the age-sex compositionof employees. Workersdo not contribute directly. Benefitsfor retirees are financedfrom pension funds, which pay 10 percent of the monthly retirementbenefits to the health insurancefund as a premiumcontribution.

7.8 The paymentsystem can be characterizedas fee-for-servicewith a cap. Providers bill the insurance fund for each item of service. Caps are set separatelyfor outpatientvisits and hospitalbed-days. On a quarterlybasis, the insurancefund reviewsthe bills submittedby providers and calculatespayments that include a bonus for bills below the cap and a penalty for bills exceeding the cap. For quality assurance, 5 percent is withhelduntil the provider passes certain quality performancestandards.

7.9 The Shenzhenscheme is managedby the ShenzhenBureau of Health Insurance, which is controlledby the city's Bureau of Health. The deputy director of the Bureau of Health serves as the director of the Bureau of Health Insurance, which has created a conflict of interest, described in more detail below.

7.10 The Shenzhenprogram has had difficultiesmeeting its stated goals of expanded coverage and reduced costs. Enrollmentand premiumcollection were problematic. Joint ventures and privately-ownedcompanies refused to participatein the compulsorysocial insurance program, which was establishedthrough regulationbut not through a new law. The Bureau of Health Insurancefound it almost impossibleto enforce the regulation. Consequently,only 5 percent (150,000) of thie3 milliontarget populationparticipated. Almost one-third of these enrollees(46,000) were previouslycovered by GIS. The majority of the remainder had been previously covered by LIS. Only 27,000 contracted workers (18 percent

60 of enrollees)in the new program were previouslyuninsured. The new program also enrolled some 8,900 retirees.

7.11 Shenzhenalso encounteredfinancial difficulties due to both adverse selectionin enrollment and fraud. Employerswho refused to enroll in the program disproportionately employedyounger workers, as shown by the fact that the average age of enrollees(36) exceeded the city-wideaverage (27) by 9 years. There is also evidencethat many employers decreased their premium contributionsby under-reportingwages. The magnitudeof such under-reportingis conservativelyestimated at 50 percent.

7.12 Shenzhenalso had difficultycontrolling hospital costs. Since the ShenzhenBureau of Health Insurance is largely controlledby the Bureau of Health, the health bureaucracy both manages and controls the financingof health facilities. Even though the insured population does not constitutea majority of patients, the hospitalsview insured patientsas a major source of revenue. Given the Bureau of Health's dominanceover the insurance operation, the conflict between providers and the insuranceadministration was resolvedmostly in favor of the providers, at the expense of cost control.

7.13 In short, Shenzhen's social insuranceprogram could not maintain financial solvency with the Bureau of Health Insurance acting in the role of both funder and provider. Because of this conflict of interest, a change is underwayto set up the health insurance administrationas an independentagency. The social insuranceprogram has also had difficultyenforcing participation by joint venturesand privately-ownedcompanies under the compulsoryinsurance regulation.

The Use of Global Budgets in Shanghai

7.14 The city of Shanghaiis a metropoliswith approximately14 millionpeople. Seven million workers are covered in principle, if not in practice, under the GIS and LIS insurance plans. Perhaps another 3 million people are covered as dependentsunder these plans. Shanghai faces problems similar to those of other urban areas--highhealth cost inflation,poor quality of services, a high proportion of uninsuredresidents, and little risk pooling.

7.15 About four years ago, Shanghaideveloped a yet-to-be-implementedplan to overhaul its health care financingsystem. GIS and LIS would be combinedinto a single city-widesocial health insuranceplan. Copaymentswould increase: insuredworkers would pay 15 percent of outpatient charges and 8 percent of inpatientcharges, up to a maximumof 1,000 yuan and 2,000 yuan, respectively. The plan includesa medical savingsaccount for individualworkers (4 percent of wages) and a social health insurancefund (15 percent of wages) with the majority of contributionsfrom the employer. The social insurance program will be managedby the ShanghaiBureau of Health. This Bureau will establisha department of social insurance, which will also handle the pension scheme. The individualaccounts will be managed by employers, and workers can withdrawfunds equivalentto the coinsurancethey pay for health services.

7.16 The long-terrngoal is to expandcoverage to workers who are employedby joint enterprises, smaller collective industries,and private enterprises. However, workers in the "Big Eight" industries(including railroads, airlines, coal, steel, and the postal service) would be permitted to set up their own systems.

61 7.17 Before implementingthis reform, Shanghaiadopted a new payment systemto control hospital cost inflation. From July 1, 1993, throughJune 30, 1994, hospitalcosts rose 53 percent over the previous year. In July 1994, Shanghaiimplemented a global hospitalbudget in which total costs were permitted to rise no more than 24 percent and drugs expenditures were permitted to rise no more than 15 percent. Prices were also reformed: fees for visits and surgeries were increasedto reflect more accuratelythe labor costs for these services; fees for CT and MRI services were reduced by 12-15percent. The hospitalcontinued to be paid on a fee-for-servicebasis. At the end of September1995, Shanghaifound that its global budget approach had kept hospitalcosts within the target rates from July 1, 1994, through June 30, 1995.

Promising Preliminary Results from the Jiujiang and Zhenjiang Experiments

7.18 A demonstrationhealth insuranceproject began in December,1994, in Jiujiangand Zhenjiang,two medium-sizedcities on the YangziRiver in Jiangxiand Jiangsuprovinces, respectively. The State Councilsponsors these demonstrations.These cities,each with about 2.5 million inhabitants,form the "cuttingedge" of urban healthfinance reform, and as such they have been visitedby more than 600 officialsfrom otherprovinces and municipalities.

7.19 In both cities,newly organized insurance centers collect Box 7.1. Wage Tax Flows FromContributors to insurance payments from enterprises Individualand CommonAccounts and public sector agencies,then commit these fundsto individual and group accountsby formula(Box 7.1). There are both individual worker and collective accounts. The Cnklbutwr: Employers10% Emplyees1% centers draw on these accountsto pay medicalfees. The personal medical savingsaccounts plus large 11% copaymentsshould encourage moderationof patientdemand yet Fns providestop-loss coverage to protect Funds: againstcatastrophic medical bills Coordinating Individual (Box 7.2). The insurancecenters Fund A% also introducedan essentialdrug list 5% 6 of 1,100Western and 500 traditional Chinesemedicines, and they reimburseonly medicinesfrom the list.

7.20 The Zhenjiangmodel includestwo additionalinnovations. First, the formerlyseparate GIS and LIS systemshave been combinedinto a singleinsurance center. Second,packaged fees are set per outpatientvisit andper inpatientadmission. The paymentrates are establishedprospectively.

7.21 All enterprisesin Jiujiangand Zhenjiangthat had GIS or LIS insurancewere asked to join this pilotstudy. By the end of August 1995,95 percentof the eligiblepopulation in Zhenjianghad joined and 93 percentof the contributionshad been collected. In Jiujiang,90 percentof the

62 ..-.-17.2.. TheThree Tiersof Financingleialth Services in the Jiujiangand Zhenjaug- ,-Experimentj . . . . n

Boththe Jiujiang-andZhenjiang experiments finance health expenditures: through three tiersof financing.The first tier :isthe individualaccount. The amountdeposited into ttis account dep,en,ds:on.awker's annual salary, A sum equivalentto. approximately6 percentof tfie '-wd'er' annualsalary is depositedperiodically into this individualaccount, with some variation by.age bracket. .Thesecondtier consistsof out-of-pocketexpenditures: for medicalbills that exceedthe m,o,untin a.workees inidividual account. .When a worker'shealth expenditures exceed the funds ,intheindividual account, the workerpays. a deductibleof up,.to5 percent-of annualincome --bef,ore-ceivingreimbursement froom the thirdtier of financing. Thethird tier is the socialrisk-pooling fund, which is designed to insureworkers aga'inst .e.inC:ial burdenof catatr,ophic The largerIlless. the expense beyond-the individual -d,dctible)jthe- more the socialrisk -pooling flid pays,with slightly, more'generous coverage-in ,, . ,j , , .;..,Zhenjiazig~ .,. .. .~.- ... ~ -. ~. -. --.s-Poiik . . . . - .. . -- - .. - eeruscvea- - : . . --..

Consider,-forexample a. worker with -n 'veag anual wage::of6,n00 uan. f e ''..,.,w'cerinrs medicalexnses in-agiven6year that'ttaiess ta6- per,ent ofhisor an,a-l L.'.'ess thlani36. yuan),then thoseeexpenses- are 'npl,,y paid'e'ndivid:aa from '. ''c,'' t.' If.i,n..thatyar tota mne,dicalexenitue o:-f56 -yua'are: crred, te-.individ- ' ..'c"ou'ntpays exO.uananmdhe'remain,ing 20 yuan come f,o.m currentacome, paid uaa deduct,i..,>ble..In,.neithe of- these twases does,te socialri5kooing- fud. c'e aypart of the:-- dicalbil'.oL 1-lowemer, if me icalexpenies:totalng .I 'yare ,ince , thesocial risk' :,,pooin fun'd:-wil,lpay ',lmo.st90 perce,n:tofthe,bill. Fo, -thii caaophic-medical' ex...... en...... se'--- .qialn o t$ two and .a half years'.inoe--'*e s'o'ial:rispoig fu-.d-liit t' wo'e' :f'na-nci,al-los,s'to: about a quarterof hisor herantnual saiaiyt .--- , ---. goverrnmentunits and 88 percentofthe enterpriseshad paid their contributions.In total, 415,000 people in Zbenjiangand 370,000people in Jiujiangare enrolledin the new plans.

7.22 Resultsfrom the first year of operationappear to be positive,as summarizedbelow, with evidenceof successfulcost containmentand reasonablecontrol of administrativecosts. (A thoroughevaluation is still neededto confirmthese results,however.) * Coveragegaps, especiallyfor teachersand for workersin deficit-riddenenterprises, have been eliminated; * The rates of over-prescriptionand relianceon expensivediagnostic tests were cut significantly; * The annual growthrate of aggregatehospital expenditures declined by 23-28 percentagepoints from the previousaverage annual rate of 33 percentbetween 1991 and 1994; * The quantityof servicesdeclined 9 percentand the bed-occupancyrate by 2 percent,when comparedto the previousyear; * Both inpatientand outpatientutilization fell for enrollees(Cai 1995a;Zhenjiang 1995; Yip 1996).

63 In early January 1996 the State Council decided to expand the Jiujiang- and Zhenjiang- style experiments to additional cities throughout China, even prior to formal evaluation. The State Council will evaluate the results in 1996.

7.23 There appear to be clear benefits from separating the funding of health services from their provision, as done in Jiujiang and Zhenjiang. This separation can encourage cost containment, efficiency, and service quality because the funding agency represents the interests of the consumers, not the health providers, and can bargain with hospitals and monitor service quality. So far, the Zhenjiang insurance center has focused on innovative payment methods for hospital services in terms of packaged fees. In the future, it must also monitor quality and appropriateness of care. In Jiujiang, the insurance center continues to pay hospitals on a fee-for-service basis, and tries to monitor appropriateness of care with retrospective claim review.

Issues in Expandingthe Jiujiang and ZhenjiangExperiments

7.24 Further evaluation of the ongoing Jiujiang and Zhenjiang experiments is needed, and systematic monitoring and evaluation need to take place as they are replicated in all provinces, to maximize the learning from all these experiments. Evaluation should include analysis of: 1) the impact of the individual accounts and deductibles on patients' demand for services; 2) the impact of payment caps or global hospital budgets on the quality of medical services; 3) how hospitals alter their behavior under the reform, including changes in efficiency, cost shifting, risk selection, and capital investments; 4) the appropriate level of wages for the individual account; and, 5) organizational changes induced by reforms, such as changes in referrals between health facilities.

7.25 In extending the demonstration projects to fifty more cities and prefectures in 1996, program managers could consider these potential improvements to the demonstration projects:

* Expand coverage to include dependents. * Protect the funds accumulated in the individual accounts and the reserves for the social risk pool by paying inflation-adjusted rates of return. * Establish modem scientific accounting and auditing methods, including public quarterly financial reports for the insurance centers. The operations of the insurance centers need to be as transparent as possible to win the confidence of beneficiaries, employers, and providers. * Include representatives of employers, workers, and providers in each city on a Board of Trustees to oversee the overall operations of the insurance program. * Direct insurance centers to pay hospitals on a case-based, packaged-fee basis categorized, perhaps, into trauma, surgery, medical, cancer treatment, pediatrics, and long-term infectious diseases.

Adjust prices for labor inputs, drugs, and diagnostic tests to end distortions that now cause much of the obvious waste in the health system.

7.26 One aspect of the insurance experiments deserves rethinking. Benefit costs and personal health accounts in the Jiujiang and Zhenjiang experiments transfer income from the frail to the healthy worker relative to the earlier arrangements. Previously, all GIS and LIS

64 wage tax contributions were available for risk pooling. A person who used no health services received no benefits while a worker with major medical costs paid nothing. Under these experiments, about half of wage contributions is deposited to individual accounts--much of which will never be spent on health care because many people never use health services. Meanwhile, very ill workers will exhaust their individual accounts, paying another deductible equal to 5 percent of their wages before the risk pool will pay a portion of the excess expenditures. In its evaluation of the experiments, the government should consider the acceptability of such outcomes, along with the cost savings generated.

Links Between Pension Reforms, Health Insurance Reforms, and Economic Modernization

7.27 Labor mobility has been extremely limited in China until recently. SOE workers, once hired, stayed with their employer for life. SOEs, in turn, met pension and health obligations to their workers on a pay-as-you-go basis. With lifetime employment, there was little need to make pensions and health care rights "portable" across employers, although problems of unfunded liabilities for pensions and retiree health benefits are already undermining the viability of pay-as- you-go finance. Further, linking pension and health benefits to employment makes workers dependent on their employers, not only for their jobs but in order to realize their pension rights and to have health insurance both while actively working and after retiring. Survival of the enterprise becomes, then, a political necessity even at the cost of open-ended public subsidy. To improve efficiency in the state enterprise sector, noncompetitive enterprises must be allowed to fail or restructure, and therefore labor must become more mobile. Delinking pension and health benefits from the enterprise helps to make this possible. The current system is now an obstacle to economic modernization, which requires labor mobility and state enterprise reform. Pension and health finance reforms are needed so that workers can transfer jobs without jeopardizing their pension and health benefits, to deal with problems of pay-as-you go finance and the cost of health care for retired workers, and to facilitate enterprise reform.

7.28 As for pensions, China appears to be moving away from a pay-as-you-go system operated by individual enterprises to a unified pension system, with a substantial part of retirement income coming from fully funded individual accounts. Benefits need to be reduced to make the new system financially sound. Coverage could be extended by gradually including all formal sector workers in urban areas and employees in large township enterprises. This reform would produce important benefits: it would delink pension administration from enterprise management, establish a level playing field for enterprise contributions, encourage labor mobility, contribute to capital accumulation, provide incentives for savings, provide poverty protection, and contribute to income security in old age (for covered workers). One issue in implementing such reforms is the cost of financing the transition, given the large outstanding unfunded liability of current pensioners in SOEs and the accrued pension rights of current workers. It appears unlikely that pension reform, in the absence of parallel reform of health finance, can meet its objectives of facilitating labor mobility and state enterprise reform. This report argues, therefore, that reform of enterprise health finance should proceed simultaneously with (and on similar principles as) pension reform.

Optionsfor BroadeningUrban Risk Pooling

7.29 In addition to these ongoing experiments with GIS and LIS, health insurance coverage needs to be broadened to included the uninsured in urban areas. First, including dependents fully under GIS and LIS arrangements would significantly expand insurance coverage in urban areas.

65 Second, over the medium to long term, some form of risk pooling needs to be developed to cover the increasing proportion of the labor force that works in both the formal and informal sectors in urban China, as well as their dependents and the elderly. As the experience from the Shenzhen scheme indicates, getting joint ventures and privately-owned companies to participate in social health insurance will be difficult. It will take many years to set up well-functioning insurance systems that cover the majority of the urban population. In the meantime, government support to public health and to basic services for the poor will help to meet the needs of the urban residents who are not covered by insurance.

7.30 Various combinations of public and private financing arrangements could be developed for risk pooling in urban areas. For the increasingly large urban population working outside the state-owned sector, the government could encourage individual participation in an expanded LIS, complemented by public support for the indigent. The result would be akin to the social insurance systems for health care that now serve much of Europe. It will be a major challenge to fully insure GIS and LIS family members as well as employees and dependents of locally-based enterprises and the growing group of mixed ownership ventures. Insurers, be they public or private, will need to be established, like the insurance centers in the Jiujiang/Zhenjiang models, or other variants.

7.31 A number of intermediate steps could be taken to support the long-term goal of establishing broad-based, equitable, and efficient urban risk pooling of health care costs. Some of these intermediate steps are already being tested in the Jiujiang and Zhenjiang insurance experiments.

- One option is to take a laissez-faire approach over the short to medium term and let private, voluntary insurance emerge, with some government regulation, to cover part of the gap in urban coverage. Employers could choose to include health insurance as part of their benefits and either self-insure or join a larger insurance pool. Individuals could join their employer's plan, if available, or voluntarily purchase private health insurance. There are problems with this model, however. Small employers would have difficulty purchasing private insurance. Workers would face the threat of losing insurance if they changed jobs, while individuals with high health risks would have difficulty purchasing insurance. * Another possible option is to require that firms offer health insurance to their workers, without requiring that employers contribute. This would at least provide the advantages of some risk pooling. * Alternatively, the government could mandate employer and employee contributions for health insurance. Initially, health care could be purchased by the firm. Ultimately, mandated contributions might be pooled in one fund, with the government contributing money to cover the indigent and elderly.

7.32 These various options for expanding risk pooling need to be carefully reviewed, especially in light of the experience in Jiujiang and Zhenjiang. Intermediate steps will be needed to phase in coverage and build up administrative and regulatory capacity. Implementing reforms will inevitably require some public financial support in terms of studies, technical assistance, and monitoring and evaluation.

66 8. RECOMMENDATIONS AND IMPLICATIONS FOR PUBLIC FINANCE

8.1 China had an enviable record in improving the health of its people between 1950 and 1980. The move toward a market economy that began in 1979 fundamentally altered China's financing and organization of health care, however. Without a coherent national health policy that appropriately defined the respective roles of the government and market in the reformed Chinese economy, the health system drifted--reacting on an ad-hoc basis to the nation's changing economic and social environment. Now China faces several major health care problems as summarized in Table 8.1.

Table 8.1. Summary of Major Health Care Issues in China

Issue Likely Causes

Health status: WhileChina's * Governmentspending for publichealth programshas declined. overallhealth statusindicators are relativelyhigh, child * Income inequalityin Chinahas grown. Some 100million rural mortalityrates have stagnated residentsremain very poor. despiterapid increasein standardsof living. * Accessto necessaryhealth care has declineddue to inadequate public financingfor the poorand lack of organizedfinancing for the rural population.

Accessto health care in China * Rapid health cost inflationmakes services less affordableto many. is inequitable * Dueto lack of risk poolingafter the collapseof the CMS,many poor and near-poorforego necessarymedical services. * Insurancecoverage is lackingfor much of the urban population includingmigrant/unregistered workers in cities and workers in private or collectiveenterprises.

Inefficiencyand wastehave * Misallocationof publicresources favors hospital services over led to many problems public health spending. includinga decline in clinical effectivenessand qualityof * Distortedprices encourageoveruse of drugs and high technology services. tests. * Duplicationof facilitiesresults from overlappinghospital systems and vertical health programs. * Epidemicprevention programs, clinics, and hospitalsall rely on user fees to financetheir operations,which encouragesoverprovision.

Health costs increasedby an * Distortedprices encourage overprovision of drugs and expensive averageof 11 percentper diagnostictests. person annuallybetween 1986-1993,in real terms. * Fee-for-servicepayment encourages overprovision. * Agingof the populationand increasesin chronic diseasesraise costs.

67 8.2 How the Chinese government addresses these health issues will strongly affect health conditions, equity and risk sharing, efficiency, and total health care costs. International experience shows that China's health care policy is at a critical juncture. If China continues on the present course, the problems will deepen and become more difficult to remedy. The United States, for example, failed to address its risk-sharing problems in the 1940s, leaving the worst health risks (the elderly, disabled, and poor) uninsured by the private market. Eventually, the government had to bear the burden for insuring the elderly, disabled, and poor, while 15 percent of the population remains uninsured today. Meanwhile, health costs escalated and consumed some 14 percent of GNP in 1995. South Korea and the Philippines did not correct their fee-for- service payment system and now face rapid inflationary pressure in health costs. Germany and Canada, on the other hand, took timely and effective action when problems appeared in the 1970s and 1980s. As a result, they were able to assure equitable access to health care for their citizens, relatively low health cost inflation, and steadily improving health conditions. Many OECD countries have undertaken health reforms in recent years that may contain some useful lessons for China, as summarized in Box 8.1.

Priorities of Government Health Spending

8.3 As a general principle, governments should not spend their limited resources for health services that individuals will readily buy for themselves. Tax revenues are difficult to collect, and government spending is often less effective at meeting individual demand than spending done by individuals themselves. These considerations underline the importance of using public sector funds for important goods and services that would not--or cannot--be purchased privately, such as cost-effective public goods for all and priority clinical services for the poor. A related principle is that government health spending and regulation should seek to leverage effective and efficient use of health spending by other providers and buyers of health services. In China, direct government spending on health, including GIS, was no more than one-quarter of total sector spending. Both individual buyers of health services, and the SOEs taken as a group, spent far more on health than did the Chinese central, provincial, and local governments.

8.4 Following these principles, to address some of the health care issues enumerated in Table 8.1, the Chinese government will need to substantially increase funding for public health programs, i.e. those activities that the government must finance if they are to be provided at adequate levels. Full funding of priority public health programs is therefore the first priority for public spending on health, as shown in Table 8.2. This report recommends increasing public funding of EPS over the medium term from its present level of 1.3 billion (in 1993) to at least 6.5 billion in 2001. This calculation is based on current cost estimates of fully funding tuberculosis control, expanded immunization programs, endemic disease control, health education, and other important programs, as well as projected cost increases to the year 2001. (As income is projected to double in real terms, EPS costs will also rise, particularly due to wage increases.)

8.5 In addition to financing high priority public health programs for all of China's population, the government's next health spending priority should be to subsidize essential health services for the poor on poverty assistance, or equity, grounds. Most of the poor and near- poor in China live in rural areas, and are concentrated in China's 592 poorest counties. This report recommends a blend of geographic targeting (aimed at poor administrative villages in China's poor counties) and universal finance of a limited number of services that particularly benefit the poor. Geographic targeting could be implemented through community-financing

68 Box 8.1. LessonsFrom Reform of OECD HealthSystems and Singapore

What can China [earnfrom the incrementalreforms that many OECD countrieshave undertakenin their health systemsover the past fifteenyears? OECDcountries have employeda wide varietyof financing,medical care providerpayment, and deliverysystem arrangements. Reforms have been aimedat achievinguniversal or near- universalaccess to care,containing costs (and improvingefficiency), and, more recently,improving consumer satisfaction. Lessonsfrom OECD countryhealth reformscan be summarizedas follows:

Extendingcoverage and risk-poolingarrangements

* With the exceptionsof Mexico,Turkey, and the U.S.A., all OECD countrieshave achieveduniversal access to health care, largelyfinanced through national and local governmentsor mandatedsocial insurance. * It is difficult to achieve universalcoverage in systemsthat rely heavilyon private,voluntary insurance. This is because private insurerswill competeon the basis of risk selection,and sub-groupsin the populationwith higher- than-averageexpected health expenditureswill not be able to obtain insurance. Privatevoluntary insurance can also impedelabor mobilityif workersare reluctantto changejobs because of a possibleloss of health insurance coverage. Someof these problemscan be amelioratedthrough regulation. * An alternativeto mandatinghealth insuranceis to mandatesavings, with the further stipulationthat the savings so generatedcan only be used to pay medicalexpenses--these are knownas "medicalsavings accounts"(MSAs). Singaporeset up a systemthat complementedMSAs with public financefor the indigentand catastrophic insuranceto cover exceptionallyhigh costs. One problemwith MSAsis that they lose many of the efficiency advantagesof risk pooling. At the same time, individualswith good health have an incentiveto spendtheir medical savings accounton health--perhapsunnecessarily--because the money is not fungible.Furthernore, MSAsand catastrophicinsurance are mostcompatible with fee-for-servicepayment methods, with consequences for cost escalationand medicallyinappropriate care. MSAshave not been successfulin controllinghealth care costs in Singapore--percapita health care costs have grownby 13 percentper annum since 1984-2 percentfaster than before the introductionof the system.

Cost containmentand efficiencyimprovements

* Increases in health expenditureshave not necessarilybrought about improvedhealth outcomes. * Fragmentedfee-for-service systems like those in the U.S. (and China),without a single payer or set of rules, have been the leastsuccessful in controllingexpenditures and guaranteeingaccess. * Consumerexpectations, new technologies,and aging populationscontinue to place upward pressureon health expenditures. * Most OECD countriesnow have surplusesof physicians,especially specialists, and hospital beds. These surplusesare a major factor in cost escalation. Once created,these surplusesare hard to reducefor politicaland other reasons, and are best not createdin the first place. * Cost containmentis a major issue in almostall OECDcountries. Empoweringconsumers and having money followpatients withinthe context of an overallglobal budget appearto be the preferredstrategies for controlling costs. Specificmedical care providerpayrnent approaches such as capitatedpayments and diagnosis-related groups (DRGs),if appropriatelyirnplemented, appear to controlcosts withoutcompromising access and quality.

Evidenceis clear fromOECD countries:fragmented systems that allowmany individualsto remainuninsured and rely principallyon fee-for-servicepayments, like China's, are inequitable,inefficient, and performpoorly in termsof health outcomes, Many countriesare paying close attentionto the refonn lessonsfrom the OECD. Sincethe 1980s,China, however, has been movingtoward less governmentsupport for priority public health activitiesfor all and clinicalservices for the poor, and more relianceon a fee-for-servicedelivery system. If current trends persist,China will move into the 21 st century with a poorly performinghealth system that is nonethelesscostly.

69 schemesor supply-sidesubsidies. If community-financingschemes can be establishedin poor areas, perhapswith technicalassistance from the government,the governmentcould channel subsidiesfor services for the poor through these schemes. China has considerableexperience with community-financingschemes. They offer many benefits, in terms of providingrisk pooling for the whole communityand efficienciesin the deliveryof care. Nonetheless,they are relativelydifficult to administer,perhaps especially in poor areas with the least administrative capacity at the village level. An alternativeor complementaryapproach would be to provide supply-sidesubsidies (subsidies to healthprograms or health facilities,in return for providing free care to the poor in the poorestcounties). Existingsubsidies to townshiphealth centers could, over time, be redirectedto the poorestcounties. It mightbe reasonableto target 25 yuan per person now, risingto 50 yuan per person in line with China's overall incomegrowth by the year 2001. The costs of these subsidiesby the year 2001 would dependon the numberof people covered. Using a rough estimate of approximately75 million people living in poor villages in China's 592 designatedpoor counties, the cost would be 3.75 billion yuan. This illustrative figure could be reduced accordinglyif the target populationgroup were smalleror if the per capita subsidywere reduced. These geographically-targetedsubsidies could be complemented by program subsidiesfor clinical servicesthat greatly benefitthe poor, perhapscosting I billion yuan per year.

8.6 In urban areas, other measuresare neededto improvethe urban poor's access to priority health services. Better fundingfor public health programswill help, but the governmentcould also consider redirectingsome of the existinggeneral hospital subsidies (4.5 billion yuan) to services for the urban poor. (Notethat the Ministryof Health arguesthat it is committedto providingsubsidies for basic salaries of hospitalemployees and that thereforereallocation is not possible.) These subsidiescould focus on public facilitiesin poor areas of the cities, on programsthat particularlybenefit the urban poor, or on individualpoor households,if individual targeting methods can be developed. Becausethere are considerablyfewer urban poor than rural poor, subsidiesto the urban poor would also be considerablylower, and are estimated here at 0.5 billion yuan.

8.7 A third priority for governmentfinance is to support reformsto prices and provider paymentmechanisms, and a fourth priority is to control investmentsand improveregional planning. These measurescan generatesector-wide improvements in efficiency. The illustrative costs of these measures have not been estimatedfor this report but are of a much smaller magnitudethan the first two priorities.

8.8 A fifth priority of public spendingon health is to promoterisk poolingamong the broader populationto provide protection,at a minimum,from catastrophichealth expenses. Table 8.2 includes the cost of a small per capita subsidyto rural communitiesthat operate community-financingschemes. Governmentincentives to rural communitiesto establish and operate community-financingschemes can be justified on efficiencygrounds: the government has an interest in seeing that efficient risk-poolingmechanisms are establishedin rural areas. This support might be providedin the form of a matchinggrant to communitiesthat agree to follow certain principlesas discussedin Chapter 6. The subsidymight be 5 yuan per person now, growingto 10 yuan by the year 2001 in line with overall incomegrowth. If 120 million members of the rural populationreceived this subsidy by the year 2001, the total cost to the governmentwould be roughly 1.2 billionyuan. Governmentsupport is also needed for reforms to improvethe efficiencyof urban insurancesystems and support expandedcoverage. As urban insuranceexperiments expand throughout China, the governmentwill need to ensure that these

70 are systematically monitored and evaluated to guide medium- to long-term policy choices. Table 8.2 includes 0.1 billion yuan in the year 2001 for these expenditures.

Table 8.2. Public FinanceImplications of ProposedHealth Initiatives: IllustrativeCosts (billion 1993yuan) 2001 1993 (Proposedor (Actual) Projected)

GDPand GovernmentTax Revenue GDP 3,451 7,500 General GovernmentExpenditures 450 1,400 Proposed IncrementalTobacco Tax 10 (20 percent,ad valorem)

Centraland ProvincialPublic Spending on SelectedPriority HealthPrograms Fully fund priority public health programs 1.3 6.5 Clinical servicesfor the poor Poor villagesin poor counties 1/ 3.7 Specificdisease programsubsidies 1.0 Urban poor 0.5 Subsidiesto promotecommunity-financing schemes 2/ 1.2 Urban health finance initiatives 3/ 0.1 Total Expenditures,Selected Programs 1.3 13.0 as share of governmentexpenditure 0.29 % 0.93% as share of GDP 0.04% 0.17% 1/ Subsidiesto the ruralpoor could either go directlyto healthprograms/facilities in poor rural areas for free clinicalservices for thepoor, or be channeledthrough community-financing schemes in poorareas, if theyexist. Forillustrative purposes, a subsidyof 50 yuanper person for the75 millionpeople in poorvillages in China's 592poorest counties was used for thiscalculation for theyear 2001. 2/ Thisreport recommends that the govemment consider a smallmatching grant that providesan incentivefor communitiesto operatecommunity-financing schemes. This small matching grant might be 5 yuanper capita now,growing to 10yuan by theyear 2001 in linewith income growth. If thesubsidy was for community- financingschemes for 120million people in ruralareas by thatyear, the costs would be about 1.2 billion. 3/ Thisestimate includes the costsof majorstudies on pricingand provider payment, monitoring and evaluation of urbaninsurance experiments, and possiblysome start-up funds for newinitiatives.

8.9 In the long run, these financing policies will yield high returns. They will generate savings and improve health conditions, especially of the poor. They will also pool risks, improve efficiency, and reduce health cost inflation while keeping the percentage of GDP spent on health care at a relatively low level. While the needs are relatively clear, the crucial question becomes: does China have the resources and will to accomplish these goals in light of other pressing economic and social issues the country faces?

Finding Funds for Public Spending on Health

8.10 Government revenues have declined in recent years as a share of GDP. To keep the budget deficit in check, the government's expenditures have been reduced from 33.8 percent of GDP in 1978 to 17.9 percent in 1994. Of the 17.9 percent of government expenditures in 1994,

71 14.1 percent were budgetary and 3.8 percent Table 8.3. China'sGovernment Expenditures in were extrabudgetary. China's budgetary International Perspective expenditures as a percent of GDP are far below those in most other countries. The Government Central central government's share of all government expendituresas government expenditures is also unusually low in China as percentof GDP expenditures shown in Table 8.3. as % of all government 8.11 China has to increase public expenditures expenditures for several urgent needs: All Countries 39.1 72.3 poverty alleviation, health, education, infrastructure, environmental protection, Industrialized 47.6 65.9 pension reform, and unemployment insurance. Countries The World Bank has recently recommended Developing 31.7 77.8 public finance reforms in China to increase Countries government revenues and enable the government to better address these urgent (Budgetary, needs. The World Bank's recommended tax 1994) measures and their revenue effect in the year (Extra 3.8 - 2000 are shown in Table 8.4. This report budgetary) takes these recommendations into account and SurcetWrld k 1996a. then evaluates their implications for the public Note: Data are averages over three years ending in 1987 finance of health care. or 1988.

Table 8.4. RevenueImplications of 8.12 With projected increases in government RecommendedTax Measuresin the Year revenue from the tax measures shown in Table 8.4, 2000 the proposed priority health programs enumerated Tax Measure Incremental in Table 8.2 should be easily affordable. However, Revenue Effect government officials are reluctant to plan for by 2000 expanded public spending on health given fiscal (Percent of constraints in recent years. If the actual increase in GDP) revenues is lower than that projected in Table 8.4, VAT 2.1 these programs could be phased in more slowly. Iniiulicma. Public funds recommended for the priority health programs would only total about 13.0 billion yuan Enterpriseincome tax 1.2 in the year 2001, or 0.9 percent of projected public Taxes on pollutants 1.0 spending. Furthermore, as discussed in Chapter 3, Payrolltaxes 0.9 there are strong public health reasons for increasing Total 6.0 the present level of taxation on tobacco products. If China does increase its tobacco tax, and directs Source:World Bank 1996a. much or all of the incremental revenue to priority health programs, the proposed health initiatives would be even more affordable.

72 Annexes

Annex 1. China's Health Expenditureand Health SystemsData

Annex 2. OutsideInfluences on Health Status

Annex 3. Deaths and DiseaseBurden in China

Annex 4. Child MortalityTrends in China

Annex 5. Health Indicatorsfrom DiseaseSurveillance Points System

ANNEX 1: CHINA'S HEALTH EXPENDITURE AND HEALTH SYSTEM DATA

75 Annex 1

Table A1.1. China: GDP, Nominal Exchange Rates and Price Deflators, 1978-1994

Nominal official exchange rate GDP: billions of yuan GDP price deflator index (annual average) current prices constant 1993 prices 1990=1 1993=1 yuan/U.S.S

1978 362.4 880.9 0.541 0.411 1.7

1979 403.8 948.2 0.56 0.426 1.6

1980 451.8 1020.8 0.582 0.443 1.5

1981 486.2 1067.4 0.599 0.456 1.7

1982 529.5 1158.6 0.601 0.457 1.9

1983 593.5 1275.2 0.612 0.465 2

1984 717.1 1471.1 0.641 0.487 2.3

1985 896.4 1669.6 0.706 0.537 2.9

1986 1020.2 1815.4 0.739 0.562 3.4

1987 1196.3 2027.2 0.776 0.590 3.7

1988 1492.8 2256.4 0.87 0.662 3.7

1989 1690.9 2348.0 0.947 0.720 3.8

1990 1853.1 2436.8 1 0.760 4.8

1991 2161.8 2664.3 1.067 0.811 5.3

1992 2663.5 3040.4 1.152 0.876 5.5

1993 3451.5 34515 1.315 1.000 5.8

1994 4500.6 3860.6 1.533 1.166 8.6

Source: China StatisticalYearbook 1995,p. 32.

.e7 Annex 1

Table A1.2. GDP, Health Spending and Population Growth, 1978-1993 (GDP and health spending in 1993 constant yuan)

Year-end GDP Total Health Health GDP Population per Spending Spending (billions (10,000 capita (billions Per Capita of yuan) people) (yuan) of yuan) (yuan)

1978 880.9 96,259 915 25.8 27

1979 948.2 97,542 972 28.7 29

1980 1020.8 98,705 1,034 31.6 32

1981 1067.4 100,072 1,067 34.2 34

1982 1158.6 101,541 1,141 39.0 38

1983 1275.2 103,008 1,238 43.0 42

1984 1471.1 104,357 1,410 47.9 46

1985 1669.6 105,851 1,577 49.0 46

1986 1815.4 107,507 1,689 54.6 51

1987 2027.2 109,300 1,855 62.5 57

1988 2256.4 111,026 2,032 71.1 64

1989 2348.0 112,704 2,083 77.8 69

1990 2436.8 114,333 2,131 84.8 74

1991 2664.3 115,823 2,300 94.2 81

1992 3040.4 117,172 2,595 103.3 88

1993 3451.5 118,517 2,912 132.1 111

Annual Growth Rates: 1978-86 1986-93 1978-93. Population 1.4% 1.4% 1.4% GDP per capita 7.7% 7.8% 7.7% Health spending per capita 8.0% 11.2% 9.5%

Source: China Statistical Yearbook 1986,p. 71; China StatisticalYearbook 1994,p. 59.; China Statistical Yearbook 1995,p. 32; Wei Ying, 1995. 77 Annex I

Table AI.3. China: National Health Expenditures, 1978, 1986, and 1993 (millions of 1993 yuan) 1978 1986 1993 Health Percent Health Percent Health Percent Expenditure health Expenditure health Expenditure health Funding Source expenditure expenditure expenditure

Government Budget 7,292 28 17,288 32 18,878 14 (excluding GIS) LIS and GIS 7,689 30 18,274 33 46,108 36 Out-of-Pocket Payment 5,268 20 14,185 26 56,106 42 Rural Cooperative 5,109 20 2,918 5 2,243 2 Medical System Other Sources 428 2 1,956 4 8,713 6 TOTAL 25,786 100 50,621 100 132,048 100 Total Health 3.0 3.2 3.8 Expenditure as Percentage of GDP Source: MOH(data providedin May 1995). HEIre-estimated 1993 out-of-pocket payments in March1996 but did not reviewearlier years.

Table AI.4. China: Government Spending on Health, Selected Years, 1978-93 (millions of 1993 yuan) Spending Category 1978 1986 1990 1991 1992 1993 Recurrent Health Budget 5,294 10,598 10,686 10,658 10,965 10,786 Recurrent Budget: 0 643 869 901 951 917 Traditional Chinese Medicine Recurrent Budget: 0 1,409 1,730 1,965 2,211 2,292 Family Planning Fund Research 160 197 132 134 177 251 Higher Education 0 0 0 899 857 899 Capital Investment 628 2,107 1,400 895 877 1,144 Other Ministries 1,209 2,335 2,113 1,226 1,406 1,415 Government Spending Subtotal .291 17288 16.930 16.680 17.443 18.878 Government Employees' 1,178 3,360 5,834 6,234 6,667 11,667 Insurance (GIS) Government: TOTAL 8469 20.648 22.764 22.914 24.1 25323 Source: MOH(provided in July 1995).

78 Annex I

Table A1.S. Cbina: Components of Recurrent Health Budget, Selected Years, 1978-93 (millions of 1993 yuan)

Spending Category 1978 1986 1990 1991 1992 1993/

Hospital Operating Expenses 1,7945 4,488 4,254 4,072 4,290 4,183 Subsidies for Health Centers 1,421 1,903 1,968 2,040 2,233 2,145 Epidemic Prevention Fund 944 1,544 1,583 1,655 1,948 1,305 (*) Maternal and Child Care Fund 0 295 401 429 536 324 (*) Pharmaceutical Control Fund 0 144 147 164 194 229 Professional Middle School Fund 264 511 503 517 590 626 Training Fund 09 100 0 108 142 156 Rural Cooperative Medical System 89 43 34 33 29 27

Fund______Kindergarten Fund 0 7 0 7 9 9 Indigent Patients' Hospital Fund 0 18 0 17 18 15 Other 783 1,662 1,789 1,610 1,856 1,919 Recurrent Health Budget: TOTAL 5,295 10,722 10,684 10,658 11,850 10,939 Source: MOH/DPF,May 1995;revisions of EpidemicPrevention Fund and Matemaland ChildCare Fundfor 1993 by HEI,March 1996.

(*) Prior to 1990,govemrnment budget data excludedrevenues from user fees;from 1991onwards, govemment data includeduser fees with budgetfigures. The budget data for EpidemicPrevention Fund and Maternaland Child Care Fund for 1993 wererevised to excludenet userfees, but 1991and 1992figures have not yet been revised. Discrepanciesin the total recurrentbudget figures between Table A1.4 and TableAl.5 arise as a resultof this inconsistencyin accountingfor sourcesof funds.

Table A1.6. China: Health Insurance Coverage and Spending, Rural Population, 1993 Health Total Health GROUP Population Incomeper Percent Services Services (millions) capita insured Spending Spending by (1993yuan) within the per capita Group (billion group (1993 yuan) 1993 yuan)

RURAL 900 750 10 60 54 Upper quartile 225 920 40 98 22 Upper-Mid quartile 225 668 0 67 15 Low-Mid quartile 225 489 0 44 10 Bottom quartile 225 361 0 31 7 Poverty Group (70) <300 0 <10 (1) TOTAL 1,200 2100 110 132 Source:Estimates based on data fromWei 1995. HEImade revised estimates of nationalhealth expenditures in January1996, adding out-of-pocket spending to totalhealth expenditure. Other sources are ZhaoZhuyan and LushengWang 1995(Iah 12, for rural income).Statistical Yearbook of China 1994provides data for urban income. HEI staff adjustedthe 1993National Health Services Survey data to estimatespending by sub-groups.

79 Annex I

Table A1.7. China: Revenues and Government Subsidies of Health Institutions Providing Hospital Services, 1993 (billions of 1993 yuan)

Source of Revenue Hospitals Township Health Total Centers

Revenuesrom: Medical Treatment 18.0 3.2 21.2 Sale of Pharmaceuticals and Others 26.8 9.2 36.0 Subtotal 44.8 12.4 57.2

Government Subsidies 4.6 2.0 6.6

TOTAL REVENUE FROM ALL 49.4 14.3 63.8 SOURCES

Source. HEI, based on MOH Statement of Health Budget and Expenditure: (Berman and others, 1995). These data differ slightly from those presented in Meng 1995.

Table A1.8. China: Average Number of Hospital Beds and Health Care Personnel, 1993 and 1994 1993 1994 Type of Hospital No. of No. of No. of Beds No. of Medical/ No. of Beds No. of Medical Personnel Technical Personnel Technical Personnel Personnel Hospitals at and above County level 137 189 145 139 192 148 Rural Township 16 21 18 114 120 17 Hospitals Other Hospitals 45 59 47 42 56 44 Source: Statistical Yearbook of China, 1995, Table 19-15, p. 667.

80 Annex I

Table A1.9. China: Utilization of Hospital Beds at County Level and Above, Selected Years, 1985-94

(Unit) 1985 1989 1990 1992 1993 1994

A. Government Hospitals Bed Turn-Over Rate Turn over per year 19.9 19.9 19.9 18.9 17.9 17.9 Bed-days in Use Days 320.9 315.9 313.9 303.9 267.9 263.9 Bed Occupancy Rate Percent 87.9 86.9 85.9 83.9 75.9 72.9 Average Length of Stay Days 15.9 15.9 15.9 15.9 15.9 14.9

B. SOE-run Hospitals Bed Turn-Over Rate Turn over per year 14.9 14.9 13.9 12.9 12.9 12.9 Bed-days in Use Days 253.9 259.9 255.9 245.9 221.9 226.9 Bed Occupancy Rate Percent 69.9 70.9 69.9 67.9 60.9 62.9 Average Length of Stay Days 16.9 16.9 17.9 17.9 16.9 16.9

C. All Hospitals Bed Turn-Over Rate Turn over per year 18.9 18.9 17.9 16.9 15.9 15.9 Bed-days in Use Days 302.9 299.9 296.9 286.9 259.9 251.9 Bed Occupancy Rate Percent 82.9 81.9 80.9 78.9 71.9 69.9 Average Length of Stay Days 15.9 15.9 15.9 16.9 15.9 15.9

Source:Statistical Yearbook of China,1995, Table 19-20, p. 671.

81

Annex 2

ANNEX 2: OUTSIDE INFLUENCES ON HEALTH STATUS

2.1 As discussed in Chapter 1, health policy is not the only important determinant of a nation's health outcomes. This annex discusses several key outside influences on a nation's health status, with specific reference to China.

The Effect of Income

2.2 The effect of per capita income on a country's health status becomes progressively weaker as incomes rise. In 1990, a doubling of per capita income from an initial level of $1,000 (adjusted for purchasing power parity) corresponded on average to a gain of approximately eleven years in life expectancy. A doubling from an initial level of $4,000 per capita, however, led to a smaller (but still substantial) increase of four years (World Bank 1993a, pp. 59-72). In China in the mid-1970s, provincial level data indicated that a 10 percent increase in income resulted in a gain of about eight months in life expectancy (Prescott and Jamison 1985). Studies undertaken at the household level similarly show a strong but declining effect of income on health indicators. This evidence points to the importance of reducing poverty as well as raising overall per capita incomes for improving health outcomes.

2.3 Rapid economic growth in China has led to rising incomes and contributed substantially to health improvements--particularly since, until recently, poverty reduction has accompanied growth. In the period following the beginning of economic reforms in China (1978-1985), the percent of the rural population living in absolute poverty declined from about 33 percent to about 12 percent (See Table A2.1). In the subsequent five years of continued high GDP per capita growth rates, however, the rural population in poverty declined by only about half of a percent overall and may have risen in some areas. In a reassessment of the 1985-90 data from four southermprovinces, Chen and Ravallion (1996) conclude that the poorest were better off than earlier analyses had indicated, although they Table A2.1. Income and Poverty in China, 1978-94 agreed progress in Percentof reducing the percentage of GDP Per Rural the population in absolute Capita GDP Per Capita Population poverty virtually ceased in Year (in 1993 Growth Rate in Absolute the period 1985-90. yuan) Poverty 1978 916 33.0 2.4 Continued rapid 7.8% (1978-85) income growth in China 1985 1577 11.9 therefore can be expected 6.0% (1985-90) to provide a base for real 1990 2133 11.5 (but modest) health 10.3%(1990-94) improvements for the 1994 3221 n.a. majority of China's population. But unless the Source: StatisticalYearbook of China, 1995, p. 32. next phase of economic World Bank (1992b p. ix) provides data on absolute property. grothaces every growth reaches the very

83 Annex2 poor, they will not have the means to TableA2.2. Illiteratesand Semi-literates,Selected escape from the sickness and Provinces, undernutrition that help keep them in 1982 and 1990. poverty. High overall economic growth rates can, however, help provide the GDPPer Illiteratesand resources needed to meet the priority Capita, Semi-literatesas a resourcesneeded tofmeetthe p riori.y (constant Percentageof the health needs of the poor. 1993 TotalPopulation'

The E.ffect of Education yuan) Region 1990 1982 1990 2.5 Education complements income Nationwide 2133 23% 18%

in enabling households to improve their *** ___ own health. Indeed, provincial data on Beijing 4707 13% 10% determinants of life expectancy in China Zbeijang 2259 24% 19% in the mid-1970s suggested that Guangdong 2424 17% 13% provincial literacy rates were more *** ___ directly associated with changes in life Gansu 1235 35% 31% expectancy than were provincial income Guizhou 849 33% 29% levels (Prescott and Jamison 1985). Shaanxi 1224 25% 20% Source: Statistical Yearbook of China, 1985, 1992, and 1993 2.6 Here, too, China's performance 'For 1982: Aged 12 and over. has contributed to progress. Table A2.2 shows the clear progress China has made in reducing the percentage of the population classified as illiterate or semi-literate between 1982 and 1990, although regional disparities remain. As with income growth, gains in education account for much past health improvement and establish a foundation for future gains.

Water, Sanitation, and Food

2.7 Cleanliness and quality of water relate directly to a broad range of diseases, while access to sanitation may be even more important to health. Relative to other countries in Asia with per capita income near that of China, China's levels of access to both water supply and sanitation are good, although differences in access between urban and rural areas persist. (Table A2.3 provides data on China and comparators.)

2.8 The quantity and Table A2.3. Accessto Safe DrinkingWater and Sanitation, quality of the food supply China and Comparators,1990 qaiyo h odspl also have a direct impact on Safe DrinkingWater Sanitation health. In extreme cases of food shortage, the lack of Country Urban Rural Total Urban Rural Total nutrients, or infections, can China 87 68 72 100 81 85 overwhelm the immune India 86 69 73 44 3 14 systems of the ______~~~~malnourished.In the period Indonesia 35 33 34 79 30 45 1959-61 China underwent Sri Lanka 80 55 60 68 45 50 such a famine, with extremely adverse Japan 100 85 96 100 100 100 consequences for mortality Source: World Bank 1994b, Table A.2. rates and health status more Note: Numbers in the table are expressed as a percent of total population.

84 Annex 2 generally. Short of famine, low levels of TableA2.4. Per CapitaNutrient Availability nutrient intake can combine with infection in China, 1950-95 to result in malnutrition, e.g. low levels of height-for-age or high levels of anemia. NationalAverage Daily per Year Capita Nutrient Availability 2.9 Table A2.4 shows China's progress Energy Protein Fat in increasing nutrient availability to its (Keal) (gm) Fat population in the period since 1949, both in (Kcal) (gm) (gm) terms of the quantity of the diet and in 1950 1742 49 22 increasing dietary quality (protein and fat). 1960' 1578 41 17 1970 2192 56 26 Fertility decline and demographic change 1980 2473 64 35 2.10 China's recent demographic history 1990 2679 66 51 compresses into a third of a century the 1995 2727 67 51 changes that took more than a century to ] 960 was the middle year of a 3-year famine in China that unfold in most of today's high-income pushedthe infant mortality rate up to around 330 per countries. China's rapid mortality decline, thousandfrom the levelof 200 achieved in 1957 (World combined with rapid growth in income and Bank 1984,p. 113). Energy availability per capita in 1957 educational opportunity, created the wasabout 2217 kcal. precondutional opportasunity eaiedf ithe Source:Piazza 1986. After1980: FAO 1994. preconditions for a sustained fertility decline. Strong family planning policies then lowered fertility rates to levels well below those actually desired by most of the population. Table A2.5 provides an overview of changes in China's key demographic variables. China's total fertility rate (TFR) declined from 5.8 to 1.9 between 1970 and 1995. The total number of births and the crude birth rate declined sharply as well, although less dramatically than TFR because of a concentration of population in the child- bearing years.

Table A2.5. Key DemographicIndicators for China: 1970,1995,and Projectionsto 2020 Year Variable 1970 1995 2020 1. Mid-year populationtotal (millions) 818.3 1201.4 1437.0 2. Lifeexpectancy at birth (years) 61.4 69.4 73.7 3. Median age at death (years) 39.8 67.4 70.2 4. Total deaths (million) 6.2 8.9 12.3 5. Crude death rate (per 1,000 population) 7.6 7.5 8.6 6. Total fertility rate 5.8 1.9 2.05 7. Total births (million) 27.3 21.4 20.3 8. Crude birth rate (per 1,000 population) 33.4 17.9 13.7 9. Percent of population over age 60 6.8 9.6 15.6 10. Percent urban population 17.5 30.3 50.8 Source: For percent urban, UN 1995; for 1970 indicators, Hill 1988; for 1995 and 2020 indicators, World Bank projections from official estimates.

85 Annex2

2.11 Declining TFR and birth rates have several important consequences for the health sector. In the short run, greater spacing between births improves the health of both mother and child. A reduction in the actual number of births--e.g.the decline in China from 27.3 million in 1970 to 21.4 million in 1995--reduces the need for obstetric care, immunization, and other child health interventions (although there is still a high unmet need for good quality obstetric and child health care in rural areas). These often constitute a significant fraction of the public effort on health in low-income countries. Finally, in the long run, declining fertility redistributes the age distribution of the population away from younger ages and toward middle and older ages (Figure A2. 1). The change in the age distribution of the population and the age distribution of death profoundly affects the pattern of demand for health services.

2.12 The rapid increase in rural to urban migration in China is another demographic factor with a major impact on health status. Rapid, sustained economic growth in coastal China since 1979 has created a huge demand for labor, mostly in urban areas. Between 1970 and 1995 the urban percentage of the population increasedrapidly from 17.5 to 30 percent (Table A2.5).

Although urban dwellers are required to be officially registered in the city where they live, in practice large numbers of rural people are gaining access to the cities. Most are not registered urban residents, however, and, therefore, they are not eligible for the food and medical subsidies, state sector employment, and insurance arrangements available to official urban residents.

2.13 This large unregistered urban populationposes several problems for the health system. Their sheer numbers put a strain on urban health facilities. Moreover, their dense living conditions, poor hygiene, and inadequate access to medical care are conducive to the spread of infection. Large numbers of male migrant workers living away from their families create strong demand for prostitution, which generates circumstances particularly conducive to the spread of sexually-transmitted diseases and HIV infection. The living conditions of the transient urban population provide an environment for incubatinga range of infections affecting that population and capable of spreading beyond it.

86 Annex 2

FigureA2.1. Evolving Patterns of Age Distribution and Mortality In China, 1970,1995,and 2020

China 1970

75+ 70-74 65-69 60-64 55.59 50-54 45-49

25-29 20-24 15-19 UW4 10-14 w 11 5-9 Eaw-i I I I 0_ =0f" I __ I _ I 20 15 10 5 0 5 10 15 20 25 30 35 40 45 50 55 Percentageof totalpopulation Percentageof total deaths

China 1995

75+ 70-74 65-69 60-64 55-59 50-54 - 45-49 40-44 35-39 i 30-34 525 25-29! C0-24 15-19,f 10-141 5-9 0-4

20 15 10 5 0 5 10 15 20 25 30 35 40 45 50 55 Percentageof total population Percentageof total deaths China2020

75 1 1 1 1 ,v1 1 18 1 70S74ua WBproecton =

60-64 _ 55-591 . 50-54 45S49 f 40-44 11|f 35-39 : 30-34 25-29 20-24 1 15-19|||llll|ll

10-14 |

20 15 10 5 0 5 10 15 20 25 30 35 40 45 50 55 Percentageof totalpopulation Percentageof totaldeaths

Source:Wodd Bank estimates and projections. 87

Anmex3

ANNEX 3: DEATHS AND DISEASE BURDEN IN CHINA

3.1 This annex presents the results from China of a forthcoming assessment of the burden of disease for major world regions in 1990 with projections to 2020 (Murray and Lopez eds., 1996). Most assessments of the relative importance of different diseases are based on how many deaths they cause. This convention has certain merits: death is an unambiguous event, and the statistical systems of many countries routinely produce the data required. There are, however, many diseases or conditions that are not fatal but that are responsible for great loss of healthy life: examples include chronic depression and paralysis caused by polio. These conditions are common, can last a long time, and frequently place significant demands on health systems.

3.2 To quantify the full loss of healthy life, the World Bank and the World Health Organization undertook a joint exercise as background to preparation of the World Bank's World Development Report 1993. Investing in Health (World Bank 1993a).' Diseases were classified into 109 categories on the basis of the WrHO'sInternational Classification of Diseases (ninth revision). These categories cover all possible causes of death and about 95 percent of the possible causes of disability. Using the recorded cause of death when available, and expert judgment when records were not available, the study assigned all deaths in 1990 to these categories by age, sex, and demographic region. For each death, the number of years of life lost was defined as the difference between the actual age at death and the expectation of life at that age in a low-mortality population. For disability, the incidence of cases by age, sex, and demographic region was estimated on the basis of community surveys or, failing that, expert opinion; the number of years of healthy life lost was then obtained by multiplying the expected duration of the condition (to remission or to death) by a severity weight that measured the severity of the disability in comparison with loss of life. Diseases were grouped into six classes of severity of disability. Class 2, for example, which includes most cases of leprosy and half the cases of pelvic inflammatory disease, was given a severity weight of 0.22, while class 4, which includes 30 percent of cases of dementia and 50 percent of blindness, was assigned a severity weight of 0.6. (The approach to disability weighting in the 1993 World Development Report was revised and updated with more extensive inputs for estimates reported here.) The death and disability losses were then combined, and allowance was made for a discount rate of 3 percent (so that future years of healthy life were valued at progressively lower levels) and for age weights (so that years of life lost at different ages were given different relative values). The value for each year of life lost rises steeply from zero at birth to a peak at age 25, and then declines gradually with increasing age. These age weights reflect a consensus judgment, but other patterns could be used--for example, uniform age weights, with each year of life having the same value, which would increase the relative importance of childhood diseases.

3.3 The combination of discounting and age weights produces the pattern of disability- adjusted life years (DALYs) lost by a death at each age. The death of a newborn baby girl represents a loss of 32.5 DALYs; a female death at age 30 means the loss of 29 DALYs; and a female death at age 60 represents 12 lost DALYs. (Values are slightly lower for males due to slightly shorter life expectancy.) The sum across all ages, conditions, and regions is referred to as the global burden of disease (GBD).

The text in this annex drawsin substantialpart directlyon pp. 26-27of this 1993study.

89 Annex3

3.4 This annex reports the values for China of estimates of the numbers of deaths by cause and of disease burden (in DALYs) for the sixth (and final) iteration of disease burden estimates for 1990. The 1993 World Development Report reported the third iteration while the fourth (Murray, Lopez and Jamison 1994) appeared in a WHO-published compendium of background papers for the 1993 report. That compendiumincludes a detailed discussion of the methods and assumptions underlying the DALY as a measure and underlying the construction of the empirical estimates. In this annex, Tables A3.1 and A3.2 report the numbers of deaths in China by age, sex, and cause (in absolute terms and in percentage terms). Tables A3.3 and A3.4 report the disease burden (in DALYs) by age, sex, and cause. This work draws fully on Chinese death registration, sample surveys, and, particularly,the invaluable data from the Disease Surveillance Points (DSPs). Table A3.5 compares the distribution of DALYs by cause in China with other regions; Table A3.6 projects those comparisonsforward to 2020 on the basis of one set of plausible assumptions about the future course of the main determinants of health. Disease burden can also be decomposed by risk factor; Table A3.7 looks at the consequences of one such factor--tobacco use--over the coming years.

3.5 The method used to compute the burden of disease can also be used to track improvements in a nation's health over time by following changes in the national burden of disease. It is to be expected that an assessment of burden by cause for China in 1995--if it were assembled--would allow careful tracking of progress from 1990 in disease control and identification of specific areas where further effort was called for.

Disease Burden Findings

3.6 Noncommunicable diseases and injuries make up a growing share of China's disease burden. The marked increase in prominence of noncommunicable diseases results mostly from the aging of the population that will result from the drop in fertility over past decades; indeed age-specific rates from most NCDs (except the psychiatric conditions) are expected to decline although, based on current growth patterns in tobacco use, rates for tobacco-related diseases may increase. In 1990, an estimated 24 percent of China's disease burden was due to communicable, maternal, and perinatal conditions, comparedto 49 percent for all developing countries (Table A3.5). By the year 2020, the share attributable to communicable, maternal, and perinatal conditions is projected to fall to just over 4 percent of the nation's total disease burden (Table A3.6) although, for this actually to occur, it will be necessary to reverse the recent pattern of increases in child mortality (documented in Annex 4).

3.7 The major causes of death in China in 1990 were dominated by noncommunicable diseases and injuries. Chronic obstructive pulmonary disease accounted for 16.1 percent of deaths, followed by cerebrovascular disease at 14.3 percent, and ischemic heart disease at 8.6 percent of deaths. Acute lower respiratory infections, which are a major killer of young children and the elderly, are the fourth major cause of death (5.3 percent). Suicides (3.9 percent), stomach cancer (3.6 percent), liver cancer (3.3 percent), tuberculosis (3.1 percent), hypertension (3.1 percent), and lung cancer (2.5 percent), make up the rest of the top ten causes of death (Table A3. l).

3.8 This order changes considerably when the DALY measure is used for disease burden, incorporating disability and the extent to which mortality is premature. Chronic obstructive pulmonary disease remains the number one cause of disease burden. But it is followed by major affective disorders, acute lower respiratory infections, "other unintentional injuries,"

90 Annex3 cerebrovascular disease, perinatal conditions, suicides, anemia, "other digestive diseases," and ischemic heart disease (Table A3.3).

3.9 China's pattern of disease burden in 1990 is distinct from other countries, both developed and developing, in several ways:

* Chronic obstructive pulmonary disease (COPD) accounts for 8.5 percent of the total disease burden in China in 1990. This is nearly twice the share in established market economies (4.8 percent) and developing countries (4.3 percent). The morbidity associated with this disease is large, with consequentdemands on health care systems and loss of productivity. Smoking is the most important proven risk factor for chronic obstructive pulmonary disease in the high-income countries but, while it undoubtedly contributes in China, other factors are likely more important. Near equal rates of COPD (Table A3.4) in males and females (despite the fact that most smoking is by males) and a concentration of particularly high rates in the northeast suggest that outdoor (industrial) air pollution and indoor air pollution (from cooking and heating fires) probably dominate the determinants in China. On this substrata of exposure, though, smoking multiplies risk--so the smoking control policies discussed in Chapter 3 will play a useful role in control.

Lliver cancer is also strikingly high in China, accounting for 1.9 percent of disease burden in China compared to 0.3 and 0.5 percent, respectively, in established market economies and all developing countries (Table A3.5). This results from chronic lifelong infections with the and perhaps the consumption of aflatoxins. Hepatitis B also contributes to China's high levels of cirrhosis of the liver. China's immunization program, which now includes , can provide primary prevention against liver cancer and cirrhosis of the liver (see Chapter 3). Unfortunately, most individuals who will die of liver cancer in the next 40 years have already been infected, so even a successful program of immunization will be unable to avert the doubling projected for liver cancer's share of disease burden by 2020 (Table A3.6).

* Relative to other countries, China is characterized by an extremely high suicide rate, particularly among rural females. In China, suicide accounts for an estimated 33 percent of all deaths of women aged 15-29, and 15 percent of all deaths of men in that same age group (Table A3.2). It accounts for a striking 5 percent of all DALYs lost among females at all ages (Table A3.4).

* Unintentional injuries also make up an unusually large share of disease burden--13 percent. Drowning is one such cause. It is a serious cause of child death (6 percent of deaths in males aged.0-4 and 25 percent in males 5-14).

* Undernutrition and intestinal nematode infections (hookworm, roundworm,and trichuris) remain unusually important in China given its level of economic development. These problems are interrelated; the nematode infections likely account for a substantial proportion of China's heavy burden from anemia and some of its burden from protein-energy malnutrition. Intervention to control the intestinal nematode infections costs little and is highly effective; continued expansion of areas receiving iodine supplementation could accelerate progress against the still major and debilitating problems of iodine deficiency disorders.

91 Tabe A3.1. Nwba of DeathsIn Cline by Age, Sa. mandCm. 1990

mm" Nam N.J.. maw NL Maims mats Fusuim Fwasu_ Frmals Fbais Finals Falrh F_rmals C,nilen Al D"t AN N.j. Al Fsmat 0-4 5-14 15-29 30-44 45-56 0-a 70+ 0-4 5-14 15-29 20-44 45-50 g0-r 70+

Total 0u _aSs 4.,2z 4,056 505 a m 347 746 1.061 lAOS 56 e3 231 233 462 e 1.807 X I.Ca?lwict. Mgnimd P5_1. 1,405 705 697 342 1a 19 35 71 74 151 m 17 s7 35 37 47 124 W Co_ndc A, NImS" A Po r 544 515 220 61 9 14 33 as 50 75 a 9 I1 23 31 5 54 1.TLbamcbst km 173 104 3 1 7 15 47 46 53 3 2 7 16 24 24 30 2 STDs eAgW 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 iL gyphift I 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b. CNmlyd 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 C Gonwrhs 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 d. OwSTDO 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3. MN 0 0 0 0 0 0 0 0 0 0 0 a 0 0 0 0 0 4. DantwIisse 93 44 46 22 1 2 1 2 5 11 26 1 1 1 1 2 14 S.Chlkhood istw 53 25 25 24 2 1 1 0 0 0 22 2 0 0 0 0 0 A Psrmds 17 9 a a I 0 0 0 0 0 a I 0 0 0 0 0

b. Poo 3 I 1 1 0 0 0 0 0 0 1 0 0 0 0 0 0

c. pM_aia 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

d. Mas_N 15 6 7 * 1 0 0 0 0 0 6 1 0 0 0 0 0 a. Tetai 1 10 6 a I 0 1 0 0 0 7 0 0 0 0 0 0 6.Msttngo 41 21 20 7 0 1 5 3 3 2 7 0 1 5 3 3 2 7.Hop"" 34 23 12 2 0 3 5 7 2 4 2 0 1 1 1 3 3 OMa.ba 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

9.Tm catotaIsr 1 1 1 o0 0 0 0 0 0 0 0 0 0 0 0 0 a Triypwa.uw 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 b.Ch bgdsica 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 c diSensonisals 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

d. Ldbstmk 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 *. Lwnphdc Ows 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

t Onchomwdt 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

10.L _..y a 0 0 0 a 0 0 0 0 0 0 0 0 0 0 0 0

1 1. Doasi 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

12. Apanme 3 1 1 1 0 0 0 0 0 0 1 0 0 0 0 0 0

13. T,sdhmu 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

14. _is nnhlstodsa 7 4 3 0 3 0 0 0 0 0 0 3 0 0 0 0 0

aAseak 4 2 2 0 2 0 0 0 0 0 0 2 0 0 0 0 a

b. T',tdgt 2 1 1 0 1 0 0 0 0 0 0 1 0 0 0 0 0

c Hootwn. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

d. OUisturtds i 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

15. ois rnt.do. 33 19 14 3 0 1 2 5 3 5 3 0 1 1 2 2 6

B. aRssrsJ.yt dk 474 220 245 132 4 3 2 5 12 71 159 4 3 2 3 9 U4

1. Lowsnasordm 467 225 241 129 4 3 2 5 12 70 156 4 3 2 3 9 63

2. Uppaqumr 5 2 2 1 0 0 0 0 0 1 2 0 0 0 0 0 1

3 Caumsna 2 1 1 1 0 0 0 0 0 0 1 0 0 0 0 0 0

4.Othwrredy 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o 0 0 0

Source: Murray and Lopez, 1996. Table A3.1. Nwunbr of D_Ihs In CNn. by Ag. SB. and Ci, tsa1

mama m aim Males Mai ms M aIm FanWs Fwnaim Fsr,lim Finales Fwnsis Furake Finfish Conclon NsDOsU Al Maim AN Finmals 0-4 5-14 15-29 30-44 45-50 0o-6s 70+ 0-4 5-14 15-20 30-44 45-569 o-w 70+

(codinuaM C.Malsmslcordlom 30 0 30 0 0 0 0 0 0 0 0 0 20 a 2 0 0 1. Hsnonhfi 12 0 12 0 0 0 0 0 0 0 0 0 a 3 1 0 0 ZSpl 1 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 3.Edsnbd 2 0 2 0 0 0 0 0 0 0 0 0 1 0 0 0 0 4. Hypuimson 1 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 B. csbI @6imf 0 0 0 0 0 0 0 a a 0 0 0 0 0 0 0 0 S. Abodon 3 0 3 0 0 0 0 0 0 0 0 0 2 1 0 0 0 7.OhsmaWnml 12 0 12 0 0 0 0 0 0 0 0 0 8 a 1 0 0 D.P_lni co-re 276 133 144 133 0 0 0 0 0 0 144 0 0 0 0 0 0 1. P6gngaiIalsboi 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2. Lowbk hwsilt 46 24 22 24 0 0 0 0 0 0 22 0 0 0 0 0 0 3.OM spIrasmdtra41 147 67 60 87 0 0 0 0 0 0 60 0 0 0 0 0 0 4.OUp paina_ 63 41 42 41 0 0 0 0 0 0 42 0 0 0 0 0 0 E.NudloinI 60 31 49 17 3 2 1 1 2 5 31 4 2 3 1 3 6 1.Prn-a_rgrmubillon 35 14 24 8 0 0 0 0 1 4 is 0 0 0 0 1 3 2 bd d5sn"dW a 4 4 3 1 0 0 0 0 0 3 1 0 a 0 0 0 3.2wnAdnA 9 5 5 3 1 0 0 0 0 0 3 1 0 0 0 0 0 4. AIund 25 8 17 2 1 1 1 1 1 2 5 2 1 3 1 2 3 5. Oum, s _ul 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 N. NcoiwnaLed 6.400 3.531 2.920 66 29 106 199 504 933 1,583 95 19 60 130 376 612 1.606 A. _gens rmeopim 1.484 924 540 5 10 30 66 245 295 250 7 5 20 53 129 151 174 1.Mouth mid omphisaW 30 25 11 0 0 1 5 7 7 5 0 0 1 2 3 3 2 2.EsophapA 1es 129 60 0 0 1 a 31 47 44 0 0 1 1 12 25 23 3.Simwsdc 319 209 1O 0 0 1 9 55 78 o6 0 0 2 9 22 34 43 4. Colorectum 63 48 37 0 0 2 4 11 14 15 0 0 1 3 9 11 13 5 Uw 293 213 0o 0 I 6 41 77 56 34 0 1 2 10 23 22 24 BPe.P ro 32 20 13 0 0 1 1 4 7 7 0 0 0 0 2 5 6 7. Trdi ebondujUamV 216 152 66 0 0 2 3 35 60 51 0 0 1 2 15 20 26 8. heknamsocw slin I I 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 9. esl 26 0 20 0 0 0 0 0 0 0 0 0 0 6 a 5 6 1oeCari 21 0 21 0 0 0 0 0 0 0 0 0 0 3 7 a a 11.Corps ulai 6 0 6 0 0 0 0 0 0 0 0 0 0 0 3 1 1 12. Owy 10 0 10 0 0 0 0 0 0 0 0 0 1 2 3 2 2 13.PFi_s 5 5 0 0 0 0 0 0 2 2 0 0 0 0 0 0 0 14.addw le 1s 4 0 0 0 0 2 6 6 0 0 0 0 0 2 2 15.LWq*mm 23 15 6 0 1 3 1 3 5 3 0 0 1 1 3 2 2 1e.Lokqia 66 34 32 3 5 6 a a 3 3 3 3 7 a 5 3 4 17. Ohsrman 115 61 65 2 2 6 9 17 12 13 3 1 3 7 14 12 14 Oth Nsopman 21 11 10 1 0 2 1 2 3 2 1 1 1 1 2 3 2 Rab@ssmsgus 60o 27 33 0 0 1 3 a a 9 0 0 1 2 a 11 12 D C. NubitonWncdoans 14 5 9 0 0 0 1 1 1 2 2 0 I 1 1 1 3

Source: Murrayand Lopez,1996. TIb A3.1. NHmbet d Des In CNn. by Age, 5e, nnd Cme., IM

NW" Mois Mmm Mai Mim Mam Mabs Finis Femab Feuis Fel_ Feuds. Fedsl. Feadma C.n.tn Al ONSO Al Mask Al FOrmais 0-4 8-14 15-26 20-44 45-56 40-` 70+ 0-4 5-14 15-21 30-44 45-5 00-40 70+4 (~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~(

0. Nssm-psdimic 06 53 46 2 2 14 10 5 5 14 2 1 7 7 5 I 1is 1.MsN q e a a 0 0 0 0 0 0 0 0 a 0 0 0 0 0 0 2.DOIKd _scvaiddm 2 1 2 0 0 0 0 0 0 0 0 0 0 0 0 0 1 lpa s5 10 5 0 0 a 4 1 1 2 0 0 1 t 1 1 1 4.Epla 12 7 S 0 0 s 1 1 0 1 0 0 2 1 1 0 1 5 Mm .p im. 1 5 4 0 0 0 0 2 1 0 0 0 0 0 0 0 0 0 .LAkVMbkue*u@ndhwdumi Z7 11 16 0 0 1 0 1 2 6 0 0 0 0 1 3 10 7. PuIim..t dmme a a 2 a 0 0 0 0 0 2 0 0 0 0 0 0 2 e.mpbsdmds 5 2 * a a 0 1 1 1 0 0 0 0 1 1 1 0 6. Daigd ndsn0 0 8 0 0 0 0 0 a 0 0 0 0 a a 0 0 10. PTlD 0 0 0 0 0 0 0 0 a O O 0 0 0 0 O a 11. 1s1 mme pa_d 2a 14 12 2 1 a 2 1 1 2 1 1 3 * 1 1 2 E. Sene oqm I 10 a 1 0 0 0 2 3 2 1 0 0 1 2 3 2 1OhAM a 4 3 0 0 0 0 1 1 1 0 0 0 0 1 I I 2z Cidweem 6 3 2 0 0 a 0 1 1 1 0 0 0 0 1 1 1 lOWahrm.nse 6 2 3 a a 0 0 1 1 1 0 0 0 0 1 1 1 F. C _dm.lr 2.U 1.322 1246 10 4 25 47 152 337 704 7 2 25 42 143 254 772 1.UlmAadch_1_deemme 1e3 6e 04 a 1 10 7 13 12 2e 1 1 6 11 19 17 37 2. Mh4c hudIimmm, 76 aes 377 0 0 5 15 53 96 217 o a 3 12 37 74 251 S s C __3.C.,mbeLnaadr 1,272 e72 6o0 2 2 7 16 96 16 3S 2 1 s 13 71 137 372 4. ffu mnycadm s 33 33 2 1 4 3 7 5 11 1 0 4 4 6 4 13 s Hypmauim n 277 151 126 1 0 1 3 20 30 69 1 0 1 2 7 20 66 6. Oeiwdsc 26 12 16 5 1 1 0 1 2 2 3 0 4 0 2 3 4 G.RPAurM. 1,530 789 741 10 1 5 10 67 204 401 14 1 2 10 47 156 5S0 1.COPD 1.432 73 696 4 0 3 7 Q6 195 465 10 0 0 a 44 131 8e0 2.MAIna 35 1 16 0 1 1 2 3 4 a 0 0 1 1 3 3 7 3.OUm ,hqrs.y 63 33 30 5 0 2 1 2 5 16 4 0 1 1 1 3 20 H. r4=mm 411 241 170 17 3 10 32 57 57 65 10 3 a 12 30 32 as 1. PsWw d.mm 33 is 14 0 0 1 1 3 4 0 0 0 0 0 2 a a 2 CkhodsotI9em 167 123 64 2 1 4 24 37 34 23 1 1 2 6 17 16 g0 3.Appmvd.s 12 7 5 0 1 3 2 1 0 0 0 1 2 1 0 0 0 4. Ow dgmm 179 02 87 15 1 3 5 16 le 33 1 1 4 5 11 12 57 I. GOao-mumy 124 73 s0 2 2 9 11 14 25 2 1 6 6 8 11 16

1. NeIftlaMpims gI a 7 42 1 2 9 a 10 t 1 17 2 1 5 7 7 6 12 Z9 llgnpmsmic h lp*qt 7 7 0 0 0 0 0 0 0 6 0 0 0 0 0 0 0 3. O6i dW-aisany 16 10 9 0 0 0 0 1 2 5 0 0 0 1 2 3 4 SU1nCrm 12 7 5 0 1 0 0 1 1 4 0 0 1 0 0 1 2 J. Musubcslo-a 36 17 10 0 0 1 0 1 5 10 0 0 1 2 2 2 11 1. PlunmdduhO 2 1 1 0 0 0 0 0 0 1 0 0 0 0 0 0 1 2. Oso.UtiUs 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 Ols1mumaMo-salsel 34 16 16 0 0 1 0 1 4 10 0 0 1 2 2 2 II

Source: Murray and Lopez, 1996. Tdble A&. Numbe d Dr_I In CM.. by Age. Sea. , Caus. IWO

ma make Maim ma_m mums M_ai Ma Fram Fe_ma Famai FenaM Famal_a FPaaM PanaM Conlnea Al D0 Al Mae Al FPnl. 0-4 5-14 15-28 30-44 45-5 o0-es 70+ 0-4 5-14 15-2s 30-44 45-59 606 70+

(cen1num K CongudaMmmxUg_ 105 51 54 40 5 5 1 0 0 0 41 5 7 1 0 0 0 1.Abd 5damd 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Z Anw=W 36 15 24 15 a 0 0 0 0 0 24 0 0 0 0 0 0 3.Anwau.bda0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4SClullp 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 & CIOE@op d 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 G.EmaphIWaisda 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 7.RPnmagu.uds I 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 G. Dom syncenY 12 6 7 3 1 2 0 0 0 0 3 1 2 0 0 0 0 9. Congeddt tow d 20 16 13 11 2 3 0 0 0 0 7 2 4 1 0 0 0 10.SplnaMlda 12 7 5 a 1 0 0 0 0 0 4 1 0 0 0 0 0 L Oral hah 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1. O alnla 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2. Padodontaldine 0 0 0 0 0 0 0 0 0 0 0 a 0 0 0 0 0 3.Edmgtiy 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4.Ow orenbOM 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 IN. hnMW 1,020 s5c0 43 75 42 154 113 o1 54 71 71 27 114 5s 50 36 75 A. U didem o2ea 400 228 71 35 99 d0 51 30 42 62 24 36 21 22 17 43 %0 1. AMr d baaddwa 135 97 38 4 a 27 26 17 9 7 3 5 9 6 a 3 4 2. Pdeo*nl e5 38 27 5 2 7 7 9 3 a 2 2 9 5 2 3 4 l Fab e5 32 33 2 1 a 4 4 5 10 4 1 1 1 3 a 10 4. Rra 24 14 10 3 1 2 2 1 0 5 2 1 1 1 1 1 4 S. Drmm 147 91 50 30 22 20 6 4 3 5 20 11 7 3 2 2 5 6.0oewrngonsitnIuJii 190 128 02 26 7 37 23 1e 9 10 27 4 o a a 3 9 IL hkawal 304 190 204 5 3 55 44 30 24 29 9 3 75 36 26 19 32 1.SuIncS-lrd 343 156 154 0 2 42 37 27 23 25 0 2 78 33 26 10 30 Z HwIdd.md 'n' 51 30 20 5 1 13 7 3 2 0 9 1 2 4 2 1 1 3. Ww 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Source: Murray and Lopez, 1996. Tabm AI. Nui_w d Dgth in Chhba by Age.Sa. ad Come. IWb. by P_n:t

_. _.. WM_ _d. Nd. Nd.. Md. F _md..Funm.. F _nag Fund. F _nd..F _nm..Fu_d.. Comiton Al Deiw Nl Mg.. M Fund.. 0-4 5-14 15-2S 30-44 45-50 a0-4S 70+ 0-4 5-14 15--S 30-44 45-SB 00-eo 70+

Tobl D.gt ('000) e.ess 4,829 4.056 506 86 279 347 748 1.01 1,8o5 565 63 231 233 462 a96 1,807 X Petcwt 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 1W.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 1.Comnncmbb, Mdwaii and Peurlal Candflor 15.8% 14.7% 17.2% 67.7% 18.1% 7.0% 10.1% 9.5% 7.0% 8.4% 70.6% 27.1% 16.0% 15.4% 7.9% 6.8% 6.9% A lrEick & Pa,uik 6.1% 6.5% 5.7% 12.0% 10.1% 501% 9.4% 8.7% &6% 4.1% 11.6% 14.3% 4.9% 10.0% a7% 5.0% 3.0% 1. TLfukob 3.1% 3.6% 2.6% 0.5% 0.8% 2.5% 5.2% 5.3% 4.3% 2.9% 0.5% 2.9% 3.1% 6.3% 5.1% 3.5% 1.6% 2. STID ududlng HIV 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% a. SypNblll 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% b. Chanyda 0.0% 00% 0.0% 0.0% 0.0% 0.0% 0o0% 0.0% 0o0% 0.0% 0o0% 0.0% 0.0% 0o0% 0o0% 0o0% 00% c. Gononte 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% d. OUw STD. 0.0% 0o0% 0o0% 0.0% 0.0% 0.0% 0.0% 0o0% 0.0% 0o0% 0o0% 0o0% 0o0% 00% 00% 00% 00% 3 HIV 0.0% 00% 00% 00% 00% 00% 0.0% 00% 00% O0% O% 00% 00% 0.0% 00% 00% 0.0% 4. Diwrt*l dbes. 1.0% 0.9% 1.2% 4.3% 16.% 06% 0.4% 0.2% O0S% 06% 50% 1.9% 0.3% 0.5% 0.3% 0.3% 0.7% &.Childhood cdW 0.0% 0.6% 0.6% 4.7% 2.3% 0.2% 02% 0.0% 0.0% 00% 4.0% 2.6% 0.2% 0.1% 0.0% 0.0% 00%

AL Peuauls 0.2% 0.2% 0.2% 1.0% 0.7% 00% 00% 00% 0.0% 00% 1.3% 0 9% 0.0% 00% 0.0% 00% 0.0% b. Polb 0.0% 0.0% 0.0% 0.3% 0.0% 00% 00% 00% 00% 00% 0.2% 00% 00% 00% 00% 00% 00% c. DlpNthmw 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 0.0% 0.0% 00% 00% 0.0% 00% d. Munl. 0.2% 0.2% 0.2% 1.3% 1.0% 0.1% 0.0% 0.0% 0.0% 0.0% 1.1% 1.3% 01% 00% 0.0% 0o0% 0.0% & TMwm 0.2% 0.2% 0.2% 1.5% 0.6% 0.2% 0.2% 0.0% 0.0% 0.0% 1.3% 0.4% 01% 0.1% 0.0% 0.0% 0.0% 6.O.wngtin 0.5% 0.4% 0.5% 1.4% 0.1% 0.3% 1.0% 0.4% 0.3% 0.1% 1.2% 0.1% 0.4% 2.1% 0.6% 0.4% 0.1% Cy, 7. Hupds 0.4% 0.5% 0.3% 0.4% 0.4% 0.9% 1.4% 10% 02% 0.2% 0.3% 0.5% 0.5% 0.5% 0.3% 0.4% 0.2% 8. mdu a 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 00% 0.0% 0.0% 0.0% 0.0% 0.0% o.0% 0.0% 9. TrqpbuI dier 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% b. TCph' oxrdeek 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0,0% 0.0% 0.0% 0.0% 0.0% 0.0% b.Chmg.s'dlemm 0.0% 0.0% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% cSd.Lomo mis 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% d. LdgIrAIuh 0.0% 00% 0.0% 00% 00% 0.0% 00% 00% 00% 0.0% 00% 0.0% 00% 00% 00% 00% 00% . Lynphok fwadW 0.0_% 0o.% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% t. rOnchooaui 0.0% 00% 0.0% 0.0% 0.0% 010% 00% 0.0% 00% 0.0% 0.0% 0.0% 00% 00% 00% 0.0% 00% 11. LDpiM 0.0% 0.0% 0.0% 0.0% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 1 1. Dungg 0.0% 0.0% 0.0% 0.0% 0.3% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.5% 0.0% 0.0% 0.0% 0.0% 0.0% 13. JT nda.ptuIl 0.0% 0.0% 0.0% 0.1% 0.5% 0.0% 0.0% 00% 0.0% 0.0% 0.1% 0.5% 0.0% 00% 0.0% 00% 0.0% 13.T,tnuch nu 0.0% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 0.0% 0.0% 0.0% 00% 14.Irimdlnrmdodes 0.1% 0.1% 0.1% 0.0% 37% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 4.7% 0.0 00% 0.0% 0.0% 0.0% b. ATdch 0.0% 0.0% 0.0% 0.0% 13% 0.0% 0.0% 00% 0.0% 0.0% 0.0% 2.9% 0.0% 0.0% 0.0% 0.0% 0.0% b. iTr mhL 0.0% 0.0% 0.0% 0.0% 15.% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 1.6% 0.0% 0.0% 0.0% 0.0% 0.0% d Hoo_ ir" 00.% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0 0% 0.0% 0.0% 0.0% 0.0% 0 0% 0.0% 0.0% 0.0% d. OkhwrItri.in 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 15&omulrdiros 0.4% 0.4% 0.3% 0.5% 0.4% 0.3% 0.5% 0.7% 0.3% 0.3% 0.5% 0.6% 0.3% 0.4% 0.3% 0.2% 0.3% B.R*Wlrd kiwloi 5.3% 4.8% 6.0% 2t.1% 5.0% 1.1% 0.5% 0.7% 1.1% 3.9% 28.2% 6.9% 1.4% 0.8% 0.6% 1.4% 3.5% 1. Low mepkidy 5.3% 4.7% 5.9% 25.6% 4.9% 1.1% 0.5% 0.7% 1.1% 3.9% 27.7% 6.8% 1.4% 0.7% 0.6% 1.3% 3.5% ZLUpperrpkiray 0.1% 0.0% 0.1% 0.3% 0.0% 0.0% 00% 0.0% 00% 0.0% 03% 0,1% 00% 0.0% 00% 00% 0o0%

Source: Murray and Lopez, 1996. Tabl A32. Numw o Ded.h hI Ch. by Ago, Sm. aid Cme. 1994 by Pe_

Him Mi. Hi. mim Him Mi.. Mi.. F_nd.s Fumdin F_ndi F_mim F_nd. Fu_di Frndl. Cond_bn ANDmI AAllml.. AN Fwmis 0-4 5-14 15-29 30-44 45-50 0-89 70+ 0-4 5-14 15-29 30-44 4S-50 0-69 70+

3. O,Ut medIs 0.0% 0.0% 0.0% 0.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 4. ofM*" 0.0% 00% 0.0% 00% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% C. Materrldcondll" 0.3% 0.0% 0.7% 0.0% 0.0% 0e0% 0.0% 0.0% 0.0% 00 % 0.0% MO% a.8% 3.5% 0.4% 0.0% 0.0% 1. Hontu 0.1% 0.0% 0.3% 0.0% 0.0% 00% 00% 00% 00% 0.0% 00% 0.0% 3.5% 1.4% 0.2% 00.% 0.0% Z S s 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.4% 0.2% 0.0% 0.0% 0.0% 3, Edwpmnn 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.5% 0.2% o.O% 0.0% 0.0% 4. Hypwpt.rim ^ 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.1% 0.0% 0.0% 0.0% 5. b d kbr 0.0% 0.0% 0o0% 0.0% 00.% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% t Abon 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.7% 0.3% 0.0% 0.0% 0.0% 7. OUm nuw 0.1% 0.0% 0.3% 0.0% 0.0% 00% 0.0% 0.0% 0.0% 00% 0.0% 0.0% 3.4% 1.3% 0.1% 0.0% 0.0% D. Pulriouidiolq 3.1% 2.7% 3.5% 26.3% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 25.4% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 1. PWr INdlao 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 00.% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 2 Low bi,f ndgt 0.5% 0.5% 0.5% 4.8% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 3.9% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 3 BiEh phyio a-d trarna 1.7% 1.4% 2.0% 13.2% 0.0% 00% 0.0% 00% 0.0% 0.0% 14.2% 0.0% 0.0% 00% 0.0% 00.% 0.0% 4. 0U p.dril 0.9% 0.9% 1.0% 8.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 7.4% 0.0% 0.0% 0.0% 00% 0.0% 00% E. NlNk 0.9% 0.6% 1.2% 3.3% 3.0% 0.6% 0.2% 0.1% 0.2% 0.3% 5.4% 5.9% 0.9% 1.2% 0.2% 0.4% 0.3% 1.P,shn-ngr mdrujrib3n 0.4% 0.3% 0.6% 1.7% 0.1% 01% 0.0% 0.0% 0.1% 0.2% 3.4% 0.1% 0.1% 0.0% 0.0% 0.1% 0.2% 2 iodkudibdIry 0.1% 0.1% 0.1% 0.6% 0.7% 0.1% 0.0% 00% 0.0% 0.0% 0.5% 0.8% 0.1% 00% 00% 0.0% 00% 3.Vtaam Adibcy 0.1% 0.1% 0.1% 0.e% 1.5% 0.1% 0.0% 0.0% 0.0% 0.0% 0.6% 2.0% 0.1% 0.0% 00.% 0.0% 0.0% 4. Anwmbi 0.3% 0.2% 0.4% 0.3% 0.7% 0.5% 0.2% 0.1% 0.1% 0.1% 0.9% 2.9% 0.5% 1.1% 0.2% 0.3% 0.2% 5. CoUw r bhru 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0 0.0%.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 00% 11.N o,_anunbable 727% 73.1% 72.2% 17.4% 33.5% 37.8% 57.3% 79.6% 87.9% 87.7% 16.9% 30.4% 34.6% 5s.9% 81 3% ee 1% 69.0% AMiigr-t nwpim. 1a5% 19.1% 13.3% 10% 11.6% 107% 24.7% 332x 27.a% 13.9% 1.2% 79% 8.6% 227% 279% 21.% 9.6% 1. Mouihwid omp.ym 0.4% 0.5% 0.3% 0.0% 0.2% 0.4% 15% 0.9% 0.6% 0.3% 00% 0,0% 0.4% 0.8% 0.7% 0.4% 0.1% ZEgophmgLs 2.1% 2.7% 1.5% 0.0% 00% 0.2% 1.9% 4.2% 4.4% 2.4% 00% 0.0% 0.3% 0.4% 2.5% 3.5% 1.3% 3.8tomch 3&8% 4.3% 27% 00% 0.2% 0.4% 27% 7.4% 7.3% 3.7% 0.0% 0.3% 10% 3.9% 4.8% 4.9% 2.4% 4. CQkrdAn 0.9% 1.0% 0.9% 0.0% 0.2% 0.8% 1.3% 1.4% 1.3% 0.9% 0.0% 0.4% 0.4% 1.5% 1.9% 1.6% 0.7% 5. L1e 3.3% 4.4% 2.0% 00% 0.7% 2.0% 11.7% 10.3% 5.2% 1.9% 00% 0.9% 0.7% 4.1% 4.9% 3.1% 1.3% e.Parcm 0.4% 04% 0.3% 0.0% 0.0% 0.2% 0.3% 0.5% 0.e% 0.4% 0.0% 00% 01% 0.1% 0.4% 0.7% 0.3% 7. TmNw6BwrohmLwA" 2.5% 3.2% 1.6% 0.0% 0.5% 0.7% 0.9% 4.7% 5.7% 2.8% 0.0% 0.0% 0.3% 0.9% 3.3% 2.9% 1.5% 8. Miara and othrwIdn 0.0% 00% 0.0% 00% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 00% 0,0% 00% 0.0% 00% 0.0% 0.0% 9. Bre 0.3% 0.0% 06.% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 00% 0.0% 01% 27% 1x% 07% 0.3% W.CerV13 0.2% 00% 0.5% 0.0% 00% 0.0% 0.0% 00% 00% 00% 00% 00% 0.2% 1.1% 1.4% 0.9% 0.3% 11.Co,pwiirl^ 0.1% 00% 0.2% 00% 00% 00% 0.0% oO% 0.0% 0.0% 00% 0.0% 0.1% 0.2% 06% 0.2% 01% 12 Owy 0.1% 00% 0.3% 0.0% 00% 00% 0.0% 0.0% 0.0% 0.0% 0.0 0.3% 04% 0.8% 0.7% 0.3% 0.1% I&Pfoie 0.1% 0.1% 0.0% 00% 00% 00% 00% 0.0% 0.2% 0.1% 00% 0.0% 00% 0.0% 0.0% 0.0% 0.0% 14. BIddw 0.2% 0.3% 01% 00% 00% 0.1% 01% 0.3% 0.6% 0.3% 0.0% 0.0% 0.0% 0.0% 0.1% 0.2% 0.1% 15.Lyrrpho. 0.3% 03% 0.2% 00% 1o0% 0.9% 0.2% 05% 0.5% 0.2% 0.0% 0.0% 0.3% 0.3% 0.6% 0.2% 0.1% 1. LukwnIa o7% 0.7% 0.8% 0.6% 5.% 2.9% 1.7% 0.6% 0.3% 0.2% 0.5% 4.1% 3.2% 2.8% 1.2% 0.5% 0.2% 17. Olucem 1.3% 1.3% 1.3% 0.3% 2.9% 2.0% 2.5% 2.3% 1.1% 0.7% 0.6% 1.9% 1s% 3.1% 3.0% 1.7% 0.8% OltwN.oplnwm 0.2% 02% 0.2% 0.2% 05.% 0.5% 0.3% 0.3% 0.3% 0.1% 0.2% 0.9% 0.3% 0.5% 0.3% 0.4% 0.1% x

Source: Murray and Lopez, 1996. T bl A3.2. Numbhc cd Doihs hi Chins by Age. Sax, and Ce. 1990. by P:mwu

md.. -Nd Nd.. Md1_ mdim N Md. F1und.._ Femae Fendes Fn. Foud. Found. FondF Ckwnmi2 M Death Al Nd.. Al Foud. 0-4 5-14 t5-20 30-44 45-59 a0-89 70+ 0-4 5-14 15-20 30-44 45-59 60-89 70+

S. O1bi. melu. 0.7% 0.6% 0.e6 0.0% 0.4% 0.3% 0.8% 0.8% 0.7% 0.5% 0.0% 0.5% 0.5% 0.7% 1.4% 1.6% 0.7% X C. NulbonW ndsln. 0.2% 0.1% 0.2% 0O% 0.2% 0.1% 0.2% 0.1% 0.1% 0.1% 0.3% 0.7% 0.5% 0.5% 0.1% 0.1% 0.2% D. Nou-pyhNift 1.1% 1.1% 1.1% 0.5% 2.2% 4.8% 2.9% 0.7% 0.5% 0.5% 0.3% 1.4% 3.0% 2.9% 1.1% 0.9% 1.0% 1. MUsordNsdw diorder 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 00% 0.0% 0.0% 0.0% 0.0% 2I d dor idwdIWmdiv 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 0.O% 0.0% 0.0% 0.0% 00% 0,0% 0.0% 0.0% 0o1% 3.Psyd 0.2% 0.2% 0.1% 0.0% 0.0% 0.9% 1.0% 01% 0.1% 0.1% 0.0% 0.0% 0.5% 0.6% 0.2% 0.1% 0.1% 4.Ephp 0.1% 0.1% 0.1% 0.0% 0.4% 1.1% 0.4% 0.1% 0.0% 0.0% 0.0% 0.2% 0.9% 0.% 0.1% 0.1% 0.0% 5*Abohddpmou 0.1% 0.1% 0.0% 0.0% 0.0% 0.2% 0.7% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 01% 0.0% 0.0% 0.0% * AkhdnranddCwduiw1ia 0.3% 0.2% 0.4% 0.1% 0.3% 0.2% 0.1% 0.2% 0.2% 0.3% 0.1% 0.1% 0.2% 0.2% 0.3% 0.4% 0.6% 7.Poutdn's di..a. 0.1% 0.1% 0.1% 0.0% 0.0% 0.% 0.0% 0.0% 00% 01% 0.0% 00% 00% 00% 0.0% 0.0% 01% ° ILAmI. s 0.1% 0.0% 0.1% 00% 00% 00% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0% 0.3% 0.2% 0.1% 0.0% 0. Dpig dqwd- 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 00% 0.0% 0.0% 0.0% 10. PTSD 0.0% 00% 00% 00% 00% 00% 0.0% 00% 0.0% 0.0% 0.0% 00% 00% 0.0% 0.0% 0.0% 0.0% 11.Otwn o-p.ycNgb 0.3% 0.3% 0.3% 0.3% 1.5% 2.3% 0.5% 0.1% 0.1% 0.1% 0.2% 1.0% 1.3% 1.1% 0.1% 0.2% 0.1% E.Sor organ 0.2% 0.2% 0.2% 0.2% 0.0% 0.1% 0.0% 0.3% 0.3% 0.1% 0.1% 04% 0.1% 0.2% 0.4% 0.4% 0.1% 1.(Imonas 0.1% 0.1% 0.1% 0.0% 0.0% 01% 0.0% 0.1% 0.1% 01% 0.0% 0.% 0.0% 0.0% 0.1% 0.1% M.O% 2. Ciaab 0.1% 0.1% 0.1% 0.1% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.1% 0.2% 0.1% 0.0% & Othw r. orgou 0.1% 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 0.4% 0.1% 0.2% 0.1% 0.2% 0.0% F. C _udlomcsju 26.9% 27.4% 30.7% 1.9% 4.6% 102% 13.5% 25.7% 31.8% 39.0% 1.3% 3.5% 11.0% 16.1% 31.0% 36.5% 42.7% kD 1.Rhtumni hol d.mm 1.8% 1.4% 2.3% 00% 0.7% 3.5% 2.0% 1.7% 1.2% 1.4% 0.1% 1.1% 3.7% 4.7% 4.2% 24% 2.0% z nb h.oudh.m 0.6% 6.0% 9.3% 0.0% 0.0% 1.9% 4.4% 7.1% 9.0% 120% 0.0% 0.0% 1.3% 5.1% 5.0% 10.a% 13.9% a C _ubgviurdisea_ 14.3% 13.9% 14.8% 0.5% 1.9% Z5% 5.2% 13.1% 17.5% 19.9% 0.3% 1.0% 22% &8% 15.4% 19.7% 20.6% 4. kilamnn cardhe 0.7% O.7% 0.6% 03% 0.7% 1.5% 1.0% 0.9% 0.5% 0.6% 0.3% 0.7% 1.0% 1.7% 1.3% 0.5% 0.7% 5& Fouoion 3.1% 3.1% 3.1% 0.1% 0.2% 0.5% 0.8% 2.7% 3.4% 5.0% 0.1% 0.0% 0.4% 0.7% 1.6% 2.9% 5.3% 6. Odhrcarim 0.3% 0.2% 0.4% 1.0% 0.9% 0.3% 0.0% 0.2% 0.2% 0.1% 0.5% 0.7% 1.6% 0.1% 0.5% 0.4% 0.2% a RosIry 17.2% 16.3% 18.3% 1.9% 1.5% 1.0% 3.0% 9.0% 19.3% 27.2% 2.5% 1.2% 0.6% 4.1% 10.2% 19.6% 20.3% I.CCPD 16.1% 153% 17.1% 0.6% 02% 1.0% 2.1% a3% 1&4% 25.% 1.7% 0.2% 0.1% 13% 0.5% 189% 27.a% 2. Adhua 0.4% 0.4% 0.4% 0.1% 0.8% 0.3% 0.5% 0.4% 0.4% 0.4% 0.0% 0.4% 0.3% 0.6% 0.6% 0.5% 0.4% ao01 mq*iwy 0.7% 0.7% O.7% 1.1% 0.5% 0.6% 0.4% 0.2% 0.4% 1.0% 0.7% 0.6% 0.4% 0.2% 0.2% 0.4% 1.1% K. Digesti 4.0% 50% 4.2% 3.3% &9% 3.0% 9.3% 7.6% 5.4% 3.6% 3.3% 4.3% 3.6% 5.2% 6.5% 4.6% .6% 1.Piaubrwdues 0.4% 0.4% 04% 01% 0.1% 0.2% 0.4% 0.4% 0.4% 0.6% 0.1% 0.1% 0.2% 02% 0.3% 0.4% 0as% 2.Cblh cIUofd w 2.1% 2.0% 1.6% 0.3% 0.7% 1.4% 6.9% 50% 3.2% 1.3% 0l% 1.0% 1.1% 25% 37% ZS% 1.1% 3 APwdbII 01% 0.1% 0.1% 0.0% 1.7% 1.0% 0.5% 0.1% 0.0% 0.0% 0.0% 1.7% 0.6% 0.6% 0.1% 0.0% 0.O% 4.OUhK g.dlw 20% 1.9% Z1% 2.9% 1.4% 1.0% 1.5% 2.2% 1.7% 1.9% 3.1% 1.4% 1.0% 20% 2.4% 1.7% 2.0% .rilou-wkwy 1.4% 1.S% 1.2% 0.3% Z0% 3.2% 2.% 1.6% 1.3% 1.5% 0.3% 1.0% Z1% a4% 1.6% 1.6% 0.0% 1. N.pluN4pft& 1.1% 1.2% 1.0% 0.2% ZO% 31% 2.3% 1.3% 1.0% 0.9% 0.3% 1.8% zO.% 3.2% 1.4% 1.2% 0.0% B. du4gnpmd bphuhypho 0.1% 0.1% 0.0% G.0% 0.0% 0.0% 0.1% 0O.% 0.0% 0.3% 0o% 0O% 0.0% 0.0% 0.0% 0.0 0.0% 0oWoo-whwy 02% 0.2% 0.2% 0O% 0.0% 0.1% 0.1% 0.2% 0.2% 0.3% 0.0% 0.0% 0.1% 0.2% 0.4% 0.4% 0.2% SWbnDiseae 0.1% 0.2% 0.1% 0.1% 0.7% 0.1% 01% 0.1% 0.1% 0.2% 01s 0.0% 0.4% 0.1% 0.1% 0.1% 0.1% J. Miujo-i.Iid 0.4% 0.4% o.s% 0.0% 0.5% 0.Q% 0.1% M1% 0.4% 0.6% 0.0% 0.4% 0.9% 0.8% 0.4% 0.3% 0.6% 1. Rhuxndocd aouUwl 0.% 0.0% 0.0% 0o.0 0.0% QOS 0.0% 0.0% Q OS 0.0% 0O.% O.OS 0.0% 0.0% o.0% 0.0% 0O.% Z O d. 0.0% 0.0% 0.0% 0.% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0o.% 0.0% 0.0% 0.0% 0.0% 0.0% 3 OlhwmcLcio-duiid 0.4% 0.3% 0.4% 0.0% 0.S% 0.2% 0.1% 0.1% 0.4% 0.S% OL% 0.4% 0.S% 0.o% 0.4% 0.6% 0.6%

Source:Muffay and Lopez, 1996. Tabl AS2 Numwr d Deha hI Chka by Age.Son. aa CauM 1900, by Pm-t

M1 Mime Mi. MIW. MOMS MMI. Maim F _mii Fani.. F _ni. Fwnik. F _nim F _I Fnmn ConcRion M DudhaDi..s Al AMFunim 0-4 5-14 1S-29 30-44 45-59 00-a9 70. 0-4 5-14 15-29 30-44 45-59 -89 70.

K CorgwiaJbrtwmiiIi. 1.2% 1.1% 1.3% 7.9% 5.3% 2.0% 0.2% 0.0% 0.0% 0.0% 7.3% 7.3% 2.8% 0.6% 0.1% 0.0% 0.0% 1.Abdonm lvlmdldet 0.0% 0.0% 00% 01% 00% 00% 00% 00% 00% 00% 0.1% 0.0% 0.0% 00% 00% 00% 00% 2. AnuWpuh 0.4% 03% 0.0% 3.0% 00% 0.0% 0.0% 00% 0.0% 00% 4.2% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 3. Auomct iresia 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 4. Ca llp 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 5.Cadpiim 0.0% 0.0% 0.0% 0.1% 00.% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 0.0% dEoophug.eMii 0.0% 0.0% 0.0% 0.0% 0.0% aO% 00% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 0.0% 7.Rwialagwmn 0.0% 0.0% 0.0% 0.1% 00% 00% 0.0% 00% 00% 00% 0.0% 0.0% 0.0% 00% 00% 00% 0.0% 8. Domwrdom. 0.1% 0.1% 0.2% 0.5% 1.1% 0.6% 0.1% 0.0% 0.0% 0.0% 0.5% 1.8% 0.9% 0.2% 0.0% 0.0% 0.0% 9.Congwiih_tddhume 0.3% 0.3% 0.3% 2.2% 2.4% 1.0% 0.1% 0.0% 0.0% 0.0% 1.2% 2.7% 1.e% 0.3% 0.0% 0.0% 0.0% 18.9pIh*bIId 01% 0.1% 0.1% 1.2% 1.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.8% 1.7% 0.0% 0.0% 0.0% 0.0% 0.0% LOralah 0.0% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 00% 0.0% 1. DEian luS 0.0% 0.0% 0.0% 0.0% 0.0% 0 0.0%.0% 0.0% 0.0% 0.0% 000% O0.0.0% 0.0% 0.0 00% 00% Z PrkodVl di_mm 0.0% 0.0% 0 00.0% 0.0% 00% 00% 0.0% 00% 0.0 00% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 3. Edwtuli 0.0% 0.0% 00% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 00% 0o0% 4.Olhmraouihsil 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0o% 0.0% 0.0% M. h*mw 11.5% 1Z2% 106% 14.9% 48.4% 55.2% 325% 10.9% S.1% 3.9% 12.5% 42% 49.4% 24.7% 10.8% 5.2% 4.2% A I _*Ugimu 7.0% &3% 5.% 14.0% 44.5% 35.4% 19.9% 0.9% 2.8% 2.3% 11.0% 37.7% 1t6% 9.1% 4.8% 2.5% 24% 1. Mdor Milo a:c=klrs 1.S% 20% 0.9% 0.8% 6.8% 9.7% 8.0% Z2% 0.8% 0.4% 0.5% 8.7% 4.1% Z% 1.6% 0.5% 0.2% 2. Poib*i 0.7% 0.8% 0.7% 1.0% 1.8% 2.3% 2.0% 1.2% 0.3% 0.3% 0.3% 3.4% 4.0% 2.3% 0.5% 0.4% 0.2% 3. F_ 0.7% 0.7% 0.8% 0.5% 1.1% 22% 1.1% 0.0% 0.5% 0.5% 0.0% 1.2% 0.4% 0.4% 0.7% 0.7% 1.0% 4. Fh. 0.3% 0.3% 30.% 0.5% M8% 0.8% 0.5% 0.1% 0.0% 0.3% 083% 1.5% 0.4% 0.3% 0.2% 0.2% 0.2% 6 Diomg 1.% 1.9% 1.4% 6.0% 25.4% 7.3% 1.8% 0.0% 0.3% 0.3% 4.5% 17.0% 3.2% 1.2% 0.5% 0.3% 0.3% a. OIh1IWSdIifrk*mi Z1% Z7% 1.5% 5.2% &8% 131% 6.5% 21% 0.9% 0.8% 4.7% 5.9% 3.5% Z4% 1.3% 0.4% 0.5% B. h l 4.4% 3.9% 50% 0.9% 3.9% 19.8% 1Me% 4.0% 2.3% 1.0% 18.% 4.8% 33.6% 15.0% 6.0% Z7% 1.7% 1. &11-hilted 39% 33% 4.5% 0.0% 2.8% 161% 10.5% 3.7% Z2% 1.a% 0.0% 2.7% 32.8% 14.0% 5.8% 2.5% 1.7% Z Hmk dmand vlm 0.8% 0.8% 0.5% 0.9% 1.1% 4.0% Z1% 0.4% 0.2% 0.0% 1.8% 21% 1.0% 1.8% 0.4% 0.2% 0.1% & w 0.0% 0.0 0.0% 0.0% 0.0% 0.2% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% o.o% 0.0%

M

Source: Murray and Lopez, 1996. Table A3.3. 0urdan o atses,in Ctmnaby Age, Sex, nd Caue, 1990 fn DALYs)

Mals Males Melee Males Males Feles Feeles Femals Feeals Fmales Concdaon Ai ANMeale Al Fenales 0-4 5-14 15-44 45-59 60+ 0-4 5-14 15-44 45-5W i0+ TotlOAlLY 200,407 108,852 99,755 24,400 5,416 40,202 17,520 20,101 20,510 5,002 37.327 13,324 17,125 1.Canuscable, MalWnelrd Pell X Condron 50,448 23.938 20.508 13,074 2,132 4,703 1,701 1.488 15,788 2,000 6,201 1,199 1,243 & lasous endP e-'c 15,022 9,390 7,232 2,880 1,157 2,403 1,247 918 2,821 1,108 1,885 710 101 t. Ttharausd 4,155 2,440 1,700 94 27 900 796 830 107 0 753 410 350 2.STDeeakxg lIl 107 23 84 2 0 19 1 I 2 1 79 0 I A.Syphie 7 4 2 1 0 2 1 1 1 0 1 0 1 b. Cthwda a9 7 62 0 0 7 0 0 0 1 el 0 0 c.Gmnonte 31 12 19 1 0 10 0 0 1 0 18 0 0 d. Oth rST0 0 0 0 0 0 0 0 0 0 0 0 0 0 3.IY 3 3 0 0 0 3 0 0 0 0 0 0 0 4. aeal Dtasse, 3,85 1,m72 1.013 847 150 550 104 114 1,008 142 483 U 102 S. CltkhoodClalr 2,254 1,190 1,064 1,029 103 Sl 0 2 044 88 20 3 1 e. p ^ss 774 398 377 351 47 0 0 0 332 45 0 0 0 b. Polo 360 209 158 195 2 12 1 0 140 1 9 0 0 cDlpth* 7 4 3 3 0 0 0 0 3 0 0 0 0 d. 1asle 517 264 253 219 34 9 1 0 211 33 8 1 0 o. Ttnm 590 315 275 201 19 29 4 1 253 9 II 1 1 S. *sii 1,201 505 0a0 350 23 190 52 35 348 21 170 49 38 7. Hlspd 626 420 197 62 12 212 111 31 52 11 72 23 40 a. Make 58 30 26 4 9 14 2 1 4 a 13 2 1 O D.TTrmFclCslCr 243 199 43 a 24 155 10 4 2 5 32 3 2 0 a Triywacrael 0 0 0 0 0 0 0 0 0 0 0 0 0 b. Cheg'sOlees 0 0 0 0 0 0 0 0 0 0 0 0 0 c Se _eleasSss 27 17 9 0 I 9 5 3 0 1 4 3 2 d. Lalelneisl I I 0 0 0 0 0 0 0 0 0 0 0 a.L LympohacFleeld 214 181 33 6 22 146 0 1 1 4 20 1 0 t Oncho0 ciSs 0 0 0 0 0 0 0 0 0 0 0 0 0 10. LOpy 7 3 3 0 1 2 0 0 0 1 2 0 0 1. De0uue 29 14 le 3 10 0 0 0 4 12 0 0 0 12.IEJapaneseE'h 479 245 234 135 ea 19 2 1 134 00 18 2 1 13.Tracho 347 98 249 0 0 22 34 42 0 0 57 79 113 14.knlnlNul 1,383 714 008 9 062 43 7 3 9 510 40 a 3 a.Ascas 648 333 312 9 319 4 1 0 9 290 4 0 0 b. TeAle 017 319 208 0 318 1 0 0 0 296 0 0 0 cJaolssfl 119 52 se 0 14 38 7 3 0 13 30 6 3 d. Othwaeelnal 0 0 0 0 0 0 0 0 0 0 0 0 0 15.OlhWs O f55 555 400 114 53 227 118 55 118 44 139 50 60 B.rlapisatybteSons 12,378 5.755 6,023 4,844 303 190 87 325 5,700 297 195 51 a00 1. LOfeyrP_oy 11,924 5,533 0,362 4,738 200 183 85 321 S,058 207 153 49 290 2. UppLArsramoy 145 09 77 45 4 13 3 4 54 4 12 2 4 3.OMMaa 300 154 155 01 92 1 0 0 se 0 0 1 0 0 4.OthwrP_bsry 0 0 0 0 0 0 0 0 0 0 0 0 0

Source: Murray end Lopez1996. Table A3.S Burden Of D1m. In Cltna by Age. Sax, wed Cause. tcs On DALYS) (00

Mals Mele Male Male Melee Fwaks Funlee Finas Fiamde Fale. Cesjon Al AN Mele AN Fenale 0-4 5-14 15-44 45-50 40+ 0-4 5-14 15-44 45-50 00+ (m4n6nued) C.M CerdMon za2. 0 2,681 0 0 0 0 0 0 0 2,501 40 0 1.H _nhp 306 0 3se0 0 0 0 0 0 0 305 It 0 2. ISep 473 0 473 0 0 0 0 0 0 0 472 1 0 2SEc1uuyia 66 0 as 0 0 0 0 0 0 0 64 2 0 4. lypm84aln 23 0 23 0 D 0 0 0 0 0 22 1 0 5. Wb td Labor 576 0 576 0 0 0 0 0 0 0 578 0 0 e. Abedon 76 0 76 0 0 0 0 0 0 0 74 2 0 7.OIJierMemel 1,006 0 1,0W 0 o0 0 0 a 0 0 23 0 O.PetndCondm 10,279 4,953 5,326 4,063 0 0 0 0 5,326 0 0 0 0 E. N l8i 9,547 4,840 4,707 1,411 672 2,165 367 225 1,660 675 1,536 302 242 1.Pro8n-Enuen,ehilamn 2X024 au 1.167 799 3 11 5 20 1.153 3 12 3 17 2. ldneDeldenc 491 251 240 214 22 5s 0 0 204 21 15 1 0 3. VkwuAn ADDe6dwy 340 174 165 115 52 a 0 0 Ilt 4e 7 0 0 4.Anua a.001 3,576 3,115 264 595 2,131 361 204 363 603 1,6c0 30S 226 5.owut0Un 0 0 0 0 0 0 0 0 0 0 0 0 0 U. No.nIxI,w0le 121.248 53,361 57.667 6,179 1.ee8 23.348 14.140 17,626 6.,365 I.3m 23,457 11,046 1S66 A malgr4Nt pleeaF 16.076 ll,292 6,75 177 364 3,237 3,a63 3,630 230 194 2,120 2,125 2.115 1. moull oredphwyO 567 369 lee 7 0 191 lO9 83 0 I as so 41 Z Es.*Aw* I.773 1,237 635 0 0 163 460 574 0 0 47 164 306 3.S1sch 3,s3o 2.157 1,1C3 0 a 262 677 969 0 a 323 36 40 o 4.CleiVPaaun 1.029 577 452 0 8 200 175 194 0 6 125 153 0sa h ~ LIUvr 3,966 3,010 976 7 24 1,00 1,109 we 7 23 265 362 2 6 . Pce 2 192 107 0 0 42 62 87 1 0 12 30 64 7. Tmc;he.dbwl%AAAmg 2,005 1.440 636 7 1s 157 541 729 0 0 61 256 301 Me84fIlMM sun 14 8 a 0 0 3 2 3 0 0 2 2 2 S.Broad 322 0 302 0 0 0 0 0 0 0 178 138 79 o.cgAb 272 0 272 0 0 0 0 0 0 0 83 109 so I1.CcLe ud 61 0 S1 0 a 0 0 0 0 0 22 53 1s I2.ty 176 0 176 0 a 0 0 0 5 7 * 7 51 27 1& PfOe00 37 37 0 0 0 a 5 30 0 0 0 0 0 14.Uledket 170 136 34 0 a 19 33 63 0 0 3 7 as 15. LpWph4i 346 243 l06 0 32 104 64 54 0 0 69 45 22 l. LaiAaa 1.613 653 7eo0 9 193 42a 94 3S 106 1oo 423 e 45 17. duerCa, 1,654 Sol6 57 95 420 202 150 lt 46 312 226 167 kI Othw9e@ple- 873 454 419 00 30 1le 93 52 tz 36 16a 75 6 C oD __eeMeUa 1,064 531 563 a 14 164 I7 1 aee 1 3 143 164 215 D. NS*lmEesaln. 5m 157 419 20 14 Q 35 25 126 30 106 as 31 E.ehu -pas c 2,516 12,665 15,660 243 364 11,200 1.145 62 1SO 31 12,31 1,427 3o 1. M w_Mr aod 2.t75 4,74S e.23 0 a 4.050 sse 142 0 0 7.072 6s6 ao Z _NAS&d e b.rder. 3,713 l1A 1.706 0 0 1,795 se 32 0 0 1,665 57 36 3. PFyd 2,644 1,450 1.103 0 0 1,426 1S 14 0 0 1.1te 1s 10 4. Ep n 490 433 8 so 322 44 17 16 62 201 29 is 5 AbhIdDen_ 1,549 1,435 111 0 0 1.331 C2 15 0 0 106 6 I

Souice: Murray and Lopez 1996. Tbb A3.3. Bu.. of Msoe In CMbl by Age, Sn. ad Cam, 1M0 *n MYs

11les Mab* Male ma" mal Faina Fwnale F.ale Finale F_aFe Co"ll" 2A Min Al Fuslw 0-4 5-14 15-44 45-US 00+ 0-4 5-14 15-44 45-U 6`4

G.AWMn drUNWd _uwft i,m 723 803 63 25 es 172 365 62 19 62 165 466 7.P _nd C_mi. 97 40 46 0 0 0 20 28 0 0 o ls 20 a. umpi, dwal 234 146 186 0 0 128 14 a 0 0 1s0 19 a 9. Drq npla 18 170 le 0 13 151 5 1 0 1 17 I 0 1e PTD 42 167 282 a 29 tli 10 2 13 46 186 14 3 IIlObasIs-Ccmp wvCox 2,361 1.057 1,324 0 101 867 47 22 0 126 1.093 74 20 12.Pdc0Deolaw 1,141 396 745 0 35 306 52 0 0 35 626 72 12 13. Othw" xo-psdldabc 1,617 904 713 154 93 61 23 18 106 42 497 32 34 F. SW"Orgi 2,051 a23 1,228 44 0 Of 312 360 34 1 175 529 474 l. Gb a 4o05 260 545 7 0 30 135 82 9 0 70 346 110 2. Cauin0 1.032 497 535 17 0 42 IsO 283 3 0 46 153 333 3. Oahw3w0eOra 213 co 147 20 0 10 22 15 23 15 59 28 23 G. Ca,doea_r 22.882 12,198 10,663 334 158 2.727 3.285 5,60 248 62 2.541 2,553 8.251 1. PAiUwalcHtD,oh_ 2.361 1016 1.346 0 28 54e 215 225 28 30 BIG 346 32 2.Z Hc_naDhOle 6.004 3,271 2.733 0 0 614 8 1.731 0 0 445 as I.34 3. C_admina r I2aae 0.1 5.818 4,604 78 62 1.086 1,81 3.032 6O 24 578 1.261 2088 4. h gmidoryCarda 1.300 879 821 50 28 361 133 on 52 20 334 lie 101 5. Olw Cagic 2.306 1,316 1,079 197 40 148 320 61o 106 16 267 155 931 H. PAwo4y 22.265 12,066 10,197 810 376 2,840 2,665 5,347 701 240 2,222 2,030 4.014 1. COPD 17,610 0,s53 8.256 141 7 1,663 2,460 5,052 325 5 1,405 1,857 4,e84 o 2. Aslhna 2,755 1,538 1,216 173 340 76a 160 97 109 211 6n 136 es t-i 3. oUure.tAGO_ 1.701 077 723 50s 28 206 6e 166 357 25 140 37 1n 1.02esSn 10,252 s,ess 4,554 1.456 18 1,66 1.432 973 1,573 129 1.147 943 762 1. PapI Ulw l)W. 461 264 177 10 6 121 76 72 1 4 es 40 se 2- 0n of gm LMr 31060 2.121 958 74 23 8es 710 440 24 24 277 349 286 3. AppaWkWo 359 213 147 5 66 140 a 3 s 43 91 6 2 3. Od_wng- 6,352 3.061 32271 1,367 84 542 630 456 1,533 66 714 546 418 J.Ga.M-Usky 2.4ss 1,563 93 76 75 573 566 272 73n1 442 166 In 1.Nuphs$MW- 1.88 048 714 51 75 530 145 147 73 SI 360 104 117 2.Z - Pm a H_ 442 442 0 7 0 a 364 63 0 0 0 0 0 3.OurGi le-UWn_y 3s6 174 222 is 0 35 57 62 0 0 73 2 a7 K. hk_ hID-S _6a 2,837 1,206 2331 0 29 320 414 443 22 20 933 850 so I reesalddA0d8sd. 66s 217 44 0 1 U4 es 35 0 S 174 173 94 2.OsIeallW6s 2.1o3 76n 1,418 0 0 In 2m7 20 0 0 515 8a 313 &ONWRa a-sheolo 716 253 466 0 28 of 40 11a 21 14 244 7 n LCa.qmiaAbrugnea 8,242 3,050 3,193 2.67 172 166 5 0 2,761 17 253 8 0 1.h dm*dmaddsG 23 13 10 13 0 0 0 0 10 0 0 0 0 t Amwu.phum 1.314 805 a 5 0 0 0 0 8 0 0 0 0 3.Aan _adi 5 3 2 3 0 0 0 0 2 0 0 0 0 4. CM p . 2 44 36 44 0 0 0 0 36 0 0 0 0 5.C lhpdhl 28 1 13 13 0 0 a 0 13 0 0 a 0 EL p 9 al.d 4 2 2 2 0 0 0 0 2 0 0 0 0 7.NA g ds 1 9 9 5 0 0 0 0 9 0 0 0 0 L D6nsW _na 76 474 405 36 so 2 0 343 43 64 3 0

Souce: Murray and Lopez 1996. Tablh A.. Surdsn d O 3sesIn Coins by Ag. Sim and Caus, 1860 ,n DALY.)

MaMe mais mais SaM. Mahs Feats Fsmsks Fdals Feans. Fens Co ter Al An aks Al Fenats 0-4 5-14 15-44 45-5s 80+ 0-4 5-14 15-44 45-5 410+

9.Co.30d1h51¶de"" 2,603 1,306 1,237 1,190 76 O6 2 0 1.030 64 140 2 0 1O. Spl bds u3 412 471 376 34 2 0 0 420 39 3 0 0 1O. low 309 180 120 114 26 40 1 0 77 26 25 0 0 SQ S111 1,015 u00 514 1e8 38 65 go 113 162 35 62 01 124 I. DuCaSs 563 219 254 18 34 44 23 11 18o 31 41 20 II t Pdodmtlsmas 33 17 16 0 0 15 1 1 0 0 14 1 I 3. Edasnla, 356 178 182 0 0 0 74 102 0 0 0 as 112 4. O,WOFls8h" 42 9 33 1 1 8 0 0 1 4 28 a 0 MOSs. 3es 174 1Ig 39 43 26 29 37 43 0 III 18 23 1. btpah 35,712 21,333 15,379 4,363 2,399 12,091 1,8 we 4,348 1,001 7,870 1,078 es2 A. UdtunlrS 2e.970 17,091 9,879 4,164 2,283 8,936 1,223 505 3,907 1,403 3,403 GM 411 1.MOWS VslidsAdse 4278 3,025 1,251 152 332 2,116 314 109 111 314 824 152 49 t PSsong 1,544 662 e82 223 67 3e0 133 47 70 94 445 36 3e 3. Fol 4,485 2,502 1.884 603 340 1,303 164 92 781 230 50e 153 161 4. RF. 68o 308 204 122 73 ¶51 19 20 ad 109 56 19 25 5. D¶ndn 4,425 2,72 1,83 1,019 819 830 so 30 805 402 316 33 36 6.Oth UerLhASorlk*mias 11,559 7,454 4,105 2,042 632 4,055 526 ¶9s 2,074 317 1.300 212 112 B h tS...l 8.742 4,242 5,500 Is 136 3,154 482 301 351 127 4,286 475 2M1 1.Sag-kUdo 8,076 3,232 4,046 0 90 2,430 419 284 0 ee 4,089 447 286 2- mldoaw dvi@bl_o 1,638 904 854 180 45 0o0 43 17 351 62 190 27 16 O3. w_ 26 20 0 0 20 0 0 0 0 0 M1

Soure: Murray and Lopez 1996. Tabl A3.4. Buden d Disa n Ch by Age, Scm, indCa, 1900 (n DALYs) by p_ nt

_im_ Mass Maim im. aim F.iuaa Fuuakm Fema_m FII_. Fmnlms cmtmi Al A im AN F_im. 0-4 5-14 15-44 4b-b9 e o+ 0-4 5-14 15-44 45-5O 0*+

TOaWDALYsCOOO) 208 108,052 99,755 24,405 8,418 40,202 17,525 20,101 2G.519 5,052 37,327 13.324 17,523 X

Port 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

anclPwIrld Cordorm 242% 22.0% 26.6% 58.8% 33.2% 11.8% 9.7% 7.3% 59.5% 41.1% 16.5% 9.0% 7.1% A. lriacimmand Pamitt 7.5% 7.7% 72% 10.9% 18.0% 6.0% 7.1% 4.6% 10.6% 21A9% 5.1% 5.4% 4.0% 1. Tikacukbi 2.0% 2.3% 1.7% 0.4% 0.4% 2.2% 4.6% 3.1% 0.4% 1.7% 2.0% 3.1% 2.0% 2. STD* skiriV HIV 0.1% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 02% 0.0% 0.0% a. Syphimis 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% b. Chmndim 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.0% 0.0% c.Gonont 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% d. Owr STDs 0.0% 0.0% 0o0% 0o0% 0o0% 0o0% 0.0% 0.0% 0o0% 0.0% 0o0% 0.0% 0.0% 3. HV 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 4. Dicnho Disas 1.8% 1.e% 1.9% 3.5% 2.4% 1.4% 0.0% 0.6% 4.1% 2.8% 1.3% 0.7% 0.e% 5. Cthdh CkdCter 1.1% 1.1% 1.1% 42% 1.0% 0.1% 0.0% 0.0% 3.6% 1.7% 0.1% 0.0% 0.0% a. Perkma 0.4% 0.4% 0.4% 1.4% 0.7% 0.0% 0.0% 0.0% 1.3% 0.9% 0.0% 0.0% 0.0% b. Palo 0.2% 02% 02% 0.8% 0.0% 0.0% 0.0% 0.0% 0.5% 0.0% 0.0% 0.0% 0.0% c. Dptha 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% d. MLaaaa 02% 02% 0.3% 0.9% 0.5% 0.0% 0.0% 0.0% 0.8% 0.0% 0.0% 0.0% 0.0% 4> a. Tehr 0.3% 0.3% 0.3% 1.1% 0.3% 0.1% 0.0% 0.0% 1.0% 02% 0.0% 0.0% 0.0% a. Mm*ig1i 0.6% 0.8% 0.6% 1.5% 0.4% 0.5% 0.3% 02% 1.3% 0.4% 0.5% 0.4% 02% 7. HFpatra 0.3% 0.4% 02% 0.3% 02% 0.5% 0.0% 0.2% 02% 0.2% 0.2% 0.2% 02% 8. mhala 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 02% 0.0% 0.0% 0.0% 9. Trpiuml Ck.tor 0.1% 0.2% 0.0% 0.0% 0.4% 0.4% 0.1% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% a. TrIVPnoST_ 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% b. Chagaa' Dma_ 0.0% 0.0% 0.0% 0.0% 0 0% 0 0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% c. Sct _omiui 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% d. Lalhl'nwm 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% a. LynrrphbcFifbis 0.1% 02% 0.0% 0.0% 0.3% 0.4% 0.0% O.0% O.0% 0.1% 0.1% 0.0% 0.0% t. n0 1 _omimah 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 10. Lepros 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0,0% 0.0% It .DOigm 0.0% 0.0% 0.0% 0.0% 02% 0.0% 0.0% 0.0% 0.0% 0.2% 0.0% 0.0% 0.0% 12. JapwisaaEnc"at*4 02% 02% 02% 0.0% 1.4% 0.0% 0.0% 0.0% 05% 1.6% 00% 0.0% O.O% 13. Tratoia 02% 0.1% 02% 0.0% 0.0% 0.1% 02% 02% 0.0% 0.0% 02% 0.6% 0.0% 14. IrAIntstl Nwutodas 0.7% 0.7% 0.7% 0.0% 102% 0.1% 0.0% 0.0% 0.0% 12.0% 0.1% 0.0% 0.0%

a. A*u . 0.3% 0.3% 0.3% 0.0% 5.0% 0.0% 0.0% 0.0% 0.0% 59% 0.0% O.O% 0.0% b. TrWhr8g 0.3% 0.3% 0.3% 0.0% 5.0% 0.0% 0.0% 0.0% 0.0% 5.9% 0.0% 0.0% 0.0% c. HoalwnTI 0.1% 0.1% 0.1% 0.0% 02% 0.1% 0.0% 0.0% 0.0% 0.3% 0.1% 0.0% 0.0% d. Oltar Iridmeinl 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 00% IIS. Othw ri 0.5% 0.5% 0.4% 0.5% 0.8% 0.6% 0.7% 0.3% 0.4% 09% 0.4% 0.4% 0.3%

Source:Wharray and lpez 1996. TabA3.4. Bwdn D _ hinCli by Agm,S.. d C , 199 ( DALYs)by p_c:t

bWh Ibis. Ibis LORI= Shims F_y.i.. F_mm Fwnmie Fwy.i. F.n.lm Cmnonn Al Al aim. ANFrnmI_ 0-4 5-14 15-44 45-50 00+ 0-4 5-14 15-44 45-5 ao

B. Ruspqu Irinciio 52% 5.3% 6.6% 19.8% 4.7% 0.5% 0.5% 1.6% 21.8% 5.9% 0.5% 0.4% 1.7% 1.LowurReIpaIr 5.7% 5.1% 6.4% 19.4% 32% 0.5% 0.5% 1.6% 21.3% 4.1% 0.5% 0.4% 1.7% 2. Upp-R..pi y 0.1% 0.1% 0.1% 0.2% 0.1% 0.0% 0.0% 0.0% 02% 0.1% 0.0% 0.0% 0.0% 3. OWsMsM 0.1% 0.1% 0.2% 02% 1.4% 0.0% 0.0% 0.0% 0.3% 1.7% 0.0% 0.0% 0.0% 4. OthmrRemp.aiuy 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 0.0% C. Mayri Coardgg,g 1.3% 0.0% 2.6% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% a.9% 0.3% 0.0% 1. Hwnoanhou 02% 0.0% 0.4% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 1.0% 0.1% 0.0% 2. Sopss 02% 0.0% 0.5% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 1.3% 0.0% 0.0% 3. Ecbmns 0.0% 0.0% 01% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.0% 0.0% 4. Hyp pimo 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 5. Obmtnctd Labor 0.3% 0.0% 08 % 0°0% 0°0% 0°0% 0.0% 0°0% 0°0% 0.0% 1.5% 0.0% 0.0% 6. Aborlon 0.0% 0.0% 0.1% 0.0% 00.% 0.0% 0.0% 0.0% 0.0% 0.0% 02% 0.0% 0.0% 7. Othwr aslwrul 0.5% 0.0% 1.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 2.8% 02% 0.0% D. PwwkimWCorsl*iau 4.9% 4.6% 5.3% 20.3% 0.0% 0.0% 0.0% 0.0% 20.1% 0.0% 0.0% 0.0% 0.0% E. Nritral 4.8% 4.5% 4.7% 5.8% 10.5% 5.4% 2.1% 1.1% 7.0% 13.3% 4.1% 2.9% 1.4% 1. P.oMn-EnrW gmmk.*norn 1.0% 0.8% 12% 33% 0.0% 0.0% 0.0% 0.1% 4.3% 0.1% 0.0% 0.0% 0.1% 2. lodmD0eftimy 02% 0.2% 02% 0.9% 0.3% 0.0% 0.0% 0.0% 0.8% 0.4% 0.0% 0.0% 0.0% 3. Vwn A Deft "y 02% 0.2% 02% 0.5% 0.8% 0.0% 0.0% 0.0% 0.4% 0.9% 0.0% 0.0% 0.0% U_ 4.AAnm 32% 3.3% 3.1% 1.2% 9.3% 5.3% 2.1% 1.0% 1.5% 11.9% 4.0% 2.9% 1t3% 5. Otwr NLirtid 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% II. Norsimsmble 58.2% 58.3% 58.0% 25.3% 29.4% 56.1% 80.7% 88.7% 24.1% 27.5% a2.8% 82.9% 89.0% A. Malig.Nut Neopism 8.7% 10.4% 6.8% 0.7% 6.0% 8.1% 22.0% 18.1% 0.9% 3.8% 5.7% 16.0% 12.1% I. MoihndOraphwynx 0.3% 0.4% 02% 0.0% 0.1% 0.5% 0.6% 0.4% 0.0% 0.0% 0.2% 0.4% 02% 2. Esopglm 0.9% 1.1% 0.5% 0.0% 0.0% 0.5% 2.7% 2.9% 0.0% 0.0% 0.1% 1.4% 1.7% 3. Staouh 1.8% 2.0% 1.2% 0.0% 0.1% 0.7% 5.0% 4.9% 0.0% 02% 0.9% 2.8% 2.8% 4. C*boryRtun 0.5% 0.5% 0.5% 0.0% 0.1% 0.5% 1.0% 10% 0.0% 02% 0.3% 1.1% 0.9% 5. Lier 129% 2.8% 1.0% 0.0% 0.4% 3.0% 6.7% 3.0% 0.0% 0.5% 0.8% 2.7% 1.6% 6. Parrna 0.1% 0.2% 0.1% 0.0% 0.0% 0.1% 0 4% 0e4% 0.0% 0.0% 0.0% 02% 0.4% 7. Tiuh.BrrhvAm Q 1.0% 1.3% 0.6% 0.0% 0.2% 04% 3.1% 3.8% 0.0% 0.0% 0.2% 1.9% 1.7% 8. Mealnomahand othw Sidn 0.0% 0.0% 00% 00% 0.0% 0.0% 0.0% 0.0% 00% 0.0% 00% 0.0% 0.0% 9. inhi 02% 0.0% 0.4% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.5% 1.0% 0.4% 0. Cmvi. 0.1% 0.0% 0.3% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 02% 0.8% 0.S% 1. Coxas 1*0,ri 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.4% 0.1% 12. Owy 0.1% 0.0% 02% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 02% 0.4% 02% 13. Promn . 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 14.Bkiddr 0.1% 0.1% 0.0% 0.0% 0.0% 0.0% 02% 0.4% 0.0% 0.0% 0.0% 0.1% 0.1% IS. Lynfho 02% 02% 0.1% 0.0% 05% 0.3% 0.3% 03% 00% 0.0% 0.1% 0.3% 0.1% 16.L.ukms 0.8% 0.8% 0.8% 0.4% 3.0% 1.1% 0.5% 02% 0.4% 2.0% 1.1% 0.6% 0.3% X 17. O0tucrou 0.9% 0.% 02% 02% 1.5% 1.0% 1.5% 0.8% 0.4% 09% 0.8% 1.7% 1.0% W

Source: Murray and Lpez 1996. Tob. A3.4. Bdan d Db_ In Ohm by Age. SU, mo Cms. IO1 fh DAMLV by pI uI

kk abbe slow mbe Room Fwrmks Fms Fanuke Fmmo wi th_ AJ Ml _ibb M Fmui. 0-4 5-14 15-44 45-60 00+ 0-4 5-14 1S-44 45-t0 go+

B. Otw NsKIMuM 0.4% 0.4% 0.4% 0.4% 0.5% 0.5% 0.5% 0.3% 03% 0.7% 0.4% 0.0% 0.3% C. 0113s 1>am 0.5% 0.5% 0.0% 0.0% 02% 0.4% 1.0% 0.8% 0.0% 0O% 0.4% 1.4% 12% D. N_twmWEniafr 02% 0.1% 0.4% 0.1% 02% 02% 02% 0.1% 05% 0.0% 0.5% 0.2% 0.2% E. Nmo-p b-1l 142% 12.0% 15S% 10% 6.0% 27.9% 6.5% 3.4% 0.7% 7.1% 340% 10.7% 56% 1. M0rAffbwDWiMWr 6.2% 4.4% 82% 0.0% 0.0% 10.1% 3.2% 0.7% 0.0% 0.0% 18S.% 8.7% 1 f% 2.B 4pa JMrr Dibodnr 1.8% 1.8% 1.8% 0.0% 0.0% 4.5% 0.0% 02% 0.0% 02% 4.5% 0.7% 0.2% 3.PuFhom 1.3% 1.3% 1.2% 0.0% 0.0% 3.5% 0.1% 0.1% 0.0% 0.0% 3.1% 0.1% 0.1% 4.Epil 0.4% 0.5% 0.4% 0.1% 1.4% 0.8% 0.3% 0.1% 0.1% 1.8% 08% 0.2% 0.1% 5. AbOimDpwdm 0.7% 1.3% 0.1% 0.0% 0.0% 3.3% 0.5% 0.1% 0.0% 00% 0.3% 0.0% 0.0% 8. A inw'sandadmo dwrwti 0.7% 0.7% 0.8% 0.3% 0.4% 02% 1.0% 2.0% 02% 0.4% 02% 1.2% 28% 7. Pakbars Dhw_ 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% 0.0% 0.1% 02% 8. Mi*& 1mom 02% 0.1% 02% 0.0% 0.0% 0.3% 0.1% 00% 0.0% 0.0% 0.4% 0.1% 00% 9. n 4gD .ro.4ap 0.1% 02% 0.0% 0.0% 02% 0.4% 0.0% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 1O.PTSD 02% 02% 0.3% 0.0% 0.5% 0.3% 0.1% 0.0% 0.0% 0.% 05% 0.1% 0.0% ll. 0bm-C"w.bmDimdr 1.1% 1.0% 13% 0.0% IA% 22% 0.3% 01% 0.0% 2.5% 2.9% 0.0% 02% 12. Paat DOr 0.5% 0.4% 0.7% 0.0% 00% 0.3% 0.3% 0.0% 0.0% 0.7% 1.7% 0.5% 0.1% 13. 0Ow1w-a r-pqylbbf 0A% 08% 0.7% 0.0% 1.4% 1.5% 0.1% 0.1% 0.4% 0.% 123% 02% 02% O F. Su_ Ougsn 1.0% 0A% 1.2% 02% 0.% 0.2% 1.8% 19% 0.t% 0.3% 0.5% 4.0% 2.7% a. DbLEan 0.4% 02% 05% O.OS 0.0% 0.1% 0.8% 0.4% 0.0% 0.0% 02% 2.0% 0.7% 2. 1N.c*s 0.5% 0.5% 0.5% 0.1% 00% 0.1% 0.9% 14% 0.0% 0.0% 0.1% 1.1% 1S% 3.OthwSu_Or,n 0.1% 0.1% O.t% 0.1% 0.0% 00% 0.1% 0.1% 0.1% 0.3% 02% 02% 0.1% G. .adibw 11.0% 112% 10.7% 1.4% 2.5% 0.8% 18.7% 28.3% 0S% 1.% 00% 192% 300% t. RhKu#ru Hem mn 1.1% O% 1.3% 0.0% 0.4% 1.4% 12% 1.1% 0.1% 0.0% 1.7% 2.e% 1S% 2. mt Hb_ OD _m 2S% 3.0% 2.7% 0.0% 0.0% 15% 5.3% 8.0% 0.0% 0.0% 12% 4.9% 9.3% 3.C _~macaimDoma 52% 5.4% 4S% 0.3% 1.0% 22% 9.6% 15.1% 02% 0.5% 2.4% 9.0% 152% 4. H'diutxyCarim 0.0%% 0.% 0.0% 02% 0.4% 0S% 0.8% 05% 02% 0.4% 02% 0S% 0A% 5. OdwCd 1.1% 12% 1.1% 02% 00% 0.4% 1.8% 3.0% 0.4% 0.4% 0.7% 12% 3.0% H.R Sby 10.7% 11.1% 102% 3.4% SS% 7.1% 152% 26.0% 3.0% 4.7% a.0% 152% 2L0% i. COPD 85% 8O% 83% 0A% 0.1% 4.6% 14.0% 253% 12% 0.1% 3.8% 13.% 208% 2. A1T 13% 1A% 12% 0.7% 5.3% 1S% 0O% 0.5% 0.4% 42% 1.J% 1.0% 0.5% S. OuhrwPA ims 0% 0% 0.7% 2.1% 02% 0.5% 0.4% OA% 1S% 05% 0.4% 0.3% OA% 1.Oi_adhe 4S% 52% 4.0% 0.0% 2.6% 4.1% 82% 4J% 5S% 2.% 3.1% 7.1% 4.% I. Palb L*wDim 02% 0.3% 02% 0.0% 0.1% 0S% 0.4% 0.4% 0.0% 0.1% 02% 0.3% O0% 2.CIhoulsdtlheLkw 15% 2.0% 10% 03% 0.4% 22% 4.1% 22% 0.1% 05% 0.7% 2.0% 1.% S. m _ 02% 02% 0.1% 02 02% 0.3% 0.0% 02% 02% 0% 02% 0.0% 0.% 3. Oth igu_e 3&0% 2J% 33% 5.% 1.3% 1.3% 3S% 2.3% 5% 12% 1S% 4.1% 2.4%

Source: bIrray and Lqxcz 1996. Table A2.4. Budmn of D _eumn Chim by A4e 8aw, rd Cue IWO1Cmn DALYs) by pa :

lales Maim mai. Malbe Mam Funaiss Fwnslm Faim FiI. Funeh Cmtian A8 Al 1Mm_ M F _alm 0-4 5-14 15-44 45-50 60+ 0-4 5-14 15-44 4-510 60+

J. Gwito-Urinrry 1.2% 1.4% 0.9% 0.3% 1.2% 1.4% 3.2% 1.4% 0.3% 1.0% 1.2% 1.4% 1.1% 1. Nqp v#i4N as 0.8% 0.9% 0.7% 0.2% 12% 1.3% 0.8% 0.7% 0.3% 1.0% 1.0% 0.8% 0.7% 2. BergnProdtc Hypwbreo 02% 0.4% 0.0% 0.0% 0.0% 0.0% 2.1% 0.3% 0.0% 0.0% 0.0% 0.0% 0.0% 3. 0WrGerwto-LUn,y 0.2% 0.2% 0.2% 0.1% 0.0% 0.1% 0.3% 0.3% 0.0% 0.0% 02% 0.8% 0.4% K MuwLio-SkeIub 1.7% 1.1% 2.3% 0.0% 0.4% 0.8% 2.4% 22% 0.1% 0.4% 2.5% 8.4% 2.9% 1. Rhal1idA nttvmis 0.3% 02% 0.4% 0.0% 0.0% 02% 0.8% 02% 0.0% 0.1% 0.5% 1.3% 0.5% 2. Osieim wt 1.0% 0.7% 1.4% 0.0% 0.0% 0.4% 1.6% 1.4% 0.0% 0.0% 1.4% 4.4% 1.6% 3. Otlw Muscto-klstal 0.3% 0.2% 0.5% 0.0% 0.4% 02% 0.2% 0.6% 0.1% 0.3% 0.7% 0.7% 0.C% L. Cornwuid Abrmielibes 3.0% 2.8% 3.2% 11.0% 2.7% 0.5% 0.0% 0.0% 10.4% 3.4% 0.7% 0.0% 0.0% i. Abdeng wall dAd 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 2.A,wcaphfI 0.6% 0.5% 0.8% 2.1% 0.0% 0.0% 0.0% 0.0% 3.0% 0.0% 0.0% 0.0% 0.0% 3. Anora1al atrow 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 4. Cet up 0.0% 0.0% 0.0% 02% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 6. SCIApabl 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 8. EKphagamlis.i 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 7. Rm ag 0.0% 0.0% 00% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 8. Down syndrom 0.5% 0.5% 0.5% 1.7% 0.8% 0.1% 0.0% 0.0% 1.3% 0.9% 02% 0.0% 0.0% O 9. Corngwr hel dmae 1 2% 1.3% 1.2% 4.9% 12% 0.2% 0.0% 0.0% 3.9% 1.3% 0.4% 0.0% 0.0% _ 10. Spir bd 0.4% 0.4% 0.5% 1.5% 0.5% 0.0% 0.0% 0.0% 1.% 0.8% 0.0% 0.0% 0.0% 11. Ot1 0.1% 02% 0.1% 0.5% 0.4% 0.1% 0.0% 0.0% 0.3% 0.5% 0.1% 0.0% 0.0% MO.Oraltleh 0.5% 0.5% 0.5% 08% 0.5% 02% 0,6% 0.6% 0.7% 0.7% 02% 0.7% 0.7% 1.Dst Cariae 0.3% 0.3% 0.3% 0.8% 0.5% 0.1% 0.1% 0.1% 0.7% 0.0% 0.1% 02% 0.1% 2. Paixkrftl D _mn 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 3. EdwtuIwn 02% 02% 02% 0.0% 0.0% 0.0% 0.4% 0.5% 0.0% 0.0% 0.0% 0S% 0.0% 4. OlUhrOmlHwlt 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0,0% 0.0% 0.0% 0.1% 0.1% 0.0% 0.0% N. Othr 02% 02% 02% 02% 0.7% 0.1% 02% 0.2% 02% 0.0% 0.3% 0.1% 0.1% Ul. lroxime 17.8% 19.6% 15.4% 17.8% 37.4% 30.1% 9.8% 4.0% 10.4% 31.4% 205% 8.1% 3.9% A. ULkttkrtla 12.9% 15.7% 9.9% 17.1% 35.3% 222% 7.0% 2.5% 15.1% 28.9% 9.1% 4.5% 2.3% 1. Mow Vele*Adkw. 2.1% 28% 1.3% 0.8% 52% 5.3% 1.8% 0.5% 0.4% 02% 1.7% T.A% 0.3% 2. Polmg 0.7% 0.8% 0.7% 029% 1.0% 1.0% 0.6% 02% 0.3% 1.8% 12% 0.3% 0.2% 3. Fal 22% 2.4% 19% 2.5% 5.3% 3.5% 09% 0.5% 3.0% 4.5% 1.5% 1.1% 0.9% 4. F 03% 0.4% 0.3% 0.5% 1.1% 0.4% 0.1% 0.1% 0.3% 2.1% 0.2% 0.1% 0.1% 5. Dooi 2.1% 2.0% 1.7% 42% 12.8% 2.1% 0.4% 02% 3.3% 8.0% 0.9% 02% 02% 6. 0St.umerom kl,a 5.8% 6.9% 4.1% 8.4% 90.% 10.1% 3.0% 1.0% 7.8% a.3% 3.7% 1.0% 0.6% B.bwebml.r 4.7% 39% 5.5% 0.8% 2.1% 7.8% 2.6% 1.5% 1.3% 2.5% 11.4% 3.0% 1.0% 1. Sal- led 3.9% 3.0% 49% 0.0% 14% 0.1% 2.4% 1.4% 0.0% 1.3% 10.9% 3.4% 1 % 2.Fk nVkdt 0.1% 0.9% 0.7% 0.8% 0.7% 1.7% 02% 0.1% 1.3% 12% 0.5% 02% 0.1% 3.Vr 0.0% 0.0% 0.0% 00% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% OA%

Source: iray aod Lq4ez199t Tablb AS.5. DALYsby Ctuse, as a Pcrcenag of Totals,Chins and Oer Regin.. 1990

Formwrly liddle Sociliist Establkhed AN # Sub-Sahran OtherAsia Latin Eastwn Economies Mwket Devloping w CondWiton World China Atica India and Islands Ameriea Crssent of Europe Economies Counfies Toii DALYs (COCO) 1.379,238 208.407 295,294 287,739 177,671 98,285 150,849 62.200 98,794 1.218,244 Dy Pwc.nt 1.Commniwcable, matmrn, perinatl, and nAritionalcondlitons 43.9 24.2 65.9 56.4 44.7 35.3 47.7 8.8 7.1 48.7 A. lnictious andprasitic dies... 22.9 7.5 42.5 28.9 22.3 17.6 20.2 2.7 2.8 25.6 Tubercuils 2.8 2.0 3.4 4.8 3.1 1.8 1.7 0.B 0.1 3.1 STDs wcudig HIV 1.4 0.1 2.1 1.9 2.3 1.2 0.5 0.6 0.4 1.5 HIV I dsclon 0.8 0.0 2.8 0.1 0.1 1.1 0.0 0.1 1.3 0.8 Denrhel disases 7.2 1.8 10.9 10.2 7.7 5.5 9.8 0.4 0.2 8.1 Chidhtod-dustr diseass 5.2 1.1 10.3 6.4 4.5 3.4 5.7 0.1 0.0 5.8 Bac.rii rrenirgits and rrnlngococceria 0.5 0.6 0.3 0.5 0.5 0.5 0.5 0.4 0.2 0.5 oMalris 2.3 0.0 9.2 0.4 1.4 0.5 0.3 0.0 0.0 2.6 Troplca dusbr deses snd hprosy 0.8 0.1 1.9 1.2 0.4 0.8 0.2 0.0 0.0 0.0 Intestinernemabds hfVecons 0.4 0.7 0.2 0.3 0.9 0.7 0.1 0.0 0.0 0.4 Otr Inlectiousand paranltc duesee 1.7 1.2 1.4 3.1 1.5 2.1 1.4 0.6 0.5 2.7 B. Respiratry h.i$ctk 8.5 5.9 10.5 11.9 8.7 4.9 10.7 2.0 1.4 9.4 Lowerrespiratorydecwborm 8.2 5.7 10.2 11.4 8.5 4.7 10.4 1.9 1.2 9.1 Other repWratry Iifections 0.3 0.2 0.2 0.5 0.3 0.2 0.3 0.1 0.1 0.3 C. Maternalcondilions 2.2 1.3 3.2 2.6 2.3 1.7 2.4 0.9 0.3 2.4 Obs*uced lbo 0.5 0.3 0.6 0.5 0.5 0.4 0.7 0.2 0.2 0.5 Abortin 0.4 0.0 0.6 0.6 0.4 0.5 0.2 0.3 0.0 0.4 Othrrnabrnal conditon 1.3 0.9 2.1 1.5 1.3 0.9 1.5 0.4 0.1 1.5 D. Parkul conditons 6.7 4.9 6.5 8.8 6.9 7.4 9.7 2.2 1.8 7.3 E Nubltioml dosicWus 3.7 4.6 3.2 4.2 4.4 3.7 4.7 1.0 0.9 4.1 Proaen-energy ulne*Liron 1.5 1.0 1.8 1.8 1.7 1.7 2.4 0.2 0.1 1.7 VlHmin A deficency and lode defincy 0.4 0.4 0.4 0.4 0.5 0.3 0.7 0.0 0.0 0.4 ArnmIa 1.8 3.2 0.9 2.1 2.3 1.7 1.6 0.7 0.7 1.9 Otw nuStlonal deficnenes 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0

Source: Murray and Lopez,1996 Table A3.5. DALYsby Cause,as a Percentageof Totls. China and OthwrPAgln, 1990

Formerly Middle Socalist Eablihd Al Sub-Sahwran Othr Asia LaUn Eatern Economis Mwrket Developing Conditon World China Arica India and Islands America Crescnt of Erope Economies Countrele

II.Nonconmrwlcb dlmase 40.9 58.2 18.8 29.0 40.9 48.2 39.3 72.6 81.0 36.1 A. Mmiignmntneoplasm 5.1 8.7 2.1 2.5 5.1 4.5 2.4 11.7 15.0 4.0 Sbnimch cer 0.6 1.6 0.1 0.2 0.4 0.4 0.2 1.6 1.1 0.5 Coonandrectumncanwrs 0.3 0.5 0.1 0.1 0.3 0.2 0.1 1.1 1.6 0.2 Liwr cnncer 0.5 1.9 0.3 0.1 0.5 0.1 0.1 0.3 0.3 0.5 Trachea, ronchus and lungcncers 0.6 1.0 0.1 0.1 0.5 0.3 0.2 2.6 3.0 0.4 Breatcacr 0.3 0.2 0.1 0.2 0.3 0.4 0.1 0.8 1.4 02 Ccrvkutrlcaerlw 0.2 0.1 0.1 0.3 0.3 0.4 0.1 0.3 0.2 02 Lymphom and meLAitcmyeloma 0.2 0.2 0.2 0.1 0.2 0.3 0.1 0.4 0.7 0.2 Ledckmn 0.3 0.8 0.1 0.1 0.5 0.3 0.2 0.5 0.6 0.3 Other miligantnnopksms 2.0 2.4 1.1 1.3 2.2 2.0 1.2 4.1 6.0 1.6 B. Dkbets malius 0.8 0.5 0.2 0.8 0.7 1.5 1.0 1.1 2.4 0.7 C. Neuro-peychisric cnditons 10.5 14.2 4.0 7.0 10.8 15.9 8.7 17.2 25.0 9.0 Urtipolrnmjcr depedon 3.7 6.2 1.5 2.8 3.8 4.3 3.0 5.0 6.8 3.4 8ipobr dsorder 1.0 1.8 0.4 0.8 1.1 1.2 0.9 1.3 1.7 1.0 SdcIoplven 0.9 1.3 0.2 0.6 1.3 1.3 0.9 1.4 2.3 0.8 Akoholuse 1.2 0.7 0.6 0.3 1.1 3.9 0.2 2.8 4.7 0.8 Demetand ohr CNS diwtders 0.6 0.7 0.1 0.3 0.5 0.6 0.2 1.5 2.9 0.4 Drug use 0.4 0.1 0.1 0.0 0.5 1.1 0.6 0.9 1.5 0.S Epilpsy 0.4 0.4 0.2 0.3 0.5 0.7 0.3 0.6 0,5 0.4 Otr nauo-psychistic condoirn 2.2 2.9 0.9 1.8 2.1 2.8 2.5 3.7 4.6 2.0 D. Ghucoo andc 0ach 0.8 1.0 0.7 1.0 0.9 0.6 0.6 0.1 0.1 0.8 E.Crdiovcuawd _eses 9.7 11.0 3.9 8.1 10.1 7.9 11.1 23.2 18.6 82 R himtc heat dbe_ 0.4 1.1 0.2 0.5 0.1 0.2 0.5 0.6 0.2 0.5 chawmichFil d_sas 3.4 2.9 0.8 3.5 2.2 3.0 3.5 11.4 9.0 2.5 CerubroecsLw di_ea 2.8 5.2 1.6 1.5 2.5 2.5 1.6 7.2 5.0 2.4 X IrnIhnmaury hit d _eames 0.7 0.6 0.5 0.6 1.3 0.5 1.2 0.7 0.7 0.8

Source: Murrayand Lopez, 1996 Table AS.5. DALYs by Caus, as a Pwrcentage of Totals, China and Other Reglons. 1990

Formerly Middle Socialist Established All x

Sub-Saharan Other Asia Latin Eastern Economies Market Developing Lo Condition World China Africa India and Islands Anerica Crescent of Euvope Economis Counwrle (confinued) Oter cardbvascular diseases 2.3 1.1 0.9 2.0 4.0 1.7 4.2 3.3 3.7 2.1 F. Respiratory dis 4.4 10.7 2.6 2.6 2.7 4.0 4.2 4.8 4.8 4.3 COPD 2.1 8.5 0.6 0.9 0.7 1.0 0.9 1.7 2.3 2.1 Asthma 0.8 1.3 0.5 0.5 0.8 1.0 0.6 0.8 1.3 0.7 Otherrespiratorydiseases 1.5 0.8 1.5 1.3 1.2 2.0 2.7 2.2 1.3 1.5 G. Digestive diseases 3.4 4.9 1.8 2.2 4.7 3.8 4.2 4.4 4.4 3.3 Cirrhois oftewliwr 1.0 1.5 0.2 1.0 1.3 1.2 0.5 1.2 1.6 0.9 Ow erdigesdivedeases 2.5 3.4 1.6 1.2 3.4 2.6 3.7 3.2 2.8 2.4 H. Musculo-skeletel deseses 1.4 1.7 0.4 0.5 1.2 3.1 0.6 4.4 4.2 1.0 Rioumatoid artitis 0.2 0.3 0.0 0.1 0.1 0.6 0.1 0.8 0.9 0.2 ) Osboartrtltis 1.0 1.0 0.3 0.4 0.9 2.1 0.4 3.2 2.7 0.7 Oter musculo-skelebtl diseases 0.2 0.3 0.0 0.0 0.2 0.4 0.1 0.4 0.5 0.1 1.Congenitl anomalies 2.4 3.0 1.3 2.9 2.3 2.7 2.7 2.2 2.2 2.4 J. Oral conditons 0.5 0.5 0.1 0.4 0.7 1.0 0.9 0.8 0.9 0.5 K Othernoncommunicable diseases 2.0 2.1 1.6 0.9 1.7 3.3 2.9 2.8 3.4 1.8 Ill. Injuies 15.1 17.6 15.4 14.6 14.4 16.4 13.0 18.7 11.9 15.2 A.Unlnitnal ijnjuies 11.0 12.9 9.3 13.0 12.1 11.9 6.8 12.9 8.7 11.1 Motor-vehicleaccidents 2.5 2.1 1.9 2.1 2.7 4.1 1.7 4.4 4.4 2.2 Pisonings 0.5 0.7 0.4 0.3 0.6 0.2 0.2 1.5 0.3 0.4 Falls 1.9 2.2 0.7 3.5 2.3 1.7 1.1 1.8 1.4 2.0 Fres 0.9 0.3 1.2 2.0 0.3 0.3 0.5 0.3 0.3 0.9 Drownings 1.1 2.1 1.1 0.9 1.6 0.9 0.6 1.0 0.3 1.2 Othr unkfwionari Injwkis 4.1 5.5 4.0 4.2 4.7 4.8 2.7 3.9 2.2 4.3 B. Intntioml hmjwles 4.1 4.7 6.0 1.5 2.3 4.5 6.2 5.8 3.2 4.1 Self-frinfllcdnjiuies 1.4 3.9 0.2 1.0 1.1 0.6 0.9 2.6 2.2 1.2 Vilence 1.3 0.8 2.2 0.5 0.9 3.2 0.8 1.4 1.0 1.3 War 1.5 0.0 3.6 0.0 0.3 0.7 4.5 1.8 0.0 1.5

Source: Murrayand Lopez,1996 Table AS.6. DALYsby Caus, as a Prcwnag of Totals, China and Other Regions, 2020

Formerly Middle Socaiset Established AN Sub-Sahwan Ohr Asia Lain Eastern Economie Lket Developing Conditon World China Afkica India and Islands America Crescent of Ewope Economies Counirie Totl DALYs OOs) 1,388,836 220,667 329,566 236,741 165,978 107,639 167,710 63,534 97,000 1,228,302 By Pcnt 1.Comwu*abie, msrl.l, prinatl, 20.1 4.3 39.8 24.4 16.5 12.6 19.9 3.0 5.2 22.2 and n1u1nil cornilon A. lfeclousandpdeafJcdiseaes 12.9 1.4 28.5 17.3 9.7 7.4 8.4 1.0 2.9 14.3 Tuberculdois 3.1 0.4 6.7 6.7 1.2 0.5 0.5 0.1 0.1 3.4 STDssaludhng HIV 0.6 0.0 1.0 0.7 0.9 0.4 0.2 0.2 0.2 0.6 HiV'hIdiln 2.6 0.1 4.4 4.6 3.0 2.9 0.2 0.2 2.3 2.8 Dbrrhel d _ea.e 2.7 0.3 5.5 2.5 2.2 1.5 4.2 0.1 0.1 3.0 Chidhood-dclier d _sses 2.0 0.2 5.1 1.5 1.2 0.9 2.4 0.0 0.0 2.3 ecBi.rmeliir s and rnIngooocen a 0.1 0.1 0.2 0.1 0.1 0.1 0.2 0.1 0.1 0.1 Mwa 1.1 0.0 4.5 0.1 0.3 0.1 0.1 0.0 0.0 1.3 Trqplol_cku desses andr pcosy 0.2 0.0 0.5 0.2 0.1 0.2 0.1 0.0 0.0 0.0 irkiss nmde lons 0.1 0.1 0.1 0.1 0.2 0.2 0.0 0.0 0.0 0.1 Oer hw cos and pac d _sase 0.5 0.2 0.6 0.7 0.4 0.5 0.6 0.2 0.2 0.7 B. RPAlkory khecUrm 3.2 1.1 5.4 3.2 2.8 1.5 4.6 0.8 1.2 3.4 L'reepkajoryk bclon 3.1 1.1 5.3 3.1 2.8 1.5 4.5 0.8 1.1 3.3 Othe repiralory b*cln 0.1 0.0 0.1 0.1 0.1 0.0 0.1 0.0 0.1 0.1 C. Manal condWios 0.3 0.1 0.6 0.3 0.3 0.2 0.4 0.1 0.0 0.3 Obducsd inbour 0.1 0.0 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.1 Abordin 0.1 0.0 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.1 O0e nUAKdl condiions 0.2 0.0 0.4 0.2 0.2 0.1 0.3 0.1 0.0 0.2 D. Pe condIkions 2.5 0.9 3.7 2.4 2.2 2.3 4.6 0.7 0.7 2.7 E NuS'Itiosldellclnea 1.2 0.9 1.5 1.2 1.4 1.1 1.9 0.3 0.4 1.3 Probln-energy manurltion 0.6 0.2 1.0 0.4 0.5 0.5 1.0 0.1 0.0 0.6 Vih*n A deficwncy and odhie decency 0.1 0.1 0.2 0.1 0.1 0.1 0.2 0.0 0.0 0.1 krwffda 0.5 0.7 0.4 0.6 0.8 0.5 0.6 0.2 0.3 0.6 x 0l nutwlloria delicienles 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 L

Soure: Murray and Lopez,1996. Table A3.6. DALYs by Cause, as a Percentage of Totals, China and Other Ragions, 2020

Formerly Middle Socialist Established All X Sub-Saharan Other Asia Latin Eastern Economies Market Developing w Condition World China Africa India and Islands America Crescent of Europe Economies Countries (continued) II. Noncommunicablediseases 59.7 79.3 31.9 56.5 66.3 68.1 59.6 79.7 84.7 56.7 A. Malignantneopasmns 9.9 18.7 4.5 7.1 11.6 8.5 5.3 16.1 17.3 9.0 Stomachoncer 1.2 3.7 0.3 0.5 1.0 0.8 0.4 2.4 1.2 1.1 Colonand rectumcancers 0.6 1.0 0.1 0.2 0.6 0.5 0.2 1.4 1.9 0.4 Livercancer 1.1 4.2 0.7 0.1 1.3 0.1 0.2 0.4 0.4 1.2 Trach, bronchusand lungcancors 1.8 3.4 0.2 1.5 1.9 1.3 1.2 4.4 4.6 1.5 Brest cancer 0.4 0.3 0.2 0.4 0.5 0.8 0.3 0.8 1.3 0.4 Cervixuteri cancer 0.4 0.2 0.3 0.6 0.6 0.7 0.2 0.3 0.2 0.4 Lymphomasand muliple myeloma 0.4 0.3 0.4 0.2 0.4 0.4 0.2 0.5 0.7 0.3 Fl LoLesmia 0.4 0.9 0.1 0.2 0.6 0.4 0.4 0.5 0.6 0.4 F-. tOfwermalignant neopbasms 3.7 4.8 2.2 3.3 4.7 3.6 2.3 5.5 6.5 3.3 B. Dibetes mellitus 0.8 0.4 0.2 0.8 0.9 1.6 1.0 0.7 2.1 0.7 C. Neuro-psychidtic condWxons 14.7 15.4 8.5 12.6 17.4 21.6 14.9 16.4 25.4 13.7 Unipolr mapr depression 5.7 7.3 3.5 5.6 6.6 6.4 6.0 5.0 6.8 5.6 Bipobadhorder 1.5 1.9 1.0 1.5 1.8 1.7 1.7 1.3 1.6 1.5 Schizoptvenim 1.2 1.2 0.3 1.1 2.0 1.8 1.8 1.3 2.0 1.2 Akohol use 1.7 0.8 1.3 0.6 1.8 5.6 0.3 2.7 4.5 1.4 Demwntaand other CNS disorders 1.1 1.3 0.2 0.7 1.1 1.2 0.4 1.9 4.5 0.7 Druguse 0.6 0.1 0.3 0.0 0.8 1.5 1.1 0.8 1.4 0.5 Eplpsy 0.3 0.2 0.2 0.3 0.3 0.4 0.3 0.3 0.3 0.3 Ofler neuro-psychidric conditons 2.7 2.7 1.6 2.7 2.9 2.9 3.4 3.0 4.2 2.6 D. Gkuom and cabractB 1.6 1.7 1.3 2.7 2.3 1.2 1.3 0.1 0.1 1.8 E. CardovacuLwd _eaes 14.7 16.3 6.0 18.4 15.6 13.2 17.7 26.1 19.4 13.8 Flow.htic hewt dissese 0.5 1.1 0.2 0.8 0.1 0.2 0.4 0.5 0.1 0.5 bchaerdc hert diesse 5.9 4.7 1.5 9.3 5.1 5.6 7.5 13.3 9.8 5.2 Cerebroyscubr dease 4.4 8.2 2.6 3.4 4.5 4.3 2.8 7.7 5.1 4.2 lnfinnatory hedrtidseaes 0.8 0.6 0.6 1.0 1.2 0.6 1.4 0.7 0.6 0.9

Source: Murrayand Lopez, 1996. Table AS.6. DALY* by Caum. a a Percntage of Totmb. China and Othr Regions, 2020

Formwrly Middle Sociaiit Estblished Ail Sub-Saharan Other Asia Latin Eastern Economies Mwrket Developing Condition World China Afica India and Islands America Crescent of Europe Economies Cournlzes (continued) Oter cardioscular diseases 3.1 1.6 1.1 4.0 4.7 2.5 5.6 3.8 3.7 3.0 F. Respiralorydhesses 7.3 16.3 4.5 6.4 4.3 6.3 6.6 8.1 5.3 7.4 COPD 4.1 14.5 1.4 2.8 1.7 2.5 2.2 3.3 2.9 4.3 Asthma 1.0 1.2 0.7 0.9 1.1 1.3 0.9 0.9 1.1 0.9 Ower reapiratory disee 2.2 0.6 2.4 2.7 1.5 2.5 3.4 3.9 1.3 2.2 a Digesdivediseases 3.5 3.5 1.8 2.5 6.5 4.6 3.7 3.9 5.2 3.4 Cirrhoeisottheliver 1.2 1.4 0.3 1.3 2.2 1.7 0.7 1.2 1.9 1.1 Oer digesl dis 2.3 2.1 1.5 1.2 4.3 2.9 3.0 2.7 3.2 2.3 H. Muscilo-skeblal dlseees 2.2 2.8 0.8 1.1 2.6 5.5 1.2 4.5 5.0 1.9 Rhoumaold ardritrs 0.4 0.5 0.1 0.1 0.3 1.0 0.2 0.8 1.2 0.3 w..) Osteoerttvitis 1.7 1.9 0.7 1.0 2.2 4.2 0.9 3.4 3.5 1.5 Other muaclo-skeleWi diseses 0.1 0.2 0.0 0.0 0.2 0.3 0.1 0.2 0.4 0.1 1.Congenit anomalles 2.2 1.9 2.2 3.2 1.8 1.7 3.3 1.2 0.9 2.4 J. Oral condWtons 0.9 0.7 0.3 0.8 1.5 1.3 1.8 0.8 1.0 0.9 K Ot ernoncommunicLble disese 1.9 1.6 1.9 0.9 1.8 2.8 2.9 1.7 2.9 1.8 IlI. Injuies 20.1 16.4 28.3 19.1 17.2 19.3 20.5 17.4 10.1 21.1 A.Unintenflonralnjwies 13.0 11.0 15.4 16.4 13.6 13.2 9.8 11.6 6.9 13.6 Motor-vehicleaccidenrs 5.1 4.8 5.1 6.5 5.2 6.4 3.6 5.2 3.6 5.2 Poesonings 0.5 0.5 0.6 0.4 0.6 0.1 0.3 1.2 0.2 0.5 Falls 1.5 1.4 0.9 2.6 2.0 1.4 1.2 1.4 1.1 1.6 Fres 0.8 0.2 1.5 1.9 0.2 0.2 0.5 0.2 0.2 0.9 Drownings 0.9 1.0 1.2 0.8 1.1 0.7 0.7 0.7 0.2 1.0 Otter unintentonal injuies 4.2 3.1 5.9 4.2 4.5 4.3 3.4 2.9 1.6 4.4 B. Intentional injwies 7.1 5.4 12.9 2.8 3.6 6.2 10.7 5.7 3.2 7.5

Selt-Mintlcedlnjuies 1.9 4.5 0.4 1.8 1.8 0.9 1.6 2.6 2.3 1.8 D Violerce 2.3 0.8 4.9 0.9 1.4 4.4 1.3 1.3 1.0 2.4 X War 3.0 0.0 7.6 0.1 0.5 0.9 7.8 1.8 0.0 3.3 L

Source: Murray and Lopez, 1996. An4ex 3

Table A3.7. Deaths and Burden of Disease Attributable to Tobacco Use, China, 1990 and Projections to 2020

1990 2020

Deaths attributableto tobaccouse 800,000 2,200,000 As percent of total deaths 9.2% 16.0%

Years of life lost attributableto tobacco use 5,800,000 23,400,000 As percentof total yearsof life lost 4.9% 18.0%

Years of disability lost attributableto tobacco use 2,300,000 12,000,000 As percentof total yearsof disabilitylost 2.6% 13.2%

DALYslost attributableto tobaccouse 8,100,000 35,400,000 As percent of total DALYslost 3.9% 16.0%

Source: Murrayand Lopez,1996.

114 Annex 4

ANNEX 4: CHILD MORTALITY TRENDS IN CHINA

4.1 The purpose of this annex is to review trends in under-five mortality in China over the last three decades, and to put these trends in the context of trends in other developing countries. The contrast between trends occurring before and after China's economic liberalization, which began in 1978, is of particular interest--though results may lack stability as the record since liberalization is short. The basic idea of the analysis, is first to estimate China's under-five mortality rate (U5MR) or the probability of dying between birth and age five (expressed per 1,000 live births) for time points separated by five years from 1960 to 1991. The changes from each point to the next are then related to the changes over the same period in per capita income. The relationship between child mortality rates and income growth in China is then compared to the average relationship estimated for almost 90 countries, as a way of comparing China's performance with that of developing countries as a whole.

Data Sources

4.2 Prior to 1980, demographic data for China were scarce, and estimating fertility and mortality rates was very difficult. Since 1980, however, data have poured out, and the problem has been lack of analysis rather than lack of data. Indirect estimates of child mortality based on child fatalities as a proportion of all births, classified in five-year age groups of women, are available from China's 1982 and 1990 censuses, and also from the 1982 One-in-a-Thousand Fertility Survey and from the 1987 One-in-a-ThousandPopulation Sample Survey. Direct estimates based on birth histories from the 1988 National Survey on Fertility and Birth Control are available for the years 1982 to 1987. Direct estimates are also available from the 1990 census, based on births and deaths in the year before the census, and from the 1992 Fertility Sampling Survey (Jiang and others 1995) 2

4.3 The data described above provide estimatesof child mortality in China from about 1970 to 1992. To reconstruct demographic trends in China from the early 1950s to 1982, Hill (1988) used then-availabledata, including estimates of child mortality from 1964 to 1982. The estimates from the various surveys and those arrived at by Hill are shown in Figure A4.1. In interpreting Figure A4. 1, it is important to note that the indirect estimates obtained for time points closest to a given survey almost always exaggerate the true level of child mortality. This is probably due to selection: the estimate for the time point closest to the survey is based on summary fertility histories of women aged 15-19, at which age the women who already have children are likely to come from the most disadvantaged social classes. Hence the sequences of indirect estimates all turn up sharply at their right-hand ends, but these upturns should be ignored. The remaining points are reasonably consistent, though there does seem to be a tendency for the indirect estimates based on reports of older women to indicate too low a level of mortality, probably as a result of omission of dead children. Child mortality appears to have fallen rapidly until the early 1980s, and then to have stabilized or even risen slightlyto a U5MR of about 45 per 1000 live

IThis Annexwas preparedby KennethHill (Johns Hopkins University) and AkikoMaeda (World Bank). 2 The authorsare gratefulto Dr. GriffithFeeney of the East-WestCenter Population Program for providing materialfrom the 1988and 1992surveys.

115 Annex4

FigureA4.1 Trendsin China'sUnder-Five Mortalit Rate, 1960-91 and Projectionsto 1995

,, 200

150

I

0 O- 19 60 1965 1970 1975 1980 1985 1990 19 5 Year

births. The estimatesof Hill (1988)appear to have exaggeratedthe rate of declineand to have underestimated U5MR.

Methods

4.4 A trend line has been fitted to the observationsfollowing methodologyemployed by Hill and Pande (forthcoming) for UNICEF. Observationsof a particular type are given weights according to a priori expectations of accuracy (for exarnple, an indirect estimate based on reports of womenaged 15-19is givena weightof zero, while a direct estimatebased on births in the last five years is given a weightof one). The naturallogarithm of the U5MRis then regressedon the year, usinglinear splines to capturepossible changes in trendsover time. Splinesare definedso the slopeof the regressionline can changeat each point that the sum of the accuracyweights reaches a multipleof five. The specificmodel fitted for Chinais

In(U5MR) = Bo+ B1 (Date-1960)+ B2(Date-1977.9)+ B3(Date-1980.7)+ B4(Date-1983.5)

(or 0, if negative) (or 0, if negative) (or 0, if negative) (or 0, if negative)

+ B5 (Date-1986.5)

(or 0, if negative)

4.5 Thus the rate of declinecan changeat 1977.9, 1980.7, 1983.5,and 1986.5-dates that are drivenby the densityof observations.The model R2 is 0.82.

116 Annex4

Results for China

4.6 The fitted equation allows the Table A4.1. EstimatedRates of Changein Under- rates of change in under-five mortality to Five Mortality, China, 1960-91 be estimated for each time period that Annual emerged from the analysis (Table A4. 1); Changeinu this information can generate estimates for TimePeriod Estimatedn USMR(%) the level of under-five mortality as indicated in Table A4.2. 1960.0to 1977.9 -6.52 -6.52 4.7 The model indicatesthat China's 1977.9 to 1980.7 -1.11 -7.63 U5MR declined rapidly until mid-1983, 1980.7to 1983.5 -1.13 -8.76 but has increased slightly since mid-1986. The model is used to estimate values for 1983.5to 1986.5 6.74 -2.02 mid-year 1960, 1965, 1970, and so on. 1986.5to 1991 2.92 0.9 Estimates are shown in Table A4.2 and depicted graphically in Figure A4. 1, which includes a projection to 1995.

Child Mortality and Income: Other Countries

4.8 The modeling strategy used above to obtain smoothed values for China has been applied to nearly Table A4.2. Estimatesor 90 other developing countries (Hill and Pande 196C-90 forthcoming). The estimates from the models have Year EstimatedUnder- been combined with information on per capita income FiveMortality Rate to examine the relationship between change in U5MR Flv 173 from one time point to another (for instancefrom 1960 173 1980 to 1985) to change in per capita GDP over the 1965 144 same time period. Change in U5MR from one time 1970 115 point to the next is only calculated if each time point 1975 85 is supported by data. In other words, no trends are 1980 calculated only on the basis of extrapolationsfrom the 1980 country-specific model. Information on per capita 1985 44 income has been taken from the Penn World Table 1990 44 Mark 5.6 (NBER, 1994). Two measures of income- real per capita GDP in constant 1985 dollars (RGDP), and real per capita GDP in current internationalprices (CGDP), both adjusted for purchasing power parity (PPP)--have been used. Table A4.3 shows the growth rates in income used for China and the (unweighted) mean of its value for the other countries in the analysis. The model relates changes in U5MR to changes in per capita income, adjusted by the U5MR values at the beginning of each observation. By introducingdummy variables, the model stratifies the observations into one of the three categories of U5MR at the beginning of the observation period: (i) U5MR < 85 ("low U5MR"); (ii) 85 < U5MRI< 177 (middle U5MR); and (iii) U5MR > 177 (high U5MR). The model equation is:

0.21n(U5MR,+ 5/U5MRiL) = aO + a{O.21n(RGDPi,,,+5RGDPj,)} + a2Dummny(low_U5MRJ,)+ a3 Dummy (middle_U5MR,t,)

117 Annex4

Table A4.3. Income GrowthRates, China and Other Countries, 1960-90

Real Per Capita GDP Growth Rates, in Constant1985 Dollars,International Prices/' Time Period China Other Countriesin Sample (samnplesize) 2 1960.0-1977.9 3.1% 3.8% (70) 1977.9-1980.7 5.1% 1.5% (73) 1980.7-1983.5 5.7% -0.2% (74) 1983.5-1986.5 3.0% 0.2% (63) 1986.5-1990 1.0% 0.7% (66) Source:'Me Penn WorldTable (Mark 5.6), November 20, 1994. Incomegrowth rates adjusted for PurchasingPower Parity (PPP) tend to be lowerthan rates not adjustedfor PPP. 2 Unweightedmean was used to calculatethe average growth rate for 'other countries."

Summary Statistics:

Coefficient t-statistic p> It a, -0.0173 -8.190 0.000 C1 -0.0884 -2.423 0.016 a2 -0.0348 -10.689 0.000 a 3 -0.0187 -6.375 0.000

4.9 Since the model relates changes in U5MR to changes in per capita income, it can be regardedas a cross-sectionalmodel with country-fixed effects. Thuscross-country differences in tastes for child survival, or differences in knowledge about child survival, or differences in health infrastructurethat persistacross the periodstudied, will not affectthe results. By stratifyingthe observationsinto low, middle,and highU5MR groups, the modelalso adjustsfor the differential effectsof initialU5MR on the relationshipbetween changes in U5MRand changesin per capita income.

4.10 UsingRGDP, the modelfinds a smallbut significantnegative relation between changes in U5MRand incomegrowth. The coefficientson the dummyvariables are also significant, indicatingstatistically significant differences between groups with high, medium, and low U5MRs. FigureA4.2 showsthe scatterplot (for a totalof 332 country/observations)of rate of changein U5MRagainst rate of changein income,with the predictedtrend linesfor the three categoriesof initialU5MR levels. The Figurealso showsthe observationsfor Chinaby time period(1960 is the changefrom 1960to 1965,and so on). For 1960-65,China's performanceis closeto the ratespredicted by the regressionequation. For 1965-69,1970-75, 1975-80, and 1980-85,China's observationsfall belowthe regressionline (middleU5MR category), implying a

118 Figure A4.2. Relation Between Change in Cbild Mortality and Change in Per Capita Income, China and Other Countries, 1960-89

0.05

Cinu, 19851 899 0_00 ______I

0~~~~~~~ 0

°'°°Predic |edfor High U5MR

* * A; 4 .0 5 . _ _ctedformiddk _____ U5MR. _Pred

IZ -0.05 - - ______China,64 o1960- Prediredor low USMR

C 196. \ 0

-0.10------~~~~~~~~~~0 0 *HIw USMR o China, 10j7519 79 aMWdleUSMAR 0 o 0 China

-0.1S i 40.15 -0.10 -0.05 0.00 0.05 0.10 0.15 Growth Rate In Per Capita kome n Annex4 faster than expected decline in child mortality given the growth of income. In 1985-90, however, China's performance is well above the regression line (low U5MR category), indicating a slower than expected decline in child mortalitygiven the income gain and the initial U5MR level.

4.11 Figure A4.3 presents the results for China alone in bar graph form.

Figure A4.3: Deviation of China's Annual Rate of Decline in Under-SMortality Rates fromRates Predictedby GrowthRates in Income

0.060 _

0.040

1 ~~0.030t Deviation 0.020 1 l from 0.010 predicted rates .0 -. 10 1960-64 -0.Olo 1965-69 1980-84 1985-89 -0.020 -0.030 1970-74 -0.040 l 1975-79

Note: Negativedeviations indicate that China'sunder-five mortality rates weredeclining morerapidly than would be predicted-- in other words,China was doing well. Positive deviations(e.g. 1985-89)point to poorperformance -- the declinewas slower than would be expected.

Conclusions

4.12 China achieved rapid reductions in U5MR from the early 1960s to the early 1980s, with rates of decline in excess of 6 percent per year. The pace of decline appears to have slowed, possibly to a standstill, from about 1983 onwards, however.

4.13 When rate of change in U5MR is related to rate of change in per capita income, China appears to have achieved faster rates of decline than income growth alone would have predicted from 1960 to 1985, but the rate of decline for the period 1985-90 was much slower (it actually increased slightly) than income growth alone would have predicted. It would be rash, however, to draw strong conclusions from a single period: although U5MR trends in China and their relation to income gain are suggestive of less effective health care provision, more detailed work on time series and provincial differentialswould be necessary in order to draw firm conclusions.

120 Annex 5

ANNEX 5: HEALTH INDICATORS FROM DISEASE SURVEILLANCE POINTS SYSTEM

Annex 5 Table I

Table A5.1. Age-Specific Mortality Statistics, Urban and Rural Areas, 1990-94 Total Urban Rural 1990 E0 (years) 71.2 73.9 70.5 IqO 0.024 0.009 0.027 5qO 0.034 0.013 0.039

4 5q15 0.101 0.085 0.106 1991 E. (years) 70.8 73.4 70.3 1q0 0.033 0.019 0.035 5qO 0.043 0.026 0.045 ______45qls 0.104 0.083 0.110 1992 E. (years) 70.7 75.2 69.6 1qo 0.037 0.016 0.040 5qo 0.043 0.018 0.048

______= 4 5q,s 0.098 0.073 0.106 1993 Eo (years) 71.5 75.2 70.7 lqO 0.034 0.013 0.038 5qO 0.037 0.013 0.042 45qljl 0.096 0.073 0.102 1994 Eo(years) 72.2 75.5 71.4 1qo 0.036 0.017 0.040 5qO 0.037 0.017 0.041 45q15 0.091 0.071 0.097 Notes: Life expectancy at birth, E,, is the number of years that a person born in a given year could expect to live, given the age-specific mortality rates for that year. 1qo refers to the probability of dying between birth and age 1. The child mortality rate, 5qo , refers to the probability of dying by exact age 5. The adult mortality rate, 45q,5, is the probability of dying between ages 15 and 60.

Source: Chinese Disease Surveillance Points System. Data are presented for the period 1990-94 because the sampling base is only nationally representativefor those years.

121 Annex 5

Annex 5 Table 2

Table A5.2. Male Age-Specific Mortality Statistics, Urban and Rural Areas, 1990-94 Total Urban Rural 1990 E. (years) 69.2 72.2 68.5 1qO 0.025 0.011 0.027 5qO 0.035 0.014 0.039 45qI5 0.118 0.101 0.123 1991 E. (years) 68.9 71.1 68.5 IqO 0.032 0.020 0.033

5qO 0.042 0.028 0.044 45q,5 0.123 0.103 0.129 1992 E, (years) 68.9 73.1 67.8 lqo 0.035 0.017 0.038

5qO 0.042 0.018 0.046 45qI5 0.117 0.088 0.125 1993 E, (years) 69.5 73.1 69.0 .lIqO 0.033 0.015 0.037

5qO 0.037 0.016 0.041

4 5qI 5 0.117 0.088 0.124 1994 E, (years) 70.2 73.0 69.4 Iqo 0.034 0.018 0.038 5qo 0.037 0.018 0.041 45ql' 0.110 0.088 0.118 Notes: Life expectancy at birth, E , is the number of years that a person born in a given year could expect to live, given the age-specific mortality rates for that year. iqo refers to the probability of dying between birth and age 1. The child mortality rate, 5qo, refers to the probability of dying by exact age 5. The adult mortality rate, 45q,5, is the probability of dying between ages 15 and 60.

Source: Chinese Disease Surveillance Points System. Data are presented for the period 1990-94 because the sampling base is only nationally representative for those years.

122 Annex 5

Annex 5 Table 3

Table A5.3. Female Age-Specific Mortality Statistics, Urban and Rural Areas, 1990-94 Total Urban Rural 1990 Eo (years) 73.3 75.7 72.7 lqO 0.023 0.008 0.026

5 qO 0.034 0.012 0.036 45qI5 0.083 0.044 0.116 1991 Eo (years) 73.5 75.9 72.1 ,qO 0.025 0.018 0.038 5qO 0.034 0.025 0.047 45ql5 0.084 0.062 0.090 1992 Eo (years) 72.7 77.4 70.8 lqO 0.039 0.015 0.053 5qO 0.045 0.017 0.059 45ql5 0.079 0.056 0.085 1993 E0 (years) 73.6 77.6 72.8 lqO 0.036 0.011 0.041 5qO 0.037 0.011 0.043 45q,5 0.074 0.058 0.079 1994 E0 (years) 74.5 78.2 73.5 1qO 0.037 0.015 0.042 5qO 0.037 0.015 0.042 0.075 4 5 q15 0.070 0.053 Notes: Life expectancyat birth, E., is the numberof yearsthat a personborn in a givenyear could expect to live,given the age-specificmortality rates for that year. lqo refersto the probabilityof dying between birth and age 1. The child mortality rate, 5qo, refers to the probability of dying by exact age 5. The adult mortality rate, 45q,5, is the probability of dying between ages 15 and 60.

Source: Chinese Disease Surveillance Points System. Data are presented for the period 1990-94 because the sampling base is only nationally representativefor those years.

123 Annex 5

Annex 5 Table 4

Table A5.4. Male Age-Specific Mortality Statistics, Rural Areas, by Income Quartile, 1990-94 Income quartile Lowest Lower middle Upper middle Highest 1990 E. (years) 64.3 68.5 68.7 68.9 IqO 0.037 0.034 0.025 0.024 5qO 0.076 0.045 0.033 0.035 45ql5 0.175 0.122 0.123 0.122 1991 E, (years) 65.0 68.7 68.6 69.4 IqO 0.095 0.035 0.026 0.026 5qO 0.107 0.046 0.037 0.033

4 5q, 5 0.126 0.126 0.134 0.125 1992 E. (years) 66.2 69.3 66.7 68.9 . IqO 0.072 0.041 0.036 0.029 5qO 0.084 0.050 0.043 0.035 45q15 0.140 0.121 0.134 0.117 1993 E. (years) 67.4 70.7 68.7 68.2 1q0 0.055 0.044 0.034 0.030

5 qO 0.063 0.050 0.035 0.035 45q,5 0.145 0.116 0.122 0.131 1994 E. (years) 66.8 69.6 70.1 69.3 Iqo 0.087 0.053 0.031 0.026 5qo 0.090 0.053 0.032 0.031

45 q15 0.146 0.114 0.113 0.122 Notes: Life expectancy at birth, E,, is the number of years that a person born in a given year could expect to live, given the age-specific mortality rates for that year. lqo refers to the probability of dying between birth and age 1. The child mortality rate, 5qo, refers to the probability of dying by exact age 5. The adult mortality rate, 45q15, is the probability of dying between ages 15 and 60.

Source: Chinese Disease Surveillance Points System. Data are presented for the period 1990-94 because the sampling base is only nationally representative for those years.

124 Annex 5

Annex 5 Table 5

Table A5.5. Female Age-Specific Mortality Statistics, Rural Areas, by Income Quartile, 1990-94

1ncome quart(ae Lowest Lower middle Upper middle Highest 1990 E0 (years) 69.4 72.5 73.1 73.1 lqO 0.032 0.034 0.026 0.021

5qO 0.104 0.026 0.031 0.035

45 q15 0.169 0.084 0.083 0.081 1991 E. (years) 72.8 72.6 72.7 72.2 IqO 0.016 0.040 0.030 0.031

5qO 0.032 0.051 0.037 0.036

4 5 q 15 0.117 0.089 0.091 0.086 1992 E, (years) 68.0 72.0 70.8 73.2 lqO 0.067 0.050 0.039 0.031 5qO 0.073 0.058 0.047 0.036

4 5 q,5 0.116 0.078 0.092 0.080 1993 E. (years) 69.0 73.6 73.2 72.8 lqO 0.078 0.051 0.034 0.031

5qO 0.078 0.052 0.035 0.035 45q,5 0.092 0.072 0.079 0.080 1994 E,, (years) 67.6 72.4 74.9 73.6 IqO 0.099 0.060 0.032 0.028 5qO 0.093 0.056 0.032 0.030 L45ql 150.096 0.072 0.074 0.075 Notes: Life expectancy at birth, E,, is the number of years that a person born in a given year could expect to live, given the age-specific mortality rates for that year. lqo refers to the probability of dying between birth and age 1. The child mortality rate, 5 qO, refers to the probability of dying by exact age 5. The adult mortality rate, 45q,5, is the probability of dying between ages 15 and 60.

Source: Chinese Disease Surveillance Points System. Data are presented for the period 1990-94 because the sampling base is only nationally representativefor those years.

125 I

I Background Papers and Consultant Reports Prepared for This Study

Akin, John S., and Paul Hutchinson. 1995. Health Insurance and the Rural Poor in China. Mimeo. Washington D.C.: World Bank.

Berman, Peter, Gilles Fortin, Vernon Hicks, William McGreevey, and J. Brad Schwartz. 1995. China's National Health Accounts. Washington D.C.: World Bank.

Cai, Renhua. 1995a. Trial of Reforming the Employee Medical Care Insurance System in Zhenjiang. Ministry of Health. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Chen, Xiaoming. 1994. Importance of Drug Income to Cross-Subsidize Hospitals' Losses. Journal of Health Resources, II. pp. 27-30.

Chen, Xiaoming. 1995. The Analysis of Hospitals/Doctors Prescribing Behavior. Shanghai Medical University. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Cross, Harry E. 1995. Health Care and Poverty in China: Issues in Providing Basic Services to the Poor. Mimeo. Washington D.C.: World Bank.

Gu, Xingyuan, Baogang Shu, Bo Cha, Jie Yu, Wei Deng, 1996. Chinese Mortality And Life Expectancy. Mimeo. Shanghai Medical University.

Hammer, Jeffrey S. 1996. Setting the Context of Health Care Finance in China. Mimeo. Washington D.C.: World Bank.

Hou, Yan and Heyu Zhou. 1995. The Relationship Between Health Capital Investment and Health Operating Expenses. The State Planning Commission. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Hsiao, William C. 1996. Notes on Health Reform in China. Mimeo in preparation.

Hu, Haobo. 1995. Report about High-Tech Medical Equipment (HTME). Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Hu, Shanlian. 1995. Pricing Policy of Health Services in China. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Hu, Teh-Wei. 1995. Cigarette Taxation in China: Lessons from International Experience. Mimeo. Draft prepared for the Human Development Department, Washington, D.C.: World Bank.

Jin, Shuigao. 1995a. Case Study on Health Finance of Shanxi, Jiangsu, and Guizhou Province. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

127 Jin, Shuigao. 1995b. Trends in the Infant Mortality Rate and Life Expectancy Change, 1949- 1955. Mimeo.

Jin, Shuigao. 1995c. Trends in the Prevalence and Mortality for Tuberculosis and Schistosomiases, 1980-1994. Mimeo.

Luo, Wujin. 1995. The Health Financing Study in Rural Area of China. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Mao, Zhenzhong. 1995. Studies in Sichuan Rural Health Insurance Experiments. West China Medical University in Chengdu, Sichuan. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Meng, Jianguo. 1994. Pilot Reform of the Separation Between State-Owned Large Enterprises and Their Hospitals. Ministry of Health. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Meng, Jianguo. 1995. Financial Analysis of Medical Facilities. Ministry of Health. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Wei,Ying. 1995. An Introduction to Health Financing Patterns in China. National Health Economics Institute, Beijing.

Yang, Hui. 1995. Political Research on Re-establishing and Improving the Chinese Rural Cooperative Medical Care System. Beijing Medical University. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Yip, Winnie. 1995. The Role of Government in Health Care Financing in China. Consultant Report to the China Health Care Financing Mission, April 10 - May 6, Washington, D.C.: World Bank.

Yip, Winne. 1996. Notes on the Zhenjiang Demonstration Project.

Yu, Dezhi and Xinming Liu. 1995. China's Economic Reform in 1980s and its Impact on Health Care Services. Ministry of Health. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Zhao,Yuxin. 1995. Policy Issues on Governmental Funding to Health Work in the People's Republic of China. National Health Economics Institute.

Zhao, Zhuyan and Lusheng Wang. 1994. Present Ways to Raise Funds for Health Care in the Countryside of China. National Health Economics Institute. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Zhenjiang, City of. 1995. Progress Report on the Pilot Health Insurance Reform. Mimeo. Zhenjiang, Jiangsu.

128 Bibliography

Ad Hoc Committeeon Health Research Relatingto Future InterventionOptions. 1996. Investing in HealthResearch and Development. Geneva: World Health Organization.

Akin, John S., and Paul Hutchinson. 1995.Health Insuranceand the Rural Poor in China. Mimeo. WashingtonD.C.: World Bank.

Barnum, Howard,Joseph Kutzin, and Helen Saxenian.1994. Incentivesand ProviderPayment Methods.Population, Health and NutritionDivision, Washington, D.C.: World Bank.

Berk, M. and J. Monheit. 1992. The Concentrationof Health Expenditures,An Update. Health Affairs, pp. 145-158.

Berman, Peter, Gilles Fortin,Vemon Hicks, WilliamMcGreevey, and J. Brad Schwartz.1995. China's NationalHealth Accounts. WashingtonD.C.: World Bank.

Bloom, Gerald, XingyuanGu, and ShenglanTang. 1995. FinancingRural Health Services in China in the Contextof EconomicReform. Journal of InternationalDevelopment 7, 3, pp. 423-441.

Cai, Renhua. 1995a.Trial of Reformingthe EmployeeMedical Care InsuranceSystem in Zhenjiang.Ministry of Health. Paperpresented at a Health Care Seminar in Beijing, China, 2-5 May.

Cai, Renhua. 1995b. An ExploratoryDiscussion of ChineseWorkers Medical Insurance Reform. IHPP Sixth Meeting of Participants.Beijing, 2-7 October.

Cardoso,Eliana. 1995.Can China RescueIts State-OwnedEnterprises? Mimeo. Washington D.C.: World Bank, EA2/EA3.

Chang,Ying, FengyingZhai, WenjunLi, Keyou Ge, Daxun Jin, and M. de Onis. 1994. Nutritional Statusof PreschoolChildren in Poor Rural Areas of China. Bulletin of the WorldHealth Organization72 (1), pp.105-1 12.

Chen, Hu Lin. 1993.Cooperative Medical Systemin China.International Journal of Health Services 23 (4), pp. 733-742.

Chen, Jie. 1994. Evaluationof TechnologyCost and Diffusionand Policy Implications: Health Social Science,III.

Chen, Jie. 1994. Evaluationof TechnologyCost and Diffusionand Policy Implications: Health Social Science,III.

Chen, Shaohua,and MartinRavallion. 1996. Data in Transition:Assessing Rural Living Standardsin SouthernChina. Mimeo. WashingtonDC: World Bank, PRDPH.

129 Chen, Xiaoming. 1995. The Analysis of Hospitals/Doctors Prescribing Behavior. Shanghai Medical University. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

China Network for Training and Research in Health Economics and Financing. 1994a. Financing and Organization of Health Care for the Poor Rural Population in China: Report on Project Activities. Paper presented at the HIP Asia Meeting, Olongapo City, May 16-17.

China Network for Training and Research in Health Economics and Financing. 1994b. Paper Catalogue of Financing and Distributing in Medical Care and Health Finance in China. (Mimeo provided through EDIHR).

China Network for Training and Research in Health Economics and Financing. 1995a. Summary of the State Government Policy Forum on Health Care Financing and Organization in Poor Rural Areas of China. Mimeo. Beijing: Health Economics Institute.

China Network for Training and Research in Health Economics and Financing, Harvard School of Public Health. 1995b. Study of Health Care Financing and Organization in Impoverished Rural Areas of China. International Seminar on Financing and Organization of Health Care for Poor Rural Population in China, Fragrant Hills Hotel, Beijing, 8-10 October.

China Network for Training and Research in Health Economics and Financing, Harvard School of Public Health. 1996. Report to UNICEF and IHPP on the Study of Thirty Poor Countries.

China, Government of. 1993. China Social Sector Policies. Extracts from the Decision of the CPC Central Committee on Issues Concerning the Establishment of a Socialist Market Economic Structure, Mimeo trans. 14 November.

China Tuberculosis Control Collaboration. 1996. Result of directly-observed short-course chemotherapy in 112842 Chinese patients with smear-positive tuberculosis. The Lancet Vol. 347, Feb 10, 1996.

Chinese Journal of Hospital Management. 1993. # (9).

Chu, David K. W. 1994. Economic Development and the Health Care System in Hong Kong. Health Policy 28. pp. 211-234.

Clayton, S., H. Yang, J. Guan, Z. Lin, and R. Wang. 1993. Hepatitis B Control in China: Knowledge and Practices among Village Doctors. American Journal of Public Health 83, pp.168 5-168 8 .

Cretin, S., Albert P. Williams, Jeffrey J. Sine, Dechun Luo, Dexiao Zhang, Yuanqiu Shi, Shuqin Yang, and Tingwei He. 1995. China Rural Health Insurance Experiment. IHPP Sixth Meeting of Participants. Beijing, 2-7 October.

130 Cretin, S., N. Duan, A.P. Williams,Jr., X. Gu, and Y. Shi 1990.Modeling the Effect of Insuranceon Health Expendituresin the People's Republicof China. Health Services Research,October 25 (4), pp. 667-683.

Cross, Harry E. 1995. Health Care and Povertyin China:Issues in ProvidingBasic Servicesto the Poor. Mimeo. WashingtonD.C.: World Bank.

Dai, Zhicheng. 1995. China's Public Health Work;Achievements and Prospects. IHPPSixth Meeting of Participants.Beijing, 2-7 October.

De Geyndt,Willy, X. Zhao, and L. Shunli. 1992.From BarefootDoctor to Village Doctor in Rural China. World Bank TechnicalPaper 187. Washington,D.C.: World Bank.

De Geyndt,Willy. 1994. Managingthe Quality of Health Care in DevelopingCountries. Human Resourcesand Social DevelopmentDivision Asia TechnicalDepartment. Washington, D.C.: World Bank.

Feng, Xueshan,Shenglan Tang, Gerald Bloom, MalcolmSegall, and XingyuanGu. 1995. CooperativeMedical Schemesin ContemporaryRural China. Social Scienceand Medicine 41, 8, pp. 1111-1118.

Food and AgricultureOrganization. 1994. FAG Annual Yearbook.

Fu, Wei and others (ChinaNetwork of Trainingand Research in Health Economicsand Financing). 1995. Chinese Poor Rural Areas Demand-SideHealth Financing Systems Research. Mimeo. July.

Greenhalgh,Susan. 1994.The SocialDynamics of Child Mortalityin Village Shaanxi.The PopulationCouncil Research DivisionWorking Papers, No. 66. New York.

Griffin, CharlesC. 1992. Health Care in Asia, A ComparativeStudy of Cost and Financing. World Bank Regionaland SectoralStudies. Washington, D.C.: World Bank.

Gu, Xingyuan. 1992.Infant MortalityRate in 8 Provincesin China. Chinese Journalof Health Statistics,9, (3), p. 7.

Gu, Xingyuan. 1995. Impacts of Research on Policy:Financing of Health Care in Poor Rural China. IHPP Sixth Meeting of Participants.Beijing, 2-7 October.

Gu, Xingyuan, Kaining Zhang, and J. Kaufman. 1993.A Proposalto Investigatethe Financing, Provision,and Utilizationof ReproductiveHealth Services in Rural China. (July 1993- July 1996). ShanghaiMedical University.

Gu, Xingyuan,Gerald Bloom, ShenglanTang, YingyaZhu, ShouqiZhou, and XingbaoChen. 1993. FinancingHealth Care in Rural China:Preliminary Report of a Nationwide Study. Social Science and Medicine,36, (4), pp. 385-391.

131 Gu, Xingyuan, Shenglan Tang, and Suhua Cao. 1995. Financing of Health Services in Poor Rural China: A Case Study in Donglan County. Department of Health Statistics and Social Medicine, School of Public Health, Shanghai Medical University, Shanghai.

Gu, Xingyuan, Shenglan Tang, Gerald Bloom, and Henry Lucas. 1995a. Health Expenditures and Finance in Three Poor Counties of China. Institute of Development Studies Working Paper 21. Brighton, England: University of Sussex.

Gu, Xingyuan, Shenglan Tang, Gerald Bloom, and Henry Lucas. 1995b. A Strategy for Health Sector Reform in Rural China. Chinese Primary Health Care 9, 8, pp. 1-49. [Special issues with many articles on rural health care reform].

Gu, Xingyuan and Shenglan Tang. 1995. Reform of the Chinese Health Care Financing System. Health Policy 32, pp. 181-191.

Gu, Xingyuan, Baogang Shu, Bo Cha, Jie Yu, Wei Deng, 1996. Chinese Mortality And Life Expectancy. Mimeo. Shanghai Medical University.

Guiyang People's Provincial Hospital. 1995. Revenue and Expenditure Tables for 1994. Provided to Mission in Guiyang on April 24.

Guizhou Health Bureau, Provincial Health and Anti-Epidemic Preventive Station. 1995. "The Development of Guizhou Provincial Health and Anti-Epidemic Preventive Station and the Existing Difficulties," Paper presented to Mission in Guiyang, April 24.

Guo, Shuqing. 1994. Present Situation, Experiments and Prospects of China's Social Insurance System Reform. Department of Comprehensive Programming and Experimenting, State Commission for Restructuring the Economic System (SCRES). Washington, D.C.: IMF Institutional Seminar on Public Sector Pension Reform, 8-12 August.

Hammer, Jeffrey S. 1996. Setting the Context of Health Care Finance in China. Mimeo. Washington D.C.: World Bank.

Henderson, Gail, M. Goldman, and D. Simon. 1988. Issues in the Modernization of . Science and Technology in China. Harvard University Press. Cambridge, Massachusetts.

Henderson, Gail, J. Akin, Zhiming Li, Shuigao Jin., Haijiang Ma, and Keyou Ge. 1992. Equity and Utilization of Health Services: Report of an Eight-Province Survey in China. Paper Prepared for Presentation at the Annual Meeting of the Association for Asian Studies (April). Washington, D.C.

Henderson, Gail, Shuigao Jin, John Akin, Zhiming Li, Jianmin Wang, Haijiang Ma, Yunan He, Xiping Zhang, Ying Chang, and Keyou Ge. 1995. Distribution of Medical Insurance in China. Social Science and Medicine 41, 8, pp. 1119-1130.

Hill, Kenneth. 1988. Demographic Trends in China from 1950 to 1982. World Bank Discussion Paper No. 22. Washington D.C.: World Bank.

132 Hill and Pande. 1996 forthcoming. Trendsin Child Mortality in the DevelopingWorld 1960to 1995. New York: UNICEF.

Hong, Wang. 1995. Analysisof Demand in Poor Rural China:Implications for Financingand OrganizingHealth Care. Mimeo. InternationalSeminar on Financingand Organization of Health Care for Poor Rural Populationin China, FragrantHills Hotel, Beijing,8-10 October.

Hong, Phua Kai. 1991. Privatizationand Restructuringof Health Services in Singapore.Institute of Policy Studies OccasionalPaper No. 5. Singapore.

Hong, Phua Kai. 1993. Health Cost ContainmentPolicies in Singapore.National University of Singapore,Dept. of Community,Occupational and FamilyMedicine. Singapore.

Hou, Yan and Heyu Zhou. 1995. The RelationshipBetween Health Capital Investmentand Health OperatingExpenses. The State PlanningCommission. Paper presentedat a Health Care Seminarin Beijing,China, 2-5 May, 1995.

Hsiao, W. 1984. Transformationof Health Care in China. The New EnglandJournal of Medicine 310 (14), April 5. pp. 932-936.

Hsiao, William C. 1995b. The ChineseHealth Care System:Lessons for OtherNations. Social Science and Medicine 41, 8, pp. 1047-1055.

Hsiao, William C. 1995d. CommunityHealth Care. Mimeo. InternationalSeminar on Financingand Organizationof Health Care for Poor Rural Populationin China, Fragrant Hills Hotel, Beijing, 8-10 October.

Hsiao, WilliamC. 1996.Notes on Health Reform in China. Mimeo in preparation.

Hu, Haobo. 1995. Report about High-TechMedical Equipment (HTME). Paper presentedat a .Health Care Seminarin Beijing,China, 2-5 May.

Hu, Shanlian. 1995. Financingand Cost of 15 SelectedItems of MCH Servicesin the Poor Counties of China. IHPP Sixth Meeting of Participants.Beijing, 2-7 October.

Hu, Shanlian. 1995a. Policy ImpactExercise Disseminating Research Findings on Policy: Financingand Organizationof Health Care for Poor Rural Populationin China. IHPP Sixth Meeting of Participants.Beijing, 2-7 October.

Hu, Shanlian. 1995b. PricingPolicy of Health Servicesin China. Paper presentedat a Health Care Seminar in Beijing,China, 2-5 May.

Hu, Shanlian. 1996. China Health SectoralReport: Problemsand Prospects. Final Meeting on Social Sector Issues in Asian TransitionEconomies. Manila:Asian DevelopmentBank.

Hu, Shanlian,Songtao Liu, Yan Ma, YunjieSong, Ming Wu, JialinJiang, and LeiyaHan (China Networkof Trainingand Researchin HealthEconomics and Financing).1995. Financing and Costof 15 SelectedItems of MCH Servicesin the Poor Countiesof China.Paper 133 presentedat the IHPPResearch,Writing and DisseminationWorkshop, Washington, D.C., March 1-9.

Hu, Teh-Wei. 1988.Diffusion of WesternMedical Technology in China since the Economic Reform. International Journal of Technology Assessment in Health Care 4, pp. 345-358. CambridgeUniversity Press.

Hu, Teh-Wei,Jushan Bai, TheodoreE. Keeler, Paul G. Barnett,and Hai-YenSung. 1994. The Impact of CaliforniaProposition 99, a Major Anti-SmokingLaw, on Cigarette Consumption. Journal ofPublic Health Policy 15, 1, pp. 26-36.

Hu, Teh-Wei,Elizabeth Li, and Zhi Hua Lin. 1995. InsuranceCoverage and Out-of-Pocket Medical Expendituresamong Chinese Urban Workers. Mimeo. BerkeleyCA: Universityof California,School of Public Health.

Hu, Teh-wei, Hai-YenSung, and TheodoreE. Keeler. 1995. ReducingCigarette Consumption in California:Tobacco Taxes vs an Anti-SmokingMedia Campaign. AmericanJournal of Public Health 85, 9, pp.1 218 - 1222.

Hu, Teh-Wei,Qiufang Ren, TheodoreE. Keeler, and Joan Bartlett. 1995. The Demand for Cigarettesin Californiaand BehavioralRisk Factors. Health Economics4, pp. 7-14.

Hu, Teh-Wei,Theodore E. Keeler, Hai-YenSung, and Paul G. Barnett. 1995. The Impact of CaliforniaAnti-Smoking Legislation on CigaretteSales, Consumption,and Prices. Tobacco Control 1995 4 (Supplement I ) pp. S34-38.

Hu, Teh-Wei. 1995. CigaretteTaxation in China: Lessons from InternationalExperience. Mimeo. Draft Preparedfor the HumanDevelopment Department, Washington D.C.: World Bank.

Hu, Teh-Wei.(date unknown)Medical Technology Transfer in Major Chinese Teaching HospitalsWith SpecialReference to DiagnosticEquipment. University of California School of Public Health. Berkeley.

Hu, Zu-liu. 1994. Social Protection,Labor MarketRigidity, and EnterpriseRestructuring in China. IMF Paper on Policy Analysisand Assessment.Washington D.C.: IMF, Fiscal Affairs Department.

Jalan, J. and Ravallion,M. 1996. Transientpoverty in rural China. Unpublishedpaper, Policy Research Departmentof the WorldBank.

Jamison,Dean T. 1985.China's Health Care System:Policies, Organization, Inputs and Finance, in Good Health at Low Cost. eds. Scott B. Halstead,Julia A. Walsh, and Kenneth Warren, eds. RockefellerFoundation. New York.

Jiang, Z., E. Zhang, S. Chen, and H. Yu. 1995. Statisticsof 1992Fertility SamplingSurvey in China. Beijing: China PopulationPublishing House.

134 Jie, Chen. 1994. Evaluation of Technology Cost and Diffusion and Policy Implications. Health Social Science, III, pp. 4-5.

Jin, Shuigao et al. 1994. Preliminary Analysis on the Impact of Economic Development on Health Status in Rural Areas, in Food, Nutrition and Health Status of Chinese in 7 Provinces, 1990, China, Chunming Chen and Zongming Shao, eds. China Statistical Publishing House, Beijing.

Jin, Shuigao. 1995a. Case Study on Health Finance of Shanxi, Jiangsu, and Guizhou Provinces. Paper presented at a Health Care Seminar in Beijing, China 2-5 May.

Jin, Shuigao. 1995b. Trends in the Infant Mortality Rate and Life Expectancy Change 1949- 1955. Mimeo.

Jin, Shuigao. 1995c. Trends in the Prevalence and Mortality for Tuberculosis and Schistosomiases, 1980-1994. Mimeo.

Jin, Shuigao. 1995d. An Evaluation on Smoking-induced Health Costs in China (1988-1989). Biomedical and Environmental Sciences 8, pp. 342-349.

Kaji, G. 1993. Health Challenges for China. (Draft speech on the occasion of the Seminar commemorating ten years of operational collaboration between China and the World Bank in the health sector.) Mimeo. China and Mongolia Department. Washington, D.C.: World Bank.

Kan, Xuegui. 1990. Village Health Workers in China: Reappraising the Current Situation. Health Policy and Planning; 5(1): pp. 40-48.

Keidel, Albert. 1995. China: Regional Disparities. Mimeo. Washington D.C.: World Bank, China and Mongolia Department.

Kessel, R. 1958. Price Discrimination in Medicine. Journal ofLaw andEconomics 1, pp. 1-19.

Khoman, Sirilaksana. 1995. Rural Health Financing; Thailand's Experience. Mimeo. Presented at IHPP International Seminar. Beijing, 8-11 October.

Krieg, Renate, and Monika Schadler, eds. 1994. Social Security in the People's Republic of China. Hamburg: Workshop on Social Security in the PRC, 5-7 Feb 1993.

Li, Heng [Betsy Heng Li]. 1993. A Study of Current Government Health Insurance System and its Expenditure in China. Mimeo. Beijing: Chinese Medical Association.

Li, Jiali. 1994. China's Family Planning Program: How, and How Well, Did It Work? The Population Council Research Division Working Papers, No. 65. New York.

Li, Zhiming. 1994. Employment-Based State Medical Insurance and Job Selection in China: A Switching Simultaneous Equation Model. Chapel Hill, NC: PhD diss., University of North Carolina, Chapel Hill, Dept. of Health Policy Administration.

135 Li, Zhisui. 1994. The Private Life of Chairman Mao. New York: Random House.

Libo County Hospital. 1995. Report to the World Bank Team on County Hospital Financial Operations, Paper prepared for World Bank Mission and presented in Libo City, Guizhou Province on April 27, 1995.

Lin, Justin Yifu, Fang Cai, and Zhou Li. 1994. Why China's Economic Reforms Have Been Successful: Implications for Other Reforming Economies. (Preliminary Draft). Beijing University, Duke University, Australian National University, and The Chinese Academy of Social Sciences. May.

Liu, Gordon, X. Liu, and Q. Meng. 1994. Privatization of the Medical Market in Socialist China: A Historical Approach. Health Policy 27, pp. 157-174.

Liu, Yuanli. 1995. Inefficient Use of Limited Resources in China's Poverty Areas: Problems and Policy Implications. Mimeo. International Seminar on Financing and Organization of Health Care for Poor Rural Population in China, Fragrant Hills Hotel, Beijing, 8-10 October.

Liu, Yuanli, William C. Hsiao, Qing Li, Xingzhu Liu, and Minghui Ren. 1995. Transformation of China's Rural Health Care Financing. Social Science and Medicine 41, 8, pp.1085- 1093.

Liu Yuanli, Shanlian Hu, Fu Wei and William C. Hsiao. 1996. Is Community Financing Necessary and Feasible for Rural China? Cambridge, MA: Harvard University School of Public Health.

Liu, Xingzhu and William C. Hsiao. 1995. The Cost Escalation of Social Health Insurance Plans in China, and its Implication for Public Policy. Social Science and Medicine 41, 8, pp. 10 95-1 101.

Lok, Sang Ho. 1995. Market Reforms and China's Health Care System. Social Science and Medicine Special Issue. Aspects of the Medical Care System in the People's Republic of China, 41, 8, pp.10 65-107 2.

Luo, Wujin and Shanlian Hu. 1995. Study of Health Care Financing and Organization in Impoverished Rural Areas of China. IHPP Sixth Meeting of Participants, Beijing, 2-7 October.

Luo, Wujin, Leiya Han, Wei Fu, Ming Wu, Yunjie Song, and Jialing Jiang. (China Network of Training and Research in Health Economics). 1995. Study of Health Financing and Organization in Poor Rural Areas of China. Paper presented at the IHPP Research, Writing and Dissemination Workshop, Washington, D.C., March 1-9.

Luo, Wujin. 1995. The Health Financing Study in Rural Area of China. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

136 Mao, Zhenzhong. 1995. Studies in Sichuan Rural Health Insurance Experiments. West China Medical University in Chengdu, Sichuan. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

McEwan, Peter J. M., ed. 1995. Social Science and Medicine, Special issue. Aspects of the Medical Care System in the People's Republic of China, 41, 8, pp. 1041-1201.

Meng, Jiachi (Director, Qiannan Health Bureau). 1995. A Report Outline to the World Bank Research and Investigating Group to Raise Funds of the Chinese Health Department. Paper and personal communications presented to Mission in Qiannan Prefecture, April 24- 29.

Meng, Jianguo. 1994. Pilot Reform of the Separation Between State-Owned Large Enterprises and Their Hospitals. Ministry of Health. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Meng, Jianguo. 1995. Financial Analysis of Medical Facilities. Ministry of Health. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Ministry of Health (MOH) 1991. Selected Edition on Health Statistics of China, 1978-1990. Beijing: Center for Health Statistics Information.

Ministry of Health (MOH). 1993a. Medical and Health Services in China. Edited by the Foreign Affairs Bureau of the Ministry of Health. Beijing, December.

Ministry of Health (MOH). 1993b. Background Papers for the International Seminar on Health Reform under Socialist Market Economy. MOH Department of Planning and Finance, Beijing.

Ministry of Health (MOH). 1993c. Proposal for WHO Support for a Field Study on Rural Health Insurance Systems in the People's Republic of China 1994-1995. Beijing.

Ministry of Health (MOH). 1995a. Department of Planning and Finance. Unpublished Tables on Health-Care Finance 1978-1993. Mimeo. Beijing: MOH\DOPF.

Ministry of Health (MOH). 1995b. National Strategic Plan for the Prevention of AIDS and STD in the People's Republic of China 1996-2000. Beijing. November 1995.

Murray, Christopher L. 1994a. Quantifying the Burden of Disease: The Technical Basis for Disability-Adjusted Life Years. Bulletin of the World Health Organization, 72 (3), pp. 429-445.

Murray, Christopher L. and Alan D. Lopez. 1994b. Quantifying Disability: Data, Methods and - Results. Bulletin of the World Health Organization, 72 (3), pp. 481-494.

Murray, Christopher L. and Alan D. Lopez. 1994c. Global and Regional Cause-of-Death Patterns in 1990. Bulletin of the World Health Organization, 72 (3), pp. 447-480.

137 Murray, Christopher L., and Alan D. Lopez, eds. 1994d. Global Comparative Assessments in the Health Sector. Disease Burden, Expenditures and Intervention Packages. Geneva: WHO.

Murray, Chistopher L., Alan D. Lopez, and D. T. Jamison. 1994. The Global Burden of Disease in 1990: Summary Results, Sensitivity Analysis and Future Directions. Bulletin of the World Health Organization 72 (3), pp. 495-590.

Murray, Christopher L. and Alan D. Lopez, eds. 1996. The Global Burden of Disease and Injury Series. Vol 1. Harvard School of Public Health. Cambridge, MA: Harvard University Press.

Novotny, Thomas, E. and Richrad Peto. 1988. Estimates of Future Adverse Health Effects of Smoking in China. Public Health Reports, September-October, Volume 103 (5), pp. 552- 553.

Organization for Economic Cooperation and Development (OECD). 1995. OECD Health Data: A Software Package for the Cross-National Comparison of Health Systems. Paris: OECD.

Palacios, Robert J. 1995. Reforming Enterprise-Based Social Services in China. Mimeo. Washington, D.C.: World Bank.

Parker, R.L. Comments on maternal and child health needs and strategies for the 1990s in China. Unpublished paper, n.d.

Penn World Table (Mark 5.6): 1994 (Quarterly Journal of Economics, May. 106: 327-68.) Computer disk supplement available from National Bureau of Economic Research, Cambridge, MA. This disk is based on: Summers, Robert and Alan Heston. 1991. Penn World Table (Mark 5): An Expanded Set of International Comparisons, 1980-1988.

Perkins, Dwight. 1994. Completing China's Move to the Market. Journal of Economic Perspectives 8, 2, pp. 23-46.

Peto, Richard. Tobacco: UK and China (Letter). Lancet 1986; ii: 1038

Peto, Richard, Alan Lopez, Jillian Boreham, Michael Thun, and Clark Heath Jr. 1994. Mortality from Smoking in Developed Countries 1950-2000. Oxford: Oxford University Press.

Piazza, Alan. 1986. Food Consumption and Nutritional Status in the PRC. Boulder and London: Westview Press.

Picard, John, Lechun Xie, and Qingsi Zheng. 1992. Cost-Effectiveness of Managing Acute Childhood Diarrhea in Yunnan Province, People's Republic of China. WHO Research Reports, pp. 419-24.

138 Popkin,Barry. M., Keyou Ge, FengyingZhai, X. Guo, HaijiangMa, and N. Zohoori. 1993. The NutritionTransition in China:a Cross-SectionalAnalysis. European Journal of Clinical Nutrition, 47, pp. 333-345.

Prescott,N. and Dean Jamison. 1983.Health Sector Finance and Expendituresin China. Population,Health and NutritionTechnical Notes, GEN 14. Washington,D.C.: World Bank.

Prescott,N. and Dean Jamison. 1984.Health Sector Finance in China. WorldHealth Statistics Quarterly 37 (4), pp. 3 87-402.

Prescott,N. and Dean Jamison. 1985.The Distributionand Impact of Health Resource Availability in China. International Journal of Health Planning and Management 1, pp. 45-56.

QiannanPeople's Hospital. 1995.Report on FinancialSituation of QiannanPeople's Hospitalfor the Delegationof the WorldBank, Public Health and Finance Officials,Paper presentedby Qiu Jinquanto Mission,April 25.

Qiannan Prefecture,Health Bureau,MCH Center. 1995. A Brief Introductionto the Health Care Center for Womenand Childrenin QiannanPrefecture in Guizhou.Danyun City, Guizhou, April 26.

Qiannan Prefecture,Health Bureau, Sanitationand Anti-epidemicStation. 1995. A Report on the Current Situation,Problems, Difficulties with Sanitationand EpidemicPrevention of the QiannanPrefecture, Danyun City, Guizhou,April 25.

Ravallion,Martin, and ShaohuaChen. 1994. An Update on Rural Poverty in China. Mimeo. WashingtonD.C.: World Bank, PRDPH.

Ren, X. S. 1994. Infant and Child Survival in Shaanxi,China. Social ScienceMedicine 38 (4), pp. 609-620.

Schieber, GeorgeJ., Pierre Poullier,andLeslie M. Greenwald. 1994. Health System Performancein OECDCountries: 1980-1992. Health Affairs, Fall, Volume 13, (4), pp. 100-112.

ShanghaiMedical University,Faculty of Health Sciencesof, and Universityof Central Queensland(Rockhampton, Australia), Department of Health ServicesManagement. 1993. MultidimensionalComparisons and DevelopmentStrategies of ChineseHealth Care. A ResearchProposal. July.

ShanghaiMedical University,School of Public Health. 1994.Financing Rural Health Services in China: Promotionand Evaluationof InnovativeSchemes with SpecialReference to Poor Areas and Householdsfor the InternationalHealth Policy Program. Shanghai,May.

139 Shanghai Medical University, School of Public Health, Department of Health Statistics and Social Medicine. 1994. Series of Preliminary Studies of Three Poor Counties (Donglan, Shibing, and Xunyi), Team led by Professor Gu Xingyuan.

Shi, L. 1993. Health Care in China: A Rural-Urban Comparison After the Socioeconomic Reforms. Bulletin of the World Health Organization 71, 6, pp. 723-36.

Sidel, V. W. 1993. New Lessons from China: Equity and Economics in Rural Health Care. American Journal ofPublic Health, 83 (12), pp. 1665-1666.

Sine, Jeffrey J. 1994. Demand for Episodes of Care in the China Health Insurance Experiment. RAND Graduate School Ph.D. Dissertation. Santa Monica, California.

State Statistical Bureau. 1985, 1992, 1993, 1994, 1995. Statistical Yearbook of China. Beijing: China Statistical Publishing House.

State Pharmaceutical Administration of China. 1995. Management of China's Pharmaceutical Market and Its Problems. Mimeo. Paper presented at IFPMA Conferences, Providing Quality Medicines for Health Care, Bangkok, December 6-8.

Sun, Aimin. 1995. Rural . IHPP Sixth Meeting of Participants. Beijing, 2-7 October.

Sung, Hai-Yen, Teh-Wei Hu, and Theodore E. Keeler. 1994. Cigarette Taxation and Demand: an Empirical Model. Contemporary Economic Policy 12, pp. 91-100.

Tang, Shenglan, Gerald Bloom, Yueshan Feng, Henry Lucas, Xingyuan Gu, and Malcolm Segall. 1994. Financing Health Services in China: Adapting to Economic Reform. Research Report 26. Brighton, Sussex, England: Institute of Development Studies.

UNICEF. 1989. The State of the World's Children 1989. Oxford: Oxford University Press.

UNICEF. 1992. Children and Women of China: A UNICEF Situation Analysis. Draft, 10-31, pp. 140-141.

UNICEF. 1995. Children and Women of China: A UNICEF Situation Analysis. Draft, 15-26.

United Nations. 1995. World Urbanization Prospects: The 1994 Revision. New York: United Nations.

Wang Shusheng, Xuesheng Li, Jun Yu, Hui Yang, Tuohong Zhang, Licheng Zhang, Shuo Zhang, Yide Yue, Guy Carrin, and Aviva Ron. 1996. The Reform of the Rural Cooperative Medical System in the People's Republic of China. Geneva: World Health Organization.

Wei, Ying. 1995. An Introduction to Health Financing Patterns in China. National Health Economics Institute (Beijing). Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

140 Wei, Ying. 1996. Memo on Revised National Accounts for 1993. Washington D.C.: The World Bank.

Wei, Yu and Minghui Ren. 1995. The Important Issue of Enterprise Reforms: Health Care Insurance System. Mimeo. Boston MA and Beijing: Gerontology Center of Boston University and China's Ministry of Health.

World Bank. 1984. China. The Health Sector. A World Bank Country Study. Washington, D.C.: World Bank.

World Bank. 1990. China Long-Term Issues and Options in the Health Transition (in two volumes.) Report No. 7965-CHA. Washington D.C.: World Bank. [This gray-cover report was summarized in the 1992 red-cover report (1992a), which did not include most annex materials included here.]

World Bank. 1990a. China Between Plan and Market. Washington D.C.: World Bank.

World Bank. 1990b. China: Reforming Social Security in a Socialist Economy. Report No. 8074-CHA. Washington D.C.: World Bank.

World Bank. 1991. Lessons of Tax Reform. Washington D.C.: World Bank.

World Bank. 1992a. China: Long Term Issues and Options in the Health Transition. World Bank Country Study. Washington D.C.: World Bank.

World Bank. 1992b. China: Strategies for Reducing Poverty in the 1990s: A World Bank Country Study. Washington, D.C.: World Bank.

World Bank. 1993a. World Development Report 1993: Investing in Health. Oxford and New York: Oxford University Press for the World Bank.

World Bank. 1993b. China Rural Health Workers Development Project. Staff Appraisal Report. July 12, 1993. Report 11404-CHA Washington D.C.: World Bank.

World Bank. 1994a. China: GDP Per Capita. Report 13580-CHA. Washington D.C.: World Bank.

World Bank. 1994b. World Development Report 1994: Infrastructure for Development. Oxford and New York: Oxford University Press for the World Bank.

World Bank. 1995a. Technical Annex to the Memorandum of the President: China Fiscal Technical Assistance Project. Report No. P-6552-CHA. Washington D.C.: World Bank.

World Bank 1995b. Memorandum of the President of the International Bank for Reconstruction and Development and the International Development Association to the Executive Directors on a Country Assistance Strategy of the World Bank Group for the People's Republic of China. Report No. 14454-CH4. Washington D.C.: World Bank.

141 World Bank. 1995c. Investing in People. Washington D.C.: World Bank.

World Bank, 1995d. World Development Report 1995. Workers in an Integrating World. Oxford and New York: Oxford University Press for the World Bank.

World Bank, 1995e. Staff Appraisal Report: China Disease Prevention Project, Nov. 8, 1995. Report No. 14912-CHA. Washington, D.C.: World Bank.

World Bank. 1996a. The Chinese Economy: Fighting Inflation, Deepening Reforms. Washington D.C.: World Bank.

World Bank. 1996b. China: Pension System Reform. Washington, D.C.: World Bank.

World Bank. 1996c. Poverty Reduction and the World Bank: Progress and Challenges in the 1990s. Washington, D.C.: World Bank.

Xu, Wenbo. 1995. Flourishing Health Work in China. Social Science and Medicine 41, 8, pp.1043-1045.

Yang, Hui. 1995. Political Research on Reestablishing and Improving the Chinese Rural Cooperative Medical Care System. Beijing Medical University. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Yang, H. 1994. More Individual Management of Funds Anticipated: Insurance System Targets Pension Reform. China Daily. October 27.

Yaung, C. 1989. Current System of Health Care Financing and Delivery in R.O.C. and its Challenge for Future Development. International Symposium on Health Care Systems. Taipei, Taiwan, Republic of China.

Yip, Winnie. 1995. The Role of Government in Health Care Financing in China. Consultant Report to the China Health Care Financing Mission, April 10-May 6. Washington D.C.: World Bank.

Yip, Winnie. 1996. Notes on the Zhenjiang Demonstration Project.

Young, Mary Eming. 1983. A Study of 's Activities in China. Doctoral Thesis. The Johns Hopkins University, Baltimore, M.D.

Young, Mary Eming. 1989. Impact of the Rural Reform on Financing Rural Health Services in China. Health Policy 11, pp. 27-42.

Yu, Dezhi. 1992. Changes in Health Care Financing and Health Status: The Case of China in the 1980s. Innocenti Occasional Papers Economic Policy Series 34. December.

Yu, Dezhi and Xinming Liu. 1995. China's Economic Reform in 1980s and its Impact on Health Care Services. Ministry of Health. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

142 Yuen, P. P. 1992. Private Medicine in Socialist China: A Survey of the Private Medical Market in Guangzhou. International Journal of Health Planning and Management, Volume 7, pp. 211-221.

Zimring, F. and W. Nelson. 1995. Cigarette Taxes as Cigarette Policy. Tobacco Control, 4 (supplement 1) pp. 525-533, Appendix Table.

Zhan, Shaokang. 1995a. Drug Use and Drug Expenditure. IHPP Sixth Meeting of Participants. [draft paper circulated for comment] Beijing, 2-7 October.

Zhan, Shaokang. 1995b. Financing of Health Services in Poor Rural China. IHPP Sixth Meeting of Participants. Beijing, 2-7 October.

Zhao, Yuxin. 1995. Policy Issues on Governmental Funding to Health Work in the People's Republic of China. Health Economics Institute. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Zhao, Zhuyan and Lusheng Wang. 1995. Present Ways to Raise Funds for Health Care in the Countryside of China. Health Economics Institute. Paper presented at a Health Care Seminar in Beijing, China, 2-5 May.

Zheng, Xiang and Sheila Hillier. 1995. The Reforms of the Chinese Health Care System: County Level Changes, the Jiangxi Study. Social Science and Medicine 41, 8, pp. 1057- 1064.

Zhenjiang, City of. 1995. Progress Report on the Pilot Health Insurance Reform. Mimeo. Zhenjiang, Jiangsu.

Zhou, Y.S. et al. 1989. Analysis on the Infant Mortality Rate in China. Journal of Chinese Demographic Sciences, 3, pp. 35-46.

Zhu, Naisu, Zhihua Ling, Jie Shen, J.M. Lane, and Shanlian Hu. 1989. Factors Associated with the Decline of the Cooperative Medical System and Barefoot Doctors in Rural China. Bulletin of the World Health Organization 67, 4, pp. 431-41.

143