India Policy and Finance Strategies for Strengthening Primary Health Care Services

Transcription

1 Public Disclosure Authorized Report No IN India Policy and Finance Strategies for Strengthening Primary Health Care Services May 15, 1995 Public Disclosure Authorized Population and Human Resources Division South Asia Country Department II (Bhutan, India, Nepal) (*mav Public Disclosure Authorized Public Disclosure Authorized

2 ABBREVIATIONS AND ACRONYMS AIDS ANM ARWSP CCHFW CGHS CHC CMO DALY(S) DC DGHS DMS ESIS FP FW GDP GIDR GOI HIV ICDS ICMR IM IMR IMRB MCH MOHFW NCAER NGO NHP NIPFP NLEP NMEP NMNP NSS NTCP O&M OPD ORG PHC PPP SC/ST STD SUBC TB UP VHG WB WDR Acquired Immunodeficiency Syndrome Auxiliary Nurse Midwife Accelerated Rural Water Supply Program Central Council of Health & Family Welfare Central Govermnent Health Scheme Community Health Center Chief Medical Officer Disability Adjusted Life Years District Collector Directorate General of Health Services District Medical Superintendent Employees State Insurance Scheme Family Planning Family Welfare Gross Domestic Product Gujarat Institute of Development Research Government of India Human Immunodeficiency Virus Integrated Child Development Services Indian Council for Medical Research Initiating Memorandum Infant Mortality Rate Indian Market Research Bureau Maternal & Child Health Ministry of Health and Family Welfare National Council for Applied Economic Research Non-government Organizations National Health Policy National Institute of Public Finance & Policy National Leprosy Eradication Program National Malaria Eradication Program National Minimum Needs Program National Sample Survey National Tuberculosis Control Program Operations and Maintenance Out Patient Department Operations Research Group Primary Health Center Purchasing Power Parity Scheduled Caste/Scheduled Tribe Sexually Transmitted Disease Sub Center Tuberculosis Uttar Pradesh Village Health Guide World Bank World Development Report

3 ACKNOWLEDGEMENTS This report has been prepared by a team led by Tawhid Nawaz. Major contributions were made by Shreelata Rao-Seshadri, Salim Habayeb and Damianos Odeh. Peter Berman of the Harvard School of Public Health was the principal field investigator and V. J. Ravishankar acted as field coordinator for the various sectoral pieces that were produced by Indian research institutions. The new research carried out for the sector study was undertaken by the following institutions: National Institute of Public Finance and Policy (Dr. V.B. Tulasidhar), National Council of Applied Economic Research (Dr. Prem Vashishtha), Gujarat Institute of Development Research (Dr. Pravin Visaria and Dr. Anil Gumber) and the Foundation for Research in Community Health (Dr. Ravi Duggal). The peer reviewers were: I. Porter, K. Subbarao and D. Jamison. The report benefitted from comments from J. Salop, R. Anderson, K. Uchimura, R. Zagha, V. Kozel, V. J. Ravishankar, A. Measham, 1. Pathmanathan and M. Chatterjee. The document was produced by Nischint Bhatnagar and Jane Mukira. The report is endorsed by Richard Skolnik, Chief, Population and Human Resources Division and Heinz Vergin, Director, South Asia Country Department II (Bhutan, India and Nepal). The extensive collaboration of the Ministry of Health and Family Welfare (MOHFW), Government of India is acknowledged. The report was discussed at two workshops: one in Hyderabad organized by the Administrative Staff College of India in November, 1994, and the other in Jaipur organized by the Indian Institute for Health Management Research (IIHMR) in March, Participants at the Jaipur workshop, who discussed the report extensively, included senior officials from MOHFW, Secretaries of Health and Directors of Medical Services from 11 State Governments, representatives from SEARO-WHO, the Asian Development Bank, IIHMR, Gujarat Institute of Development Research, National Council of Applied Economic Research, Centre for Enquiry into Health and Allied Themes, All India Institute of Hygiene and Public Health and the Indian Institute of Management at Ahmedabad. MOHFW provided a summary of the main conmments made at the Jaipur Workshop. These have been incorporated into the present version of the report.

4 I

5 INDIA POLICY AND FINANCE STRATEGIES FOR STRENGTHENING PRIMARY HEALTH CARE SERVICES TABLE OF CONTENTS PAGE NO Executive Summary... Table on Main Findings & Recommendations... v-xiv xv-xxii INTRODUCTION A. Background... 1 B. Terms of Reference for the Study I C. Structure of the Report... 2 D. Definitions and Reference Points OVERVIEW OF THE HEALTH SECTOR A. Mortality, Morbidity, Fertility and the Burden of Disease... 5 B. National Health Policy C. Administrative, Budgeting and Financing Structure PUBLIC EXPENDITURES ON PRIMARY HEALTH CARE SERVICES A. Introduction B. Public Expenditures on Health C. Government Financing for Health D. Trends in the Level and Composition of Government Spending E. Impact of Government Spending on Health Status F. The Effect of Adjustment on Government Health Spending PUBLIC PRIMARY HEALTH CARE SERVICES: STRUCTURE. INPUTS, PERFORMANCE AND ESTIMATES OF FINANCING REOUIREMENTS A. The Development of Health Services B. Inputs for Public Sector Health Programs at Primary Level C. The Financial Implications of Input Gaps PRIVATE HEALTH EXPENDITURES AND PROVISION OF PRIMARY HEALTH CARE SERVICES A. Introduction B. Private Health Expenditures and Primary Health Care C. Private Provision of Health Services... 72

6 6. PUBLIC AND PRIVATE SECTOR ROLES IN MEETING THE DEMAND FOR HEALTH CARE A. Introduction B. Outpatient Treatment of Illness C. Preventive Care D. Treatment in Hospital SUMMARY OF MAIN CONCLUSIONS AND RECOMMENDATIONS FOR ACTION A. Introduction B. Increase Government Expenditure on Primary Health Care C. Improve the Structure of Public Financing for Primary Health Care D. Ensure Adequate Financing of Public Primary Health Care Facilities at the State Level E. Increase Supplementary Central Funding to Needy States F. Mobilize Resources through Financing Innovations G. Strengthen Private Primary Health Care through Improvements in Service Quality H. Encourage Private Sector Participation in Preventive and Promotive Health Services I. Improve Sectoral Effectiveness and Efficiency by Strengthening the Referral System J. Improve Understanding of the Health Needs of Women Beyond Those Covered by the Family Welfare Program K. Strengthen GOI/MOHFW Planning and Analysis Capacity L. Strengthen Training; Increase Awareness of Health-Related Issues; and Initiate Community Participation in Existing Health Education Programs M. Need for Further Analysis BIBLIOGRAPHY CHAPTER 6 APPENDICES Appendix 1: Appendix 2: Demand for Non-hospital Illness Care and Determinants of Expenditure: Multivariate analysis Hospital use and Expenditure: Multivariate Analysis STATISTICAL ANNEXES Annex 1: Annex 2: Annex 3: National Statistics of Health Care Financing and Provision Detailed Tables on Finance, Inputs, and Performance of Primary Health Care in Four States Detailed Tables on Health Care Use and Expenditures Based on NSS 42nd Round ( ) from Five Major States

7 TABLES 2.1 India: Burden of Disease in Goals for Health and Family Welfare Programmes to the Year National Health Spending: An Estimated "Sources and Uses" Matrix Growth of Health Sector Expenditure Classified by Inputs Growth of Expenditures by Sources: Total and Health Per-Capita Health Expenditures, 1970s to 1980s An Estimate of the Financing Requirement at State Level to Complete the Rural Primary Care Delivery System Household Out-of-Pocket Health Expenditure and Primary Care Percent of Total Household Expenditure to Health and to Non-Hospital Treatment by Expenditure Quintiles Differences by Age and Sex in the Probability of Use of Non- Government Providers for Non-Hospitalized Illness Treatment Public/Private Services, Poverty and Specific Diseases User Expenditure for Illness Treatment by Expenditure Quintile Cost per Treated Episode as a Percent of Monthly Household Consumption Expenditure (Users Only) Access to Free Care for Illness Treatment Illness Treated by Non-Government Providers Outside of Hospital for Total Population and Insured Population Role of Government Providers in Routine Preventive Care Expenditure Per Episode for Hospitalization by Expenditure Quintile and Type of Provider Access to Free Hospital Beds

8 iv FIGURES 2.1 India: Burden of Disease by Age and Cause The Structure of Government Health Financing National Health Spending Estimated Composition of Government Health Budget Shift in Input Composition Center and State Shares in Different Components of Government Health Budget: Changing Composition of Health Spending Infant Mortality and Per Capita Health Spending in Indian States Channels Through Which Structural Adjustment Affects Health Spending Composition of Center's Fiscal Adjustment Measures Differential Impact on Health Budgets for States Grouped by Income Levels Weighted Average of Percentage Changes in Real General Health Expenditures The Completion of Facility Norms in Sample Districts Gaps in Total Staffing Relative to Norms in Existing Government Health Facilities: 12 Districts, 4 States Distribution of Qualified Staff at CHCs and PHCs Compared to Norms Qualitative Assessment of Adequacy of Supply of Sample Drugs Facilities and Staff Levels Relative to Norms Against Overall Performance Score Relationship Between Imbalance in Inputs and Overall Performance Health Expenditure as a Percent of GDP: Asian Countries (1990) Treatment of Illness Outside of Hospital by Provider Type a Use of Non-Government Services by Different Expenditure Quintiles b Government and Non-Government Providers Hospital Treatment of Illness: Roles of Government and Non-Government Facilities Non-Government Hospital Use by Different Expenditure Quintiles BOXES 1. India's Main Public Health Programs Key Findings on Public-Private Sector Roles in Meeting Health Care Demands

9 INDIA POLICY AND FINANCE STRATEGIES FOR STRENGTHENING PRIMARY HEALTH CARE SERVICES EXECUTIVE SUMMARY SECTORAL BACKGROUND AND THE CHALLENGE OF THE FUTURE 1. India has made considerable progress in the last several decades in expanding its public health system and reducing the burden of disease. The Government has established a health policy based on the primary health care approach to provide free curative and preventive health services to a large section of the population. The National Health Policy (NHP) of 1983 has set targets for improving the health status of India's population and for reducing fertility. An extensive infrastructure has been developed during the last decade for the public provision of primary health care services. Partly as a result of these efforts, the infant mortality rate has been reduced from 137 per 1,000 live births in 1970 to 90 per 1,000 live births in 1991 and life expectancy at birth has increased over the same period from 50 to 61 years. 2. Yet, the present health care system has serious problems with respect to access, efficiency, effectiveness and quality, despite the high level of overall private and public expenditures on health. Key health indicators remain low when compared to some other Asian countries, such as Sri Lanka and China. These problems prevent the health system from achieving desirable outcomes. Moreover, India faces a formidable challenge in providing health care services to its people for several reasons: There is still an important unfinished agenda for addressing childhood and maternal morbidity and mortality, and communicable diseases. These health problems take a heavy toll on individuals and on society as a whole, but are largely preventable. The leading sources of Disability-Adjusted Life Years (DALYs)' lost, as estimated by the World Development Report (WDR, 1993) for 1990, were maternal and perinatal causes (35 million DALYs), respiratory infections (33 million DALYs), diarrheal diseases (28 million DALYs), and tuberculosis (11 million DALYs). In addition, the risks of infection from communicable diseases, such as malaria and leprosy, remain high. * New health problems such as AIDS and drug resistant forms of several communicable diseases, including malaria and tuberculosis (TB), have emerged. * The composition of the Indian population is changing towards an increasing fraction of older individuals, and as such the country must deal with a rising incidence of noncommunicable diseases, such as cardiovascular diseases and cancers. In 1991, cardiovascular diseases and cancers accounted for the loss of about 28 million and 12 million DALYs respectively. This will mean that India will have to simultaneously deal with high rates of both communicable and non-communicable diseases, and that the DALY is defined as a unit for measuring both the global burden of disease and the effectiveness of health interventions, as indicated by reductions in the disease burden (World Development Report, 1993).

10 vi Government will have to bear some of the burden of either providing for or financing the budgetary needs for addressing this problem. * Without careful attention to the provision and financing of primary health care services in general, and preventive and promotive services in particular, India risks spending an increasingly large share of its GDP on health while still failing to get sufficient return for that investment, somewhat like the pattern in the United States. * Financial constraints facing India are likely to affect Government health spending both at the state and central levels. 3. The above factors suggest that it is very important for India and its constituent states to review their approach to the provision and financing of health care and develop a coherent framework for meeting present and emerging health needs. The existing health care system is founded on the notion of universal primary health care, that is, the Government can provide and finance a wide range of health care services for a large section of its population. However, experience has shown that this approach has not worked and cannot constitute the basis for meeting the health needs of the future in an efficient, effective and equitable manner. 4. In the light of these conclusions, this report reviews some of the key challenges for the development of the health sector in India and provides recommendations for addressing the main constraints with respect to the effectiveness and efficiency of India's primary health care services. The report synthesizes the findings of several years of collaborative sector work between the World Bank and the Government of India and builds on the study "India: Health Sector Financing: Coping with Adjustment and Opportunities for Reform", that was completed in It provides a detailed analysis of private health care, evaluates the adequacy, efficiency, and effectiveness of current expenditures in the health sector, identifies and quantifies the resource gaps for delivering public primary health care services, and provides initial evidence of the impact of structural adjustment on the provision of health care. However, since the initiation of background work and primary data collection for this sector work, new themes have emerged regarding the development of the health sector in India on which the Government and the Bank have broad agreement. These include, in particular, a focus on state level health reform and state health systems development since the states account for about three-quarters of health expenditures in the public sector. This report covers some aspects of state level health reform and provides direction on further health sector work based on an emerging dialogue between the Bank and the Government of India. POLICY AND FINANCE ISSUES RELATED TO PRIMARY HEALTH CARE SERVICES 5. Two issues are most significant when considering further development and financing of primary health care in India. The first concerns the role that the public sector will play in both the provision and the financing of health care. The second concerns the amount of funds which the public sector will make available for health in India and how those funds will be allocated. These issues are discussed briefly below. There are a number of institutional issues that are also very important for further health sector development. Those are treated in the text of the report.

11 vii The Roles of the Public and Private Sectors in the Provision and Financing of Health Need to be Revised. 6. There is a difference in emphasis in the type of health services provided by the public and private sectors. Government provided services are by far the dominant source of preventive health care, such as immunization, ante-natal care, and infectious disease control services in both rural and urban areas. Private providers are dominant when it comes to the provision of ambulatory care for acute illnesses, or illnesses not requiring hospitalization, even among the poor. This is despite substantial public investment during the past ten years in establishing an enhanced publicly financed health care system. 7. The demand pattern for inpatient treatment in hospitals is, however, sharply at variance with that of ambulatory care. Government provision to meet this demand reflects a high degree of equity. In contrast with outpatient or ambulatory care, where equity is much more of an issue, treatment of inpatients reflects the fact that the lowest expenditure quintile is more likely to receive subsidized care than the higher expenditure groups. In almost all states, for both rural and urban areas, hospitalized episodes reported by the lower expenditure classes were more likely to be treated in Government facilities. Most of those using Government hospitals report not paying any bed charges, although it is more common to pay for specific additional services, such as diagnostic tests and surgery. For these services, private hospitals are far more costly on average than public facilities. On average, about 60% of all hospitalized cases are admitted to public facilities. 8. The above findings confirm earlier evidence on this subject in a rigorous manner and suggest a number of important implications for further development of the health sector in India: * Public involvement in the provision of health must build upon the fact that the private sector is currently the largest provider of health services overall and is likely to remain so in the near future. In particular, the Government needs to re-examine the role of the public and private sectors with respect to provision versus financing of health services. Greater opportunities for public financing of health services provided by the nongovernmnent voluntary sector need to be examined as well. * The extensive involvement of the private sector provides an opportunity to focus public expenditure on primary health care, especially preventive and promotive services, which are the most cost-effective and best serve the needs of the poor. * Public provision for inpatient care at both rural and urban hospitals in general reflects a high degree of equity and lower costs compared to services provided at private facilities. This implies that public hospitals providing first referral services at the secondary level are a key input for the Government's package of basic health services. * Since the private sector provides a wide range of health services, it is imperative that Government create an environment which encourages the private sector to provide costeffective services of acceptable quality.

12 viii Public Spending on Health Needs to be both Increased and More Efficiently Allocated to Provide Satisfactory Outcomes. 9. Overall health spending is sizable. but the contribution of public spending on health is inadequate. Total health spending in India in 1991 accounted for about 6 percent of GDP, which is about US$13 per capita or Rs. 320 ( prices). As a percentage of GDP, this is a higher level of spending on health care than in other Asian countries such as China, Indonesia, Thailand, the Philippines, Pakistan, Bangladesh and Sri Lanka (World Development Report, Table A9), most of which have better health outcomes than India. Public health spending, however, at about 1.3% of GDP is in the middle range of low income countries -- less than in China, Sri Lanka, Bangladesh and Pakistan but higher than in Indonesia and the Philippines. Moreover, the portion of GDP and Government expenditure devoted to disease control programs is very low compared to many other countries at India's level of per capita income. 10. Of the total health spending in India, 75% or Rs. 240 per capita in 1991 prices is out-of-pocket spending of private households and about 3 % is contributed by corporate or third party insurance (Table 1). The Government's overall contribution, including center, state and municipalities, accounts for 22% or Rs. 70 (US$2.5) per capita ( prices). State and local Governments contribute about 16% (Rs. 50 per capita), most of which is the contribution of the state Governments (15%). The central Government contributes about 6% (Rs. 20 per capita). Table 1: National Health Spending: An Estimated "Sources and Uses" Matrix (in percent of total expenditure)* Sources State & Central Local Corporate/ Uses Government Government 3rd Party Households Total Primary Care (7.3) (9.5) (1.3) (81.7) (100) Curative (.7) (6.0) (1.6) (91.7) (100) Preventive and Promotive Health (43.9) (29.4) (26.7) (100) Secondary/Tertiary Inpatient Care (2.3) (21.7) (6.4) (69.6) (100) Non-service provision n.a n.a 2.5 Total * Note: Row percentages are in parenthesis

13 ix 11. Three important observations related to the above follow: (i) public spending on health is inequitable across states, and compensatory central financial measures need to be taken to improve the health status of people in those states; (ii) a disproportionate burden of household out-of-pocket spending falls on the poor; and (iii) the paucity of funds for operations and maintenance has an adverse effect on the quality and effectiveness of health services. 12. Public Spending on Health is Inequitable Across States and Fiscal Constraints Will Add to the Equity Problem if Compensatory Measures are not taken. As noted above, state and local Governments account for about 72 % and the central Government about 25 % of total public expenditures on health. Center and state roles have shifted marginally during the 1980s, with the center's contribution increasing moderately. There is, however, significant variation in health spending and its composition across states. Public spending on health overall is significantly lower in the poorer states, where the population has a lower health status. The ratio of per capita spending on public health between the highest and lowest state was 7:1. States where the health status has improved relative to others have been those with higher levels of per capita income, suggesting that overall state Government resource constraints persist as a major impediment to improving health spending in the poorest states. 13. Central allocation of health resources in the past has not been in proportion to the needs of individual states as indicated by socio-economic and health indicators. Some of the centrally-funded communicable disease programs, including the National Malaria Eradication Program, which is the largest one, are funded on a matching basis by state and central budgets. However, the poorer states are unable to come up with sufficient matching funds to mnake optimum use of these programs. These states are least able to mobilize state resources but are most in need of supplementary central allocations. 14. Moreover, there is a distinct possibility that significant cutbacks and imbalances at the state level due to stabilization policies are likely to be introduced into a system that in fact needs to be expanded and strengthened. Stabilization can affect Government health spending through: reductions in central plan scheme allocations; reductions in central untied transfers to states; reduced Government revenue at state level; and autonomous state reductions in actual health spending in response to their general fiscal constraints. Sizeable reductions in health spending have not yet appeared, although where reductions have occurred they have affected poorer states and disease control programs the most. 15. The Burden of Household Out-of-Pocket Expenditures for Health Care Fall Disproportionatelv on the Poor. In addition to its large share in overall health spending, private spending is dominant in the provision of primary care services, accounting for 82% of the total (Table 1, column 4). If primary care services are disaggregated by curative and preventive care, the share of private spending for curative care is about 92%, while only 27% is accounted for by preventive and promotive services. Private household expenditure is also dominant for secondary/tertiary inpatient care at 70%, although it is somewhat less so than for primary care services. Extrapolating the share of direct household spending from its contribution to national health spending, it is estimated that household ambulatory curative care spending accounts for about 50% of national health expenditures.

14 x 16. Surveys undertaken in five states show that the burden of household health expenditure falls disproportionately on the poor and on rural populations. 2 Out-of-pocket expenses for serious illnesses are large and disproportionately affect the poor. Household health spending in rural areas averaged 5% of total consumption expenditure, while in urban areas it averaged only 2.3%. This inequity in the financial burden of health care on the poor and those living in rural areas is particularly related to household out-of-pocket spending for ambulatory care. 17. Expenditure per episode for private providers averaged times the cost to patients of consultations at Government facilities. For in-patient care, privately hospitalized episodes are much more costly than those in public facilities on average, with state averages ranging from 1.3 to 9 times higher. 18. Public Spending on the Operation and Maintenance of Health Programs is Inadequate and Constrains the Ouality and Effectiveness of Health Services. Inefficiencies related to operations and maintenance of health programs occur due to two reasons: (i) inappropriate allocation of health spending; and (ii) low level of funding of non-salary recurrent costs. Overall, about 45 % of the total health sector budgets of the center and states is spent on curative care and health facility operations (Chapter 3, Fig. 3.2). This reflects well on the prioritization of public investment on health since this figure is often well over 60% in a country at India's level of per capita income. Preventive and promotive services come next with 30%, of which about equal shares go to prevention and control of communicable diseases, and family planning and immunization. In addition, 7% of the budget is allocated for insurance for central Governnent employees (ESIS) and organized industrial workers (CGHS); 9% for research, education and training of doctors and paramedical staff; and 9% for capital investment and administration expenses. The category labeled "medical head" by the Government, which comprises hospital based services, some inputs which support primary care, and medical education and research accounted for about 60% of health expenditure in FY93. Since the mid-1980s through the FY93 budget estimates, "the medical head" category and FW grew, while expenditures on disease control programs declined in real per capita terms. Within the "medical head" category, expenditure on medical education and research increased from 10% in to 14% in It is apparent from this allocation pattern that there is considerable scope for reallocation of resources by major categories. For example, spending on medical education could be reduced in order to provide additional funds for preventive and promotive services. 19. The second inefficiency related to operations and maintenance of health programs results from inadequate resources for non-salary recurrent costs. Because of the massive infrastructure that is already in place, a sizeable operations and maintenance budget is needed to provide an adequate level of services. This is, however, not available since salaries continue to consume an increasing share of resources. Salary shares, based on data from 12 sample districts and 8 municipalities, indicate that they are in the range of 70-80%.3 Salaries have grown faster than all other components of health spending, averaging 10% growth annually in real terms, 2 Visaria, P. and A. Gumber, "Utilization of and Expenditures on Health Care in India, " Gujarat Institute of Development Research, Vashishtha, P. et al, 'Survey of Primary Health Care", National Council of Applied Economic Research; 1994.

15 xi compared to 5% or less for other inputs. In other words, the proportion of salaries in total Government health spending has increased at the expense of non-salary maintenance expenditure. As a result, funds for operations and maintenance purposes have fallen from nearly 30% to less than 20% of the budget since the mid-70s. The fact that less than 20% is allocated for operations and maintenance, including drugs, is a cause for serious concern. Lack of availability of medicines and other supplies at publicly-managed facilities, especially in rural areas, is a major cause for the low quality of care provided at primary health facilities. This is an important reason for the lack of demand for some services. RECOMMENDATIONS 20. In response to the systemic problems discussed above, the report makes the following specific recommendations. Coordinate and Integrate the Roles of Public and Private Sectors in the Provision of Primary Health Care Services. 21. A major recommendation of this sector report is that the roles of the public and private sectors in the provision and financing of primary health care services need to be rethought. A strategy needs to be developed by the Government that takes into consideration the existing levels of private provision of services. The Government's health care strategy should be to: (i) re-assess the role of the public and private sectors with respect to provision versus financing of health services in areas of preventive and curative care; (ii) encourage the private sector to continue to play the important role that it is already playing in the delivery of health care services; (iii) create an environment which encourages the private sector to provide costeffective services of acceptable quality; (iv) monitor and regulate private care provision, such as licensing and certification; and (v) increase public expenditures on preventive and promotive care services. 22. Following this strategy, the Government would need to target its own resources to where critical gaps exist so as to increase the return from the substantial private spending on health services. Those areas of primary health care services which are already being provided by the private sector, but for which there are excess demand and positive externalities, could be supported by selective increments in public provision. Such areas include treatment of acute illness and routine curative care for priority diseases of children and adults. The focus of public spending, however, would be in areas where private investment is negligible such as preventive health care, infectious disease control and limited clinical interventions providing inpatient treatment. The public sector should provide enhanced support to a basic package of public health measures and clinical services that will reduce the burden of disease in a cost-effective manner. This will mean scaling back on some current public investments, such as public support of medical education and tertiary care, and allowing the private sector to play a greater role in those areas. 23. The Government would also need to reduce the burden of out-of-pocket spending on the poor. While maintaining the emphasis on cost-effectiveness, it could be possible for the Government to improve equity aspects of public sector health spending. For example, if public sector health spending, which has been estimated at about Rs. 70 per capita in 1991 prices, were targeted only to the bottom forty percent of the population in terms of income, it would mean a

16 xii considerably larger per capita health spending (about Rs. 115) on that segment of the population. Since targeting poor people on the basis of household income is difficult administratively, the Government could target districts which are characterized by low income and poor social indicators. 24. Primary health care, especially preventive and promotive services, is an area in which the private and NGO sectors could become more involved through the following actions: (i) encouraging the central and state Governments to develop incentives and schemes to finance, train, and integrate private providers in case-finding, diagnostics, referral treatment and monitoring for priority problems such as TB, STDs, ARI, diarrhea, malaria, leprosy and high risk pregnancies; (ii) increasing public support for voluntary agencies in health in such areas as social marketing of essential drugs and contraceptives, and behavior-changing health education activities; and (iii) contracting-out services to the private sector where possible, especially support services, in order to cut costs and increase efficiency. Re-evaluate Financing Priorities for Public Expenditures on Health. 25. Increase Government Expenditure on Primary Health Care. In the light of fiscal realities faced by central and state Governments, it would be very difficult to obtain sizeable increases in health expenditures overall. The Goverrunent must, however, redirect or increase spending on primary health care, especially preventive and promotive services. The amount of money currently allocated is inadequate to meet the primary health care needs as defined by Government of India norms and the WDR (1993) recommendation of an essential clinical and public health package. 26. For the Government to achieve its objectives of providing minimal essential services at primary health care facilities that have been established, it would require that expenditures on primary health care services be increased from its current level of 0.65% to at least 1.0% of GDP. There is a strong argument for providing at least this level of additional funds for primary health care services to make optimal use of physical infrastructure that has already been built. In addition, it is critical for the Government to provide a basic package of clinical interventions to reduce the burden of disease in a cost-effective manner. Alternatively, providing the WDR package of essential public health and clinical services, which is estimated to cost US$12 per capita in low income countries, would cost several times more than the approximately US$2.5 per capita that is currently being spent by the Government. Even if these services are provided somewhat less expensively in India, it would still require substantial amount of additional funids. 27. The additional resources for a basic package of primary health care services and limited clinical interventions could be achieved by: (i) redirecting funds from tertiary hospitals to primary and secondary health care services, particularly preventive and promotive aspects; (ii) substituting public funds with private funds in secondary and tertiary hospitals by instituting means-tested user-charges; (iii) implementing full cost recovery from private and Governmentsubsidized insurance schemes as well as by enhancing non-tax schemes; (iv) reducing public subsidies for medical education; and (v) increasing the overall central and state health budgets taking into consideration the Government's stated public health priorities and fiscal constraints. At the least, expenditures on preventive and promotive services should be protected from fiscal cuts engendered by the stabilization program.

17 xiii 28. Increase Supplementary Central Funding to Needy States. To accomplish its health aims and to enhance inter-state equity, there is a need to develop mechanisms to provide increased supplementary central funding to the poorest states in cases where alternative sources of revenue are limited. Thus, supplementary financing should be provided to those states most in need, but only when these states are already taking credible steps to improve their finances. 29. Strengthen Financing Strategies at the State Level. There is a critical need to focus on state level financing issues, including overall resource adequacy as well as input mix. In addition to augmenting state health budgets, a strengthened finance strategy applying new Government approaches to financing is also needed. It is best to envisage a period of substantial experimentation during which new approaches to financing can be adapted to Indian conditions. In particular, these would include: user charges in urban tertiary and referral hospitals, financing of non-government providers, and allowing private insurance and ESIS reimbursement at full cost for publicly-provided services. 30. Mobilize Resources through Innovative Financing. Innovative financing at the hospital level should address the fact that administrative responsibility and financial accountability are artificially separated between the responsible Government agency and hospitals. The lack of appropriate management arrangements and financial authority to act means that there are few incentives for hospitals and their staff to improve hospital operations and quality of services. State Governments should undertake necessary administrative, regulatory and legal actions to ensure that greater devolution of autonomy to hospitals be allowed so that they are able to retain income generated by them. These measures, which allow each hospital to retain most of the income it generates through user charges, would go a long way to ensure that more funds are available for drugs and medicine and for operations and maintenance purposes. Such measures would strengthen service delivery management by improving the implementation capacity of these hospitals. User-charges could be implemented with the principle that they would: target the receipts, particularly on non-salary recurrent costs; charge for amenities such as private beds at hospitals; and charge for procedures that are low in cost-effectiveness in order to pay for those interventions which are high in cost-effectiveness. 31. Strengthen GOI/MOHFW Planning and Analysis Capacity. GOI/MOHFW's planning and analysis capacity could be strengthened through the following actions: (i) implementing systematic analysis of Government health priorities based on disease burden, cost-effectiveness of health interventions, and current public and private sector health care coverage (the ongoing study in Andhra Pradesh is a useful contribution to this); (ii) establishing a substantial operations research program to determine essential input norms for health facilities and programs and linking this to planning and budgeting. An alternative scenario, if such work cannot be supported by the health policy and finance unit in MOHFW, could be to establish a formal linkage with an external unit, such as the National Institute of Public Finance and Policy; (iii) providing a cohort of Indian experts and Government officers with advanced training in health financing, including short- and long-term training, both local and international; (iv) expanding the role of the policy/finance unit to include design, review, and evaluation of financing innovations experiments; (v) accelerating current efforts to develop a standard health information system by enhancing capacity in MOHFW through management information related to inputs and outputs by facility and program; and (vi) drafting an appropriate strategy for urban primary health care which includes financing projections.

18 xiv 32. Other specific recommendations of the report are presented in the policy matrix on Maeor Findings and Actions Recommended. NEED FOR FURTHER ANALYSIS 33. This sector study has covered a number of major themes on health sector financing issues. There still remain several health sector issues in India where further work would be beneficial for policy analysis. A few suggestions follow: * Issues related to performance and quality of health care at the state level. Since threequarters of health expenditures are incurred at the state level, a study of key issues related to strengthening performance and quality of health care at the state level is a logical extension of this report. The outcome of such a study would be to suggest actionoriented recommerldations for implementing a coherent strategy and a policy reform package for the development of the health system at the state level, adding to state level issues discussed in this report. * Financing and provision of primary care for poor populations in urban areas. Urban health needs are likely to increase substantially in coming decades. Poor urban populations face a very different environment from that prevalent in rural areas. A study that would include the role of public financing and provision in urban environments could be very useful in the light of an active private sector. Differences by size of cities, levels of current public and private provision facing consumers, and the role of regulation and quality control could also be analyzed. - Efficiency in public hospitals. Public hospitals remain a major part of the public expenditure programn in health and a very important factor constraining expanding finance for primary care. An assessment of the current level of efficiency of public hospitals, the potential gain to the Government from improving efficiency and the cost implications of quality improvements could be analyzed. * The potential of new approaches to financing public hospitals. The potential for user charges and private insurance to finance public hospitals while maintaining protection for the poor could be studied. In addition, the lessons learned from Government grants to NGOs and their potential for expanded coverage of key programs to the poor could also be analyzed. * Primary care and the unqualified practitioner. The study could analyze the nature of practice of undocumented private primary care providers, the role of these providers in expanding primary care coverage, and the terms of regulation, quality control, and pricing structure. * A follow up could be undertaken to the sector report on Issues in Women's Health in India, with a focus especially on the disease burden on women.

19 INDIA: POLICY AND FINANCE STRATEGIES FOR STRENGTHENING PRIMARY HEALTH CARE SERVICES Maior Findinps and Actions Recommended OBJECTIVES ISSUES ACTION Strengthen the capacity of state health systems About three-quarters of public spending on Develop a coherent strategy that will: to deal with the evolving burden of disease. health is accounted for by the state budgets and review the financing and provision of health the states are primarily responsible for care in the public and private sectors; address implementing various health programs. both the preventive and curative aspects of However, a coherent strategy with regard to the health care in a cost-effective and efficient development of a health system at the state manner; and analyze coverage of health care level is lacking. Budgetary issues, including issues more broadly than has been covered in overall allocation to the health sector and this report. optimal allocation between different tiers of the health system, remain unaddressed; health Undertake further sector work on key planning capacity to assess sectoral needs based operational issues related to the performance on epidemiological monitoring is weak; and quality of health care at the state level. institutional capacity, including surveillance development, continues to be neglected; persistent technical and qualitative inefficiencies adversely affect the performance of the health care system; and the role of private health sector in delivering quality health care remains underdeveloped. x Increase government expenditure on primary India is in the middle range of low income. Increase spending by state and central care through reallocation of existing resources Asian countries in terms of the percentage of governments on primary health care from the and allocation of additional funds. GDP spent by government on health (1.3%). current level of 0.65% to about 1.0% of GDP. The share of government health spending on This could be achieved by: primary health care is 43%. Nonetheless, given India's substantial health needs and the link (i) redirecting incremental resources almost between health care and poverty alleviation, a entirely to primary and secondary health care, significant increase in govemment primary health care expenditure is needed, especially on preventive and promotive health services. particularly preventive and promotive aspects; (ii) substituting public funds with private funds in secondary and tertiary hospitals by instituting means-tested user-charges; (iii) implementing full cost recovery from private and govemment-subsidized insurance schemes as well as enhancing non-tax revenues; (iv) reducing public stibsidies for medical

20 OBJECTIVES ISSUES ACTION education; and (v) increasing the overall central health budget against the background of the present fiscal constraints. At the least, protect expenditures on preventive and promotive services from fiscal cuts engendered by the stabilization program. Improve the structure of public financing for The existing fiscal and administrative structures Initiate, jointly through the Ministries of primary health care. for primary health care are complex and impede Finance, and Health and FW, a substantial effective financing and accountability for local review of the fiscal structure and procedures in area management, programs, and health the health and FW sectors. facilities. The structure of two departments (Health and Family Welfare), plan/non-plan Review and reform the role of central, expenditures, center-state financial transfers and state, and local govemment financing in the jointly financed schemes is ineffective in: public health sector in terms of national health (i) assuring essential inputs for health facilities policy objectives and provision of basic inputs. and vertical schemes; (ii) correcting inequities in health expenditure between states and Develop program budgeting tools at the increasing spending in states with the worst central and state levels to monitor and assess health indicators; (iii) providing flexibility and expenditure for important schemes. accountability to local officials; and (iv) supporting essential monitoring of program Develop flexible decentralized financing inputs and outputs. tools at the state and district levels to allow local administrators to respond to local needs. Develop fiscal tools to enable greater experiments with resource reallocation, cost recovery, and financial incentives to NGOs and other private providers. Review, at the state level, state governments' fiscal structures and procedures and implement recommendations. At the state level, ensure adequate functioning Much of the planned system of rural primary. Prioritize and ensure that state govemments and improve efficiency of public primary health health care facilities has been created, with the maintain sufficient funds in their non-plan care facilities. lowest, most peripheral facilities being most health budgets to provide adequate and timely complete. In contrast, essential staff and supply supply of essential inputs to existing facilities. inputs are below stated govemment norms or The following package is recommended for adoption:

21 OBJECTIVES ISSUES ACTION adequate levels, and are most lacking in the more peripheral facilities. Financing the provision of staff, drugs and other inputs to norms would require about 26% increase above current public sector health expenditures. (i) Review staff norns and ensure that necessary nursing care can be provided; (ii) Allocate at least Rs. 50,000/- per annum at current prices for drug purchase to each PHC; (iii) Rationalize personnel policies to ensure adequate staffing of posts at rural PHCs; (iv) Ensure that doctors provide two years of rural service as a pre-condition for eligibility for admission to post-graduate medical courses; (v) Provide for staff quarters where critically needed, and electricity and water; (vi) Undertake studies to decide an optimal annual maintenance budget for each PHC; (vii) Ensure better communication arrangements; (viii) Provide regular training for medical/paramedical staff in health management/health economics; and (ix) Strengthen sub-centres by providing an additional worker for looking afker general health care. Following the adoption of the above package, consider and carefully evaluate whether financing the expansion of physical infrastructure to meet established norms is critical in the light of limited resources.

22 OBJECTIVES ISSUES ACTION Increase supplementary funding to needy states. Central allocation of health resources has not. Develop mechanisms to provide increased been in proportion to the needs of individual supplementary central funding to the poorest states as indicated by socio-economic and states in cases where alternative sources of health indicators. Some national disease control revenue are limited. Supplementary financing programs are funded on a matching should be provided to those states most in need, basis, but the poorer states are unable to come but only when these states are already taking up with sufficient matching funds to make credible steps to improve their finances. optimum use of these programs. Moreover, because of stabilization policies there is a distinct possibility that significant cutbacks and imbalances are likely to be introduced into a system at the state level that in fact needs-to be expanded and strengthened. Strengthen private primary health care by Private fee-for-service providers are easily. Strengthen Government capacity to register, improving the quality of services provided by accessible to the rural and urban population and certify, regulate and monitor private health care the private sector, particularly private are heavily used by the poor for ambulatory provision, especially qualifications of doctors practitioners and drug suppliers. care. However, many of these providers are and other medical personnel and the quality of unqualified, and they may cause substantial their services. Central and state Govemments harm as well as good. should enact legislation to register nursing homes, private clinics/hospitals and ensure minimum standards of care by providing appropriate guidelines. x Increase Government capacity to control pharmaceutical supply and dispensing, as well as to provide training and information on drug use to primary care providers and general populations. Increase public support for voluntary agencies in health in such areas as social marketing of essential drugs and contraceptives, and behavior changing health education activities. Contract out services to the private sector where possible, especially support services to cut cost and increasefficiency.

23 OBJECTIVES ISSUES ACTION Encourage the private sector to contribute Government is the major provider of preventive Develop incentives and schemes to finance, more to preventive and promotive health and promotive health care services, but its train, and integrate private providers in casecare services. coverage is very low (20%); the private sector finding, diagnostics and treatment for priority contributes mainly to ambulatory care illnesses problems such as TB, STDs, ARI, diarrhea, but hardly anything for preventive and malaria, leprosy and high risk pregnancies; promotive care. Therefore, neither the public nor the private sector is contributing enough to Publish, through MOHFW, a quarterly the coverage or improvement in quality of newsletter to disseminate information about preventive and promotive care. Given that such schemes in the various states, and other investments in preventive and promotive care innovative activities involving private sector services yield high social returns and benefit participation. from externalities, a strategy to enhance the private sector's contributions to national health goals is needed. Set up Referral Committees at the District Improve sectoral effectiveness and efriciency by Public hospitals are a major source of level to coordinate/manage referral between strengthening the performance of the referral treatment, reach the poor, and appear to be primary care and secondary level diagnosis, system at the state level. progressive in their financial subsidy. However, treatment and care. Referral Committees would: most admissions to public hospitals do not receive prior primary care. The referral system (i) issue administrative directives that x does not function well. Each tier operates as an would specify procedures to be followed in x independent entity providing similar levels of order to make the referral system effective and care and resource utilization is poor. acceptable to the community; (ii) develop referral protocols that specify the types of conditions that should be referred either for investigation or treatment at higher levels; (iii) develop clinical management protocols to provide guidelines and standards for the management of common conditions by doctors who do not have post-graduate qualifications in that specialty; and (iv) provide incentives for priority treatment to those patients using the referral system, such as reduction in user fees and shorter waiting time for diagnosis and treatment of such patients.

24 OBJECTIVES ISSUES ACTION Mobilize resources by financing innovations The proposed reforms in health policy and Develop innovative national health financing through user charges and private insurance finance would require a variety of new areas of schemes to provide grants to state and local where possible. action in the health sector. Central and state governments, NGOs, and private sector governments have limited experience in these organizations to develop, test, and evaluate new areas. It is besto envisage a period of approaches to financing. substantial experimentation during which new approaches can be adapted to Indian conditions.. Implement user-charges with the principle In particular, these include: user charges in that they would: target the receipts, particularly urban tertiary and referral hospitals and on non-salary recurrent costs; charge for allowing private insurance and ESIS amenitiesuch as private beds at hospitals and reimbursement at full cost for publicly-provided charge for procedures that are low in costservices. Moreover, at the hospitalevel, effectiveness in order to pay for those administrative responsibility and financial interventions which are high in costaccountability are artificially separated between effectiveness. the responsible government agency and hospitals. Lack of appropriate management. Sort out legal/other issues that are arrangements and financial authority to act prohibiting greater devolution of autonomy to means that there are few incentives for hospitals for retaining income generated by hospitals and their staff to improve hospital them. Allow each hospital to retain most of the operations and quality of services. income it generates through user charges. This will strengthen service delivery management by improving the implementation capacity of these hospitals. x Initiate studies at national and state levels and analyze the resource implications of various kinds of innovative financing mechanisms such as user fees, insurance schemes, and additional state and local taxes targeted to improve health care provision.

25 OBJECTIVES ISSUES ACTION Improve understanding of the health needs of women beyond those covered by the family welfare program. The national average sex ratio of 927 females to 1,000 males is a matter of concem. The informnation base to understand the underlying cause of this imbalance is lacking and more analysis needs to be done. Although some progress is noted in recent years, the central and state govemments need to move more expeditiously towards a reproductive health approach that will extend beyond those aspects of women's health covered by the family welfare program. Undertake in-depth analysis of the special health needs of women at the primary and secondary health care level beyond those covered by the family welfare program. Such a study should focus on women's burden of disease, health-seeking behaviour and the resource implications of shifting to the reproductive health approach. Strengthen GOI/MOHFW and state level planning and analysis capacity. Although there has been significant progress in. Implement systematic analysis of recent years in the availability and use of govermtient health priorities based on disease information on health financing in India, and burden, cost-effectiveness of health despite efforts to create a health policy/ interventions, and current public and private financing unit in the MOHFW, the capacity for sector health care coverage. The ongoing study health planning and policy analysis at state and in Andhra Pradesh is a useful contribution to central levels remains limited. this. x X Establish a substantial operations research program to determin essential inputs norms for health facilities and programs. This should be linked to planning and budgeting. As an altemative scenario if such work cannot be supported by the health policy and finance unit in MOHFW, a formal linkage with an extemal unit, such as the National Institute of Public Finance and Policy, could be established. Provide a cohort of Indian experts and govemment officers with advanced training in health financing. This includes both short- and long-term training, both local and intemational. * Expand the role of the policy/finance unit to include design, review, and evaluation of innovative financing experiments. * Accelerate current efforts to develop a standard health information system by

26 OBJECTIVES ISSUES ACTION enhancing MIS capacity in MOHFW through management information on inputs and outputs by facility and program. Develop health care MIS at the state level. Draft an appropriate strategy for urban primary health care which includes financing projections. An integral component of strengthening policy and finance strategies for primary health care Strengthen training; increase public awareness services is to improve and introduce some. Strengthen training institutions both in of health-related issues; and initiate community complementary activities that are crucial to the public and private health sectors through participation in existing health education development of the health sector. These training of trainers and regular short and long programs. include the need to improve human resource term courses in: health management and development skills of health practitioners, raise administration; analysis of cost effectiveness of the awareness of health workers and the public various health services; hospital audits; and on health-related issues and encourage greater drug management. community participation in programs on school health and nutrition. Initiate health education activities to x improve awareness of health-related issues - among the public, politicians and health workers. Initiate community participation in school health and nutrition education programmes.

27 1. INTRODUCTION A. Background 1.1 Health is a critical investment for human resources development and poverty alleviation in India. Public policy for health has been based on an implicit assumption of health care as a basic right to which people should not be denied access due to inability to pay or other socio-economic reasons. Yet, the resources provided by the Govermment to achieve better health status through the provision of high priority primary health care services for the vast mnajority of Indians has been inadequate. Goals have been achieved only to a limited extent, despite the fact that India spends a higher percentage of its GDP on health care in comparison to other Asian countries that have achieved greater improvements in the health status of its peoples. This is because the current level of public health expenditure in India, especially spending on preventive and promotive primary health care services, is inadequate and has not kept up with the growing demands being made upon the system. If the Government is to achieve its stated objectives, which are reasonable and achievable, reorienting priorities within the health sector is critical. The challenge is two-fold in that a strategy needs to be developed that would encourage the private sector to contribute more resources to preventive and promotive health care services and the public sector to invest more resources in these activities as well as regulate and monitor private care provision. 1.2 The purpose of this report is to provide information that can assist India in improving the effectiveness and efficiency of its public health programs. It does so by focusing largely on questions related to the financing of health services. The report synthesizes collaborative sector work between the World Bank and Government of India (GOI) and proposes a set of action-oriented recommendations to improve sectoral policies on key health issues relating to primary health care services. It includes: a detailed analysis of private health care that has to date not been undertaken; an assessment and quantification of the resource gaps for delivering public health services; and initial evidence of the impact of government spending on health status as well as the effects of structural adjustment policies on health spending in the public sector. B. Terms of Reference for the Study 1.3 An Initiating Memorandum (IM) was issued on June 14, 1991, and included the following agenda: (a) estimate, for selected states the trends in total public health expenditure from state and central sources and, for a sample of districts, expenditures for district, municipal and local government bodies; (b) calculate, in real and financial terms, the gap between current input levels for primary care and planned levels according to government norms. This would include both investment and recurrent cost items; (c) estimate the funds required to complete establishment of services to meet the government's current targets, and the possible impact on primary health care performance of changes in finance levels or the composition of public sector inputs; (d) estimate for private primary health care services reported patterns of need and use of private health services and expenditure for key targeted population groups, using the 42nd round of the National Sample Survey (NSS); and (e) discuss the implications of the above analysis for policies to enhance public financing and policy for primary health care, the contribution of the private sector to primary care, improved regulation of private providers, and the implications for cost recovery in the public sector.

28 2 1.4 A Yellow Cover Report entitled "India: Health Sector Financing: Coping with Adjustment, Opportunities for Reformn" (June 30, 1992), followed that Initiating Memorandum. Following the earlier report, the Bank and GOI agreed to commission three additional studies on key issues related to primary health care services'. This report synthesizes the major findings of the sector work on health undertaken to date. C. Structure of the Report 1.5 This report covers a number of issues relating to India's health sector. Its primary focus, however, is on Policy and Financing Strategies for Primary Health Care Services in India. 1.6 Chapter 2 provides an overview of the health sector in India. The first part of the chapter describes the health status situation, as summarized in the recent burden of disease estimates developed for the World Development Report (WDR), The rest of the chapter describes the objectives of India's National Health Policy, the administrative and financial structure of the public sector, the budgeting process and center-state finance issues. 1.7 Chapter 3 summarizes the analysis of public health expenditure levels, composition and trends with a focus on primary health care services. The first part of the chapter provides an overview of both public and private contributions to financing India's health system. The rest of the chapter highlights public financing of health services, impact of government spending on health status and the effect of structural adjustment policies on government health spending. 1.8 Chapter 4 synthesizes the provision of public primary health care services. The emphasis of the chapter is on the structure of the public primary health care system, provision of inputs and performance, and financing requirements. 1.9 Chapter 5 analyzes the pattern of use of private health care services and the important role played by private health care provision on ambulatory curative services. Extensive new analysis of the NSS, 42nd Round data is undertaken in this chapter Chapter 6 focuses on consumer demand for both public and private services, including both service use and household health expenditures. Linkages between public and private service provision and between the primary level of services and hospitalization are also analyzed Chapter 7 outlines the implications for central and state Governments for policy reform and action, and for Bank lending. A number of major areas of reform are highlighted: increasing government expenditures on primary health care; improving the structure of public ' The major studies whose results have been incorporated in this report are: (a) "Survey of Primary Health Care" (Vashishtha. P et. al., National Council of Applied Economic Research, 1994); (b) "Utilization of and Expenditures on Health Care in India, " (Visaria, P. and A. Gumber, Gujarat Institute of Development Research, 1994); and (c) "Health Financing Trends in the Public Sector since the mid-1970s and the Impact of Structural Adjustment on the Health Sector" (Tulasidhar, V.B., National Institute of Public Finance and Policy, 1992 and 1993).

29 3 financing for primary care at the center and state levels; strengthening private primary care through quality improvement; encouraging private sector contributions to preventive and promotive health care; improving sectoral efficiency and effectiveness through improvements in the referral system; mobilization of resources through user charges and strengthening management and financial authority at hospitals; improving the health needs of women beyond those covered by the Family Welfare program; strengthening MOHFW/GOI planning and analysis capacity; and strengthening training, health awareness and community participation in health awareness programs. D. Definitions and Reference Points 1.12 Some key definitions and concepts which have been repeated throughout the text are provided below: (a) (b) (c) (d) (e) (f) Primary Health Care Services: It includes all ambulatory illness treatment services; routine personal preventive care such as ante-natal visits, well-baby check-ups, immunization and other personal disease prophylaxis; maternity care on an outpatient basis; and public health disease and vector control measures. The focus is on all central and state Government services and private services provided by all types of health care providers. It also includes non-profit voluntary agency providers but excludes nutrition and feeding programs and water supply projects. Preventive and Promotive Health Care Services: This is a subset of primary health care services and includes only personal ambulatory services (such as ante-natal care) and population based services (such as spraying for malaria). Ambulatorv Illness/Curative Care Services: This is a subset of primary health care services and complementary to preventive and promotive health care services. It includes personal curative services but does not include treatment as an in-patient in a hospital. In-patient Services: Services availed of at secondary or tertiary hospitals as an admitted patient. Secondary Hospitals: Non-teaching hospitals with bed capacity varying between beds. This includes bedded (community), bedded (area) and bedded (district) hospitals. Services offered include surgery, clinical and diagnostic that are more sophisticated than those provided at primary health centers (PHCs) or community health centers (CHCs). Tertiarv Hospitals: Teaching hospitals and those providing super-specialty services with a capacity generally exceeding 700 beds.

30 In addition, the report also uses the two following concepts developed in the WDR (1993). (a) Essential Clinical Package: This includes the following components: (i) pre- natal and delivery care; (ii) family planning services (these two components together constitute a Safe Motherhood Program); (iii) management of the sick child; (iv) treatment of TB; (v) case management of STDs; and (vi) treatment of minor infection and trauma otherwise known as limited care. The delivery of the minimum clinical package is estimated to cost an average of US$8 per capita annually in low-income countries (WDR, 1993; Table 5.3). This was estimated by the WDR to avert 24% of the country's burden of disease in low income countries. (b) Essential Public Health Package: This includes: (i) the expanded program on immunization, including micronutrient supplementation; (ii) school health programs to treat worm infection and micronutrient deficiencies and to provide health education; (iii) programs to increase public knowledge about family planning, health and nutrition; (iv) programs to reduce the consumption of tobacco, alcohol and other drugs; and (v) AIDS prevention program, with a strong STD component. In low-income countries, the WDR estimated that it would avert more than 8 % of the country's burden of disease at a cost of US$4 per capita annually (WDR, 1993; Table 4.7). Combined, these two packages were expected to reduce the burden of disease by approximately 32% in low income countries The combined total cost of an essential clinical package and an essential public health package is estimated by the WDR (1993) to be about US$12 annually in low incomes countries. Although the costs for these packages may vary for India, they provide a useful reference point for the analyses that follow in this report Requirements to meet GOI norms. Finally, the report uses estimates of additional resources needed to adequately meet the government's norms with regard to: (a) the funds required to complete the physical infrastructure -- the "missing" facilities -- in accordance with population per type of facility norms set by the government; and (b) the funds that would be required to fill the input gaps (drugs, essential supplies, staff, infrastructure maintenance, etc.) at "existing" facilities With respect to (a), the total finances that would be required to complete the physical infrastructure as a percentage of current annual total government health and FW spending is estimated to be about 90%. Assuming a four to five year construction period, it would imply a 20% increase in capital expenditures annually. In addition, recurrent costs of drugs, supplies and staff at the missing facilities would require about a 14% increase annually With respect to (b), recurrent costs of drugs, supplies and staff would require a 13 % increase over the current annual health and FW spending. However, this estimate does not include maintenance of capital infrastructure, which is typically 6-7%. Therefore, an increase of about 52% over the current spending level on health and FW is at least required to meet GOI's norms. Spending on primary health care services would thus have to increase from 0.65% to about 1 % of GDP to meet GOI's own standards and norms.

31 5 2. OVERVIEW OF THE HEALTH SECTOR A. Mortality, Morbidity. Fertility, and the Burden of Disease 2.1 Mortality and Morbidity. Considerable progress has been made in improving the health status of the population since India's independence in Life expectancy at birth has increased from 50 years in 1970 to about 61 years in The crude mortality rate has declined throughout the country, largely as a result of the considerable decline in the infant and child mortality rates. The infant mortality rate (IMR), a sensitive indicator of both socio-economic development and access to health services, has been reduced from 146 per thousand births in the 1950s to 110 in the early 1980s, and to 91 at the beginning of the 1990s. Inspite of these broad favorable trends, with 16% of the world's population in 1990, India accounted for 19% of all deaths in that year. The risk of death for children under 5 years of age remains high in India at 12.4%, about 30% higher than the average risk faced by the world's population and higher than in all regions of the world except Sub-Saharan Africa. The risk of death for the adult population (between 15 and 60 years of age) in 1990 is estimated to have been 21 % higher than the global average. 2.2 Moreover, comparing India with other countries in the region that started with a similar resource base several decades ago, shows that India has not fared as well as some of its neighbors. The gains in life expectancy over the past three decades, for example, have been 23 % in India compared with 60% in China and 28% in Indonesia (World Development Report, 1993). 2.3 Communicable diseases and maternal and perinatal causes continue to account for a large number of deaths in India, about 470 per 100,000 population (standardized for age), compared to only 117 in China and 187 in the world as a whole. At the same time, the gains achieved in life expectancy have resulted in proportional increases in mortality from chronic and degenerative diseases of adulthood, such as heart ailments and cancers. These trends are likely to persist, and are currently being augmented by the rapid spread of infection with HIV which has not yet appeared significantly in mortality statistics. 2.4 The annual risk of infection with Mycobacterium tuberculosis (TB) in India is 1.5 %, equal to the average for Sub-Saharan Africa and about 55 % higher than the global average. Forty years of a national program has not resulted in a sizeable reduction in either incidence or prevalence of this disease. An estimated 300,000 or more deaths are attributed to TB annually; estimates suggest that by the year 2000, India may have at least 20 million active TB patients if present rates persist, with one-fifth of those being infectious. 2.5 Diarrheal diseases account for more than half a million infant deaths annually. India has about 2.0 million or one-third of the world's leprosy victims. Annual incidence of malaria, after declining dramatically from 75 million in the 1950s to less than one million by 1970, has risen since then and is estimated at about 2 million in recent years. 2.6 AIDS, a newly emerging problem, has already spread within high-risk segments of the population, and there is increasing concern about wider spread of the disease through sexual transmission and unsafe blood transfusions. Currently it is estimated that 2.0 million people are HIV-positive in India.

32 6 2.7 The national aggregate picture of mortality hides wide disparities between different states as well as between rural and urban areas of each state. The IMR ranges from 122 per thousand live births in Orissa to as low as 22 in Kerala; the rural average for the country as a whole is 98 compared to only 58 for the urban population. There is also considerable variation by caste: the IMR for scheduled castes and tribes is higher than that for the general population, by 22% in rural areas and by 44% in urban areas. 2.8 Sizeable interstate disparities exist in the distribution of communicable diseases, Reported malaria cases are concentrated in the states of Gujarat, Madhya Pradesh and Orissa; leprosy in Andhra Pradesh, Bihar, Tamil Nadu and Orissa; filariasis is most concentrated in Bihar, Kerala and Uttar Pradesh; kala-azar is reported only in Bihar and West Bengal. Tuberculosis, in contrast, is more or less uniformly prevalent in all states. 2.9 India is one of only seven countries in the world in which women have higher mortality (at least up to age 35) and lower life expectancy than men. In general, states with the highest overall mortality levels have the greatest differentials between males and females. The excess mortality of women is greatest in rural populations and especially in the main years of childbearing. Disturbingly, there has been a 5-point decline in the sex ratio over the past decade, from 934 to 929 females per 1,000 males Fertility. In spite of being one of the first countries in the world to accept official intervention in population control as a matter of national policy, the total fertility rate remains relatively high in India. Estimated at 4 children in 1990, it is higher than in China (2.5), Indonesia (3.1) and most other Asian countries. Rural women have 4.4 children on average compared to 3.2 in urban areas. With the sole exception of Kerala, where both urban and rural women have very low (replacement level) fertility rates, all states exhibit the expected rural-urban differential, with only modest declines since On average, women in Kerala and Tamil Nadu have the lowest fertility, due to both relatively high age at marriage and relatively high contraceptive prevalence. Each woman in the poor states of Bihar and Uttar Pradesh has an average of more than 5 children. The most rapid declines in fertility in the past decade have occurred in Tamnil Nadu and Kerala; the slowest in the most disadvantaged states As mortality rates declined in the early decades after independence, the rate of population growth accelerated in India. The latest 1991 Census shows that the rate of growth has remained largely unchanged over the last two decades, at about 2.2% per year, reflecting some success in reducing fertility over that period.

33 7 Table 2.1: India: Burden of Disease in 1990 Males Females Total Rank DALYs % DALYs % DALYs Total DALYs lost, thousands 145, , , Communicable, Maternal and 70, , , Tuberculosis 6, , , II STDs , , HIV 2, , , Diarrheal Disease 13, , , Childhood Cluster 9, , , Meningitis 1, , Hepatitis Malaria Tropical Cluster 1, , Leprosy Trachoma Intestinal Helminths 1, , , Respiratory Infections 15, , , Maternal Causes 0.0 7, , Perinatal Causes 14, , , Noncommunicable 59, ] 57, [117, Malignant Neoplasms 6, , , Diabetes Mellitus , , Nutritional and Endocrine Causes 9, , , Neuropsychiatric 9, , , Sense Organ (mainly eye) 1, , , Cardiovascular 14, , , Respiratory 3, , , Digestive 5, , , Genitourinary 1, , , Musculoskeletal , Congenital Abnormalities 4, , , Oral Health , Injuries 14, , , Unintentional 12, , , [Intentional 2, , , Sources: World Development Report, Tables Bl, B2, B3. Global Comparative Assessments in the Health Sector, WHO Publication, 1994 edited by C. I. L. Murray et al., Annex Table 6.

34 Disease Burden. Assessments of the relative importance of different diseases have traditionally been based on how many deaths they cause, reflecting the availability of mortality data. However, many diseases are not fatal but are responsible for considerable loss of healthy life. Estimates of the full loss of healthy life due to different causes have been presented in the World Development Report (WDR 1993) in terms of Disability-Adjusted Life Years (DALYs) lost, based on a methodology developed by the World Bank and WHO. 2 According to these estimates, India accounted for 292 million DALYs lost in the year 1990, which is over 21 % of the global burden of disease, even higher than its share of overall mortality The estimated national burden of disease for India is summarized in Figure 2.1 and Table 2.1. The largest loss of DALYs is from the group of communicable disease and maternal and perinatal causes. This falls disproportionately on young children, especially those under five. Non-communicable diseases account for nearly as large a burden. Their incidence tends to cluster in the older age groups Table 2.1 presents the total DALY loss estimated for 1990 for the major causes. There is little difference in the total figures for males and females, although the composition of the burden differs somewhat between the sexes. Females have somewhat higher figures for major causes affecting children, such as respiratory infections and diarrheas, as well as the full burden of maternal mortality/morbidity Table 2.1 also ranks the major causes of DALY loss by their relative size. Of the top ten causes of DALY loss, 50% were in the communicable disease category and 50% were in the non-commnunicable disease category, indicating a mix of causes for the total burden of disease in India Burden of disease information alone is insufficient for evaluating public sector priorities -- the feasibility and cost of interventions must also be considered, along with the extent of private provision of services, and constraints of organizing multi-functional health facilities and programs at the primary level. At this time, only crude estimates exist for analyzing the allocative efficiency of India's health sector based on disease burden and cost- effectiveness. It is likely that substantial gains in health status as defined by the DALY index could be achieved by shifting allocative priorities towards more cost-effective interventions, but the specific feasible and effective strategies for achieving this need to be determined. A study to develop current estimates of disease burden and cost-effectiveness for the state of Andhra Pradesh is near completion. 2 The WDR (1993) defines Disability-Adjusted Life Years (DALYs) gained as 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 death or cases of disability occurring in a particular year.

35 9 Figure 2.1: India: Burden of Disease by Age and Cause 'E T 0. Age 0-4 Age 5-14 Age Age Age 65+ * Injuries C Non-communicable * Communicable, Maternal, Perinatal Causes

36 Challenges for the Future. New health challenges are likely to emerge in India over the next few decades, involving an increase in non-communicable diseases arising from the ongoing demographic and epidemiological transition. As fertility declines further, the age structure of the Indian population will shift. The proportion of people above the age of 65 will increase, and as a result, the burden of non-communicable diseases will rise. At the same time, the challenge of communicable diseases of the young and middle-aged will persist This situation will be especially visible among the poor. It is likely that India will go through an "epidemiological polarization" in which one part of the population will successfully complete a demographic and epidemiological transition while another part remains in the pretransition regime dominated by the diseases of poverty. Indeed, this situation is already present in India and accounts for much of the dilemma of its publicly provided health care. The demands of the rural and urban middle and upper classes for accessible, technologically advanced, and free clinical services compete with the still pressing need for coverage with basic disease control interventions. This conflict over public resources is likely to be exacerbated by epidemiological and demographic trends and poses a major future challenge for primary health care policy. B. National Health Policy 2.20 Since the major responsibility for government health expenditures is jointly shared between central and state governments, goals and strategies for the public health sector are established in a consultative process involving these different participants through the Central Council of Health and Family Welfare. This has been the main formal vehicle for agreeing on the structure of the public sector health care delivery system and, more recently, the National Health Policy issued in Improvement in health status has been a stated objective in development policy pronouncements in India. The government has chosen a single approach to achieve this goal: designing and creating a publicly financed and publicly managed system of health services throughout the country, from primary health centers to hospitals, to provide free curative and preventive health services to a large section of the population. Based on the principle that equitable allocation means equal access to health facilities on a per capita basis, nationwide population-based norms were set for the establishment of such facilities. The National Health Policy (1983) expanded this "supply-oriented" approach to policy setting by specifying quantitative targets for health and fertility gains and a timetable to the year 2000 for meeting them. These targets are summarized in Table 2.2. They strongly emphasize the reduction of preventable mortality and morbidity affecting mothers and young children and were closely identified with the primary health care approach. Significantly, the national health policy also recognized the need for government to "cooperate" with the private sector, although actual efforts in that direction have been limited.

37 11 Table 2.2: Goals for Health and Family Welfare Programs to the Year 2000 Goals hndic tor- Re.oned level 1. infant mortality rate Rural 136 (1978) 122 Urban 70 (1978) e0 To!a 125 (1978) b30 perinatal mortality 67 (1976) Crude death rate Around Pro-school child (1.5 yrn.) mortlity 24 ( ) C matemal mortality rate 4-5 (1976) below 2 5. Ufe expectancy at birth (yrs.) Male 2.6 (1976.1) Female 51.6 (197681) Babies with birth weight below 2500 gms. (%) la Crude birth rate Around Eftectiv couple protection (N) 23.6 (March. 1962) o 9. Net Reproduction Rats (NM) 1.48 (1981) Growth rate (annual) 2.24 (197141) Family size 4.4 (1975) Pregnant mothers receiving ante-natal care () Deliveries by trained birth attendants (%) Immunisations status (% coverage) TT (for pregnant women) TT (for school children) 10 years years DPT (children below 3 yrs.) a5 as Polio (infants) CG (infanls) 56 T 60 a DT (now school entrants 56 years) 20 so Typhoid (new school entrants 5- years) 2 70 a5 a5 15. Leprosy. percentage of disea arrested cases so outod those detected 16. TP. porcontage of arrested cases out of those detmeted 17. Mindness - Incidence of () Source: Statement of National Health Policy. GOI. Ministry of Health and Family Welfare, 1982

38 Although the states do not formally propound their own health policies, they retain a certain degree of autonomy in pursuing their own goals and objectives. Some states spend much more than others on health-related programs, included those of primary health care. Several states have initiated major actions to provide feeding and nutrition services on their own account. State governments also may promote prestige projects, such as new medical colleges and tertiary hospitals as additional expenditures. Since the capacity- of states to support the agreed-upon national policies may be affected by such actions, the reality of health policy at the state level must also be considered. C. Administrative, Budgeting and Financing Structure 2.23 Administrative Structure. The complex structure of government financing in the health sector is shown in Fig Under the federal structure of the Indian Union, public provision of social services including health and education are primarily the responsibility of state governments. The Constitution of the Republic of India (1950) includes health as part of the State List, while medical education is in the Concurrent List of public responsibilities. The center, however, exercises its discretion to initiate and fully or partially finance centrally sponsored schemes through the mechanism of specific purpose grants to the states. These include the National Family Welfare Program (FW)(which includes family planning and maternal and child health services), the National Malaria Eradication Program (NMEP), the National Tuberculosis Control Program (NTCP) and other "national" disease control programs.' The state governments have little independent say in the formulation and design of such national schemes, although they can refuse them. The states retain responsibility for implementing such schemes, which appear as schemes in the state plans. The administrative structure and budgeting process limit the ability of state governments to overcome existing differentials in the resource base There are two wings in the administrative structure of the Ministry of Health and Family Welfare: the bureaucratic or administrative wing and the technical wing. They consist of parallel hierarchies of officers. At the center, the secretariat (administrative wing) of the Ministry of Health and Family Welfare (MOHFW) and the Directorate General of Health Services (DGHS -- the technical wing) are staffed by civil servants and by medical doctors respectively. Both wings report to the Health Minister, and the managers of the two wings enjoy the same level of seniority. The Health Secretary is at the same level as the Director General of Health Services and the Joint Secretary is at the same level as the Deputy Director General. 3 National in this context impliesignificant central funding and program authority as well as a delivery structure clearly defined to the peripheralevel.

39 13 Figure 2.2: The Stmcture of Government Health Financng Union Center"e urnbed tans parvr,han M ' ( Stat ~~~~Govwnment Union MrOdHai FaW Oh rcrrnsaare 7 \ ~~Ob'er central Hospitals ~ ~ ~ ~ ~ ~ ~ ~ ~ Ste ~ MOHI ~~~~~Sae4cl wk ; MOHFW / Cenra functions/\ I,~~, Hosptals / Non-Plan ~~~~~~~~Locall Pdrimay Care OMer Fa - ifes H w d Faciliges and Progrm and Prgwra FW

40 14 Box 1: India's Main Public Health Proorams Endemic diseases account for over two-thirds of the total morbidity and mortality in India. As a resul, direct expenditure on public health programs is high, both by the central government and the states. In FY92. the combined budgeted expenditure on all India public health represented 15 % of total health expenditure, second only to hospitals (31.7%). The central government contribution is about 23 % of total all India public health expenditure, and it ranks third (12.4%), after family welfare (53.3%), and education and research (13.3%). The states share of public health expenditure is over 77%, and, in the state healthbudget, it ranks second (16%) after hospitals and other health facilities and dispensaries (61.2%). Health and health-related programs in India includes national programs aimed at controlling endemic diseases as well as health components in the family welfare, nutrition and urban-based programs. The central government funds ten programs aimed at prevention and control of the following endemic diseases: (i) leprosy; (ii) malaria; (iii) tuberculosis; (iv) guinea-worm; (v) blindness; (vi) goiter, (vii) STDs, (viii) aids, (vix) mental hith, and (x) diabetes. Leprosy, blindness, goiter and aids are 100 percent centrally funded while others are funded on a share arrangement with the states. Endemic diseases pose serious challenges to national development efforts. Beyond their toll on individual illness and death, they have insidious effects on the society and impede national and individual development. However, the programs on leprosy, malaria and tuberculosis remain among the top priorities for action in India although there are other problems to deal with, most notably AIDS and certain newly resurgent infectious diseases (see text). At present, India has 2.0 million cases of leprosy with 0.3 million new cases are detected annually. About 25 % of untreated leprosy patients suffer from deformities due to nerve damage. Suffering due to social stigma is signiticant and leprosy patients face social discrimination and the risk of loss of employment and rejection from their families and communities. Malaria is resurging as one of the leading causes of illness in India, About 2 million cases are recorded every year. Its impact is severe on infants, young children and pregnant mothers with fetal death and premature delivery. It is a serious impedimento development and its socio-economic impact affects various sectors including agriculture. The direct costs of incapacitating malaria include lost wages, cost of treatment, and the expense of travelling to seek care. It reduces productivity, augments pressure on health services, increases school absenteeism and results in lost investments in child health. The burden of tuberculosis is also staggering in India. About 1.5 % of the total population is estimated to be suffering from radiologically active tuberculosis, about 1.5 million cases are identified and more than 300,00 deaths occur every year. Because tuberculosis deaths occur in the economically productive segment of the population, further consequences of adult deaths on children and other dependents are significant. The preceding three programs remain inadequate in terms of effectiveness. The major issues fciing leprosy control include inadequacies in coverage, disability and ulcer care, detection of female patients, public awareness and social stigma. Constraints facing malaria control include inadequacies in logistical support for residual insecticide spraying, transport and accessibility in difficult terrain, non-acceptability of te spraying operation by populations at risk, vector susceptibility to insecticides, plasmodium falciparum resistance to chloroquine, organizational deficiencies including weak technical leadership, and vacancies and frequent changes at key posts. Overriding constraints facing the tuberculosis control program include weak management, inadequate utilization of health workers and sectoral resources, reliance on and abuse of radiology for diagnosis which is not cost.ffective, and the use of conventional and lengthy chemotherapy regimens which contribute to low treatment completion and cure rates. The Government of India recognized these problems and has responded by selecting the districts with high levels of endemicity for priority implementation of the revised and enhanced disease control programs, espeially for malaria, leprosy and tuberculosis. Also, for HIV/AIDS control, the government is promulgating new standards for condom quality and has initiated, with IDA assistance, a control program with a balanced multi-pronged strategy focusing on key actions for promoting public awareness, blood safety, surveillance, clinical skills and sexuallytransmitted disease control.

41 The MOHFW secretariat is further divided into two parts, the Department of Health and the Department of Family Welfare, each with its own secretary. The Secretary for Family Welfare is the senior most officer in the Department of Family Welfare and is responsible for family planning as well as maternal and child health. The Health Secretary is the senior-most officer in the Department of Health and is responsible for all other health programs under MOHFW. Both the Departments of Health and Family Welfare have their own hierarchy of additional, joint, deputy and under secretaries who look after various programs and schemes. This same administrative structure is more or less repeated at the level of each state government, with a Health Minister, a Health Secretary and a Director of Health. As mentioned above, there is also a Central Council of Health and Family Welfare (CCHFW), which includes the health ministers and secretaries from all the states. The CCHFW is the primary advisory and policy making body for health care in the country. The central government's Planning Commission also has a health cell that supports this advisory and policy making function, besides preparing the plan-financed schemes for the sector At the level of each district within every state, there is a Chief Medical Officer (CMO) in charge of rural non-hospital facilities, a District Medical Superintendent (DMS) in charge of the district hospital, and a District Collector (DC) who is the overall head of civil services in that region Budgeting in the Public Sector. Plan and Non-Plan Budgets. Budgeting and accounting of government expenditures at the central and state levels in India has been influenced by the planning process, which takes place within the framework of central and state five-year plans. The plan budget refers to all expenditures, both capital and recurrent, incurred on programs and schemes that have been initiated in the current five-year plan. Once the five-year period of any particular plan is over, the recurrent expenditure associated with the continuation of that activity is generally transferred to the non-plan budget, except for the FW program In the case of centrally sponsored programs (other than the FW program), central financing ratios refer only to the plan component of expenditure. For example, the centrally sponsored National Tuberculosis Control Program is implemented as a 50% centrally funded program; this means that central grants finance half of plan expenditures under this program, while state governments have to bear the full amount of non-plan expenditures. In effect, central grants account for much less than 50% of total government spending on tuberculosis control. The average share of central financing of communicable disease control programs is less than 25 %. Thus, central leverage is limited in its power to assure adequate state funding of these non-plan inputs-- for example, rural field staff for disease control programs Plan expenditure in the health sector accounts for about one-third of total government health spending. If the FW program, which is financed almost entirely out of the central plan budget is excluded, then the ratio of plan to total health spending is less than 20%. In other words, more than 80% of government health spending, excluding FW, is made up of committed expenditure on maintaining existing level of services, financed out of the non-plan budget. In fact, the degree of flexibility that central and state governments have over their health budgets is even more limited than this 20% ratio would indicate, since a part of plan spending is also of a committed nature. Between the center and the states, the former enjoys relatively greater degree of flexibility; about 65% of central health spending and 99% of FW spending is in the plan budget, while 86% of state health spending is in the non-plan budget.

42 Public sector health budgets at the state level, which include all non-hospital primary health care as well as state and district hospitals, are thus financed out of three distinct budget sources: (a) the state's non-plan budget that finances the recurrent cost of maintaining the infrastructure and level of services established through previous plans; (b) the state plan budget that finances schemes initiated by the state during the current five-year plan, as well as the state's component of financing centrally sponsored programs; (c) and the central plan grants that finance the central component of vertical programs. Total spending on health and FW at the state level is financed out of these three different budget sources roughly in the ratio of 68:14:18 (1990/91). The corresponding ratio in the case of drinking water supply is estimated as 28:56: The composition of the health budget of state governments by these different sources of funds is significant from the standpoint of protecting public health spending in the context of general fiscal contraction. The degree of financial constraint can be very different on these three different budget sources. Typically, the non-plan budget of each state is constrained by the overall revenue position of that state, supplemented by the statutory central transfers recommended by the Finance Commission; the state plan budget is constrained by the non-plan gap of the state and the untied central assistance to state plans, whose level is determined by the Planning Commission; and finally, the constraints on tied central plan grants are determined by the budget of the concerned central ministry. In the former two cases, inter-state differences in the degree of financial constraint can be considerable, whereas the constraint is uniform in the case of the last component Center-State Finances. The relationship between the center and state govermnents in the health sector occurs at two distinct levels. First, is the overall allocation of resources by the center's Planning and Finance Commissions to States, which constrains or provides opportunities for states' initiatives in new projects. Second, is the intra-sectoral allocations of grants-in-aid and other earmarked funds from center to state. Mechanisms used by the central government to fund health programs at the state level have the potential to reduce disparities in resources among states, and even within states. As currently organized, however, these mechanisms are not designed to overcome inter-state inequities. In the health sector, it is manifest in the following ways: (a) (b) Some central schemes depend on matching funds from the states. A few of the centrally-funded communicable disease programs, including the largest one, the National Malaria Eradication Program, are funded on a matching basis by state and central budgets. Some poorer states are unable to come up with sufficient matching funds to make optimum use of the program. It should be noted that even matching schemes often require more than 50% contribution by states, since overhead and some other recurrent costs borne by the states are excluded from the estimate of total program cost. Poorer states are least able but most in need of supplementary central allocations to these programs; The central government has gone into debt to the states. In recent years, the FW program and a few other centrally-sponsored schemes have fallen behind in their payments to the states. Therefore, the states have effectively been paying for schemes that were supposed to be centrally-funded. The Eighth Plan has allocated Rs. 5 billion to the FW program to pay for past debts to the states. However, in the

43 17 short term, it is the states that can least afford additional, often unanticipated, outlays that suffer the most; and (c) Plan schemes revert to non-plan schemes after five years. States are wary of participating in projects initiated by the central government under plan budgets, since participation implies that the state will bear the responsibility for recurrent costs in subsequent plan periods. For example, extensive construction of primary health centers (PHCs) under one plan period can become a severe liability during the following period, when all operating costs must be found within the non-plan allocation, and the center withdraws assistance. The integration of Indian Systems of Medicine doctors into PHCs, undertaken by the central government in many states in an earlier plan period, must now be supported by the states, which find themselves with additional personnel costs. Again, the better-off states are able to take advantage of plan projects to a much greater degree than are the poorer states, though they may require the effort less.

44 18 3. PUBLIC EXPENDITURES ON HEALTH AND PRIMARY HEALTH CARE SERVICES A. Introduction 3.1 Health spending in India, at 6% of GDP, is amongst the highest levels estimated for developing countries in terms of its percentage of contribution to national income. 4 In per capita terms, it is higher than in China (US$11), Indonesia (US$12) and most African countries but lower than in Thailand (US$73) and Malaysia (US$67). Comparisons of national health expenditures developed for WDR (1993), using international purchasing power parity (PPP) dollars (1$), also confirm that India is a relatively high spender among developing countries. In 1990 international dollars, India's total health expenditure per capita was I$62, compared with the Philippines at I$46, and Indonesia at I$48. However, in this comparison China's health expenditure was I$73 and Pakistan's I$66 per capita (Murray and Govindaraj, 1993). 3.2 The estimated national health expenditure in India is about Rs. 320 per capita (US$13 at the exchange rate). The largest contributors are private households (75%), followed by state governments (15.2%), central government (5.2%) and third party insurance and employer payment (3.3 %). Municipal Governments and external donors together contribute about 1.3%. 3.3 A "Sources and Uses" matrix for national health expenditure highlighting the major contribution to financing health care in India is presented in Table Overall, primary health care services account for 58.7%, secondary/tertiary for 38.8% and non-service for 2.5% of the total national health spending. Of the total spending for primary health care, 85% is for curative care services, while 15% is for preventive and promotive care services. Household out-of-pocket expenditure accounts for 82% of total primary care spending and is particularly concentrated (92%) in the curative care component of primary care services. In contrast, about 73 % of preventive and promotive health services are financed by the center (44%), state governments (29%) and private households (27%). The large central government share is due to its almost exclusive financing of the Family Welfare (FW) program. 4 Health expenditure in India is defined in this report to include both public and private health spending. Public sector spending are those elements of spending whose primary purpose is to prevent or treat disease or control fertility. This includes: central, state and local government expenditures under the budgetary heads of medical, public health and FW; public expenditures on services or insurance for government and parastatal employees and formal sector workers. Public spending on nutrition programs and drinking water supply provision are excluded. Private health spending includes those spent by households and firms on preventative and curative care and for family planning.

45 19 Figure 3.1: National Health Spending (Rs. per capita in ) PRNvate Households Municipal Govt (0.5%) Central Govt (5.2%) StatGovt (15.2%) A J Exl Donors (0.8%) Third Party (3.3%) Rs. per Capita Tta Percent of Total Central Goverunent % State Government % Municipal Government % External Donors % Third Party Insurance and % Employer Payment Private Households % TOTAL %

46 20 Table 3.1: National Health Spending; An Estimated "Sources and Uses" Matrix (in percent of total expenditure)* Sources State & Central Local Corporate/ Govemnnt Government 3rd Party Households Total Primary Care (7.3) (9.5) (1.3) (81.7) (100) Curative (.7) (6.0) (1.6) (91.7) (100) Preventive and Promotive Health (43.9) (29.4) (26.7) (100) Secondary/Tertiary Inpatient Care (2.3) (21.7) (6.4) (69.6) (100) Non-service provision n.a n.a 2.5 Total Note: Row percentages are in parenthesis

47 Private household expenditure is also dominant for secondary/tertiary inpatient care, although it is somewhat less so than for primary care services. For secondary/tertiary inpatient care, the largest contribution comes from private household funds (70%); state governments contribute 22% and the central government about 2%, and third party insurance and employer payment add another 6.4%. 3.6 While expenditure levels alone are not sufficient to understand the role of the private and public sectors for delivering health care, especially primary health care services, the overwhelming dominance of private household spending is certainly a good reason for the government to take a close and careful look at both its scope and impact. This chapter, therefore, addresses public expenditures on health with special emphasis on primary health care services. B. Public Expenditures on Health 3.7 Schemes and Programs. The estimated composition of total government health spending by different end uses is shown in Figure About 47 % of the total health sector budget of center and states is spent on curative care and health facility operations. This might seem excessive, but in fact, this figure is often well over 60% in a country at India's level of per capita income. Preventive and promotive services comes next with 30%, of which about equal shares are for prevention and control of communicable diseases and family planning and immunization. This compares well with mnany other low-income countries, where the figure is well below 10% for preventive and promotive services. The problem is less the distribution of spending by major head than the inefficiencies that result from the low level of current expenditures relative to the massive infrastructure that has been built up. Finally, 7% of the budget is allocated to insurance for central government employees (ESIS) and organized industrial workers (CGHS); 5 and 9% to research, education and training of doctors and paramedical staff. The low share of capital investment at 6.4% of the Government's health budget is due to the fact that much of the physical infrastructure has already been achieved over the past one and a half decades. I The largest insurance scheme in operation in India is the Employees State Insurance Scheme (ESIS), a government subsidized insurance plan, established in 1948, to provide benefits to the organized working class in case of sickness or employment injury. This scheme provides services through an estimated 111 hospitals and 1,384 dispensaries in Access to the ESIS is limited to a rather narrow group of workers and their dependents: employees receiving not more than Rs.2,500 per month and employed in covered factories and establishments. Including beds assigned to plan in other facilities, nearly 23,000 beds are currently available to approximately 27 million beneficiaries. The Central Government Health Scheme (CGHS), initiated in 1954, was designed to provide comprehensive medical care facilities to central government employees and their dependents. In 1990, CGHS provided care to an estimated 3.8 million beneficiaries through a network of about 300 dispensaries, 3 yoga centers and 13 poly-clinics. While the dispensaries provide the basic health care and emergency services, hospitalization is provided through central, state or municipal hospitals.

48 Services at the Primarv Level. Public expenditures at the primary level are reported under three major heads: medical, disease control and FW. Districts which are better endowed spend a larger share of total expenditure on medical than moderate and poor districts. This is due to the fact that these districts allocated a higher proportion of their recurring expenditure to hospital based services rather than to non-hospital based services. The average share of expenditure on medical for such districts is about 65%. Municipalities generally have a much higher proportion of expenditure on medical, about 79.5%. The per capita expenditure on medical is much higher than that on disease control and FW and this gap is quite conspicuous in the better-off districts and municipalities. The distribution of expenditure on disease control among various programs varies across states and districts. A common feature of most districts, however, is that malaria (NMEP) gets the highest allocation followed by Leprosy (NLEP). The allocation to individual programs under public health depends on the priority established by the state based on its epidemiological profile Inputs. Salaries and wages account for about 62% of government health spending, non-salary maintenance for about 20%, capital investments in building and machinery accounts for 7 %, and transfers to local bodies make up the remaining 11 % (average of 14 major states in , NIPFP, 1993). Since the health specific transfers to local bodies have a high share of salary support, the overall share of staff costs in the total health budgets is in fact higher than the 62% quoted above. Data from 12 sample districts and 8 municipalitie surveyed recently indicate salary shares in the range of 70-80% (NCAER, 1993) The growth of salary expenditures has been faster than the growth of all other components of health spending over the past one and a half decades. Salaries have grown at an average annual rate of about 10% in real terms, while non-salary maintenance and capital expenditures have each grown at about 5 % or less. As a result, the proportion of salaries in total government health spending, excluding the salary component of transfers to local bodies, has increased at the expense of non-salary maintenance expenditure (Figure 3.3).

49 23 Figure 3.2: Estimated Composition of Government Health Budget ( ) Capital Investment 6.4% Aministradon & other MediBal Educafon & - Hosp itals 31.8% Research 8.8% Insurance (CGHS, ESIS) 7.2% Family Welfare 14.6% Pubrlc Heaflth (Disease ~~~~~~~ ~~Control) 15.1 % Pnmary Care Faciities 13.6% Govt Health Budget (Rs. Billion) Percent of Total Hospitals % Public Health (Disease Control) % Primary Care Facilities % Family Welfare % Insurance (CGHS, ESIS) % Medical Education & Research % Aministration & other % Capital Investment % TOTAL %

50 24 Figure 3.3: Shift in Input Composition *CapiaOuuay Transems 60.0 Operaton & Maintenance 0 SWaraes ~ Salaies Operation & Maintenance Transers I Capital Outlay

51 At the state level, salaries appear to have increased much faster than operations and maintenance (O&M) expenditure, except for Punjab. The problem, seems more acute in the case of Orissa, Rajasthan, Kerala, West Bengal and Andhra Pradesh were O&M expenses have grown by less than 3% annually in real terms while salaries have grown by 10% or more (Table 3.2) 3.13 At PHC and subcenter (SUBC) levels, the extent of underfunding of recurrent costs is difficult to estimate, since there are multiple budgetary sources of financing these facilities, and the budget data are not classified according to the type of facility. However, some broad indicators are discemible. For instance, the number of SUBCs per capita doubled between 1981 and 1988, while the number of PHCs per capita more than doubled during this period. However, the per capita expenditure towards rural health facilities did not rise that rapidly; comparing average levels for the period and , per capita expenditure was 56% higher in the case of the rural component of states' medical budgets, 24% higher in the case of disease control programs and more than double only in the case of rural family planning services. In other words, while the central FW program probably ensures more or less adequate funding at the lowest level facilities (SUBCs), constrained funding of other components of central and state health budgets seem to have resulted in spreading of resources too thinly on too many facilities at the higher levels such as PHC and Community Health Centers (CHC). A more accurate and detailed discussion of the size, location, and effects of these resource gaps based on the latest survey of sample districts is discussed in Chapter Griffin (1992, Table A. 14) based on 1983 data, reported government spending on non-salary items at 42% of total spending, whereas, average non-salary spending for Asian countries in Griffin's comparative study was 38%. However, current data indicates that overall, salary expenditures account for 70-80% of total spending at district/municipalevel. Drugs and supplies generally comprise less than 15% of the total. This ranges around Rs. 5 per capita for the rural districts, but around Rs. 13 for the municipalities; this is still a very low amount, since town-based hospitals also serve the rural population. The fact that less than 20% is allocated for non-salary maintenance including drugs, is a serious cause for concern. Lack of availability of medicines and other supplies at publicly managed facilities, especially in rural non-hospital facilities, is widely perceived as one of the major factors for the low quality of care provided at such facilities and the resulting consumer preference for private practitioners.

52 26 Table 3.2: Growth of Health Sector Expenditure Classified by Inputs (Real Annual % Growth, ) Salary Operation & Capital Spending Maintenance Qty Andhra Pradesh Bihar Gujarat * 11.2 Haryana Kamataka * Kerala Madhya Pradesh * Maharashtra Orissa * 16.5 Punjab Rajasthau Tamilnadu Uttar Pradesh West Bengal All Major States * Estimated rate not significantly different from zero. Source: Tulasidhar, 1992

53 C. Government Financina for Health Central and State Governments. Central and state Governments together account for about 20 % of national health spending. Of this, state governments account for about 73 % of total government health financing, while the central government accounts for about 25 %, while a small amount is financed by urban municipal bodies. There has been some confusion in accounting for state and central roles, since central contributions to jointly-funded schemes appear in the budget as state expenditures as part of the state plan. The accounting used in this report is based on the actual source of funds Central and state Governments finance very different components of the total expenditure. Figure 3.4 breaks down the uses of funds in the budget by center and state shares. States heavily finance primary health care facilities, hospitals, disease control programs and insurance. The center, on the other hand, emphasizes FW and to a somewhat lesser extent, education and research. Capital investment is shared equally by the center and the states. The central Department of Health allocates over 45% of its budget to the central teaching hospitals and research institutions, about 15 % towards the Central Government Health Scheme (CGHS), a medical benefit scheme for its own employees, and about 35% towards the disease control programs. The Department of FW allocates about 85 % of its budget towards family planning and 15% towards maternal and child health and universal immunization The family planning and immunization programs are fully centrally financed while the disease control programs are partially financed by the center, with the state Governments required to allocate matching funds from their budgets and bear the staff costs. In either case, the concerned department of the central ministry is responsible for program design and monitoring, while the corresponding state level department is responsible for implementation. The entire expenditure on these national programs is recorded in the state budgets, while the centrally financed component is also recorded in the central budget as a grant to the states State Governments finance the bulk (97%) of curative hospital care, as well as a significant share of expenditure involved in operating the primary health care infrastructure in rural areas. Central grants partially finance the disease control programs and the centrallyfinanced "rural health" scheme under the public health head which provides some funds for operating primary care facilities. The state governments bear all other costs of non-hospital rural services.

54 28 Figure 3.4: Center and State Shares in Different Components of Government Health Budget ( ) :;ts. >4s:. 0;, O States' Share % L.60 U) 50 EU C 10 I- ~ ~ ~ ~ Ccnte's Sarc Sates CneShare % 6 I~~~~~ Primary Care Facilities Famnily Welfare j Insurance (CGHS, ESIS) Medical Education & resarch Administration & other I11 89 Capital Investment

55 Central intervention in the health sector is both through the design and operation of centrally-sponsored programs as well as support for infrastructure development. A major vehicle for the latter is the National Minimum Needs Program (NMNP), a mechanism that allows the center to influence and encourage states to spend on building up infrastructure for rural health, water supply and nutrition. The NMNP is part of each state government's own plan, but for each rupee that the state spends towards these minimum needs, it receives a matching rupee from the center as a grant. In other words, disbursements under one of the central vertical programs is tied to the states' own efforts to fulfill minimum requirements of rural health infrastructure Local Governments. Local bodies have no significant financial authority in India except in large cities. In some states, however, local bodies have a significant responsibility for managing services and implementing national or state government programs. The degree and pattern of decentralization in state-local relations exhibits wide inter-state variation. Transfers to local bodies, as a share of total state government budgets, for example, vary from over 40% (Gujarat and Maharashtra) to 15% or less (Haryana and Madhya Pradesh). For the 14 major states, the average share of transfers to local bodies was 30% of total expenditure in the second half of the eighties; the share of such transfers accounts for about 11 % of state health spending While the federal structure of government in India is based on a significant devolution of taxing powers to the states, supplemented by a statutory right to their share in major central taxes, local bodies have very limited taxing powers or statutory rights. Decentralization has taken the form of delegation of implementing responsibility with minimal or no devolution of financial powers. Thus, even in the case of Gujarat or Maharashtra, where 40% of state government expenditure is transferred as grants to local bodies, the local bodies have little or no access to any financial resources on their own; their spending is totally dependent and determined by what is transferred from the state budget. The only exceptions to this general rule are municipal corporations of cities and towns, who raise their own resources through urban land and property taxes, and spend about 20-40% of such resources on health and related services Of the total amount transferred as grants by the states to local bodies, over 95% consist of specific purpose grants to support social service facilities run by the local bodies, such as grants to support salaries of panchayat school teachers, grants to support salaries of paramedical staff in rural health centers, etc. Less than 5% consists of general purpose grants over which the local authority has flexibility of use. Such grants have remained more or less constant in nominal terms in all states, over the past four and half decades.

56 D. Trends in the Level and Composition of Government Spending Health and FW spending has grown at an average annual rate of a little over 4% in real per capita terms since the mid-1970s Health and FW have absorbed about 3-4% of total government spending throughout the past four decades, declining slowly since the early seventies from about 3.8% to about 3.3 % at the end of the eighties. This occurred despite the increase in health spending in real per capita terms because overall government expenditures rose even faster than health expenditures, growth of most of the centrally-sponsored programs decelerated and financial constraints faced by the state governments became more acute. Within the health and FW budget, the share of FW has risen at the expense of health. Figure 3.5 shows the size and composition of health-related public spending in four different time periods Although the center finances only 25% of total government health spending, the centrally sponsored programs have nevertheless acted as the major catalysts in shaping new directions in the pattern of health spending. The rapid growth of the FW program since the late seventies initiated the broad shift towards preventive and promotive health spending by the government. The growth in the Accelerated Rural Water Supply Program in the early eighties, and in the ICDS program more recently, have added to this trend, apparently in line with the emphasis on rural primary health care contained in the National Health Policy adopted in However, expenditure on disease control programs has grown less rapidly throughout this period, declining as a share of total spending on the broadly defined health sector. Thus, there is a general correspondence between the rates of expenditure growth on the major components and the stated priorities of health policy, i.e., there has been an increase in spending on those elements expected to benefit poor women and children and reduce mortality and morbidity from the health burdens they principally face. However, the specific strategy and the intended linkages amongst the major components -- medical care, public health, family welfare, nutrition, and water supply -- is unclear. 6 If nutrition and drinking water supply are included, then the growth rate is higher at 6%. In other words, spending on these health-related components of the budget has grown faster than spending on health proper. The most rapid growth has been in the case of nutrition, driven by the centrallysponsored Integrated Child Development Services (ICDS) program.

57 31 Figure 3.5: Changing Composition of Health Spending (Rs. per capita at Prices) * Nutrifa Water Supply OFamily Wuaem Pubikc Hea_t EMedical_ X _I S8 Medical Public Health Family Welfare Water Supply Nutrition

58 One of the effects of state level financial constraints has been the slowdown in achievements under the National Minimum Needs Program (NMNP), the major source of financing the expansion of rural PHCs in the state budgets. Whereas in the sixth plan period ( ), it is estimated that 83% of planned outlay on rural health under the NMNP was actually spent by the states, in the seventh plan period ( ) this ratio declined to 73% (Tulasidhar, 1992.) While the state governments faced more stringent budgetary constraints in the late eighties, the center also did not use its leverage very effectively to protect funding for primary health care in this period. The only financial leverage exercised by the center on the NMNP was the centrally sponsored Accelerated Rural Water Supply Program (ARWSP), under which the center matched its grant allocation with the states' spending on NMNP The impact of the deceleration in expenditure growth since the mid-1980s is twofold: (a) slowing down in the creation of new rural non-hospital health facilities and (b) decline in the level of recurrent expenditure per facility. Considering that by 1986, the target for rural facilities had already been achieved up to 75% in the case of PHCs and over 85% in the case of SUBCs, some slowdown in facility creation may be considered natural and not a cause for concern. The decline in recurrent expenditure per facility poses a more serious problem, since inadequate recurrent cost support for the facilities already created would undermine the capacity of such facilities to deliver services to the people Trends in Government SRending on the Medical. Public Health, and Family Welfare s. The share of medical education and research which falls within the budgetary category of "medical services" has increased appreciably, from less than 10% in the mid seventies ( ) to about 14% by the late eighties ( ). This increase has come at the expense of general curative care services under the budgetary category of "medical relief" and the Employees' State Insurance Scheme (Tulasidhar, 1992). General curative care services for rural areas has increased and in line with the stated policy objective of strengthening outreach facilities. However, the rising proportion of spending on medical education, given the relative excess of physicians to paramedical staff, is less appropriate Within the category of FW, the share of maternal and child health (MCH) has risen only marginally, from about 2.5% in the seventies to about 4.5% in the late eighties. This is in spite of the heavy emphasis laid on MCH in policy documents and priority statements of the central government and a major new initiative in the 1980s to increase immunization coverage. Monetary incentives for permanent methods of contraception (sterilization), referred to as "compensation" in the budget, continues to absorb 15-20% of spending under the FW program Trends in Plan and Non-Plan SRending. Medical and disease control programs, and FW have together consumed a fairly steady 3-3.5% share of plan budgets over the past four decades, with the share of FW growing and health declining. However, health still accounts for more than 80% of the total of combined center and state plans. Non-plan spending by the center and states together has been rising more rapidly than plan spending, especially in the recent period of deceleration of government health spending. Table 3.3 shows the growth of government expenditures at the state level in the period , by the different sources of finance.

59 The faster growth of non-plan expenditures compared to plan, both in the case of the health sector as well as overall, indicates reduced flexibility to finance additional expenditures for program initiation or expansion. State plan for health including FW allocations have grown somewhat faster than central plan grants for the sector, indicating some substitution of state for central plan funds in the health sector That state Governments have been able to partially ameliorate the impact of decelerating or declining central plan allocations in the late eighties could be interpreted as a sign of state level commitmento protect funding for the health sector. However, the capacity of state Governments to do this depends to a large extent on their overall financial position, determined by their own resources and the untied transfers made available to them by the center. In the period , as shown in Table 3.3, the states were able to expand their overall plan size by over 6% annually in real terms, and their non-plan budgets by nearly 8%, as a result of untied central plan and non-plan transfers rising by over 8 % and 11 % respectively Inter-State Variation in Expenditure Trends. There is wide variation among the different states both in terms of the level and the composition of government health spending. The level of spending is lowest in the poorest states which are also those with the lowest health status. Poor states which have low health status cannot afford to spend much on improving this status (NIPFP). The per-capita spending on health and FW in Punjab, the richest state, is three times as high as in Bihar, one of the poorest states. Table 3.4 shows the variation among the 14 major Indian states in terms of their per capita Government health spending in the late seventies and in the late eighties The ratio of the highest to the lowest level between the two periods increased from 3.0 to 3.5, showing a slight increase in the degree of inequality of per capita health spending between states. On the other hand, the lowest level of spending (Bihar) improved from 45% to 54% of the national average. There is thus no evidence of any appreciable change in the degree of variation among the states during the two periods. Bihar and Uttar Pradesh, however, continued to remain at the bottom of the ladder in both periods.

60 34 Table 3.3: Growth of Expenditures by Sources: Total and Health (Rs. Billion) Annual Growth Rate 1986/ /91 Nominal Real Total Central Grants % 5.9% Tied Grants % 2.6% Untied Plan Grants % 8.4%/e Untied Non-Plan Grants % 11.1% Total State Plan Budget % 6.2% Total State Non-Plan Budget % 7.9% Health and Family Welfare % 5.5% Central Grants % 2.6% State Plan % 3.5% State Non-Plan % 6.7%

61 35 Table 3.4: Per Capita Health Expenditures, 1970's to 1980's (Rs. at 1988/89 Prices) Rank Rank Annual Growth Andhra Pradesh (10) (9) 4.6 Bihar (14) (14) 6.1 Gujarat (6) (4) 3.6 Haryana (5) (6) 3.1 Karnataka (8) (10) 2.5 Kerala (2) (2) 3.6 Madhya Pradesh (12) (11) 5.0 Maharashtra (7) (3) 5.3 Orissa (11) (12) 4.2 Punjab (3) (1) 8.8 Rajasthan (9) (8) 3.8 Tamnilnadu (4) (5) 3.1 Uttar Pradesh (13) (13) 6.2 West Bengal (1) (7) 1.7 Average (14 major states) * refers to the budget heads of medical, public health & family welfare Note: Figures in parentheses show ranking of states Source: Derived from Tulasidhar, 1992

62 The states which significantly improved their relative position with regard to per capita health expenditures were Punjab and Maharashtra, and Gujarat to a lesser extent, all of which are among the richer states in the country. This confirms the thesis that state Government health spending is mainly resource constrained, rather than driven by need Among different components of the health budget, the inter-state variation in spending levels in the late eighties has been found to be higher in the case of disease control programs, than in the case of medical services and FW. The ratio of per-capita spending between the highest and the lowest among the 14 major states was 7:1 in the case of disease control programs while it was around 2:1 in the case of medical services and FW The general pattern of inter-state variations indicate that the central government has not utilized the mechanism of specific purpose grants very effectively to redress inter-state inequalities in health status, especially in the case of communicable disease control where there are significant externalities. E. Impact of Government Spending on Health Status 3.38 Several analysts in India have interpreted the positive correlation between per capita public spending on health and the health status of the population as evidence of the positive impact. However, the positive correlation (Figure 3.6) is not necessarily evidence of such a causal relationship. It is possible that socio-economic development is the causal factor that determines health status as well as the level of spending, but more in-depth analysis is needed to confirm this hypothesis. Establishing the impact of public spending on health status requires more rigorous analysis of data, separating the effect of socio-economic variables from the effect of public spending. Available analytical efforts pertaining to the past two decades in India, and their key findings are examined below An analysis of pooled cross-section and time-series data, covering 13 major states and 12 years ( ) in a random effects framework, has shown that the level of per capita public spending had a significant effect on postnatal infant mortality rate during this period (Tulasidhar and Sarma, in Berman & Khan, 1993). This analytical work also showed the growth of rural health facilities in that period to have had a significant effect on the demand for reproductive health care, as revealed by the proportion of cases availing of medical attention at birth. One of the conclusions of this study was that while public spending related variables explained a relatively larger variation in post-neonatal mortality, socioeconomic variables seem to play a major role in the case of neonatal mortality. Neonatal mortality refers to infant mortality during the first month after birth, while post-neonatal mortality refers to infant mortality between the first and the twelfth month after birth.

63 37 Figure 3.6: Infant Mortality and Per Capita Health Spending in Indian States 0-20 a U Rs. per capita spending on health RL per Capits IMR per 1000 Births in in 1989 (.4 as-a P.1 Andhra Bihar Gujarat Haryana Karnataka Kerala Madhya Pradesh Maharashtra Orissa Punjab Rajasthan Tarnil Nadu Uttar Pradesh West Bengal 47 77

64 Another analysis pertains to the performance of the family planning program in 17 states in the period (Jolly, 1986). This analysis found that a much larger part of the variation in program performance was explained by socio-economic variables than by program input variables: literacy rate alone explained over 50% of variation in program performance. Nevertheless, this study affirmed that the average expenditure per couple had a statistically significant impact on couple protection. One of the more important findings of this analysis was the considerable degree of variation among districts within each state, both in terms of socioeconomic standards and in terms of family planning program performance. It pointed out that the application of uniform national norms were inappropriate, and recommended devolution of responsibilities from center to states and from states to local bodies Analysis of more recent data pertain to 15 states and two years, 1985/86 and 1990/91 (Seeta Prabhu & Chatterjee, May 1993). Regression of health attainment (constructed on the basis of infant mortality rate and the crude death rate in each state) on health infrastructure and the level of per capita recurrent expenditure on health showed that: (a) the coefficient of health infrastructure development was statistically significant; and (b) the coefficient of per capita recurrent government expenditure was not significant These studies confirm the significant positive impact of public health spending on health status. However, due to inefficiencies in the structure of financing as well as measurement problems, it is difficult to draw conclusions about the benefits of increasing public spending on health at the margin. On the relationship between infrastructure constructed and health status, these analyses suggest that once the basic health infrastructure had been created, further improvements in health status have depended not only on adequate levels of recurrent cost financing but also on the appropriate packaging of staff and non-staff inputs. As the health budgets of central and state governments came under pressure in the mid eighties, and as salaries tended to crowd out other components such as drugs and supplies, the marginal productivity of government health spending seems to have declined, while the level of impact remained low. Given the continuing burden of preventable disease and mortality in India, government health spending can have a positive impact provided it is targeted towards addressing the binding resource constraints such as improving the packaging of inputs, and targeting appropriate health problems and vulnerable populations. F. The Effect of Adjustment on Government Health Spending 3.43 Mechanisms of Adjustment Effects. The present round of structural adjustment in India began with the new package of policy changes announced in July It involved a onestep devaluation of the Rupee by 23 %, a major contraction of the fiscal deficit of the central government by more than 2 percentage points of GDP during the year 1991/92, reduction of average import tariff levels in the subsequent year, 1992/93, etc. Conceptually, the process of adjustment can affect the output of government health spending by affecting: (a) the quantum of financial resources available with health ministries and departments; and (b) the unit costs of providing health care. Fiscal contraction by the central government is translated into tightening of budget constraints at the state level through different mechanisms, corresponding to the three different sources of financing health expenditures at the state level, namely the non-plan budget of the state, the state's own plan budget and the budget of centrally sponsored programs. Figure

65 shows the different channels through which the impact of adjustment is conveyed to the state level There are two kinds of pressure on the financial resources of state governments, namely: exogenous macroeconomic factors and contraction of central transfers to states. A deceleration or decline in domestic industrial output, for example, may lead to a reduction in tax revenues collected by both the center through excise duties and personal income taxes, and states through sales tax. Since a statutorily fixed proportion of central excise duties and personal income taxes are shared with the states, reductions in either of these types of tax revenue will squeeze the revenue of both the center and state Governments In addition to revenue effects, the center can: (a) reduce the quantum of untied plan grants to states and/or (b) reduce the quantum of tied plan grants transferred under one or more centrally-sponsored programs. Reductions in allocations to centrally- sponsored health sector programs are the most obvious form of squeezing the health sector and have received attention in Bank-GOI dialogue. However, the other channels of pressure, though less visible, are likely to be more significant as the funds involved are much larger in magnitude, since they affect statelevel spending Central transfers to the non-plan budget of states, consisting of tax sharing and gap filling grants, are statutorily determined by a quasi-judicial body called the Finance Commission; such transfers are therefore not at the discretion of the central Government and hence not vulnerable to contractionary pressures by the center. On the other hand, central plan transfers to states, both tied and untied, are largely at the discretion of the center and hence more vulnerable to central policy.

66 40 Figure 3.7: Channels Through Which Strctural Affects Health Spending A4justnent Exogenous reductions in center's revenue i CENTER'S EXPENDIUE- ll rplanning Firnance MOHFW </ Commisslon C- Comnon I Cuts in centrai pbn gwj schemes STATE EXPENDITURE unfd PLAN / \ ad]=w f at sxe lev estate NON-PLAN Macroeconomic effects State Health Budget on input prices Exogenous rductions revenue HEALTH SERVICE PROGRAMS

67 Untied transfers from center to states, called "central assistance to state plans", consist of 30% grants and 70% loans in the case of the 14 major states, and of 90% grants and 10% loans in the case of the special category states, which are mainly the hilly and predominantly tribal states plus the state of Jammu & Kashmir. The center is free to decide the quantum of assistance to each of the special category states, whereas the assistance to the 14 major states is distributed among them on the basis of an objective formula called the modified Gadgil formula. However, even in the case of the latter, only the inter-state distribution is formula driven; the total quantum of such assistance is at the discretion of the central Ministry of Finance States also have some discretion in how they use untied funds. For example, a reduction in central assistance to state plans may result in different levels of reduction to health spending in different states. Similarly, states exercise some discretion in their non-plan spending, and so can favor or disfavor the recurrent cost needs of the health sector In addition to the factors outlined above, there are also other macroeconomic pressures that operate, such as (i) reduction in small savings by households and (ii) the devaluation of the Rupee. A fixed proportion of collections from national savings schemes, operated by post offices and linked with tax incentives, are on-lent by the center to the states as a loan under the non-plan account; any decline in such collections would thus reduce the quantity of central loans available to the states. A major devaluation of the currency, by affecting the cost of imported inputs, especially drugs and pharmaceuticals, could affect the unit cost of health care financed by the government; even if financial allocations are maintained, the real value of such allocations could decline due to an abnormal rise in unit costs Fiscal Contraction by the Center in The patterns of fiscal adjustment in recent years are described in Figure 3.8. In 1991/92, the first year of fiscal contraction, the center reduced its deficit by 2.4 percentage points of GDP, from 8.4% to 6.0%. Reduction in central subsidies (export and fertilizer subsidies) and other components of the center's own expenditure accounted for nearly 70% of this major effort; revenue expansion, including the sale of public assets, accounted for more than 20%, while less than 10% was accounted for by squeezing transfers to the states The aggregate picture of central transfers to states in 1991/92 hides a significant variation between the behavior of transfers to the major states and to the special category states. While plan assistance to the special category states were stepped up significantly, central plan assistance to the 14 major states were sharply reduced. The general purpose transfers governed by the Gadgil formula is estimated to have more than doubled in the case of the special category states, led by Assam and Jammu & Kashmir; on the other hand, these transfers declined by 8% in nominal terms (over 20% in real terms) in the case of the 14 non-special states.

68 42 Figure 3.8: Composition of Center's Fiscal Adjustment Measures (change over previous year) 2.5% O Revenue Increases /bl O Squeeze on States /a El Expenditure Cuts 1.5% 0~ % 0.5% 0.0% 91/92 91/ Budget Acual Budget Actual Budget 91/92 91/92 92/93 92/93 93M4 Budget Actual Budget Actual Budget Expenditure Cuts 1.4% 1.6% 0.5% 0.3% - Squeeze on States i, 0.6% 0.2% 0.5% 0.1% 0.5% Revenue Increases /b 0.2% 0.6% - 0.1% 0.3% /a Tax Share, grants and net loan disbursements /b Before tax devolution to states

69 In 1992/93, the second year of adjustment, the central government budgeted a further reduction of 1 percentage point in its deficit, from 6% to 5 % of GDP. About 60% of this budgeted adjustment was to come from reduced flow of resources to the states; only 40% was to be achieved through reductions in the center's own expenditure, while revenues were budgeted to maintain their share of GDP. The actual outcome during this year turned out to be vastly different from the budget targets. Overall deficit reduction was less than 0.5 percentage point, and almost all of this came from reducing the center's own expenditure; the targeted squeeze on states did not materialize In addition to the reduction in untied plan transfers in the case of the 14 major states, the central budget for 1992/93 also witnessed a sharp reduction in allocations for the centrally sponsored disease control programs, led by a cut of nearly 40% in the National Malaria Eradication Program. However, these cuts in the central programs were restored during the course of the year as part of the effort to provide a social safety net In 1993/94, the third year of adjustment, the central budget envisaged a further reduction of 0.8 percentage point in the fiscal deficit, from 5.5% to 4.7% of GDP. What is especially worrying is the hefty cut in net transfers to the states envisaged in the 1993/94 budget. Again, the central fiscal deficit turned out to be 7.3%, much higher than planned. There is, however, some fear that the actual squeeze on untied transfers to the states may turn out to be harsher than the budgeted amounts in the coming years Sharp reductions in the untied plan assistance to states reduces the ability of state governments to maintain spending on their own plan programs. Whereas in the late eighties, state governments were able to partially make up for shortfalls in central plan allocations by stepping up their own plan allocations, such a response would be ruled out in the 1990s because of fiscal pressures. Further, the impact of reduced central assistance to state plans would also affect poorer states more acutely than the better off ones, because it is precisely the poorer states that are more dependent on central assistance to maintain their plan size Thus, despite the restoration of budget cuts in centrally sponsored health programs in 1992/93, and increased allocations for these programs in 1993/94 and 1994/95, an effective social safety net has not been provided to the poor who have continued to rely on private providers. If increased central allocations to vertical programs are accompanied by a sharp reduction in general purpose transfers to states, this would disproportionately affect the poorer states, reducing their ability to take advantage of many of these schemes which require matching allocations from state budgets. This would also reduce the effectiveness of central expenditures through the lack of complementary inputs financed by the states Macroeconomic Effects. In addition to the actions by the central Government to reduce its deficit, certain macroeconomic factors have also added to the problem of protecting state government health budgets in these years of adjustment. Among the most significant of these factors are: (a) a decline in small savings collections since 1991/92; and (b) a deceleration and stagnation of domestic industrial output since 1992/93, although there has been some improvement more recently.

70 Net collections under national small savings fell in 1991/92; as a result, the amount on-lent to the states also declined by almost 0.5 % of GDP. The financially more dependent states such as Orissa and West Bengal were the most acutely affected by this development. Small savings collections have remained low in the following year, and are likely to decline further as more diversified portfolio choices become available to households This fiscal contraction has led to demand recession. The growth of domestic industry has decelerated since 1991, although there has been change lately. Declining tax revenues as a share of GDP, as well as further decline in small savings collections, would put pressure on the non-plan budgets of state governments One macroeconomic factor that has generally had a major impact on health budgets in other countries undergoing adjustment, namely currency devaluation and rising import prices, has not so far played a significant role in India. Although the Rupee was devalued by over 20% in July 1991, and has declined in value even further since then, the average rise in drug prices has remained below the overall inflation rate. Two reasons have been cited for the remarkable stability of drug prices in India: (a) India's external sector is still small in relation to the size of its economy and health inputs are largely manufactured within the country; and (b) the government controls the prices of many essential drugs, allowing only marginal profits on their manufacture (Tulasidhar, 1993). The ongoing structural adjustment and liberalization policy may alter some of these conditions; if price controls are removed or relaxed, there could be a sudden spurt in drug prices. There is already some evidence of substantial increase in the case of some essential drugs (Rane, 1993). Drug prices in 1995 are showing substantial increases Evidence of Impact on Health Budgets. Compiled data on actual health spending by central and state governments combined are available for the first year of adjustment (1991/92), while estimates for the second year are only available in individual state budget documents. The compiled data for the first year shows that total government health spending fell only slightly, by about 2% in real terms; however, this was a combination of a sharper decline of 4% in the case of medical and disease control programs and an increase of 7% in the case of the FW program. The real decline was over 5 % in the case of spending on the disease control programs (Tulasidhar, 1993) What is even more striking from the evidence available for 1991/92 is the fact that the health budgets of poorer states have been squeezed to a greater degree than richer states (Figure 3.9). Spending on the prevention and control of communicable diseases fell by 7 % in real terms in the case of the 5 poorest states (Bihar, Madhya Pradesh, Orissa, Rajasthan and Uttar Pradesh), compared to 5% in the case of middle income states and 4% in the case of the richest states. This corroborates the hypothesis that in a climate of declining untied transfers from the center, the poorest states are the ones least likely to take advantage of centrally-sponsored programs that require matching allocations from state budgets Evidence from a recent survey of 3 sample rural districts from each of 4 selected states (Figure 3.10) shows that the real decline in the level of recurrent health spending in 1991/92 was 16 % in the case of West Bengal, 7 % in Uttar Pradesh and 3 % in Tamil Nadu; there

71 45 was a real increase of 5 % in the case of Gujarat (Table 3.1, NCAER Report 1993). This survey also shows that in the states which experienced the sharpest decline in recurrent spending, nonsalary expenditures fell even more steeply than salaries, although there is considerable variation between different districts in this regard. The average real decline in non-salary rural health expenditures in West Bengal was more than 17% In Orissa, one of the poorest and financially most dependent states, the government responded to the fiscal contraction by consciously substituting central plan funds for state plan funds in the social sectors. In the case of medical and disease control programs, while central plan grants more than doubled in 1991/92, the state plan budget was cut by more than 30%, maintaining the level of overall spending from all sources at the same level of the previous year in nominal terms (Ravishankar, 1993) Evidence of impact on actual health spending by states in the second year of adjustment (1992/93) is available in some selected cases, including West Bengal and Tamil Nadu (Ravishankar, 1993). In the case of West Bengal, while funding from all sources declined in the first year, there was a partial recovery in non-plan and central plan allocations in the second year, while the state plan allocation continued to decline. Overall health expenditure from all sources recovered by only 3% in real terms, after having declined substantially in real terms in the previous year. In the case of Tamil Nadu, a relatively better off and independent state, spending on social sectors was protected in the first year in spite of stagnant central plan allocations. However, in the second year of adjustment, while central allocations continued to stagnate in real terms, the state was able to increase only its non-plan budget, while its own plan budget declined by 8% in real terms Thus, to date there is some evidence of negative effects on health spending. It suggests that the "indirect" effects of central adjustment policies, especially those affecting state plan and non-plan spending levels and the propensity of states to spend on health, are more significant than the "direct" effects of central co-funding directly of state plan schemes. It is likely that the "indirect" as well as "direct" pressures will increase in the next phases of adjustment. Special attention is needed to assure that states, especially poor states, maintain adequate funding for basic health inputs.

72 46 Figure 3.9: Differential Impact on Health Budgets for States Grouped by Income Levels /a (Index of Real Exp. with 1989/90 = 100) X = _~~~~~~~~~~~~~1 OPift Shiles PooresstSlae l990 0 MM& mlnm ih / Richest States (Punjab, Haryana, Gujarat & Maharashtra) Middle Income States (Andhra Pradesh, Karnataka, Kerala, Tamil Nadu & West Bengal) Poorest States (Bihar, Madhya Pradesh, Orissa, Rajasthan & Uttar Pradesh) /a Expenditure on medical and public health (Revenue Account)

73 47 Figure 3.10: Weighted Average of Percentage Changes in Real General Health Expenditures FY92 from FY91 (in 91 prices); Four Major States c s2~1 'EE c ~~~ -5 Urban Areas

74 48 4. PUBLIC PRIMARY HEALTH CARE SERVICES: STRUCTURE, INPUTS, PERFORMANCE AND ESTIMATES OF FINANCING REQUIREMENTS A. The Development of Health Services 4.1 Background. The Bhore Committee Report (1946) was the first plan for an Indian National Health Service. It envisaged the construction of a massive publicly managed health infrastructure that would require the government's allocations to health as a percent of GDP to be increased three-fold over existing allocations. However, it did not foresee an important role for the private health sector. This plan conceptualized the basic health unit as a 75 bed hospital for every 10,000 to 20,000 population; 30 such units were to be supported by a 650 bed secondary unit with 140 doctors and 180 nurses; and each district was to have 5 such secondary units supported by one tertiary hospital with 2,500 bed capacity. Besides this structure, special programs for major diseases were recommended. The first and second Health Ministers' Conference held after independence accepted the recommnendations of the Bhore Committee, but maintained that lack of financial resources prevented their implementation. The First Five-Year Plan ( ) continued to pursue the same line of argument. Although subsequent Five-Year Plans stopped referring to the Bhore Committee report, the concept of targeted populations norms such as for SUBCs, PHCs, CHCs and various secondary and tertiary hospitals remained ingrained in these Plan documents. 4.2 During the first three Five-Year Plans ( ), the vertical disease control programs were given high priority. An important emphasis of these plans was on eradicating malaria and small-pox. The former was almost successful although in recent years there has been significant resurgence of malaria, and the latter was completely successful, resulting in eradication of smallpox. Since the mid-sixties, there has been a shift in focus as the public programs began to concentrate on fertility reduction through permanent methods of family planning. These programs received significant support from external donors, especially in the earlier periods. 4.3 The vertical programs were carried out initially by "single purpose" workers, specially trained for control of individual diseases and for family planning. Recognizing the advantage of integrating different types of preventive and curative services, the Fifth Five-Year Plan ( ) stressed the reorientation of existing workers into "multi-purpose" workers. Basic health workers for malaria and tuberculosis, vaccinators and FW health assistants were designated as Health Workers (male), and formally given responsibility for implementing all of the vertical program activities simultaneously. The auxiliary nurse midwives (ANMs) working in the Family Planning program were renamed Health Workers (female) and given responsibilities for FW activities, including both family planning and maternal and child health. 4.4 In response to the Alma Ata Declaration, the central government also established the Village Health Guide (VHG) scheme in the late 1970s. One VHG was to be trained for every 1,000 rural population. These workers received a very small monthly honorarium. The scheme expanded rapidly and large numbers of VHGs were trained, although a few states rejected the offer of central matching funds to establish the scheme. Over the years, however, the effectiveness of VHGs has been questioned and inadequate attention has been paid to retraining,

75 49 supervision and support. Some state governments also faced legal challenges as VHGs sought to be regularized as government employees. While the scheme still exists on paper, in many states it has largely disappeared as a functioning part of the public provision system. 4.5 Primary Care Facilities and Schemes. Two concepts of the Bhore Committee Report that were pursued in the initial Five-Year plans were PHCs and special disease control programs. PHCs were initially to have 3 or 4 emergency beds, with one or two doctors; and one PHC was planned for every 100,000 to 200,000 population. Over the next two decades, this target was reduced to one PHC per 30,000 population; and each PHC was to support 6 SUBCs. Each SUBC was planned to serve a population size of 5,000 people. This structure of rural primary health care was to be supported by one hospital located in each district headquarters. 4.6 The current policy on public facility development calls for establishment of a CHC for every 100,000 population. These CHCs are mainly multi-functional outpatient facilities. However, they are planned to include 30 inpatient beds and a minimal cohort of specialist medical staff. These facilities bridge the capacities of the larger district and sub-district hospitals with the purely outpatient and extension/promotion functions of the PHCs and SUBCs. 4.7 During the Sixth Plan ( ), new norms were laid down for population coverage and staffing of SUBCs, PHCs, and CHCs. These were one CHC per 100,000, one PHC per 30,000 (20,000 in hilly, tribal and backward areas), one SUBC per 5,000 people (3,000 in tribal and backward areas). The achievement to-date is, on average, about one CHC per 450,000, one PHC per 40,000 and about one SUBC per 6,700 population. 4.8 There is also considerable inter-state variation in the primary health infrastructure. Among the 14 major states, the average population served by a PHC in 1991 varied from a low of 6,500 in Punjab (almost five times the national norm) to a high of 40,000 in Madhya Pradesh (33 % less than the national norm). The average population per SUBC varied from a low of 3,770 in Karnataka to 5,900 in West Bengal. 4.9 Public Sector Provision of Hospitals. Three levels of government -- central, state and municipal -- have made large investments in developing public hospitals. The total number of public hospitals increased from about 2,500 in 1951 to over 3,200 in In 1990 this increased to over 4,500. Bed capacity in the public sector grew by more than 60% during the same period, totaling 425,358 by The public investment program in hospitals, however, reflects some variation across the states in terms of population-bed ratios. Kerala and Maharashtra have the lowest populationbed ratios, a population of about 1,100-1,200 per bed. At the other extreme are Bihar and Haryana, with population-bed ratios of about 3,800 and about 3,400 respectively.

76 B. Inputs for Public Sector Health Programs at Primary Level' This section of the report analyzes and assesses whether sufficient and/or timely inputs in the form of facilities, staff, and drugs and medicines were available to adequately perform the primary health care tasks assigned to the public sector. It also assesses the quality of inputs and whether they were used efficiently. Analysis was undertaken for a representative sample of districts and municipalities in four major states. In order to assess the adequacy of the available inputs and the extent to which they are in place, real numbers of facilities and staff were compared with Government of India (GOI) norms, where those existed. The gap between norms, sanctioned inputs, and inputs actually in place were compared. In the case of drug supplies, where no government norms have been established, interviews with facility managers were used to gauge their impressions on supply adequacy Norms for Public Provision. GOI norms exist for some, but not all, dimensions of public sector primary level services. For example, facility norms exist for the number of facilities required to serve rural populations; 8 and staffing norms Results presented in this section are based on a study commissioned for this sector report. The study was undertaken by the National Council for Applied Economic Research (NCAER) entitled, "Survey of Primary Health Care" P. Vashishtha et. al. (1994). * Staffing Norms. Government of India norms for hospitals are available depending on the bed size of the hospital. These provide details on the number of general and specialist medical staff, required nursing staff, and equipment. Since this study did not examine hospital services in detail, these norms were not used and are not presented here. Staffing norms exist for CHCs, PHCs, and SUBCs. The total number of staff in a district depend on both the extent to which facility norms have been met and the adequacy of staffing of the existing facilities. These norms are: Type of Staff Norm (i) CHC Doctor 4 Paramedical 13 Administrative 8 (ii) PHC Doctor 2 Paramedical 7 Administrative 7 (iii) SUBCs Paramedical 2 (I male,l female MPW)

77 51 exist for the staffing pattern of hospitals, CHCs, PHCs, SUBCs. 9 However, there are no quantity or expenditure norms for drugs and supplies. Although individual states may budget and spend on a "per facility" basis (e.g., allocating Rs. 5,000 per year per PHC for drugs), these levels are based on the amount of funds available, rather than on any analysis of supply requirements relative to health care demand or need Norms have also not been specified for the disease control programs. Despite their apparent "vertical" funding with central grants, these programs are in practice embedded in the overall primary health care delivery system. Staff working in these programs are included in the norms for multipurpose workers posted to SUBCs and the paramedical staff posted to PHCs and CHCs, in practice, are likely to have multiple responsibilities. Drugs and supplies may be provided separately in part, and in part integrated into general facility supplies, for which in any case no norms exist. The FW program also does not provide norms for its inputs, although as a 100% centrally-financed scheme it seems to be marginally better off in assuring staff and supply inputs The general conclusion is that while it is possible to partially assess the adequacy of various inputs relative to norms for the overall levels of facilities and staff, it is not possible to disaggregate input levels for specific public health programs in order to assess their adequacy or link to performance. It reflects the diffuse accountability in public programs where central allocations may be tied to specific vertical schemes, but the provision and monitoring of funds and inputs at field level is managed differently. This is a significant constraint to monitoring and assessing specific program activities and needs to be addressed Resource GaDs in Real Terms. Resource gaps at public facilities can be estimated by comparing the actual inputs available for primary health care services in districts and municipalities with the levels of provision already sanctioned and, where possible, official norms exist. The gap between actual and "sanctioned" resources represents those inputs for which budgets have been approved, but which remain unfilled. Using this definition of resource gap, three types of input measures were evaluated: facilities, staff, and drugs and supplies For facilities, those reported in place were compared with government norms. For staff, the actual staff in place for CHCs and PHCs with sanctioned staff positions and official norms were compared. Medical, paramedical, and non-medical staff were distinguished. Subcenter based male and female multipurpose workers were included in the calculation of staff adequacy for PHCs. For both of these measures, the gaps in both real and financial terms, using cost levels were calculated. Since no norms were available for drugs and supplies, results of interviews with facility managers on the adequacy of supplies are reported. Facility Norms. Govermnent of India norms are as follows: Hospitals/beds: CHCs PHCs SUBCs No norms 1/100,000 population 1/30,000 (1/20,000 in hilly/tribal areas) 1/5,000 (1/3,000 in hilly/tribal areas)

78 Facilities. Figure 4.1 presents the current level of facilities and beds in the sample districts as a percentage of the official population norms taking into consideration the higher level of provision for tribal districts Real gaps in facilities are greatest for the CHCs in all states and all districts, and least for SUBCs in most districts. On average across all districts for CHCs, 22% of those required were found in place. For PHCs, the figure was 94%, and for SUBCs, it was slightly more than the norm at 104%. The more complete SUBC system reflects the central Government financial assistance provided by the FW Program for SUBC development, while most PHC and CHC development is financed through the states, although with some co-financing from central funds through the Minimum Needs Program. There is also some evidence within the states that the more developed districts are better off in terms of health facilities The overall picture in terms of facilities confirms the successful efforts in the 1980s to establish the more peripheral elements of the delivery system. To get a sense of the overall facility gap, an average weighted by the population coverage of each unit was calculated. The weighted average over all districts is 99%. That is, the total number of facilities available is roughly equal to the number of facilities required by the norms, although there might be more of some and less of others in individual districts. Of course, this calculation is dominated by the very large number of SUBCs required. The overall average represents a range of 74 to 155%, with four of the twelve districts below 90% No norms exist for facility provision in urban areas. The government hospitals usually located in towns and cities also provide outpatient services which substitute for some of the similar functions of CHCs and PHCs Since hospitals are concentrated in urban areas, hospital bed provision is of course much higher than in the rural districts, and the numbers of government non-hospital facilities are quite low.

79 53 Figure 4.1: Completion of Facility Norms in Sample Districts CHCs U) 90% 80% 70% _ 450% 2 _ 20% % E 0% -~ I ii i j Gujarat Tamil Nadu Uttar Pradnh j West Bengal 0 PHCs 160% 140% _ 120% 100% 160% - Gujarat L Tamli Nadu j Utbar Pradesh j Went Benga I 180% Sub-Centres 140% - 120% %_ j nt Gujud i Tamil Nadu UtterPradnsh Wetngal

80 Staff. Government health personnel were divided into three groups: physicians, paramedical staff including medical and public health, and support staff including non-medical, administrative personnel. For the rural districts, most physicians and paramedical staff were mainly found posted outside of hospital service. In contrast, in the municipalities, non-hospital staffing in public facilities was a much smaller component of total staffing for both physicians and paramedics. However, there were far more of both physicians and paramedics per 1,000 population in municipal than in rural districts. In reviewing these data, it is important to keep in mind that states have considerable autonomy in staffing rural facilities, so that they may choose not to adhere to GOI norms in terms of facility design or staffing Because norms exist for staffing patterns at CHCs, PHCs, and SUBCs, it was possible to calculate the gap in personnel for different types of facilities as well as for different levels of staff. Figure 4.2 presents the main findings in terms of overall staffing gaps at different types of facilities. Gaps are defined here as the percentage of staff positions according to norms that are filled. Figures for sanctioned but unfilled posts as a percent of the norms are also given. All types of staff positions are treated equally in Figure 4.2. PHC figures include the staffing required at SUBCs although only for those SUBCs already established The main conclusion is that the pattern of staff provision is the inverse of that found for facilities. Results point to significant problems in staffing the more peripheral rural facilities. While the most peripheral facilities have been physically completed relative to norms, they have the lowest level of staffing relative to norms, followed by the more central PHCs and the CHCs. On average, additional PHCs have 59% of posts filled, main PHCs 93%, and CHCs 142% of the norm of posts filled. For additional PHCs, the addition of sanctioned but unfilled posts was sufficient to approximate norms in only one of the districts--for all others there is a considerable absolute gap of as much as 50% of the normative level of staffing. For main PHCs, the absolute gap is much lower, as more than half the districts meet or exceed norms by combining filled and sanctioned positions. In the CHCs, staff levels exceed norms in six out of ten districts which reported establishing CHCs. The contract between facility and staff provisions is shown clearly in Figure The bias against peripheral facilities in overall staffing persists in the distribution of the most qualified staff as well (Figure 4.3). For physicians, actual staffing levels exceed norms only for CHCs. Actual paramedic staff levels exceed norms only in the main PHCs. When actual and sanctioned posts are combined, CHCs exceed the norms for physicians by about 50%. Main PHCs are relatively adequately staffed in relation to norms, with an excess mainly for center-based paramedics. As in the overall staffing patterns, additional PHCs have the lowest levels of staffing adequacy, with actual staffing for physicians at 50% of the norm and field paramedical staff at 66% of the norm.

81 55 Figure 4.2: Gaps in total Staffing Relative to Norms in Existing Government Health Facilities - 12 Distrist, 4 States 600% 500% 400% * Sanctioned but Unfilled o OFilled Positions :z 300%_ 0 200% CHCs 100% - 10o% V m ~ ~ ~ U I Gujarat Tamil Nadu Uttar Pradesh West Bengal ~Main PHCs 160% _ Sanctioned but Unfi led 140% 0 Filled Positions-- -ldotn_ 120%F 100% 6, s_ nouuu ~ ~~ _e 2 9_ Z80% 60% % 20% EE.2 < X R) < S 9 i 2 1i i p E GuJart Tamil Nadu ; UttariPradesh - West Bengal Additional PHCs 100% 90% Sanctioned but Unfilled- 80% -_"""" - - OFilied Posiin 70% - 60% Z 50% o) 40%- 30%- 20% 10%- 0% F Gujarat F Taml Nadu Utr Pradesh WeSt Bengal

82 56 Figure 4.3: Distribution of Qualified Staff at CHCs and PHCs Compared to Norms 140% 120% 100% 380% c -3 60% I ~~~~~~~~~~~~~~~~~~~CSanctioned but unfilled posktions 0 U ~~~~~~~~~~~~~~~~~~~~~Filled positions 40% 20% 0% * w s 0 c j (, E = ^~~~ E ~~ ' E U U I Eso 2 X X -L 0 0. U~~CL u 0

83 Drugs and Supphies. Supplies of drugs have often been reported to be inadequate in field studies of government health facilities (e.g., ICMR, 1989 and ORG, 1989). Moreover, the sources and flows of drugs and supplies for specific facilities and programs are also confusing. State governments purchase drugs and supplies for health facilities from non-plan funds. Drugs provided to districts might find their way to CHCs, PHCs, and SUBCs to varying degrees, depending on the distribution decided by the District Medical Officer. Centrallyfinanced schemes also pay for drugs and supplies for specific programs, such as, TB Control, Malaria Control, Family Planning, and Maternal and Child Health. It is unclear to what extent there is duplication in supply between these different sources of financing, or whether supply is combined but distribution may be uneven due to a "trickle down" approach of letting higher level facilities release drugs to lower level ones In the absence of official norms, the study relied on the qualitative assessment of medical officers at PHCs. Medical officers were asked to evaluate the adequacy of the supply of a sample of drugs in four categories: (i) general medicines, (ii) emergency medicines, (iii) vaccines and supplements, and (iv) contraceptives. For each drug, they were asked to rate supplies as: (a) absent or out of date (=0); (b) inadequate supply (= 1); adequate supply (=2). The score for all drugs in each category was then summed. A total score was also calculated for each district. A score above 50% indicates a mix of adequate and inadequate supply. The results for main and additional PHC areas combined are presented in Figure The total scores for all PHC areas studied ranged between 56 and 85% of the maximum. There is some evidence of lower scores overall for additional PHCs in West Bengal, but the differences between additional and main PHCs are small in the other states. Between categories of drugs, supplies are clearly more adequate for vaccines/supplements and contraceptives, less so for general medicines, and lowest for emergency medicines. For the latter, most scores were below 50%, suggesting a serious undersupply problem Inputs. Finances and Performance. Analysis of input, finance and performance data highlight that adequacy of overall resources are very important for improving the functioning of India's public health system. Data collected in the rural districts included a variety of reported output measures from government sources. Output measures included different types of output related to medical care, public health programs, and family planning and maternal and child health services.

84 58 Figure 4.4: Qualitative Assessment of Adequacy of Supply of Sample Drugs 100% 80% up 60% f l t E40% 20% 0% Total Scores General Medicines Emergency Vaccines and Contraceptives Medicines Supplements Average for all districts

85 Annual levels of output for different schemes in the areas of medical care, public health, and FW were collected for each of the study districts. For medical care, an index of outpatient equivalent contacts per 1,000 population was used, counting each inpatient admission as equivalent to four outpatient contacts. For public health, output rates per 1,000 population for different indicators from the TB, Malaria and Leprosy Control Programs were used since these programs each produce a number of different outputs, making it difficult to identify any one or even several unique measures. For FW, tetanus immunization for pregnant women and the couple protection rate with family planning were used To develop an overall performance score, each district's performance on each indicator was compared with all other districts, and a score of 1 given if below average and 2 if above average for the sample of districts. This is a relative perfornance measure. It indicates whether the activity level in a district is above or below the average for all districts, not whether the activity is adequate or sufficient. The average score in each district for each major area of health activity was also calculated and then averaged to get the overall performance score. This is a relative measure of overall service activity or output, roughly equally weighted between medical, public health, and FW programs The relationship between inputs and specific program performance measures was examined for all the different output estimates. The plots and simple regression are shown in detail in Statistical Annex 2. Although some of the plots were suggestive of links, none of the estimated regressions were statistically significant. Thus, the results of this study do not indicate any strong or consistent link between spending levels overall and performance measures Further analysis was undertaken of the relationship between different types of real inputs and performance. Figure 4.5 shows the plots of facilities and staff levels in existing facilities relative to norms against the overall performance score For facilities, after the removal of one outlier value there appeared to be a strong positive relationship between a weighted index of facility availability relative to norms and the performance score suggesting that access to clinics can make a difference. Performance continues to rise for districts with facility levels well above the norm. This mnay indicate the presence of excess capacity across the board or be interpreted to mean that the norms may be too low. However, for staff levels in existing facilities, there was no evidence of a link. This is consistent with an overall picture of low personnel productivity and excess staff capacity, despite the gap between

86 60 Figure 4.5: Facilities and Staff Levels Relative to Norms Against Overall Performance Score O ~1.8 X e o 0 1 piff Sfacilts Ft12.34". FaclUdh inorn (wtd) I,.. ] pwf tstaffnom ? t t3.95) F o N.S. 2 ~~~~~~~~~~~~a IA 1.6 a I7 l.a ~ ~ _ a a pert drug OA O.46 am am 0.66 t I bud e"*.e.

87 61 staffing levels and norms. For drug adequacy again there appears to be some positive relationship, although it was marginally significant statistically (p <.16). These results suggest that real inputs may be more important than spending levels in determining output, and that the current high level of staff salaries relative to other inputs may have a substantial cost in real output The 1992 Health Financing Study noted the potential importance of better bundling of inputs for measuring the productivity of public health resources. To explore that hypothesis, the staff and drug scores for the sample districts were compared. The absolute value of the difference between these two scores could be interpreted as a measure of input balance--the higher the difference the greater the imbalance between staff posts filled and drug adequacy. As shown in Figure 4.6, that measure also has a strong positive relationship with the overall performance measure While these results are exploratory and should be interpreted with caution, they suggest some interesting conclusions about the link between inputs, finance and performance. The link between inputs and performance is more consistent than that between finance and performance. This implies that the allocation of funds between various inputs may be as important as the total amount of money available, and in particular the adequacy of facility. There is also evidence that the input mix--at least between staff and drugs--is related to output levels. C. The Financial Implications of Input Gaps 4.36 Financial costs can be estimated for two of the three types of input gaps identified relative to official norms. These are for facilities and staff, and are discussed below The Financing Gap for Primary Health Care Facilities. Figure 4.1 earlier presented the facility gap in terms of numbers of facilities in place as a percentage of the desired number as expressed in GOI norms. The investment and recurrent costs of this "gap" between the existing and the completed system of rural health facilities can be estimated with the average capital cost of each type of facility, the average annual salary cost of the full complement of required staff, and an estimate of the average annual expenditure on drugs, supplies, and maintenance. This latter figure was estimated as 30% of the total annual salary cost for each facility, since norms are not available.

88 62 Figure 4.6: Relationship Between Imbalance in Inputs and Overall Performance 0 0.1' o CL A= 0-4tr * I C 0.5 E r t Overall performance score imbalance porf t (4.11A,2.825) F

89 As noted earlier, the bulk of the "missing" facilities are CHCs. These also have the highest capital cost. Thus these estimates of the finance gap associated with incomplete facility provision are dominated by the required capital investment. When compared with current levels of health spending at the district level, completing the rural delivery system would require a onetime expenditure generally equal to several times total annual government health expenditure level. The recurrent cost burden of such new facilities would average about 21 % of current spending The Financing Gap for Staffing of Existing Primary Level Facilities. The figures for staff vacancies relative to norms and the gaps between actual staff and sanctioned positions were also costed by applying the average annual salary cost for each class of personnel. This was done for CHCs and PHCs, but not for public hospitals located in districts Across all the study districts on average, the staffing gap was estimated to be equivalent to about 18% of total recurrent health spending or 33% of non-hospital salary expenditures. That is, a one-third increase in current salary spending for non-hospital services would finance the missing personnel for existing facilities at current salary rates. Of this gap, more than 60% reflects sanctioned but unfilled positions. These are positions for which budgetary resources have been allocated. The gap, net of sanctioned but unfilled posts, is equal to about 12% of current non-salary hospital expenditures. In per capita terms this averages Rs. 7.3 (about US$0.25) per person per year. As shown above with the gaps in real staff positions, the bulk of the cost would be incurred in filling posts in the more peripheral PHC facilities For the municipalities, the study also estimated the cost of filling currently sanctioned but unfilled positions. In seven of the eight areas, this equaled less than 5 % of current expenditures Estimates of State and National Level Real Input and Finance Requirements for Primary Health Care. The figures above for facilities and staff were extrapolated to state level estimates and combined in a single calculation of the financing gap for government primary level health care. The average per capita figures were used for each district in the sample, and then expanded to state level by multiplying by the population. In the absence of better estimates, the financing requirement for improving drugs and supplies at existing facilities has been estimated as a doubling of current spending levels. Table 4.1 shows these different components of the financing gap for primary health care and the state level total, with the latter compared with current levels of spending The capital investment required to meet existing norms is the dominant figure in the state level estimates. It is estimated to be Rs. 15 billion (US$484 million at current exchange rates). The weighted average total finance gap as a percentage current annual total government health and FW spending is estimated to be about 90%, varying between 57% in Gujarat and 163% in UP. If this investment expenditure were prorated over a five-year plan period, it would imply a 20% increase in government spending.

90 Looking at the recurrent cost burden, the figures are sizeable. Completing the staffing and strengthening the drug supplies at existing facilities could be accomplished for a 13 % increase in current spending levels annually. Funding the recurrent costs of the "missing" rural health facilities would add another 13.5%. A 26% increase in spending levels would be quite large at a time when states are facing acute fiscal pressures. It would raise public spending on health from about 1.5% to 1.9% of GDP. This is a more relevant long-run financial estimate, since facility construction is mainly a one-time expense in the short- to medium-term. However, when capital maintenance is factored in to these estimates the amounts of funds required would be even greater. However, this would require a more detailed analysis of the adequacy of existing facilities and maintenance needs of the current and completed network of facilities. This task would be an important contribution to completing estimates of recurrent finance needs of the public system.

91 65 Table 4.1: An Estimate of the Financing Requirement at State Level to Complete the Rural Primary Care Delivery System Gujarat Tamil Nadu Uttar Pradesh West Bengal Totl Finances required to complete facilities 606,697 2,132,585 9,051,237 3,393,423 15,183,942 (capital) Rs. 'OOOs Finances required to staff and supply new 352, ,568 1,395, ,542 2,588,985 facilities (annual recurrent)* Rs. 'OOOs Finances required to fill staff of existing 397, , ,472 43,914 1,374,692 facilities (annual recurrent) Rs. 'OOOs Finances required to double existing 89, , , , ,126 budget for drugs and supplies at primary level (annual recurrent)" Rs. '000s Total annual recurrent financing required 840, ,354 2,291, ,728 4,658,803 Rs. 'OOOs Gujarat Tamil Nadu Uttar Pradesh West Benpl Weighted Average Total financing required in per capita terns (Rs.) Total financing required as percent of 57% 78% 163% 901Y Y current annual total H&FW expenditure Capital financing as percent of annual 24% 56% 130% 74% t9.6% current total H&FW expenditure Recurrent cost financing for new facilities 24%A 10% 12% 13% 13.5% as percent of total H&FW expenditure Recurrent cost financing to complete 9% 8% 21% 3% 13.0% existing facilities as percent of total H&FW expenditure * Calculated at the average salary cost for normnative staffing level for each type of facility, plus 30% of salary cost to cover drugs and supplies and other overhead. ** Calculated by averaging per capita drug expenditure on non-hospital services over the sample districts and multiplying by the current estimate of nural population for the state.

92 66 5. PRIVATE HEALTH EXPENDITURES AND PROVISION OF PRIMARY HEALTH CARE SERVICES A. Introduction 5.1 In Chapter 3 it was noted that while the level of total health spending at about US$13 per capita was low in absolute terms, the health sector's contribution to national income in India was higher than in most developing countries at similar levels of per capita income. Despite the historical and legal emphasis on the government's role in the health sector in India, expenditure data clearly emphasize the dominance of non-government spending. Private sector expenditure in India is estimated to be about 75% of total health spending. In comparison with the other Asian countries shown in Figure 5. 1, the percentage of private to total health spending is comparable with Thailand, which has an absolute per capita spending level four times greater than that of India. Government health spending in India, on the other hand, is in the middle of the range reported for lower income Asian countries. It is higher than Indonesia and the Philippines; and lower than China, Sri Lanka and Pakistan. 5.2 A 1991 national household expenditure survey carried by the National Council for Applied Economic Research estimated that per capita household out-of-pocket spending in India was Rs 240, which is about 75% of total national health expenditure.'" Other estimates also point to the conclusion that private, and especially private out-of pocket expenditure, is of major significance in financing health care in India. 5.3 Analysis of another data source, the NSS 42nd round, collected in was also carried out by the Gujarat Institute of Development Research (GIDR) as part of this sector report." Data were analyzed from the five states of Gujarat, Maharashtra, Tamil Nadu, Uttar Pradesh, and West Bengal. The survey estimated household health care use and expenditure for a nationally representative sample, with separate estimates for inpatient treatments in hospital, ambulatory illness care treatment, and preventive and family planning services for mothers and children. It provides an opportunity to understand in more detail the allocation of private spending and its distribution in the population. Annex 3 provides a more detailed discussion of the characteristics of the NSS data and presents other supporting statistical tables. 10 This estimate was extrapolated from a single-round single season national cross-section sample to an annual figure. The estimate for private corporate spending was also based on a limited sample of companies. " The GIDR study is entitled "Utilization of and Expenditure on Health Care in India A Study of Five States." Gujarat Institute of Development Research ( Visaria, P and Gumber, A. 1994).

93 67 Figure 5.1: Health Expenditure as a Percent of GDP: Asian Countries (1990) 6 5 ' ~~~~~~~~~~~~~~~~~~~~~~~ UPrivate 2 0. S C (A 0 U *,0.,i,MM> Source: WDR Table AS

94 68 B. Private Health Expenditures and Primary Health Care 5.4 The "sources and uses" matrix shown in Table 3.1 of Chapter 3 showed that household out-of-pocket expenditure accounts for about 75 % of total national health expenditures. As a share of primary care expenditure, which is about 60% of total national health spending, the share of household out-of-pocket spending is even higher at 82%. If primary care is further separated between curative, and preventive and promotive health services, the share of household out-of-pocket spending for curative care is even higher at about 92%. For preventive and promotive services the share of out-of-pocket spending is only 27%. 5.5 Private household expenditure is also dominant for secondary and tertiary level inpatient care, although it is somewhat less than for primary care services. The share of out-ofpocket household spending for inpatient care at secondary/tertiary facilities is about 70%. 5.6 The Contribution of Out-of-Pocket Spending on Out-Patient Services. In industrialized countries, hospitalization accounts for the largest share of health expenditure and little of it is financed directly by households. In India, the pattern is reversed. About two-thirds of household out-of-pocket health spending in India is on ambulatory or outpatient services and one-third on inpatient care. This pattern is especially relevant to policy on primary health care, since private practitioners dominate in the provision of outpatient services, much of which substitutes for services which are supposed to be available through government providers as part of public sector primary health care programs. 5.7 As shown in Table 5.1, 65% and 60% of household health spending in rural and urban areas respectively goes towards ambulatory illness treatment. Extrapolating from the national estimate that 75% of total health spending is direct household spending, these figures imply that household ambulatory care spending accounts for approximately one half of total national health expenditures. 5.8 Unfortunately, there is little basis for estimating spending on primary care from private firms, or estimating household out-of-pocket spending on preventive and promotive services. Both are likely to be modest. Thus, approximating primary care expenditures with the available evidence suggests that it accounts for about 60% of total national health expenditures and that four-fifths of that expenditure is from household out-of-pocket sources. 5.9 Out-of-pocket Spending, Private Health Care, and the Poor. The burden of out-ofpocket spending falls disproportionately on the poor and that burden is mainly for primary illness care. This is shown in Table 5.2.

95 69 Table 5.1: Household Out-of-Pocket Health Expenditure and Primary Care Gujarat Maharashtra Tamil Uttar West Weighted Nadu Pradesh Bengal Average. Rural Areas Percent of out-of-pocket to 62.1% 63.6% 73.6% 58.6% 74.1% 64.9% expenditure to non-hospital treatment Percent of illness episodes to 68.9% 78.2% 70.9% 90.9% 83.6% 81.7% private providers Percent of total out-of-pocket 48% 55% 68% 51% 65% 56.40% spending to private non-hospital treatment Urban Areas Percent of out-of-pocket to 61.1% 60.0% 68.8% 57.1% 62.5% 60.8% expenditure to non-hospital treatment Percent of illness episodes to 81.5% 75.1% 68.5% 85.0% 77.7% 78.9% private providers Percent of total out-of-pocket 53% 51% 62% 46% 54% 51.58% spending to private non-hospital treatment Source: Derived from NSS 42nd round.

96 TabWe 2: Percent of Total Houseold Expenditure to Halt and to Non-mlospital Treatment by Expenditure Quintiks Gujarat Maharashira Uttar lrradesh 'Ilamnil Nadu West Bengal Weighted Averages Health Non- Health Non- I lealith Non- I lealith Non- Health Non- Health as Nonas % of Hosp. As as%of llosp. As as % of l losp. As as % of llosp. as % of Hosp. As %of Hosp. As Total % of Total % of Total % of Total As % of Total % of Total % of Exp. Health Exp. Health Exp. Hlealth Exp. Health Exp. Health Exp. Health (avg). (avg). (avg). (avg). (avg). (avg). Rural Areas III IV V Total Urban Areas IV V Total Non-Hospital (bottom 40%) as a percent of (top 40%) flujigrat Maha nhtr Utar Pradesh Tamil Nadu Wesl Bengeal Wkig b Rural Urban Source: Derived from NSS 42nd round.

97 On average, 5% and 2.3 % of total household consumer expenditure in rural and urban areas respectively were for health expenditures. In almost all cases, the percentage of household spending on health was highest in the lowest expenditure quintiles, reflecting that the burden of out-of-pocket spending was regressive and imposed a heavier burden on the poor. This trend was even stronger when household spending on ambulatory illness care was examined separately. Ambulatory care accounted for a larger portion of household health spending in the lower expenditure quintiles in both rural and urban areas Approximately half of household out-of-pocket expenditure is payment to private ambulatory care providers in the states analyzed (Table 5.1). The figures are fairly similar in both urban and rural areas, despite the fact that one might expect higher percentages in cities accompanying the greater concentrations of private providers likely to be found there. Most illness care contacts in all expenditure quintiles are with non-government providers. These contacts generally involve out-of-pocket payments about times higher than when contacted with government providers. These estimates suggest a consistent pattern that household health spending is mainly for private ambulatory care providers and that the burden of this cost falls disproportionately on the poorest households Private Expenditures on Inpatient Treatment. Overall inpatient care accounts for about one-third of total out-of-pocket spending. About 70% of in-patient care expenditure is estimated to go to non-government facilities (Table 3.1 of Chapter 3). Household health expenditure data show that hospitalization is both a smaller part of household health spending than ambulatory care, as well as having a much "flatter" distribution as a proportion of total household expenditures. This reflects several factors. Poorer households may be much less likely to seek inpatient treatment, and when they do, they are much more likely to use public facilities which require lower out-of-pocket expenditures In fact, government facilities provide approximately 60% of all episodes of inpatient treatment in India, and out-of-pocket costs are well below those in non-government facilities. Average state-level out-of-pocket costs for hospitalization in government facilities ranged from 11-74% of the cost reported for hospitalized episodes in non-government facilities Inpatient care is the predominant expenditure for other private sources of financing, such as private firms, which make payments directly for their employees or pay for private insurance. Again, these do not account for a large share of total spending. Both ESIS and CGHS are substantially financed by employer and employee contributions. ESIS maintains a sizable network of hospitals (111) and dispensaries (1,400). Estimates of the composition of ESIS and CGHS expenditures were not available for this report.

98 72 C. Private Provision of Health Services 5.15 Private Provision of Hospital Care. Although, Government facilities dominate the hospital sector, accounting for 44 % of all facilities and 71 % of all beds in national statistics there has been some shift towards private provision of hospital facilities in recent years. In the two decades since 1970 there has been rapid growth of private hospitals, whose numbers increased at an annual rate of over 13%, from about 750 in 1974 to an estimated 5,650 in There are indications that the growth of private hospitals has been more rapid since For the most part, the growth of private sector hospitals has taken the form of relatively small for-profit or trust hospitals and so-called nursing homes which have small inpatient facilities. National data shows that private hospitals include a large number of facilities with less than 10 beds which can hardly be considered hospitals in the complete sense. In 1990, the average bed strength of private hospitals in 1990 was estimated at only 31, compared to the average bed strength of 94 at public hospitals However, a recent effort to enumerate private hospital beds in Andhra Pradesh state found almost twice as many as were reported in official statistics, suggesting that we may still have an incomplete picture of the size of this sector. In recent years, the corporate sector has invested heavily in large and highly sophisticated facilities catering to the urban middle and upper class patients. Evidence suggests that the number of nursing homes, even in rural towns, has rapidly expanded. There is rapid and highly visible growth of for-profit hospitals in major urban centers It is intriguing to note the wide differences in private hospital provision across the states in India. This varies from less than 2% of all beds in Jammu and Kashmir to more than 60% in Kerala. These differences cannot be explained simply by variations in income and education levels. They probably reflect more complex historical and cultural determinants as well as response to variations in public investment patterns (Berman and Rannan-Eliya, 1993) Private Provision of Ambulatory Curative Care. Data on the total number of "doctors" or "medical practitioners" are not available in India. Government figures indicate a national average of one privately practicing physician for every 3,500 people. These figures do not include the often-illegal private practices of publicly employed doctors, although these may not be as widespread as in other countries. Official data provide information only on the qualified, allopathic practitioners -- the MBBS or MD physician. "Registered" or "licensed" medical practitioners also exist with lower qualifications. But there are no numbers available and in most states registration or licensing has not functioned for almost two decades.

99 Part of the reason why data on the number of doctors is unavailable is the longstanding uncertainty about what constitutes a "doctor" or "practitioner" and the quantities and types of practitioners available. Private providers include a wide array of qualified, less-thanqualified, and unqualified practitioners. Most of them practice an eclectic form of medical care, combining allopathic and one or several "traditional" forms of medical care. Most are unregistered, unlicensed, and unregulated, although there are numerous gradations of legal practice which vary from state to state There has been little change in the percent of physicians in full private physician practice over time. For the other components of private practice, information is very limited. A recent study of the "rural private practitioner" in Uttar Pradesh estimated that there may be as many as 1.25 million practitioners in India, or one for every 600 people in rural areas (IMRB, 1993). While this is very tentative estimate, it highlights a fact that is obvious to a careful observer, i.e., there is widespread access to private, fee-for-service health care throughout India, although the quality of that care is likely to be very low in rural areas. It may be that such practitioners are even more numerous in urban areas, serving the growing numbers of the urban poor The IMRB used an innovative technique to identify private practitioners -- asking village respondents to name the "doctors" they knew, then interviewing all those mentioned. The survey shows them to be young (under 40), male, and modestly educated. About half of the practitioners had received some formal health care training, but in most cases this was inadequate for a full qualification. They had on average more than 10 years of practical experience, saw 11 patients a day, and more than 90% of them used clinical methods from allopathy and at least one other system of medicine. The survey found these practitioners doing "fee-for-service" treatmnent, although they reported mainly charging for medicines plus a profit margin and not for "consultations." They both prescribed and dispensed medicines Most studies of health care utilization and spending in India have not adequately investigated the role of these private practitioners. The National Sample Survey results for private care provision reported in this chapter are a typical example of this problem, coding responses for "private doctor" without distinguishing between qualified and unqualified practitioners. More careful micro studies have found that these "less than fully qualified" practitioners account for the vast majority of treatment contacts and expenditures in both urban and rural areas Other Tvpes of Providers. While little is known about the typical private practitioner, there may be other sources of private medical care even less well documented. Drug sellers and pharmacists commonly diagnose and prescribe as well as dispense drugs. It is unclear to what extent they are included in the response "private doctor" on surveys. In some parts of India, private diagnostic facilities such as radiology and laboratory testing are now appearing even in small rural towns. It is not known to what extent these are also increasingly functioning as private treatment facilities.

100 The main issue with private practicing doctors in India is that there is a wide array of qualified and less-than-qualified practitioners. Most are unregistered, unlicensed and unregulated. These private practitioners are often perceived by the medical profession to be providing care of poor technical quality, sometimes endangering their patients, and to be exploiting the poverty and ignorance of their patients.

101 A. Introduction PUBLIC AND PRIVATE SECTOR ROLES IN MEETING THE DEMAND FOR HEALTH CARE 6.1 Public and private sector provision of health care influence each other, potentially as both substitutes and complements. The NSS, 42nd round carried out in , included detailed investigations of health care use and expenditures for hospitalization, non-hospital illness care, and some preventive/public health services for a representative national and state level probability sample. The sector report has undertaken an extensive analysis of the NSS data from Gujarat, Tamil Nadu, Uttar Pradesh, West Bengal and Maharashtra.' Results are presented in two major sections. The first, dealing with ambulatory and preventive care; and the second with hospital-based care. The quantity and type of illness events are described, as are health care utilization and spending patterns for both public and private sector provision.' The analysis shows that there are significant differences between the states. Of the five states studied, the public system in U.P. seems the least successful. Household spending on hospital care is high and there is little differentiation by expenditure class or public and private facilities. For ambulatory treatment, private providers dominate the market and actual expenditures in public facilities are also high in comparison to the private sector. In contrast, neighboring West Bengal has little private hospital provision and low cost public services which differentially benefit the poor. However, the primary care system in West Bengal does not seem as successful as that of other states such as Tamil Nadu. Some of these major differences can be associated with state-level spending decisions, both in terms of the total amount spent as well as the allocation of spending between types of facilities. 6.4 A summary of the key findings relating to public and private sector roles in meeting health care demand is presented in Box 2. This is a synthesis of the information analyzed earlier in Chapters 3, 4 and 5 and provides the basis for the discussion on public-private partnership which is the focus of this chapter. The chapter concludes with a discussion of the linkages between ambulatory care, preventive services, and hospital treatment and the implications of this analysis for health policy for India. 12 This chapter is based on a report which was prepared by the Gujarat Institute of Development Research as a contribution to this sector study. The study is entitled 'Utilization of and Expenditure on Health Care in India, A Study of Five States." (Visaria, P. and A. Gumber, 1994). 13 Annex 3 provides extensive statistical tables of the survey results, along with a description and discussion of the survey design.

102 76 Box 2: Key Findings on Public-Private Sector Roles in Meeting Health Care Demand * Government provided services are the major source of inpatient care in all the states studied. In contrast, non-government providers -- mainly for-profit, fee-for-service practitioners, provide the bulk of outpatient and ambulatory care, the curative care component of primary health care. Although the NSS data do not provide details identifying these private practitioners, other studies suggest that most of them are inot fully qualified, andh that they practice an eclectic form of treatment with little quality control or regulation. * E t 0 Government providers are the major source of preventive care in rural areas, although coverage remains low overall with several dimensions of routine preventive care for mothers and children. In urban areas, coverage is higher, and there is a larger private sector role. One important exception has been the dramatic increases in immunization coverage achieved by government provision. * Private primary level treatment is the dominant source of care for a number of diseases targeted by public disease control programs in both rural and urban areas. Tuberculosis has the highest level of government coverage for the illnesses identified and that reached only 50 percent of reported cases. *;; tit: Government primary care services do not appear to be well targeted to the poor. Curative services:are not:highly subsidized. In contrast, public hospitals provide real subsidies to the poor, although these subsidies are also not always very well targeted to the poor only. The public hospital system: seems to be the better perfborming part of the public delivery system, at least, as far ias clinical care in concerned. * ; There are significant linkages between hospital and primary care providers -- the two components of the medical care system -- and between different types of hospitals. This report has not distinguished between different types of secondary and tertiary level hospitals and acknowledges their varying equity focus. These linkages have important implications as well for the efficiency and effectiveness of the health care. system in India. These include: (i) One third to a half of all hospital admissions do not receive any prior treatment. Those that do mainly use private practitioners. The poor performance of the public primary care system and the lack of involvement with private, practitioners is likely to be associated with inefficient and inappropriate use of public hospitals; (ii) Diseases for which mass public health outreach programs exist such as TB control, malaria control, diarrheal disease control and safe motherhood/mch make up a large part of the caseload of public hospitals. Most of these public health programs have low coverage in terms of finding and: treating patients on an outpatient basis. Even publicly funded outpatient TB treatment reaches only half of those reporting the disease, and coverage ffor other diseases is much lower. Improving the coverage and effectiveness of public health outreach programs could both reduce the demand for public hospital services and improving outcomes, efficiency, and equity. However, cutting back on public hospital provision without improving services at the primary level would probably further disadvantage the poor in terms of both health and cost; and (iii) Despite public subsidies for hospital care, out-of-pocket expenses for the serious illnesses are large anrd have an impact on the poor. In a serious illness episode, families might pay fully for private ambulatory care, then go to a public hospital where they mnight receive a firee or highly subsidized day charge but still pay for other services as well as for items not availablelat a public hospital. After discharge they may again pay fully for private followup treatment. The. total costs of treatment are much higher if private:hospitals are used. ::::..:..:: : ::.:: :.*.-....

103 77 B. Outpatient Treatment of Illness 6.5 Use of Government and Non-Government Health Care for Ambulatory Treatment of Illness Outside of Hospital. Government providers include public hospital outpatient departments, and various types of public clinics such as CHCs and PHCs, as well as services provided by non-profit charitable facilities."4 Non-government providers include private doctors, nursing homes, and private hospital out-patient departments. It is not possible to distinguish between qualified and unqualified private doctors -- all were referred to as "private doctor" -- nor were multiple responses allowed for a single illness episode. Use of purely traditional practitioners was reported to be low. 6.6 Figure 6.1 presents the percentage breakdown for the five states between government and non-government providers. These figures exclude cases for which no treatment was sought. 6.7 In all the states more than two-thirds of treated illness episodes in rural areas were taken to non-government providers, mainly private fee-for-service practitioners and hospitals. The highest proportion of private service use was in Uttar Pradesh, the state with the lowest public sector health expenditure. 6.8 In urban areas, the proportion of episodes reporting non-government service use was slightly lower than the rural figure in four of the five states, the exception being Gujarat. 6.9 The pattern of use of government and non-government treatment sources by age and sex is also given in Table 6.1. In almost all states and in both urban and rural areas there is a slightly higher probability of children under five years of age using non-government providers for treatment. This is contrary to what one would expect from the emphasis of public programs. In the case of use patterns by sex, there is little difference in most states except Tamil Nadu, where females have a much lower propensity to use private providers in rural areas. In urban areas, females have only a marginally higher rate of utilization in all states except Gujarat, where the female utilization rate of private providers is significantly higher than male The demand for non-governmentreatment has been reported to have a positive income elasticity which increase with income levels. This would suggest that the proportion of cases taken to private providers should increase steadily with household resource levels. Figure 6.2a and 6.2b present the patterns of use for five quintiles of monthly per capita household consumption expenditure. The upper part of the table shows the proportion of episodes reported from each expenditure class taken to government and non-government providers. The lower part shows the composition of patients at public and private providers by expenditure class. 14 Charitable facilities were included with government in these calculations, although technically these are non-government organizations, often receiving public subsidies. They account for a very small part of total utilization.

104 78 Table 6.1: Differences by Age and Sex in the Probability of Use of Non-Government Providers for Non-Hospitalized Illness Treatment (percent of all treated episodes) Gujarat Maharashtra Tamil Nadu Uttar West Weighted Pradesh Bengal Average (by pop.) Rural Areas < > Males Females Urban Areas < > Males Females Source: Derived from NSS 42nd round.

105 79 Figure 6.1: Treatment of Illness Outside of Hospital by Provider Type Rural Areas _on-govt Non-Govt. 70 * Govt Gujarat Maharashtra Tamil Nadu Uttar Pradesh West Bengal Urban Areas 10 - _ 80 0 E Non-Govt.~ 70 * Govt.~ 60-- LU ~ _ Gujarat Maharashtra Tam il Nadu Uttar Pradesh West Bengal

106 In rural areas, Tamil Nadu and West Bengal show a significant gradient of increasing demand for non-government treatment with income, while Gujarat, UP and Maharashtra do not. In urban areas, Maharashtra shows this positive effect with consumption expenditure Figure 6.2b presents the pattern of utilization in terms of the mix of patients seen by different types of providers. Non-government providers do not predominantly serve higher income patients. Overall, there is some evidence of a bias in non-government care towards the better-off, and a surprising degree of use of non-government providers by the lowest quintiles in the income distribution The pattern of service use between government and non-government providers for ambulatory care can be broken down by specific self-reported diagnoses, although there is no way of clinically verifying these diagnoses. These results are presented in Annex 3 for the main groups of diseases in ICD-9, with some further breakdowns for the largest category, i.e., infectious and communicable diseases In rural areas in all states, the vast majority of infectious (category I)" diseases, which should be a major focus of public services, are treated by non-government providers. One exception is TB, where the proportion of cases treated by non-government sources ranges from 32.3% in Maharashtra to 58.1% in U.P. For malaria and dysentery, both diseases which are the subjects of national control programs, all states reported more than 60% of cases and sometimes as high as 95 % of cases treated by non-govermnent providers In urban areas, the same picture emerges for malaria and dysentery. However, the treatment patterns for TB are even more varied, ranging from 22% non-governmentreatment in West Bengal to 58 and 60% U.P. and Gujarat The pattern of use of government and non-govermnent services for specific diseases and expenditure quintiles is given in Table 6.2. TB, malaria, and dysentery are presented as examples of diseases with large public control programs, although other diseases can also be analyzed in a similar fashion. For malaria and dysentery treatrnent, non-government services are dominant for almost all expenditure classes in both rural and urban areas. As in previous tables, the picture for treatrnent is more complex. In Gujarat, Maharashtra, and West Bengal, public services are more successful in focusing on the poor for TB treatment than in U.P. and Tamil Nadu. 5 The Government of India classifies diseases in Categories I-XIII. Category I includes dysentery, typhoid, TB and malaria.

107 81 Figure 6.2a: Use of Non-Government Services by Different Expenditure Quintiles Rural Areas: Use of Non-govemment Services by Different Expenditure J Q~~~~~~~~~Cuintiles 100 go Gujaat Mahehraslr ~---- Uttw Pradesh 40 Tamd Nad I 30 - vst We Bnal 20 V li 10 o 1 _! _I _._I Expenditure Quintiles Urban Areas: Use of Non-government Services by Different Expenditure Quinties 100-7Guirat _- _- _ - Maharaahtra Uttw Pradsh TamE Nadu West Bengal I w 10 I II: j ii III IV V Expenditure Quintibs

108 82 Figure 6.2b: Government and Non-Government Providers Rural Areas, Government Providers Rural Areas, Non-Government Providers I 60t 601 Z 2~~~~~~~~~~~~~~~~~~~o~ 20r 2 20 E ~~~~~~~~~~E 0 0 I 1 I II1 IV V I 11 III IV V Expenditure Quintiles Expenditure Quintiles I~~~~~~~~~~~~~~~~~~ Urban Areas, Government Providers o Urban Areas, Non- Government Providers 100~ ~~~~~~~~~~~~~~~~0 8Ot ~~~~~~~~~~~~~ 20.9~ ~~04 I II ~III IV V I II III IV V Expenditure Quintiles Expenditure Quintiles

109 83 Table 6.2: Public/Private Services, Poverty and Specific Diseases (row percentage) ts',jart Mabazsahtra Uttar Pradesh Tamil Nad West Bengal G NG G NG G NG G NG G NG Rural Areas TB All II III IV V Malaria All I III IV V Dysentery All I I III IV V Note: I is lowest expenditure class, V is highest

110 84 Table 6.2: Public/Private Services, Poverty and Specific Diseases (row percentage)(cont'd) G1Aj Maharashtra Uttar Pradesh Tamil Nadu West B ennal G NG G NG G NG G NG G NG Urban Areas TB All II III IV V Malaria All II III IV V Dysentry All II III IV V Note: I is lowest expenditure class, V is highest. Source: Derived from NSS 42nd round.

111 Thus, private providers dominate ambulatory illness care. The preeminent role extends to many of those services, specific diseases, and population groups thought to be the main emphasis of public provision. One possible exception to this is TB, where public and private provision appear to share almost equally the case burden of the population. There is evidence, from the majority of states sampled, of bias towards the poor in public service provision. However, in the case of TB, with a major national program offering free treatment, one might be surprised that even half the cases seek non-government treatment and that a sizeable number of these are from the poorer population, especially in U.P., the most populous state Household Expenditure on Ambulatory Illness Care. Chapter 3, 4 and 5 discussed some of the findings related to aggregate household health expenditure patterns on illness care and hospitalization. Additional information is available on the total reported expenditure per episode as well as the amount spent directly on the provider or "source of treatment".' The cost to users of health care by public and private providers. User expenditure on treatment is estimated by episode of illness and averaged by type of provider. The survey did not record multiple provider use for a single episode. Table 6.3 presents a breakdown for government and non-government providers of the average expenditure per episode in different expenditure classes in both urban and rural areas These data provide some striking insights into both the overall costs and distributional characteristics of both public and private health care. For rural areas, there are large differences in the average cost of an illness episode, with Tamil Nadu and West Bengal reporting much lower costs for both public and private treatment than most of the other states, and U.P. and Gujarat reporting higher costs. In all states except U.P., the overall average cost of illness treatment taken to a public sector provider is well below that for a private provider. In rural U.P., the area with the lowest public expenditure and coverage levels, government services were reported to be more costly than private. Excluding U.P., public sector ambulatory care costs ranged from about 20 to 70% of the cost of treatment at a non-government provider There is evidence of a weak positive link between overall household consumption expenditure levels and the average expenditure per treatment in public services, but this also is apparent to some extent for non-government services. Further evidence on income elasticities is presented in the attachments to this chapter. 16 Travel costs and wages foregone was explicitly not included in this analysis and costs for drugs may or may not have been included.

112 86 Table 6.3: User Expenditure for Illness Treatment by Expenditure Quintile (average Rs. expenditure per episode) C1arai MaharachtUattar Prdesh Tanil Nadu West Bengal G NG G NG G NG G NG G NG Rural Areas Expenditure Quintile I II III IV V Avg/All Urban Areas Expenditure Quintile i III IV V Avg/All Note: I is lowest expenditure quintile, V is highest. Source: Derived from NSS 42nd round.

113 In urban areas, cost per episode was generally higher for both public and private providers than in rural areas. The average expenditure per episode for government services was lower than non-government in all the states, ranging from 24 to 67 % of the private treatment cost levels. Looking at the bottom three expenditure quintiles, treatment expenditure levels show no consistent positive relationship with total expenditure, although much more is being spent, on average, in the upper quintiles than in the lower ones Is ambulatory care affordable? Table 6.4 summarizes the cost per treatment episode as a percent of monthly household consumption expenditure by expenditure class. For government services, expenditure on illness care was almost always under 10% of monthly total household consumption spending, with the exception again of U.P. In U.P. the high cost reported for public service use means that % of monthly household expenditure can go to one episode of routine illness care -- a high figure. The percentage of household expenditures spent on government care is fairly constant across expenditure quintiles, neither redistributive nor regressive with regard to household economic status For private treatment, health expenditures may be 50% or more greater than the level of public treatment costs. However, in some states, there is not such a large difference. In Gujarat, expenditure per episode for private treatment can be a quarter or more of monthly household spending. In most of the other areas it ranged between 10 and 20% Access to free care. Ambulatory care provided through CHCs and PHCs is generally provided without charge. In some locations, nominal charges are made at government dispensaries and hospital outpatient departments. Charges may be levied for medicines, supplies, and diagnostic tests of various kinds Most patients pay something for treatment consultations. Table 6.5 summarizes the results on use of free care. For ambulatory treatment, most of the free care at least involved the provision of free medicines. About a fifth of patients reported receipt of free drugs. Typically more than 80 % of them received this from government providers Table 6.5 shows the proportion of respondents in each expenditure class who received free medicine. In Tamil Nadu and West Bengal, the poor have benefitted from the free provision of drugs in both rural and urban areas. This is also true of urban Maharashtra. Elsewhere, however, this subsidy has been less successful in targeting the poor.

114 88 Table 6.4: Cost per Treated Episode as a Percent of Monthly Household Comsumption Expenditure (users only) XiaBt Maharashtra Uatt Tamil Nau WestBengal Pradesh G NG G NG G NG G NG G NG Rural Areas I II III IV V All Urban Areas I II in IV v All Note: MTHE = mean total household expenditures. I is lowest expenditure class, V is highest. Source: Derived from NSS 42nd round.

115 89 Table 6.S: Access to Free Care for Illness Treatnent Gujarat Maha- Uttar Tamil West rashtra Pradesh Nadu Bengal Rural Areas % received free medicine % of those w/free medicine who went to govt. service Urban Areas % received free medicine % of those w/free medicine who went to govt. service % of free medicine by SES Rural Areas I II III IV V Urban Areas I II III IV V Note: I is lowest expenditure quintile, V is highest. Source: Derived from NSS 42nd round.

116 The role of insurance. Insurance coverage is very limited in India. Most of it is available through the Employees State Insurance Scheme (ESIS), which provides mandatory coverage for staff of formal sector enterprises with more than employees." 7 For rural and urban areas, insurance coverage, including ESIS and other types of insurance such as employer payment was highest in Gujarat, at about 5 % (rural) and 12 % (urban). Tamil Nadu and U.P. had the lowest levels for the rural population and U.P. the lowest levels for the urban areas. Coverage was highest for those in the upper expenditure quintiles, as expected Table 6.6 compares the utilization patterns of those who are insured with the entire population. The insured population was much more likely to use government services than the general population in all states and in rural and urban areas. This may represent a subsidy to a relatively better off population through their use of public facilities. However, it is not clear to what extent insured respondents would report the use of an ESIS-owned facility as "government". This is because these facilities are run as private hospitals, although they benefit from government subsidies Treatment cost and specific tvpes of illness. The household expenditure per episode on treating specific types of self-reported illness is given in Annex 3. Some of these cells are based on small numbers of reported cases, so the levels and differences should be interpreted cautiously. For infectious diseases, higher costs are incurred in treating tuberculosis than most other types of illness and this holds true for both government and non-government providers. Reported TB treatment in rural areas ranges from a low of Rs. 60 per episode in Tamil Nadu to Rs. 258 in Uttar Pradesh and in urban areas from Rs. 24 in Tamil Nadu to Rs. 204 in U.P Across the broad categories of diseases reported, the differences in per episode expenditures are not very large. Nor is there large and consistent difference between the cost of public and private providers. C. Preventive Care 6.32 The line between "curative" and "preventive" care cannot always be drawn precisely. For example, the role of private practitioners in treating several communicable diseases for which vertical government programs exist was highlighted in the preceding section. Some of these public health programs are mainly curative. TB, malaria, and STD control programs include major components of case-finding, diagnosis and treatment. The "preventive" aspect is in terms of secondary prevention and reducing the spread of infection. For the diseases addressed by such programs, the dominance of private providers in treatment is evidence of the limited coverage achieved by public programs. 7 Employees with wages under Rs. 1,600 per month are covered; 10 employees or more in establishments using power; 20 employees or more in certain other kinds of establishments.

117 91 Table 6.6: Mllness Treated by Non-Government Providers Outside of Hospital for Total Population and Insured Population (percent of all episodes treated) Gujarat Maharashtra Tamil Nadu Uttar West Pradesh Bengal Rural Areas Total population Insured population only Urban Areas Total population Insured population only Source: Derived from NSS 42nd round.

118 Other public interventions are more uniquely preventive in orientation. These include maternal and childhood immunizations, routine screening and education contacts for pregnancy, the post-partum period, and early childhood; and vector control interventions in the community. Provision of trained birth attendance, or access to childbirth in hospital may also be included in this list National survey data provide some evidence of the coverage and sources of care for many such services. The NSS 42nd round estimated immunization coverage, but the figures represent the early years of the national Universal Immunization Program, a period of rapidly expanding coverage, and they are unlikely to be an accurate reflection of current coverage levels. Public providers were and still are the major source of immunization, especially in rural areas Government providers are by far the dominant source of routine preventive care contacts in rural areas and generally account for more than half those reporting such care in urban areas (Table 6.7). One exception is the use of routine post-natal care in cities, where on average private providers account for about two-thirds of all those reporting such services. However, the overall levels of coverage for these interventions is generally low in rural areas Estimates are also available on the role of public providers in birth attendance. As shown in Table 6.7, most births are at home in rural areas, while most are in hospitals and nursing homes in urban areas. The role of government personnel, excluding trained traditional birth attendants in home deliveries, is very limited in both rural and urban areas. In contrast, government hospitals are the dominant site for facility-based births, mainly in cities. D. Treatment in Hospital 6.37 The linkages between public and private hospitals as sources of illness care and their implications for equity, efficiency, and health impact, are important issues for public policy. The NSS collected data from a random sample of households report, at least, one hospitalization during the year preceding the survey. Unlike the outpatient care sample, which excluded households who reported a hospitalization in the initial enumeration and so is not an unbiased sample of those who were ill but not hospitalized, the hospital care sample is a true random sample of the rural and urban households which experienced a hospitalization in the last year.

119 93 Table 6.7: Role of Government Providers in Routine Preventive Care Gujarat Maharashtra Uttar Pradesh Tamil Nadu West Bengal Weighted Average Routine Pediatric Care Rural % Registered % Government Urban % Registered % Government Antenatal Care Rural % Registered % Government Urban % Registered % Government Post-natal Care Rural % Registered % Government Urban % Registered % Government Source: National Sample Survey Organization, "Child and Maternity Care." 42nd Round, July 1986-June 1987, Number 368, Department of Statistics, New Delhi.

120 Use of Government and Non-Government Hospital Services. In sharp contrast to the picture for ambulatory care, inpatient treatments are predominantly in government facilities in both rural and urban areas (see Figure 6.3). Only in Maharashtra does use of private hospitals exceed that of government facilities. In the other states, excluding West Bengal, public hospitals account for about 60% of all hospitalizations. This contrasts with the more than two thirds of cases taken to private practitioners for ambulatory illness care. In West Bengal, 92% of rural and 76% of urban hospitalizations are in public facilities. Hospital care remains largely a function of government in most of India Public sector hospital care also seems well targeted to the poor, and this is more true in rural areas than in urban areas. Figure 6.4 shows the proportions of hospitalized cases in different expenditure quintiles that used non-government facilities. In rural areas there is a consistent pattern of higher probability of public facility use in the lower quintiles. In the urban samples, this is also the case, with the exception of U.P., where the bottom quintile patients had the lowest probability of using government hospitals Do patients come directly to public hospitals or do they make use of a functioning referral system? The low level of use of public outpatient facilities makes this an important question. For both rural and urban areas, between one-half and two-thirds of hospitalized cases had some clinical consultation prior to hospitalization. This means that a third to a half of all hospitalizations experience no clinical referral. This is a striking failure of both the public and private primary care system For those who have a consultation prior to hospitalization, most do so with a private practitioner. This is to be expected given the findings presented earlier about the low level of use of public sector ambulatory care The main conclusion is that, private providers are the major source of initial consultation for those using public hospitals. This important link between ambulatory and hospital care and public and private provision has received little attention Illnesses Treated at Public and Private Hospitals. The self-reported diseases that comprise hospitalized cases are reported in Annex 3 for all cases and for government and nongovernment facilities. Infections comprise the largest category of illnesses treated in hospital in all states and for the urban and rural populations, making up about 25% of all cases for rural respondents and 20% for urban. Within the infection category, TB and dysentery are the most frequent diseases reported.

121 95 Figure 6.3: Hospital Treatment of Illness: Roles of Governmnent and Non-Government Facilities Non Govt. * 30 20f 10 ~ ~~~~~ 40 ~~~~~~~~~~~~Govt ~~~~ < 80 ~ * a~~~~~~~zc C - ~~~~~~~~E Urban Areas

122 96 Figure 6.4: Non-Government Hospital Use by Different Expenditure Quintiles Rural Aras: Non-Govemment Hospital Use Pattenms by Different Expenditure Quintiles 100 j ~~~~~~~~~~~~~~~Gujarat ~~~~~~~~~~~Maharashtra Uttar Pradesh Tamil Nadu * ~ - -West Bengal lv V Expenditure Quintiles Urban Area: Non-Govemment Hospital Use Pattems by Different Expenditure Quintiles 100-3go 100 r Gujarat 170 ff, Maharashtra 50 -Uttar Pradesh 40 ~ ~ ~ ~ ~ ~ ~ ~~~~ Tamil Nadu 30 ~~~~~~~~~~~~~--.West Bengal 10 0 I i IV V Expenditure Quinties

123 Both government and private hospitals predominantly treat infectious disease (category I) cases. Hospitalization for TB cases mainly takes place in public facilities for both the rural and urban populations. Mental and nervous disorders (category V and VI)" are a significant component of public hospital activity in rural areas and this type of ailment tends to be brought to government facilities. There is some evidence that hospitalization for pregnancy and childbirth (category XI)`9 favor non-government facilities. Overall, however, strong patterns of segmentation of types of ailments between public and private hospitals across the states or rural/urban division do not exist. When these specific disease categories for which there are major public health programs are disaggregated by expenditure quintiles, no strong distributional patterns appear, although sample sizes are small in each cell Expenditure on Public and Private Hospital Care by Households. Hospital stays typically result from more severe illnesses and require higher levels of expenditures for an episode of illness. However, they are also much rarer events. The lower burden of hospital expenditure on the poorer classes reported in Chapter 3 reflects the successful insurance function of public provision of hospital care. However, it is difficult to determine the extent to which the poor spend less because they simply cannot afford the direct and indirect costs of hospitalization Nonetheless, hospital-based treatment is costly to households. Table 6.8 shows the actual rupee expenditure on an episode of illness requiring hospitalization for each expenditure class excluding time and travel expenses. These data are actual expenditures, already reflecting household decisions on whether to use hospital care and how much of which kind of care to use. They also do not control for other factors, such as severity of illness (see multivariate analysis included as attachments to this chapter). Data suggests that a real financial subsidy to the poor from more extensive goverrnent provision of hospital care are being provided In most states, on average, a hospitalized illness costs the equivalent of a month's household consumption expenditure or more. Private hospital care clearly places a higher burden on households -- in most cases more than double that of public care. In some states, the burden on poor households is much greater than on the better-off. This is particularly visible in the rural and urban areas of U.P. and to some degree in Gujarat and Maharashtra. In U.P., the rupee expenditure on hospitalization in public facilities is sometimes greater than on private facilities and in most cases is not that different. This is quite different from the situation reported for the other states, where those hospitalized in public facilities reported much lower levels of spending per episode. In contrast, in West Bengal, with little private hospital care available, the burden on the poorer classes is clearly lower than in other states and lower than the upper expenditure classes in the state. " Government of India classification: Category V - Mental; Category VI - Nervous. '9 Government of India classification: Category XI - Pregnancy.

124 98 Table 6.8: Expenditure per Episode for Hospitalization by Expenditure Quintile and Type of Provider (Rs. per episode) GuAiarat Maharashtra Uttar Eradesh Tamil Nadu West Rengal Exp. Quintile G NG G NG G NG G NG G NG Rural Areas I III IV V Average All Urban Areas III IV V Average All Note: I is lowest expenditure class, V is highest. Source: Derived from NSS 42nd round.

125 Payment to providers and use of free care. Payment for hospital care can be divided into the proportion paid to the provider and other additional payments. Although expenditure on hospitalization in public facilities is clearly less costly in total to households, the provision of free or highly subsidized public hospital services accounts for less than a third of household spending per episode in rural areas. In contrast, when private facilities are used, payment to providers makes up more than two thirds of the total cost. The same pattern is found in urban areas Hospital charges include per diem fees and fees for specific services such as medicines, diagnostic tests, surgery, etc. Provision of a "free bed" was reported most frequently. With the exception of higher figures in West Bengal, about 40-60% of inpatients benefitted from free beds and virtually all of these were in public hospitals. When access to free beds is calculated for expenditure quintiles, in almost all cases there is evidence of this subsidy being directed successfully towards the poor (see Table 6.9). However, in some states less than 60% of the poorest patients in public facilities benefit. In U.P. and in rural areas of Tamil Nadu and West Bengal, there is also substantial subsidy to the higher expenditure classes and the degree of targeting is modest Expenditure on hospitalization and specific diseases. The rupee expenditure per hospitalized episode for specific self-reported disease categories shows that the range of costs per episode for public facilities is much smaller than that reported for private hospitals. The public subsidy does operate across the board for a wide variety of conditions in both rural and urban areas. Expenditure on treating infectious diseases in public hospitals is relatively modest compared with private facilities, especially in urban areas. For example, private hospitalization of a case of TB in urban areas can cost from two to ten times as much as in public facilities. Unfortunately, it is not possible from the available data to fully interpret these differences, since we have no information on the quality and outcomes of treatments at different types of hospitals The presence of insurance coverage generally reduces expenditures more significantly in urban than in rural areas. Gender and age do not show consistent or strong effects on expenditure.

126 100 Table 6.9: Access to Free Hospital Beds Gujarat Maharashtra Uttar Tamil West Pradesh Nadu Bengal % Receiving Free Bed Rural Areas Urban Areas % with Free Bed Getting Govt. Health Care Rural Areas Urban Areas % with Free Care by SES Rural Areas X II [II IV V Urban Areas II III IV V Note: I is lowest expenditure class, V is highest. Source: Derived from NSS 42nd round.

127 SUMMARY OF MAIN CONCLUSIONS AND RECOMMENDATIONS FOR ACTION A. Introduction 7.1 India has achieved significant improvements in the health status of its population and developed an extensive basic health care infrastructure, especially during the last decade. Life expectancy has steadily increased, from 50 years in 1960 to an estimated 61 years in Infant and under-five child mortality rates are estimated to be 79 and 106, respectively. The health infrastructure is progressing toward, inter alia, a community health center (CHC) for 100,000 population, a primary health center (PHC) for 30,000, a sub-center (SUBC) for 5,000, and a village health guide (VHG) for 1,000. In 1991, India had: 130,983 SUBCs; 22,065 PHCs; and 1,932 CHCs. 7.2 Despite this progress, the government's overall goal of improving the health status of its people has been achieved only to a limited extent. Coverage of preventive and promotive health services, mainly provided by the government, has been very low (20%), with the exception of immunization. The public primary health care system, especially its preventive and promotive services, is overextended in terms of its capacity to provide and maintain services at peripheral facilities; and the system is underfinanced for the range of services it is expected to provide. It lacks essential inputs for existing facilities, needs to finance new ones, and faces new health problems, such as the rising incidence of HIV infection. As a result, the primary health care system is unable to cope with the present demands being placed on it and is ill-prepared to deal with likely future needs. 7.3 Two issues are important when considering further development and financing of primary health care in India. The first concerns the amount of funds which the public sector will make available for health in India. The second concerns the roles of the public and private sectors in the provision and financing of health care. 7.4 Financing Needs. The challenge for the Government is immense if it is to achieve its stated objectives of reducing mortality, morbidity, population growth and fertility rates. These are reasonable and achievable objectives, but the reality of achieving these goals means that the government will need to reallocate resources within the health sector and find additional resources from elsewhere. If additional resources are not available, especially for preventive and promotive services, the government will need to redefine its priorities in the health sector and be much less optimistic about its objectives and goals. For example, if the government were to achieve its objective of providing coverage of basic health services to a large section of the population, it would need to: (i) allocate additional resources on primary health care on the basis of WDR estimated requirements in low income countries to meet the recommendations of essential public health and clinical packages; or (ii) allocate additional funds for expanding primary health care coverage on the basis of the government's own norms. 7.5 Providing the WDR package of essential public health and clinical services would cost about 3.6% of GDP on the basis of US$12 per capita investment, which is the amount estimated by the WDR to provide these services in low income countries. This compares to the 0.65% of GDP (US$2.5 per capita) the government currently spends on primary health care, the level estimated as necessary to meet most of the operations and maintenance expenses in

128 102 accordance with the norms set by the Government. Even if these services are provided somewhat less expensively in India, it would still require substantial additional funds. Alternatively, if the govermnent were to achieve its objectives of providing minimal essential services in accordance with its facility and service norms, expenditures on primary health care services would need to be increased from its current level of 0.65% to at least 1.0% of GDP. There is a strong argument for providing at least this level of additional funds for primary health care services to make optimal use of the physical infrastructure that has already been built. 7.6 Public/Private Roles. In addition to a key role in the provision of preventive and promotive services, the Government can ensure that private health care provision which accounts for 75% of total expenditures can play a very important role in the health sector overall. A major recommendation of this sector report is that a strategy needs to be developed by the Government that takes into consideration the existing levels of private provision of services. The Government's private health care strategy should be to encourage the private sector to continue to play the important role that it is already playing in the delivery of health care services, contribute more to preventive and promotive care services, and monitor and regulate private care provision such as licensing and certification. 7.7 Government would need to target its own resources to where critical gaps exist so as to increase the return from the substantial private spending on health services and to reduce the burden of out-of-pocket spending on the poor. Those areas of primary health care which are already being provided by the private sector, by for which there are excess demand and positive externalities, could be supported by selective increments in public provision. Such areas include treatment of acute illness and routine curative care for priority diseases of children and adults. The focus of public spending, however, would be in areas where private investment is negligible such as preventive health care, infectious disease control and limited clinical interventions providing inpatient treatments. This might also mean scaling back on some current public investments, such as public support of medical education, tertiary care and government subsidized insurance schemes, and allowing the private sector to play a greater role in those areas. 7.8 The major issues relating to the policy and finance strategies for strengthening primary care services are synthesized in this chapter and specific recommendations are made on the actions needed to be undertaken by the central Government to improve effectiveness and efficiency of primary health care in India. A number of actions that need to be undertaken by state Governments are also discussed. The following section summarizes and highlights the issues and actions discussed in the earlier chapters. B. Increase Government Expenditure on PrimarEy Health Care 7.9 While total health spending in India is 6% of GDP, the percentage of GDP spent by government on health is only 1.3%. The share of government health spending on primary health care services is 43% or about 0.65% of GDP (about US$2.5 per capita). This is clearly inadequate to meet the primary health care needs as defined by Government of India norms and the WDR recommendation of an essential clinical and public health package Recommendations for Action. In the light of fiscal realities faced by central and state governments, it would be very difficult to obtain sizeable increases in health expenditures overall. The government must, however, redirect or increase spending on primary health care,

129 103 especially preventive and promotive services. This could be achieved by: (i) redirecting funds from tertiary hospitals to primary and secondary health care services, particularly its preventive and promotive aspects; (ii) sub-stituting public funds with private funds in secondary and tertiary hospitals by instituting means-tested user-charges; (iii) implementing full cost recovery from private and government-subsidized insurance as well as by enhancing non-tax revenues; (iv) reducing public subsidies for medical education; and (v) increasing the overall central and state health budgets by taking into consideration the Government's stated public health priorities and fiscal constraints. At the least, expenditures on preventive and promotive services should be protected from fiscal cuts engendered by the stabilization program. C. Improve the Structure of Public Financing for Primary Health Care 7.11 The existing fiscal and administrative structures for primary health care are complex and impede effective financing and accountability for local area management, programs, and health facilities. The structure of two departments (Health and Family Welfare), plan/non-plan expenditures, center-state financial transfers and jointly financed schemes is ineffective in: (i) assuring essential inputs for health facilities and vertical schemes; (ii) correcting inequities in health expenditure between states and increasing spending in states with the worst health indicators; (iii) providing flexibility and accountability to local officials; and (iv) supporting essential monitoring of program inputs and outputs Recommendations for Action. The central and state Governrnents should consider the following actions for improving public financing for primary health care: (i) initiate, jointly through the Ministries of Finance, and Health and FW, a substantial review of the fiscal structure and procedures in the health and FW sectors; (ii) review and reform the role of central, state, and local government financing in the public health sector in terms of national health policy objectives and provision of basic inputs; (iii) develop program budgeting tools at the central and state levels to monitor and assess expenditure for important schemes; (iv) develop flexible decentralized financing tools at the state and district levels to allow local administrators to respond to local needs; (v) develop fiscal tools to enable greater experiments with resource reallocation, cost recovery, and financial incentives to NGOs and other private providers; and (vi) review, at the state level, state Governments' fiscal structures and procedures and implement such recommendations. D. Ensure Adeguate Financing of Public Primary Health Care Facilities at the State Level 7.13 Much of the planned system of rural primary health care facilities has been created, with the lowest, most peripheral facilities being most complete. In contrast, essential staff and supply inputs are below stated government norms or adequate levels, and are most lacking in the more peripheral facilities. Financing the provision of staff drugs and other inputs to norms would require about a 26% increase above current public sector health expenditures Recommendations for Action. State Governments need to prioritize and ensure that sufficient funds are maintained in their non-plan health budgets to provide adequate and timely supply of essential inputs to existing facilities. The following package is recommended for adoption: (i) review staff norms and ensure that necessary nursing care can be provided; (ii) allocate at least Rs. 50,000/- per annum at current prices for drug purchase to each PHC;

130 104 (iii) rationalize personnel policies to ensure adequate staffing of posts at rural PHCs; (iv) ensure that doctors provide two years of rural service as a pre-condition for eligibility for admission to post-graduate medical courses; (v) provide for staff quarters where critically needed, and electricity and water; (vi) undertake studies to decide an optimnal annual maintenance budget for each PHC; (vii) ensure better communication arrangements; (viii) provide regular training for medical/paramedical staff in health management/health economics; and (ix) strengthen sub-centres by providing an additional worker for looking after general health care. Following the adoption of the above package, consider and carefully evaluate whether financing the completion of existing physical facilities to norm is critical in the light of limited resources. State Governments would, therefore, need to evaluate the composition of plan and non-plan budgets for the health sector. E. Increase Supplementary Central Funding to Needy States 7.15 Central allocation of health resources in the past has not been in proportion to the needs of individual states as indicated by socio-economic and health indicators. Some of the centrally funded communicable disease programs, including the National Malaria Eradication Program, which is the largest one, are funded on a matching basis by state and central budgets. However, the poorer states are unable to come up with sufficient matching funds to make optimum use of these programs. These states are least able to mobilize state resources but are most in need of supplementary central allocations Moreover, there is a distinct possibility that significant cutbacks and imbalances at the state level due to stabilization policies are likely to be introduced into a system that in fact needs to be expanded and strengthened. Stabilization can affect government health spending through reductions in central plan scheme allocations; reductions in central untied transfers to states; reduced government revenue at state level; and autonomous state reductions in actual health spending in response to their general fiscal constraints. Sizeable reductions in health spending have not yet appeared, although where reductions have occurred they have affected poorer states and public health programs more Recommendations for Action. To accomplish its health aims and to enhance interstate equity, there is a need to develop mechanisms to provide increased supplementary central funding to the poorest states in cases where alternative sources of revenue are limited. Thus, supplementary financing should be provided to those states most in need, but only when these states are already taking credible steps to improve their finances. F. Mobilize Resources through Financing Innovations 7.18 The need to focus on state level resource adequacy and input mix is critical. In addition to augmenting state health budgets, a strengthened finance strategy applying new government approaches to financing is also needed. Central and state governments have limited experience in innovative financing. It is best to envisage a period of substantial experimentation during which new approaches to financing can be adapted to Indian conditions. In particular, these include: user charges in urban tertiary and referral hospitals, finance of non-government providers, and allowing private insurance and ESIS reimbursement at full cost of publiclyprovided services.

131 Innovative financing at the hospital level should address the fact that administrative responsibility and financial accountability are artificially separated between the responsible government agency and hospitals. The lack of appropriate management arrangements and financial authority to act means that there are few incentives for hospitals and their staff to improve hospital operations and quality of services Recommendations for Action. The government should consider developing national health financing schemes to provide grants to state and local Governments, NGOs, and private sector organizations to develop, test and evaluate new and innovative approaches to financing. Greater emphasis should be placed on collecting user fees. User-charges should be implemented with the principle that they would: target the receipts, particularly on non-salary recurrent costs; charge for amenities such as private beds at hospitals; and charge for procedures that are low in cost-effectiveness in order to pay for those interventions which are high in cost-effectiveness. The state Governments should sort out legal/other issues that are prohibiting greater devolution of autonomy to hospitals for retaining income generated by them. They should allow each hospital to retain most of the income it generates through user charges. This will strengthen service delivery management by improving the implementation capacity of these hospitals. G. Strengthen Private Primary Health Care through Improvements in Service Oualitv 7.21 Private, fee-for-service providers are ubiquitous, accessible to the rural and urban population and are heavily used by the poor. However, many of these providers are unqualified, licensing is weak or non-existent, and they may cause substantial harm as well as good. To date, adequate attention and resources to ensuring that quality of services are being maintained have not been provided Recommendations for Action. The public sector could monitor, register, certify and regulate private health care provision with regard to qualification of medical personnel and dispensing of approved drugs as needed. The quality of services provided by the private sector could be improved by: (i) strengthening central and state government capacity to register, certify, regulate and monitor private health care provision, especially qualifications of doctors and other medical personnel and the quality of their service; (ii) enact legislation to register nursing homes, private clinics/hospitals and ensure minimum standards of care by providing appropriate guidelines; (iii) increasing government capacity to control pharmaceutical supply and dispensing, as well as to provide training and information on drug use to primary care providers and general populations. Alternatively, the government could play a more proactive role in encouraging and promoting existing medical societies in matters of board certification and qualification of medical personnel; (iv) increase public support for voluntary agencies in health in such areas as social marketing of essential drugs and contraceptives and behavior- changing health education activities, and contract-out services to the private sector where possible, especially support services to cut cost and increase efficiency. H. Encourage Private Sector Participation in Preventive and Promotive Health Services 7.23 Government is the major provider of preventive and promotive health care services, but its coverage is very low (20%); the private sector contributes mainly to ambulatory care illnesses but hardly anything for preventive and promotive care. Therefore, neither the public nor the private sector is contributing enough to the coverage or improvement in quality of

132 106 preventive and promotive care. Given that investments in preventive and promotive care services yield high social returns and benefit from externalities, a strategy to enhance the private sector's contributions to national health goals is needed Recommendations for Action. Encourage the central and state Governments to develop incentives and schemes to finance, train, and integrate private providers in case-finding, diagnostics, referral, treatment and monitoring for priority problems such as TB, STDs, ARI, diarrhea, malaria, leprosy and high risk births. MOHFW should publish a quarterly newsletter to disseminate information about such schemes in the various states and about other innovative activities involving private sector participation. 1. Improve Sectoral Effectiveness and Efficiency by Strengthening the Referral System 7.25 Public hospitals are a major source of treatment, reach the poor, and appear to be progressive in their financial subsidy. However, most admissions to public hospitals have not received prior primary care. The referral system does not function well. Each tier operates as an independent entity providing similar levels of care and resource utilization is poor Recommendations for Action. State Governments should be encouraged to set up Referral Committees at the District Level to coordinate and manage referral between private and public primary care and secondary level diagnosis, treatment and care. Referral Committees would: (i) issue administrative directives that would specify procedures to be followed in order to make the referral system effective and acceptable to the community; (ii) develop referral protocols that specify the types of conditions that should be referred either for investigation or treatment at higher levels; (iii) develop clinical management protocols to provide guidelines and standards for the management of common conditions by doctors who do not have post-graduate qualifications in that specialty; and (iv) provide incentives for priority treatment to those patients using the referral systems such as reduction in user fees and shorter waiting time for diagnosis and treatment of such patients. J. Improve Understanding of the Health Needs of Women Beyond Those Covered by the Family Welfare Program 7.27 The national average sex ratio of 927 females to 1,000 males is a matter of concern. The information base to understand the underlying cause of this imbalance is lacking and more analysis needs to be done. Although some progress is noted in recent years, the central and state Governments need to move more expeditiously towards a reproductive health approach that will extend beyond those aspects of women's health covered by the Family Welfare (FW) program Recommendations for Action. State Governments should undertake in-depth analysis of the special health needs of women at the primary and secondary level beyond those covered by the FW program. Such a study should focus on women's burden of disease, healthseeking behaviour and the resource implications of shifting to the reproductive health approach.

133 107 K. Strengthen GOI/MOHFW Planning and Analysis Capacitv 7.29 There has been significant progress in recent years in the availability and use of information on health financing in India and in efforts to create a health policy/ financing unit in the MOHFW. However, the capacity for health planning and policy analysis at state and central levels remains limited. Recruitment of local consultants has been slow, and their position in the central Ministry has been weak Recommendations for Action. GOI/MOHFW's planning and analysis capacity could be strengthened through the following actions: (i) implementing systematic analysis of government health priorities based on disease burden, cost-effectiveness of health interventions, and current public and private sector health care coverage. The ongoing study in Andhra Pradesh is a useful contribution to this; (ii) establishing a substantial operations research program to determine essential inputs norms for health facilities and programs. This should be linked to planning and budgeting. An alternative scenario if such work cannot be supported by the health policy and finance unit in MOHFW, a formal linkage with an external unit, such as the National Institute of Public Finance and Policy, could be established; (iii) providing a cohort of Indian experts and government officers with advanced training in health financing. This includes both short- and long-term training, both local and international; (iv) expanding the role of the policy/finance unit to include design, review, and evaluation of financing innovations experiments; (v) accelerating current efforts to develop a standard health information system by enhancing capacity in MOHFW through management information on inputs and outputs by facility and program and through development of health management information systems at the state level; and (vi) drafting an appropriate strategy for urban primary health care which includes financing projections. L. Strengthen Training; Increase Awareness of Health-Related Issues: and Initiate Community Participation in Existing Health Education Programs 7.31 An integral component of strengthening policy and finance strategies for primary health care services is to improve and introduce some complementary activities that are crucial to the development of the health sector. These include the need to improve human resource development skills of health practitioners, raise the consciousness of health workers and encourage greater community participation in programs on school health and nutrition Recommendations for Action. The Government should: (i) strengthen training institutions both in public and private health sectors through training of trainers and regular short and long term courses in: health management and administration; analysis of cost effectiveness of various health services; hospital audits; and drug management; (ii) initiate health education activities to improve awareness of health-related issues among the public, politicians and health workers; and (iii) initiate community participation in school health and nutrition education programmes.

134 108 M. Need for Further Analysis 7.33 This sector study has covered a number of major themes on health sector financing issues. There remain several health sector issues in India where further work would be beneficial for policy analysis. A few suggestions follow: (a) (b) (c) (d) (e) (f) Issues related to Derformance and quality of health care at the state level. Since three-quarters of health expenditures are increased at the state level, a study of key issues related to strengthening the performance and quality of health care at the state level is a logical extension of this report. The outcome of such a study would be to suggest action-oriented recommendations for implementing a coherent strategy and policy reform package for the development of a health system at the state level, adding to state level issues discussed in this report. Financing and provision of primarv care for poor populations in urban areas. Urban health needs are likely to increase substantially in coming decades. Poor urban populations face a very different environment from that prevalent in rural areas. A study that would include the role of public financing and provision in urban environments could be very useful in the light of an active private sector. Differences by size of city, levels of current public and private provision facing consumers and the role of regulation and quality control could also be analyzed. Efficiency in public hospitals. Public hospitals remain a major part of the public expenditure program in health and a very important factor constraining expanding finance for prim ary Care. An assessment of the current level of efficiency of public hospitals, the potential gain to the government from improving efficiency and the cost implications of quality improvements could be analyzed. The potential of new approaches to financing public hospitals. The potential for user charges and private insurance to finance public hospitals while maintaining protection for the poor could be studied. In addition, the lessons learned from government grants to NGOs, their potential for expanded coverage of key programs to the poor could also be analyzed. Primary care and the unqualified practitioner. The study could analyze the nature of practice of all undocumented private primary care providers, the role of these providers in expanding primary care coverage, and the terms of regulation, quality control, and pricing structure. A follow up could be undertaken to the sector report on Issues in Women's Health in India, with a focus especially on the disease burden on women.

135 109 BIBLIOGRAPHY Bajaj, J.L. and O.P. Agarwal (ND), "State Finances: Uttar Pradesh ( )." Berman, Peter (1991), "Health Economics, Health Financing, and the Health Needs of Poor Women and Children in India," The Ford Foundation, New Delhi. Berman, Peter (1992), "India's Health Expenditure," paper prepared for NCAER- Harvard Seminar on the Health Transition in India, India International Center, New Delhi. Berman, Peter and M.E. Khan, eds. (1993), Paying for India's Health Care, Sage Publications, New Delhi. Bhaskara Rao, N., M.E. Khan, and C.V.S. Prasad (1986), "Health Sector Expenditure Differentials in India--A State and National Level Study," Operations Research Group, Baroda. Bhat, Ramesh (1991), "The Non-Government Healthcare Sector in India: Geographic Disparity and the Implications for Health Policy," Research Paper No. 63, Takemi Program in International Health, Harvard School of Public Health, Boston. Bhat, Ramesh (1991), "The Private Healthcare Sector in India: Some Policy Concerns," Research Paper No. 54, Takemi Program in International Health, Harvard School of Public Health, Boston. Bhat, Ramesh (1993), "The Private/Public Mix in Health Care in India," Health Policy and Planning 8: Bose, Ashish (1991), "Population of India: 1991 Census Results and Methodology," B.R. Publishing Corp., Delhi. Central Bureau of Health Intelligence (1990), Health Information India--1990, Ministry of Health & Family Welfare, Government of India, New Delhi. Dave, Priti (1992), "Private Sector Provision of Health Care in India," UNICEF, New Delhi. Djurfeldt, Goran and Staffan Lindberg (1975), Pills Against Poverty: A Study of the Introduction of Western Medicine in a Tamil Village, Scandinavian Institute of Asian Studies Monograph Series, Studentlitteratur, Curzon Press, London. Duggal, Ravi (1992), "Health Care Services and Financing in India," a report for the Health Financing Review Mission of World Bank. Foundation for Research in Community Health, Bombay.

136 110 Duggal, Ravi (ND), "State Health Financing and Health Care Services in India," Mimeo. Duggal, Ravi and Suchetha Amin (1989), "Cost of Health Care: A Household Survey in an Indian District," The Foundation for Research in Community Health, Bombay. Duggal, Ravi, Sunil Nandraj, and Sahana Shetty (1992), "State Sector Health Expenditures, A Database: All India and States ," Foundation for Research in Community Health, Bombay. Foundation for Research in Community Health (1987), "Health Status of the Indian People. " Foundation for Research in Community Health (1989), special number on Health and Related Statistics, FRCH Newsletter, vol. III, no. 6, Bombay. Foundation or Research in Community Health (1987), FRCH Newsletter, vol. I, no. 4, Bombay. Govinda Rao, M. and R.J. Chelliah (ND), "Survey of Research on Fiscal Federalism in India," National Institute of Public Finance and Policy, New Delhi. Govinda Rao, M. and V.B. Tulasidhar (ND), "Public Expenditure in India: Emerging Trends. " Guhan, S. (1991), "State Finances in Tamilnadu : A Survey," Madras Institute of Development Studies. Guhan, S. (1992), "Social Security for the Unorganised Poor: A Feasible Blueprint for India," disussion paper from the Symposium on Economic Growth, Sustainable Human Development and Poverty Alleviation in India, United Nations Development Programme, Indira Gandhi Institute of Development Research, Bombay. Gupta, Gouri S. (1992), "Unit Cost of Health Care Programmes at Micro Level in India," Ministry of Health and Family Welfare, Government of India. Gupta, J. P. and Murali, Indira (1989), "National Review of Immunization Programme in India," National Institute of Health and Family Welfare, New Delhi. Indian Council of Medical Research (1991), "Evaluation of Quality of Family Welfare Services at Primary Health Centre Level," An IMCR task force study, New Delhi. Indian Institute of Management (1987), "Study of Health Care Financing in India: Based on Case Studies in Maharashtra and West Bengal," a report submitted to Asian Development Bank, I.I.M.-Ahmedabad. Indian Market Research Bureau (1988), "Profile of the Rural Doctor: Report on a

137 Ill Study in Rural India," IMRB/RI/HV/40479, prepared for the United Nation's Children's Fund, Delhi. Jesani, Amar (1989), "Drugs and Pharmaceuticals at the PHC," from Study of Primary Health Centres in Maharashtra, Foundation for Research in Community Health, Bombay. Jesani, Amar and Saraswathy Anantharam (1989), "Private Sector and Privatisation in the Health Care Services," a review paper prepared for the ICSSR-ICMR Joint Panel on Health, Foundation for Research in Community Health, Bombay. Jolly, K.G. (1986), Family Planning in India, : A District Level Study, Hindustan Publishing Corporation, Delhi. Kakwani, N. and K. Subbarao (ND), "Rural Poverty in India: ," in Including the Poor, van der Gaag and Lipton, eds., Oxford University Press (?). Khan, M.E. and C.V.S. Prasad (1985), "Health Financing in India: A Case Study of Gujarat and Maharastra," Operations Research Group, Baroda. Krishnan, T.N. (1991), "Kerala's Health Transition--Facts and Factors," Harvard Center for Population and Development Studies, Cambridge. Kulkarni, Sumati (ND), "Pattern of Expenditure on Family Welfare in Maharashtra," International Institute for Population Sciences, Bombay. Mahapatra, Prasanta (1992), "Sub Allocation of Hospital Services Outlay in Andhra Pradesh, India during the Eighties. Have they been Appropriate?" Research Paper No. 67, Takemi Program in International Health, Harvard School of Public Health, Boston. Minhas, B.S. (1990), "Brief Notes on Access to Subsidized Food and Social Services in India: Preliminary Evidence from the Survey Data," Unpublished. Ministry of Health & Family Welfare (1993), "Conference of Central Council of Health & Family Welfare, Agenda Notes," Government of India, New Delhi. National Council of Applied Economic Research (1991), "Household Survey of Medical Care," sponsored by Ministry of Health, Government of India, New Delhi. National Council of Applied Economic Research (1992), "Municipal & Local Government Finance for Health in India." National Council of Applied Economic Research (1992), "Rural Household Health Care Needs and Availability, Findings--Vol. I," New Delhi.

138 112 National Council of Applied Economic Research (1992), "Rural Household Health Care Needs and Availability, Detail Tables--Vol. II," New Delhi. National Council of Applied Economic Research (ND), "Survey of Primary Health Care Services: Summary." National Institute of Health and Family Welfare (1990), "National Review of Immunisation Programme in India: Cost Analysis of Universal Immunisation Programme," New Delhi. National Institute of Public Finance and Policy (1992), "States' Financing of Health Care in India: Some Recent Trends," New Delhi. National Institute of Urban Affairs (1989), "Upgrading Municipal Services: Norms and Financial Implications," Vol. 1, Research Study Series No. 38, New Delhi. National Sample Survey Organisation (1986), "Socio-Economic Survey, Instructions to Field Staff, Volume I: Design, Concepts, Definitions and Procedures," Fortysecond round, Department of Statistics, Government of India, New Delhi. National Sample Survey Organisation (1986), "Socio-Economic Survey, Instructions to Field Staff, Volume II: Schedules of Enquiry," Fortysecond round, Department of Statistics, Government of India, New Delhi. National Sample Survey Organisation (1989), "Child and Maternity Care," Fortysecond round, Department of Statistics, Government of India, New Delhi. National Sample Survey Organisation (1989), "Morbidity and Utilisation of Medical Services," Fortysecond round, Department of Statistics, Govermment of India, New Delhi. National Sample Survey Organisation (1989), "Utilisation of Public Distribution System," Fortysecond round, Department of Statistics, Government of India, New Delhi. Operations Research Group (1991), "Study of Time Use Pattern of PHC Functionaries and Economic Cost of Primary Health Care," a study sponsored by GOI/WHO, Baroda. Potdar, R.D. and Alka Karande (1988), "Knowledge, Attitude and Practices (KAP) Study of Non-M.B.B.S. Practitioners in Bombay Slums with Reference to ARI/DD/Immunization/ORT," Bombay Municipal Corporation, Bombay. Ravallion, Martin and K. Subbarao (1992), "Adjustment and Human Development in India." Mimeo.

139 113 Ravishankar, V.J. (1989), "India: Government Expenditure on Social Services 1976/ /87," World Bank, New Delhi. Ravishankar, V.J. (1993), "India: Rationalization of Center-State Transfers," Unpublished draft. Reddy, K.N. and V. Selvaraju (1993), "Health Care Financing by Government in India: to (Growth, Structure and Priorities by Programme and Sector)," National Institute of Public Finance and Policy, New Delhi. Sahni, Ashok, ed. (1985), Financing of Health Services in India, Proceedings of the Fifth Annual Conference--1984, Indian Society of Health Administrators, Bangalore. Seeta Prabhu, K. and Somnath Chatterjee (1993), "Social Sector Expenditures and Human Development: A Study of Indian States," Development Research Group, Study 6, Department of Economic Analysis and Policy, Reserve Bank of India, Bombay. Sen, Tapas Kumar (1992), "Public Expenditures on Human Development in India: Trends and Issues," discussion paper from the Symposium on Economic Growth, Sustainable Human Development and Poverty Alleviation in India, United Nations Development Programme, Indira Gandhi Institute of Development Research, Bombay. Sisodia, Pratap (1991), "Public Expenditure on Health Care Through the State Government of Rajasthan," Indian Institute of Health Management Research, Jaipur. Sisodia, Pratap (1992), "Identification of Availability and Composition of Health Resources in the Public Sector," Indian Institute of Health Management Research, Jaipur. Social and Rural Research Institute (ND), "Rural Doctors Study," New Delhi. Social and Rural Research Institute (ND), "The Rural Private Practitioner," SRI/HV/90014, prepared for the Ford Foundation, Delhi. Srinivasan, K. and Sumati Kulkami (ND), "Do We Spend Enough on Family Planning Programme in India?" International Institute for Population Studies, Bombay. Subbarao, K. (1987), "Interventions to Fill Nutrituion Gaps at the Household Level: A Review of India's Experience," paper prepared for a Workshop on Poverty in India, Oxford University. Tendulkar, Suresh D. (1992), "Poverty Alleviation, Economic Restructuring and the Growth Process in India," discussion paper from the Symposium on Economic Growth, Sustainable Human Development and Poverty Alleviation in India, United

140 114 Nations Development Programme, Indira Gandhi Institute of Development Research, Bombay. Tulasidhar, V.B. (1993), "Allocation of Resources to Primary Health Care in India: Did Alma-Ata Make Any Difference'?" Draft. Tulasidhar, V.B. (1993), "Expenditure Compression and Health Sector Outlays," National Institute of Public Finance and Policy, New Delhi. Tulasidhar, V.B. (1993), "Structural Adjustment Programme: Its Impact on Health Sector," National Institute of Public Finance and Policy, New Delhi. The Urban Institute (1993), "The Role of Rural Doctors in India," Policy and Research Report. Vyas, V.S. and Vidya Sagar (1992), "Alleviation of Rural Poverty in the States-- Lessons of Eighties," discussion paper from the Symposium on Economic Growth, Sustainable Human Development and Poverty Alleviation in India, United Nations Development Programme, Indira Gandhi Institute of Development Research, Bombay. Wallich, Christine (1982), "State Finances in India, Volume 1, Revenue Sharing," World Bank Staff Working Papers, Number 523, The World Bank, Washington, D.C. World Bank (1989), "India: Poverty, Employment, and Social Services," A World Bank Country Study, Washington, D.C. World Bank (1991), "Staff Appraisal Report: India: Child Survival and Safe Motherhood Project," Report No IN, Asia Country Department IV (India), Population and Human Resources Operations Division. World Bank (1992), "Financing of Urban Services in India: Selected Issues," Report No IN, Country Operations, Industry and Finance Division, India Department, South Asia Region. World Bank (1992), "India: Health Sector Financing: Coping with Adjustment Opportunities for Reform," Report No IN, South Asia Country Department II (India), Population and Human Resources Operations Division. World Bank (1993), "India: Public Expenditure Review, Health Sector," South Asia Country Department II (India), Population and Human Resources Operations Division. World Bank (1993), "India: Public Expenditures in the Social Sectors, A Brief Overview," South Asia Country Department II (India), Population and Human Resources Operations Division.

141 CHAPTER 6 APPENDICES

142 116 Appendix 1 Demand for Non-hospital Illness Care and Determinants of Expenditure: Multivariate analysis 1. The Demand for Non-hospital Illness Care: Multivariate Models. The preceding discussion has described use and expenditure patterns for ambulatory illness care by public and private providers looking mainly at uni- or bivariate relationships. Clearly a variety of factors affect both the choice of provider as well as expenditure, including the availability of free treatment, seriousness of the illness and type of illness, characteristics of the ill person, and household social and economic factors. In order to sort out the net effects of key variables from among the large number of relevant influencing factors, regression analysis was done on both the choice of provider (government/non-government) and the expenditure per episode of illness. 2. Choice of Provider Model. Attachmnent I presents the results of a logistic regression on the choice of ambulatory care provider (government = 0/non-government = 1) for both rural and urban areas and the five states. Two income-related variables were included: the log of monthly household consumption expenditure and a dummy variable for households in the top expenditure quintile. In rural areas, these expenditure variables are not statistically significant in general. This reflects the results seen above, that for the rural population non-government services are the major source of care with little income gradient. The regression confirms this observation, controlling for a wide variety of other factors, including personal characteristics, type of illness, and household economic factors. In contrast, the urban models show a more significant income effect in several states, although the findings are consistent across all samples. 3. The models generally show a positive coefficient on education, both of the ill person as well as the head of household. However, this is usually not statistically significant. Insurance coverage is generally significantly predictive of government service use, except in Gujarat. 4. In terms of the types of ailments listed, the models show a strong tendency for reported TB cases to be taken to public providers, the only illness variable with consistent and statistically significant effects. For other diseases targeted by public programs, this is not the case. Dysentery and malaria dunmy variables have varying coefficients, but when statistically significant they are positive, reflecting use of private providers centrally for other factors. 5. Determinants of Expenditure per Episode. In addition to looking at the choice of provider, regressions were run using the rupee expenditure per episode of illness as the dependent variable. These are presented in Attachment 2 for both rural and urban areas. The model includes dummy variables for whether payment was made for a public or private provider. The other version omits these variables. A wide range of variables were included to control for confounding of expenditure with factors such as the type of illness, type of service used, and whether there was insurance coverage. this allows estimation of the isolated effects of key individual and household variables. 6. Both rural and urban regressions showed a consistent positive effect on the "income" variable. this was statistically significant in eighteen of twenty equations. The Scheduled Caste and

143 117 Schedule Tribe (SCIST) variable was almost always negative and statistically significant in many cases. 7. The dummy variables on whether payment was made to public or private provider are also universally positive and significant. Charges for treatment do impact on expenditures per episode for both public and private providers. In both rural and urban samples, the coefficients are almost equal in size, suggesting that the effect of choice of provider on expenditure per episode is small if payment must be made. 8. Variables representing illness duration and severity were positive and generally significant, indicating higher cost. Duration of illness was recorded up to 30 days, after which the episode was coded as "prolonged". For those episodes, cost per illness was likely to be lower, reflecting perhaps the routine maintenance of treatment for chronic illness. 9. A dummy variable for insurance coverage was included. This had mixed signs in the rural equations, and was negative (insurance reduced health expenditure) and statistically significant in two of the states. In the urban samples insurance coverage had a higher and significant negative coefficient in four of five states except Gujarat. This suggests that insurance is affecting payment levels for outpatient treatment, especially in cities. 10. In rural areas, being male almost always had a positive coefficient, suggesting higher expenditure and was significant in two states. In urban areas, the estimated coefficients on being male varied in sign and none were significant. The signs on the types of ailment variables (diseases) were not consistent across states although some of them were always statistically significant. This may reflect the fact that the type of ailment is not such an important predictor of expenditures of ambulatory care, since practitioner fees and medicine charge may vary more for reasons other than the type of disease presented in most cases.

144 118 Appendix 2 Hospital use and Expenditure: Multivariate Analysis 1. Logistic regression models using the choices of type of hospital provider (government = 0/non-government = 1) as the dependent variable were estimated for the rural and urban samples in all five states (see Attachment 3). These are "conditional" choice models, since the sample includes only households who sought hospital treatment for an illness. 2. The coefficient on monthly household expenditure is positive in all the estimated equations, but only statistically significant in some of the urban models. However, a dummy variable for whether a household is in the top expenditure quintile was positive and significant in the rural equations for three out of five states. These results indicate a fairly "income neutral" distribution of public hospital services, with some evidence of increased private hospital use at the highest income levels, controlling for a wide range of individual, household, and illness-related factors. Dummy variables related to other elements of low socio-economic status (non-sc/st and being in the casual labor force) generally have signs indicating a tendency to use public services, but these are often not statistically significant. Overall the results suggest some positive bias towards the poor in public hospital care, but one that is not very strong. 3. There is little evidence of differentiation in public-private use patterns by sex or age. Availability of insurance is associated with use of public facilities, especially in urban areas (ESIS hospitals may have been reported as government facilities). 4. The dummy variable coefficients for reported cases of TB are consistently negative and often significant, confirming the importance of public facilities in treating that disease. Accidents and trauma cases also show a strong bias towards public facilities. The coefficients for other communicable/infectious diseases (e.g. malaria, typhoid, dysentery) do not show this effect. 5. The regressions using expenditure per episode as the dependent variable (Attachment 4) are similar to the earlier models for ambulatory care. The coefficient on total household expenditure is always positive and significant in the majority of cases. In contrast to the ambulatory illness expenditure models, the coefficients of payment to providers show a large difference between public and private facilities. this suggests that private hospital use results in higher expenditures, other things being equal, whereas for outpatient treatment if payment was made the choice of public or private care made little difference in the amount paid per episode. Treatment before and after hospitalization also had a significant effect on expenditure.

145 119 Attachrment 1 Page 1 of 2 Logistic Regressions of Determinants of Choice of Treatmnent for Illness Not Requiring Hospitalization, (0 = public, I = private) RURAL AREAS PredictorS Gujarat Maharashtra Tamil Nadu Uttar Pradesh West Bengal 1. Characteristics of [II PerSOn Male *' Age @ ' Age Sq. (102).049@ N. Married "* "* Head * Education Insured 2.401* ' ' * 11. Characteristics of the Non SCST 1.777'.414* *.342' Main Source of Income Casual Labor @ Regular Labor ' Others ** LnMTHIE ' Top 20% of MPCE ihhsize @.053@ AF Edu @ " Ill. Character;stics of thc IICad of Iousehold [id. MalC * [id. Age ' Hd Edu * I15 IV TVDC of Ailment Dysentary -.764@ @ Typhoid TB ' ' ' ' ' Malaria @ '.416 Cancer NCrvous Heart @.295 Asthma Accident Constant ' % of Correctly Predicted to Observed Outcomes

146 120 Attachment 1 Page 2 of 2 URBAN AREAS Predictors Gujarat Maharasbtra Tamil Nadu Uttar Pradesh West Bengal 1. Characteristics Of l1 Person Mile Age -.079*.042* -.036" Age Sq. (10 2 ).091'.0380".0412*" N. Married @ -.573" Head " *250 Education.183@ Insured ' ' ' II. CharaCteristics of the Househld Non SCST 1.053' ' Main Source of Income Casual Labor ' ' -.400" Regular Labor Others ' LnMTHE S Top 20/% of MPCE.07S @.460*.051 HHSIZE AF Edu '.065@ Ill. Characteristics of the Head Of Household Hd. Male Hd. Age ' ' ' Hd Edu @ IV TVye of Amlment Dysentary @.551@ Typhoid TB ' S30 Malaria '.262 Cancer " Nervous Heart @ Asthma '.213 Accident ' Constant % of Cotrectly Predicted to ObservCd Outcomes For this and subsequent p<. 1 0 i lp<05 SignifiCanCe ICVclS CalCUlated Using the exponential or B

147 121 Attachment 2 Page I of 4 OLS Models of Determinants of Expenditure on Treatment of Ilnes Outside Hospital, (Dependent variable - Rs. expenditure per Illness epbsode) Gujarat Maharashtra Tamil Nadu Uttar Pradesh West Bengal Rural Areas I. Type of Health Service 1. Payment made to Health Care Provider (I -yes) Public 1.615* 1.861* 2.445* 1.332* 1.847* Private 1.537* 1.883* 2.483* 1.071* 1.823* I. Duration of Treatment 2. No. of Visits *.094*.037* Duration of.082*.065*.042*.076*.077* illness 4. Prolonged * -1.i46* * -.797* -.978* illness (> 30 days, 0/1) MIl. Type of Ailment 5. Dysentery ** Typhoid TB * * Malaria -.615* Polio.879** * Cancer * **.839* Nervous -.769* Heart * * 13. Asthma -.334@ Accident *.377** IV. Characteristics of III Person 15. Age *.010 Age Square Male.266**.l27.242* Head Not Married * -.220* -.382* 19. Whether Completed Schooling Primary Secondary Insured * *.2340

148 122 Attachment 2 Page 2 of 4 OLS Models of Determinants of Expenditure on Treatment of Illness Outside Hospital, (Dependent variable = Rs. expenditure per illness episode) V. Characteristics of the Household Gujarat Maharashtra Tamil Nadu Uttar Pradesh West Bengal 21. Non SCST HH Size @ -.047* -.014@ Ln MTHE* *.389*.253*.359* 24. Main Source of Income (0/1) Casual Regular Emp * Others Highest Education of Adult Female (0/1) Primary **.185** Secondary ** VI. Characteristics of the Head of Household 26. Male (0/1) Age @ Education of the Head (0/1) Primary ** Secondary Constant @ * * R Square No. of Cases * MTHE = Monthly Total Household Consumption Expenditures ADD SIGNIFICANCE LEVEL LABELS

149 123 Attachment 2 Page 3 of 4 OLS Models of Determinants of Expenditure on Treatment of Illness Outside Hospital, (Dependent variable = Rs. expenditure per illness episode) Gujarat Maharashtra Tamil Nadu Uttar Pradesh West Bengal I. Type of Health Service 1. Payment made to Health Care Provider (1 =yes) Urban Areas Public 1.064* 1.643* 2.606* 1.431* 1.581* Private 1.063* 1.727* 2.695* 1.193* 1.588* II. Duration of Treatment 2. No. of Visits.061*.043*.052*.029* Duration of.096*.090*.061*.098*.056* illness 4. Prolonged * * * * -.637* illness (>30 days, 0/1) III. Type of Ailment 5 Dysentery ** ** Typhoid.857*.392** *.680** 7. TB * Malaria ** Polio * 1.208* Cancer * Nervous -.717* * Heart *.364* Asthma @ Accident **.207 [V. Characteristics of III Person 15. Age Age Square @ Male Head ** 18. Not Married -.430** * Whether Completed Schooling Primary Secondary Insured * * * * -.603*

150 124 Attachment 2 Page 4 of 4 Determinants of Expenditure on Treatment in a Hospital, Gujarat Maharashtra Tamil Nadu Uttar Pradesh West Bengal V. Characteristics of the Household 21. Non SCST **.238* HH Size * -.035** Ln MTHE*.325*.377*.563*.398*.343* 24. Main Source of Income (0/1) Casual Regular Emp Others @.392** 25. Highest Education of Adult Female (0/1) Primary Secondary -.291** VI. Characteristics of the Head of Household 26. Male (0/1) Age -.009** * 28. Education of the Head (0/1) Primary Secondary Constant * R Square No. of Cases * MTHE = Monthly Total Household Consumption Expenditures ADD SIGNIFICANCE LEVEL LABELS

151 I. CHARACTERISTICS OF ILL PERSON 125 Attachment 3 Page 1 of 2 Logistic Regressions of Determinants of Choice of Treatment in Government and Non-Government Hospitals, (Dependent variable = 0 if government, 1 if private hospital) Gujarat Maharashtra Tamil Nadu Uttar Pradesh West Bengal RURAL AREAS MALE * @.788** AGE -.056* AGE SQURE (102).001@.000@ N.MARRIED -.292@ -.576@ HEAD EDUCATION * INSURED * * * II. CHARACTERISTICS OF THIE HOUSEHOLD NON.SC.ST.438@.923*.590* MAIN SOURCE OF INCOME (0/1) CASUAL LAB -.519** * REGULAR LAB * OTHERS Ln MTHE @ TOP 20% MTHE.544* * * HH SIZE AF EDU @.957** III. CHARACTERISTICS OF THE HEAD OF HOUSEHOLD MALE (011) * AGE @ HD EDU (0/1) 1.400***.940* IV. TYPE OF AILMENT DYSENTERY * * TYPHOID ** TB * * MALARIA CANCER NERVOUS ^ -.620@ HEART ASTHMA ACCIDENT -.696** -.614* -.369@ -.278@ -.903** Constant ** % of Correctly Predicted to Observed Outcomes

152 1. CHARACTERISTICS OF ILL PERSON 126 Attachment 3 Page 2 of 2 Logistic Regressions of Detenninants of Choice of Treatment in Government and Non-Government Hospitals, (Depepdent variable = 0 if govermnent, 1 If private hospital) Gujarat Maharashtra Tamil Nadu Uttar Pradesh West Bengal URBAN AREAS MALE AGE S AGE SQURE (10 2 ) l N.MARRIED * HEAD EDUCATION -.573@ INSURED * * -.862* -.866* II. CHARACTERISTICS OF THE HOUSEHOLD NON.SC.ST *.418 MAIN SOURCE OF INCOME (0/t) CASUAL LAB * -.886* REGULAR LAB ".114 OTHERS Ln MTHE *.5133* * TOP 20 % MTHE HH SIZE AF EDU.675*.418* III. CHARACTERISTICS OF THE HEAD OF HOUSEHOLD MALE (0/1) AGE HD EDU (0/1).770* IV. TYPE OF AILMENT DYSENTERY TYPHOID TB * MIALARIA CANCER NERVOUS HEART ** ASTHMA * ACCIDENT ' -. W05 Constant * * % of Correctly Predicted to Observed Outcomes ADD SIGNIFICANCE LEVEL LABELS

153 127 Attachment 4 Page 1 of 4 Determinants of Expenditure on Treatment in a Hospital, Gujarat Maharashtra Tamil Nadu Uttar Pradesh West Bengal Rural Areas 1. Type of Health Service 1. Payment to Health Care Provider = 1 if payment made Public.949* 1.673* 3.087*.895* 1.826* Private 1.617* 2.658* 4.170* 1.165* 2.722* H. Duration of Treatnent 2. Treatment.271*.483* *.479* Sought Before Hospitalization (0/1) 3. Treatment.266*.195@.294**.247*.521* Sought After Hospitalization (0/1) 4. Duration of.019*.020*.016*.017*.010* Episode M. Type of Ailment 5. Dysentery * 6. Typhoid ** TB -.348@ Malaria -.664** -.961@ ** Cancer * Nervous * * Heart Asthma -.514** * @ 13. Accident ** *.064 IV. Characteristics of ID Person 14. Age * Age Square ** 15. Male Head Not Married ** ** 18. Whether Completed Schooling Primary ** @ Secondary ** Insured * * * -.695*

154 128 Attachment 4 Page 2 of 4 V. Characterhtks of the EHousebold Edlsruaiinnts of Expmnditure on reatmnent in a Hospital, *(pmrst M1raslbtra Tami Nadu Uttwr Prudisb Weot DEapi 20. Non SCST * HH Size Ln MTHE *.348* Main Source of lxorne (0/1) Casual ** Regular Emnp -.332* * Other ** 1.010** 24. Highest FEhduaion of Adult Female (0/1) Primauy "* Secodary VI. Charactarbscs of the Eleod of EomumoKd 25. Male (0/1) Age Education of the Head (0/1) Primary * Secoxndary O **.626 constant 3.657* * 2.096* R Squae S No. of Caws

155 129 Attachment 4 Determhnants of Expenditue om Treatment in a Hospitad, Page 3 of 4 Gujarat Maharashtra Tamil Nadu Uttar Pradesh West Bengal Urban Areas I. Type of Health Servlce 1. Payment to Health Carm Provider = I if payment made Public.572* 2.008* 2.784* 1.348* 2.289* Private 1.165* 2.747* 4.363* * LI. Duration of Treatment 2. Treatment.607* *.688*.3033 Sought Before Hospitalization 3. Treatment Sought After Hospitalization (0/1) 4. Duration of.014*.013*.010*.011*.012* Episode HI. Type of Ailment 5. Dysentery * * -.884* 6. Typhoid TB Malaria -.923* Cancer 1.198* ** Nervous * -.770* 11. Heart Asthma -.742** Accident.521** IV. Characterits of Il Person 14. Age **.024 Age Square Male ** -.358** Head Not Married **.733* Whether Completed Schooling Primary ** Secondary Insured -.970* * * * *

156 130 Attachment 4 Page 4 of 4 OLS Models of Determinants of Expenditure on Treatment of Illness Outside Hospital, (Dependent variable = Rs. expenditure per illness episode) V. Characteristics of the Household Gujarat Maharashtra Tamil Nadu Uttar Pradesh West Bengal 20. Non SCST @ HH Size * * Ln MTHE * * 23. Main Source of Income (0/1) Casual Regular Emp -.287@ -.240** Others Highest Education of Adult Female (0/1) Primary * 1-.S40* Secondary ** * VI. Characteristics of the Head of Household 25. Male (0/1) ** 26. Age * Education of the Head (0/1) Primary ** Secondary Constant 3.863* * R Square No. of Cases

157 I Statistical Annexes

158 132 Annex I Page I of 22 Selected Indicators of Health and Development IMR Child Middle Beds Doctors Govemment Income CMIE per 1000 mortality school per per health per capita index of live per 1000 enrolment 100, ,000 expenditure peryear devebirths 0-4 age as % of pop. pop. per capita lopment groups age group (Rupees) (Rupees) 1988 (F) 1988 (F) 1987 (F) Andhra Pradesh 83 ( 76) 27 (26) 33 (24) (6) Arunachal Pradesh (30) (5) Assam 99 ( 97) 37 (38) 65 (62) (5) (incl. NE states) Bihar 97 (100) 38 (42) 32 (15) (7) Goa (97) (8) Gujarat 90 ( 93) 31 (33) 53 (41) (6) Haryana 90(102) 29 (34) 61 (48) (5) Himachal Pradesh 80 ( 80) 24 (22) 90 (73) (7) Jammu & Kashmir 71 (75) 25(27) 50 (36) (6) Kamataka 74 ( 68) 24 (23) 69 (54) (7) Kerala 28 ( 27) 8 ( 7) 96 (94) (8) Madhya Pradesh 121 (121) 51 (53) 46(25) (6) Maharashtra 68 (62) 22 (22) 67 (53) (5) Manipur (60) (7) Meghalaya (50) (7) Mizoram (62) (3) - 28 Nagaland (48) (5) Orissa 122 (117) 37(36) 38(27) (6) Punjab 62 ( 66) 21(24) 61 (54) (8) (incl. Haryana) Rajasthan 103 (104) 52 (57) 40 (16) (7) Sikkim (50) (6) Tamil Nadu 74 ( 75) 21 (23) 79 (65) (7) Tripura (42) (6) Uttar Pradesh 124 (126) 47 (50) 43 (23) (5) West Bengal 69 ( 65) 22 (23) 63 (55) (8) India 94 ( 93) 33 (35) 53(39) (4) (Incl.Union Temtories) NOTES: (F) This pertains to female rates for the respective variable. Only Revenue Expenditure; excludes water supply; parenthesis figures is percent of total Govemment expenditure. - Not available. SOURCES: (1) Health Informaton of India 1990, MOHFW, C8HI, GOI, ND. (2) Family Welfare Year Book, MOHFW, GOI, ND. (3) Basic Statistics, Vol. ll, States, CMIE, Bombay. (4) RBI Bullebn, Special Supplement on State Finances, April RBI, GOI.

159 133 Annex I Page 2 of 22 Global Burden of Diseases for India and Other Regions, 1990 (hundreds of thousands of disability-adjusted life years (DALY] lost) Sub- Other Lain Saharan Asia and America and Africa India China Iskiads the Caribbean Disease or injurv DALYs % DALYs % DALYs % DALYs % DALYs % Total DALYs lost Communicable, maternal and perinatul i Infectious and parasitic Tuberculosis , STlDs excluding HlIV H1V Diarrhocal diseases Childhood cluster Meningitis Hepatitis Malaria S Tropical cluster Leprosv Trachoma Intestinal helminths Respiratory infections Maternal Causes Perinamal Causes Noncommunicable s8oo Malignant neoplasms Diabetes mellitus Nutritional and endocrine Neuro-psychiatric Sense organ Cardiovascular Respiratory Digestive Genito-urinary ! Musculo-skeletal Congenital abnonmalities Oral health Injuries Unintentional Intentional

160 134 Annex 1 Page 3 of 22 Total Amount and Distribution of Current Central and State Budgets, 1991/92, in Crores (Rs. 10 Million) Cental State All India Shares Amount Percent Amount Percent Amount Percent Amount Percent Hospitals PHCs & Dispensaries Cental Government Health Scheme Employees Ste hisurance System Education & Research Adminisration Public Health Other Captial (Health) Family Planning MCHK/I Total Source: Mission estimaes based on MOH7FW Performance Budget , additonal bies on stnt budgets by Dr. V.B. Tulasidhar, tables by Ravishankar produced for the mission, and Tulasidhar (1992). Note: Family Planning includes bodh recurrent and capital spending.

161 135 Annex I Page 4 of 22 Real Growth in Spending on Medical and Public Health (Annual 10%) First Second Overall Period Peniod Period State Andhra Pradesh Assam Bihar Gujarat Haryana Kamataka Kerala Madhya Pradesh Maharashtra Orissa Punjab Rajasthan Tamil Nadu Uttar Pradesh West Bengal All States* Centre Centre + States# * Refers to only the fifteen major States. # Refers to Centre plus fifteen major States. Source: Tulasidhar (1992)

162 136 Annex I Page 5 of 22 Real Growth in Spending on Family Welfare (Annual %) Fist Second Ovemrl Period Period Period State Andhra Pradesh Assam Biba Gujarat Haryana Kaumataka Kerala Mladhya Pradesh Nlahamshtra cissa Punjab RaJasthan Tamil Nadu Utar Prmdesh West Bengal Al States Centre Centre + Staes# *Refer to only the fften numor States. # Refers to CentCr plus fifteen major States. Source: Tulasidhar (1992)

163 137 Annex 1 Page 6 of 22 Real Growth in Spending on Water Supply & Sanitation, (Annual%) First Second Overall Period Period Period State Andhra Pradesh Assam Bihar Gujaat Hhryana Karnamaka Kerala Madhya Pradesh Maharaslfhta Orissa Punjab Rajbsthan Tamnl Nadu Uuar Pradteh West Bengal AAU States Centre Centr + States# * Refers to only the fiften major Staes. # Refers to Cente plus fifteen major States. Source: Tulsidhar (1992)

164 138 Annex 1 Page 7 of 22 Real Growth in Spending on Nutrition and ICDS, (Annual %) First Seond Overall Period Period Period State Andhra Pradesh Assam Bihar Gujarat ; Haryana Kamaltka Kerala Madhya Pradesh Maharashnra Orissa Punjab Rajasthan Tamil Nadu UttarPRPdesh West Bengal All States Centre Centre + States" Refers to only the fifteen major States. # Refers to Centre plus fifteen major States. Source: Tulasidhar (1992)

165 139 Annex I Page 8 of 22 Real Growth in Health Spending since 1984/85 Medical and Prevention Public Public and Control Family HEAD/YEAR Health" Medical Health of Diseases Welfare Absolute expenditure Rs. lakh* Index = NA R B Per capita expenditure Rs Index R B Note: * Absolute amnount for the year at constant prices. ** Total expenditure. In the case of remaining items, expenditure shown is revenue expenditure.

166 Central Government Plan and Non-Plan Expenditure in Social Sectors (Rs. Billion at Current Prices) ~~~~~~~~~~l l1I2 (H E) (a El Plan Non-PIa" TOgl9i Plan NOn-PIa Te Pan Non-Plan TOdl Plan nplan Toal Plan NHon PIan 109 A Total 1N I S S la is d S oacls t0o o IS E 2031 as t 6431 C. Sal%OA 901% 2076% 00% 11.17% 2 33% 5.13% 11.19% 1.95% 467% 12.35% 254% 176Y. 1245% 251% 566% o aas%ofgdpmp 0a6% 013% 102% 0.72% 0.34% 100% 065% 0.2% a 3% 067% 034% 101% 063% 031% 002% E heaoln f anmywson,e a I IS73 F. HealS a Is as 9.10a 342 Il Al FYn4lyWevm ) G. Wal Sup4. Sania.n ho.. wnganulrbadove' pil I S WaiteSoplpenlSan.LalLm I E as%ol 3031% 1234% 3013% 2S 63% 13.14% 24 30% 4052% 1230% 3100% 27.09% 1265% 2289% 3021% 1287% 2441% 2.1 Fa *%d6 1206% 1256% 1225% 1001% 1216% 10.10% *.90% 11.63% 10.53% 0 *0% 12.40% a 12% 701% 12 43% S 79% 30 Oas % of 21.15% 01% % is % 1210% 551% _ 047% 3098% 2142% 448% is579% 1033%' 40a7% 1240% 4E 6 % dia 340% a 36% I 53% 3.49% 0 21% 1.21% 4 76% 0 24% 14S% J 45% a 13% 1 32% 376% 032% 139% S fas% OA I 10% 034% 002% 16Y% 026% 055% 171% 023% 0490% 085% 031% 0 50% 087% 031% a 050% 0 Gas%PIA I92% 010%, 06% 156% 010% 062% 065% 001% 019% 265% 011% a01% 203% 012%Y 071% 7 Eao%alOOPoip 020% 004#~ OS12 021S 004% 028% 026% 003% 030% 01l% 004% 023% 01% 004% 023% 6F FP%o GOPmp 000% 004% 013% 0.07% 004% 011% 0.00% 003% 010% 005% 004% 000% 004% 004% G as #%of GOPp a 15% % 0.11% 002% 0 13% 004% 000% 0.04% 014% 002% 0 16% a 10% 001% 012% Noea0 h.m GDPWp S M37 52S a 00 7l 01 S.ems: I. UOGma 0u.,9 OoamUs. GO. 2. F"Aft Aooalor. Deo.lSd d6 IEanOes * t.f A.D_ ~ d E _ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Q th

167 State Government Plan and Non-Plan Expenditure in Social Sectors (Rs. Bilon at current prices) /91 Plan Non-Plan Total Plan Non-Plan Total Plan Non-Plan Total A. Total B. Social Sectors C. B as % of A 19.29% 41.73% 34.84% 26.53% 38.87% 35.10% 23.39% 36.82% 31.71% D. B as % of GDPmp 0.84% 4.10% 4.94% 1.23% 4.08% 5.30% 1.04% 4.01% 5.04% E.HealthandFamllyWelfare F. Waler Supply and Sanilatlion E as % of B 18.66% 18.91% 17.21% 14.84% 14.84% 14.84% 12.86% 19.74% 18.32% 2. F as % of B 14.26% 2.49% 4.49% 16.65% 2.70% 5.93% 25.44% 2 84% 7.48% 3. E as % of A 3.60% 7.06% 6.00% 3.94% 5.77% 5.21% 3.01% 7.27% 5.81% 4. F as % of A 2.75% 1.04% 1.56% 4.42% 1.05% 2.08% 5.95% 1.05% 2.37% 5. E as % of GDPmp 0.16% 0.69% 0.85% 0.18% 0.60% 0.79% 0.13% 0.79% 0.92% 6. F as % of GDPmp 0.12% 0.10% 0.22% 0.20% 0.11% 0.31%1 0.26% 0.11% 0.38% NHoe: Slate Goveivnent Expendiure on Soral Servies Excludes he Grants fromn Central and Centrally Sponsoed schermes. Souwcs: 1. Rol Bauei on State Fhauns RSI 2. Plan Docun. PlannW Conuid"on 3. Wovd Bank Sta Ea

168 State Governmeat Plan and Non-Plan Expediture in So l Sectors (Rs. Bilion at current prices) ioe /91 Plan Non-Plan Total Plan Non-Plan Total Plan Non-Plan Total A. Total B. Socal Sectors C. B as % ofa 12.80% 19.84% 17.39% 18.13% 17.66% 17.80% 16.97% 16.96% 16,96# D. B as % of GDPmp 1.53% 4.44% 5.96% 1.95% 4.41% 6.36% 1.68% 4-29% 5.97% - E. Health and Family Welfara F. Waler Supply and Sanltalon E as % of B 27.50% 16.65% 19.43% 20.31% 14.71% 16.42% 23.50% 19.25% 20.45% 2. F as % of B 16.48% 2.50% 6.08% 15.67% 2.49% 6.52% 16.03% 2.65% 6.43% 3. E asn of A 3.52% 3.30% 3.38% 3.68% 2.60% 2.92% 3.99% 3.27% 3.47% 4.Fas%ofA 2.11% 0.50% 1.06% 2.84% 0.44% 1.16% 2.72% 0.45% 1.09% 5. E as % of GDPrnp 0.42% 0.74% 1.16% 0.40% 0.65% 1.04% 0.40% 0.83% 1.22% B. F as % of GDPrnp 0.25% 0.11% 0.3B% 0.30% 0.11% 0.41% 0.27% 0.11% 0.38%, Souwce: Aaes d~~~~~~~~~~~~~~~~~~~~

169 14 3 Annex I Page 12 of 22 Input Composition of Medical and Public Health Total Expenditure (Revenue and Capital, Plan and Non-Plan), Compared to ) State Comnuodities Salaries Capimal Salaries: Commodities Major States 31.3% 25.3% 59.0% 66.1% 9.7% 8.7% Andhra Pradesh 36.5% 22.5% 58.6% 74.1% 4.9% 3.4% Assam 29.8% 22.5% 58.6% 48.4% 15.7% 17.1% Bihar 20.5% 34.5% 54.5% 74.4% 12.5% 10.9% Gujarat 27.7% 14.7% 67.0% 68.6% 4.7% 6.9% Haryana 25.4% 24.6% 67.6% 73.6% 19.5% 5.3% Kamatalka 28.3% 21.1% 55.1% 65.1% 10.3% 5.8% Kerala 30.5% 23.0% 66.5% 71.3% 3.1% 5.7% Madhya Pradesh 29.7% 23.3% 62.8% 70.6% 7.6% 6.1% Maharashta 31.6% 25.9% 64.1% 66.2% 4.4% 7.9% Orissa 26.8% 14.4% 70.7% 80.9% 2.5% 4.7% Punjab 21.5% 29.1% 64.4% 59.6% 14.1% 11.2% Rajasthan 30.2% 19.4% 64.8% 71.0% 5.0% 9.7% Tamnil Nadu 36.6% 27.8% 51.6% 63.0% 11.8% 9.2% Uttar Pradesh 33.7% 30.0% 61.4% 56.0% 4.9% 14.0% West Bengal 34.1% 28.7% 46.1% 62.7% 19.8% 8.6% Sources: Calculated from Tulasidhar. 1992; Table 4.1 (page 78). Assumes that expenditures in "Other" category (mostly grants-in-aid to district panchayats) are allocated in the same proportion as funds spent directly by the state government. This probably underestimates the proportion devoted to salary and overestimates the proportion to commodities and capital. This information cannot be obtained at the state or national levels; it is available only in distict accounts.

170 144 Annex I Page 13 of 22 Cost Recovery in Medical and Public Health Serices (Non-ESIS) AmL. Rewvered in '8849 as % Stue Avenage of Receipts 15 Major Stes Andhm Pradesh Asum ibar Gunt Hwyana K,uamaka Kaamla Madhya Mabarushua Orissa Punjab Rajauhii 3.9t Tamnl Nadu Utar Pradesh Weon Bengal Souuxe: Tulasidhar, 1992; p.85

171 145 Annex I Page 14 of 22 Basc Rural Health Infrastructure and Some Derivations (as of September 30, 1991) Average Average Average Population Populatfon Population Served by Served by Sved by State Sub-Centre PHCs CHCs Sub-Centre PHC CHC (lakha) Andhra Pradesh Arunachal Pradesh a Assam Bihar Goa Gujarat Haryana Himachal Pradesh Jammu & Kashmnir Kamataka Kerala Madhya Pradesh Maharashtra Manipur Meghalaya Mizoram Nagaland Orssa Punjab Rajasthan Sikdkm Tamil Nadu Tripura Uttar Pradesh West Bengal A&N Islands Chandigarh D&N Haveli Daman & Diu 30 4 Delhi Lakshadweep Pondicherry Note: TOTAL PHC - Primary Health Centre CHC - Community Health Centre

172 Total Numbers of Primary Health Centres and Sub-Centres (per One Hundred Thousand Rural Population) Major Stales PCiCs Sub.CHCa PMC Sub-CHCss PHCs Sub-CHCs PPHC Sub-cHCs PHCC Sub-CHCs PPHCC Sub-CHCs Goo. Damn&Diu Andhra Pradesh Assam Baw Gu4aial ' Haana S Karnalaka A Kerala I Madhya Padush Mahatashka S Olmssa Pungab t Rajiathan Tamd Nadu l Uttat Pradesh WestBengal plhet Slale2 A,unachalPiadesh Mizorman Pondicheiry t HimachadPiadesh Jamnmu 1.95 A Kasui Manlpu MeghabyaV Nagaland Si n Tilpura * b a ALL INDIA i _ Nol: PHCs * Pdmary HeFm Cenites CHCs Community Health Centres

173 147 Annex I Page 16 of 22 Average Population covered by Key Health Personnel/Facilities in Different States, 1987 State PHC Sub-CHC LHV MPWF TBA Andhra Pradesh 43, , Assam 60, , Bihar 55, , Gujarat 59, , Haryana 45, , Himachal Pradesh 25, Jammu & Kashmir 37, , Karnataka 56, , Kerula ' 50, , Madhya Pradesh 48, , Maharashtra 32, , Orissa 37, , Punjab 7, ; Rajasthan 51, , Tarni Nadu 46, , Uttar Pradesh 50, West Bengal 34, , All India 40, , Note: PHC: Primary Health Centre Sub-CHC: Sub-Centre LHV: Lady Health Visitor, also known as Heaith Assistant MPWF: Multi-Purpose Worker - Female, also knows as Auxialiary Nurse Midwife TBA: Trained Birth Attendant ' Kerala disontinued the Dai Training programme in SOURCE: Computed fom data provided in Health Information India, 1987, Central Bureau of Health Intelligence, Ministry of Health and Family Welfare, New Delhi, India, 1988.

174 148 Annex 1 Page 17 of 22 Total Number of Hospital Beds (per One Hundred Thousand Population) Major States Goa, Daman & Dlu Andhra Pradesh Assam Bihar Gujarat Haryana 5z Z Kamataka Kerala Madhya Pradesh Maharashtra Orlssa Punjab Rajasthan Tamil Nadu Uttar Pradesh Webst Bengal atr states Arunachal Pradesh 25Z Mizoram Pondicheny Himachal Pradesh Jammu & Kashmir Manipur Meghalaya Nagaland Sikkim Tripura ALL INDIA Source: World Bank 1992

175 149 Annex 1 Page 18 of 22 Total Number of Rural and Urban Hospital Beds (per One Hundred Thousand Rural and Urban Population) Major States Rural Urban Rural Urban Rural Urban Goa, Daman & Diu Andhra Pradesh Assam Bihar Gujarat Haryana Kamataka Kerala Madhya Pradesh Maharashtra oissa Punjab Rajasthan Tamil Nadu Uttar Pradesh West Bengal COther Sowh Arunachal Pradesh Mizoram Pondicheny Himachal Pradesh Jammu & Kashmir $ Manipur Meghalaya Nagaland Sikkim Tripura ALL INDIA Source: World Bank 1992

176 150 Annex 1 Page 19 of 22 Medical Education Infrastructure in India (Allopathic Doctors, Dendsts, and Nurses) Medical C0lle2W Private Dental Nursing institutions Year Number (percentage) colleges B.Sc. General n.a na n.a n.a. 40 n.a. na. SOURCE: Health Information of India, CBHI, Govemment of India relevant years. Medical Educaton in India, CBHI. Govemment of India, Handbook of Medical Education In India, Association of Indian IVrWeA 1967

177 151 Annex 1 Page 20 of 22 Outturn of Medical Personnel in India 1951/52 to 1987 Postgraduates (Allopaths and Nurses Year Allopaths Dentists dentists) B.Sc. General ,190' , ,197' , , n.a , n.a n.a n.a n.a n.a. na. *Data on the outtum of allopaths was not received from medical colleges in , one in , two in , six in , seven in fourteen in , fifteen in , and twenty:five in Thus, the data is qrossty underrated. SOURCE Health Information of India, CBHI, Govemment of India, various years.

178 152 Annex I Page 21 of 22 Comparison of Migation of Doctors to Annual Outturn of MBBS Students, Migration Net Migration Yeas Outturn Number Percentage Number Percentage (32.35) 4650 (29.15) (2,504) (810) (730) (32.25) 4784 (29.81) (3,210) (1,035) (957) (21.57) 4813 (19.54) (4,927) (1,063) (963) (27.17) 6800 (25.66) (6,624) (1,800) (1,700) (27.67) 7950 (14.24) (11,164) (3,090) (1,590) , (33.62) (16.91) (12,670) (4,260) (2,142) , (11,718) (4,699) (2,342) , (10,383) (5,164) (2,643) * Na migration equab number of migration minus number of remued. ** Data on the outtum of allopaths was not received from two medical colleges in , one in , two in , aix in , seven in , fourteen in , fifteen in , and twenty-five in Underrated. SOURCE: 1AMR, Brain Drmin Study: Phase 1-Analysis of ordinary passports issued during ', pg. 23. CBIl Health Stastics of India', relevant years. Rele M..R. and R.S. Bali, 'Study of Brain Drain of Indian Pbys-iant-, NIHAE (und-ed)

179 153 Annex 1 Page 22 of 22 Human Resources Power Employed in Rural Areas as of end March 1990 Medical Doctors at PHCS Period up Doct at PHCs to which Number Number in Vacant information StatelU.T. sanctioned position posts relates to Andhra Pradesh :03.87 Arunachal Pradesh NIL Assam NIL Bihar NIL Goa, Daman, and Diu Gujarat Haryana Himachal Pradesh Jammu and Kashmir NIL Kamataka 945' 81Br Kerala NIL Madhya Pradesh Maharashtra Manipur Meghalaya Mlzoram Nagaland NIL Orissa Punjab Rajasthan Sikkim Tamil Nadu Tnpura NIL Uttar Pradesh West Bengal A and N Islands Chandigarh 7 7 NIL Dadra & Nagar Haveli 8 8 NIL Delhi 6 6 NIL Lakshadweep 9 9 NIL Pondicherry NIL Total / Revised figure received from Kamataka State. Revised figure received from Pondicherry in respect of doctors at P.H.C. NOTE: Figures are provisional.

180 Table 2-1: Study Sites - Main Characteristics - Districts OTHER CHARACTERISTICS Population Development Infant Percent of Agr.: Per Monetization: Urbanization: Stale 1991 Index Survival Literacy Couple Population Capita Value Per Capita Percentage (0o0) Rate Protection Having Access of Output of Bank Deposits of Population (Districts) (Make into to Safe 26 Major Outstanding to Total IMR) Drinking Crops - Av. As on Last Population Water of 3 Years Friday 12/86 Ending Guiarat 1. Panch Mahals Bhavnagar Valsad Tamil Nadu 1. Dharmapuri South Arcot Thanjavur Uttar Pradesh 1. Azamgarh Almora Meerut West Bennal 1. Puruliya Medinipur Buddhaman N.A Source: Column Census Census 5 - Family Planning Year Book, 1986, Ministry of Health & Family Welfare Census 7 - Centre for Monitoring Indian Economy, Bombay; p Centre for Monitoring Indian Economy, Bombay; p Cemus k oo

181 Table 2-1: Study Sites - Main Chbacteristics - MunicipaIitIes State Population Civic Sex Literacy Raue (1981) Workers (1991) Bank Per Sta(us Ratio % (% to total population) Offices Capita (Local Body) of local (Females per Banlc 1991 Growth Body ' per 1000 per Bank Rate Males) Male Female Total Main Maginal Total Popula- ('000 ('000) ( ) tion) Rs.) Guiarat 1. Rajkot MC Dhrangadhra M 929 N.A. N.A. N.A. N.A. N.A. N.A. Tamil Nadu I. Coimbatore M.Corp Kovilpatti M 1011 N.A. N.A. N.A. N.A. N.A. N.A. N.A. N.A. Untar Pradesh U' 1. Allahabad MC Nawabganj MB 882 N.A. N.A. N.A. N.A. N.A. N.A. N.A. N.A. West Beneal 1. Haora MC Chakdaha M 965 N.A. N.A. N.A. N.A. N.A. N.A. N.A. N.A. Revised 9/24 I/ As per the Cenu 1991, the civic satus of urban units included in this s is i ted : M Municipalit; = MB = Municipal Board; MC = Municipal Comminec; M. Corp - Municipal Corporaion. Source: Baking Statistics, Basic Statistical Returns, Vol. 18. June 1909, po dia i Februay 1991, RBI.

182 156 Annex 2 Page 3 of 28 Summary Statistics: Sample Districts and Municipalities Suae Popuiauon Development Infan Mormhy Limntcy % (Districts) 1991 Index Rate ('000) (per 1000 live birdn) DISTRICTS Gujarat PanchmahaLs Bhavnagar Valsad Tamil Nadu _ Dharmapun South Arcot Thanjavur Uttar Pradesh Azamgarh Almorah Meerut West Bengal Puruliya Mediipur Barddhaman MUNICIPALXTES L Gujarat _ Rajkot (MC) Dhrangadhra (M) 54 NA Tamil Nadu _ Coimbatore (MCorp.) _ Kovilpari (M) 78 NA Uttar Pradesh Allahabad (MC) Nawabganj (MB) 65 NA West Bengal Haora (MC) Chakdaha (M) 75 NA MCorp. = Municipal Corporation MC = Municipal Commitee M - Municipality MB = Municipal Board Sources: MR and literacy rac: 1981 Census

183 157 Annex 2 Page 4 of 28 Recurring Expenditure on Health and Family Welfare by District STATE LEVEL DISTRICT LEVEL State Per Capita Per Capita Expenditure Total Per Capita As % of State (Districts) Net Domestic Expenditure on Health and Health Health Per Capita Product on Health and Family Expenditure Expenditure Expenditure Family Welfare as % (Rs. '000) (Rs.) (Col. 5/2) (rs.) Welfare of NDP Gujarat Panchmahals 108, Bhavnagar _ 92, Valsad 120, Tamil Nadu _ Dharmapun 101, South Arcot _ 191, Thanjavur 253, Uttar Pradesh Azamgarh _ 75, Almorah 62, Meerut 155, West Bengal Puruliya _ 81, Medinipur 272, Barddha2man 220,

184 Recurnug Public Expenditure for Health md Famiy Welfare: Sources by Major Head MEDICAL' PUBLIC IIEALTISA FAMILY WELFARE SLua Centre Statc Total (D istrkics) Ceatre Sle Total Ccenle Stale Total _ Gujarat Panch Mahals Dhavnagar Valsad Tamil Nadu Dharmaburi t South Arcot Tlianjavur IQ Utlar Prudesh Azamgarh I.5 Almora I.5 Meerut co West Oengpl_ Puruliya I O.00 Mcdinipur farddbaman _ I Public health expendltures include both central and statc funds. The allocation was calculaied by estimating for each scheme the anmunl that would of central be provided funding according to the terms of central granis, hen tolalilng ibe anount of estimated central and stale conaribulow. NTCP central Por gram example, supporns if ihe 50% of the cosis of drugs and supplies, half of dhat item was allocaled to cemral funds, the rest to state. 2 Family Welfare is a I00% cenlrally funded scheme. Stales augment cenral funds lo a limiled extent in state plans. It is impossible distinac to levels differentiatl between stale at and central groups of fluns so ihe se kvel bmakdmw ar used here. oo

185 159 Annex 2 Page 6 of 28 Recurring Public Expenditure on Health and Family Welfare: Districts FACILITY SCHEME SEat Hospital- Non- Total Medical Public FaMiLY Total (Distict) Based Hospital Healt Welfare Based Row % Gujarat I I Panchmabals Bhavnagar Valsad S Tafmil Nadu Dharmavur South Arrot Thanjavur Uttar Pradesh _ Azamgarh Aimora I.0 ioo.oo Meenr West Bengal _ Puniiiya Medinipur Bardchbam Municipaliies _ Gujarat _ * Rajkot (MC) Dhra=gadhra * (M) _ I I Tamil Nadu i Coimbatore (M. Corp.) _ Koviloarn (M) _. I I Uttar Pradesh I I I I Allahabad (MC) Nawabgaj (MB) West Bengal I I 2I8_ Haora (MC) Chakdaha (M)

186 160 Anex 2 Page 7 of 28 Recurring Public Expenditure on Health and Family Welfare: Inputs PER CAPITA (Rf) % TO RECURRING EXENDITURE Siace 0 Salary Drugs & INon-SaIay Slary Drugs & Non-Saary (Dluric) Supplies Supplies Districts Gujarcmat Bhavna_r Vaisad Tamil Nadu. Dharmapri South Arcot Thanjavur Uttar Pradh _ Azamgah k Almom Me_U West Begal. Puruliya Mediipt u U Bddhaman Municipalitfes G ujarat _ I _ Rajkot(MC) Dhransadhm (M) Tamil N2da. Coinbaore (M. Corp.) Kovilpufii (M) Uttar Pradah Allahabad (MC) Rawaba (MB) West Bengal 46_ 76_ 5 _0_22_44 Haom Chskeaba (M)

187 161 Annex 2 Page 8 of 28 Table 2-7: Staffing Norm for Non-Hospital Services 9.1: PHCs and Sub-Centres I/ No. of Posts A. Staff for Sub-Centre 1. Health Worker (Female) ANM Health Worker (Male) Voluntary Worker (paid 0 Rs. 50/- per month as honorarium) -1 B. Staff for New Primary Health Centre 2V 1. Medical Officer Commnunity Health Officer (CHO) Pharmacist Nurse Mid-wife (Staff Nurse) Health Worker (Female)/ANM - I 6. Health Educator - l 7. Health Assistant (Male) Health Assistant (Female)/LHV UDC LDC Lab. Technician Driver (subject to availability of vehicle) Class IV 4 16 Note: The data presented on staffing pattern for PHCs is inclusive of the sub-centre staff. It helps interstate comparison, for in some cases (particularly in Gujarat and West Bengal) it was difficult to separate out the sub-centre staff from that of the PHC in the paramedical category. The staff at the sub-centre is treated as the paramedical 'field staff'. Thus as per the norm of Government of India, the staff for a PHC including the sub-centre under its jurisdiction, is taken as shown in 9.2. I/ Tbese norms are given in various issues of Bulletin on Rural Hea1th Statistic, quarterly publication of the Rural Health Division. Directorate General of Health Services, Minbtry of Healtb and Family Welfare. New Delhi. _I For every population in plain area and 20,000 populadon in tribal, hilly and backward areas.

188 162 Annex 2 Page 9 of 28 Table 2-7: Staffing Norm for Non-Hospital Services 9.2: PHC Staff as per Norm (Including Sub-Centers) Staff Category GOI PHC Norn PHC + Sub-Centre Designation No. Norm No. 1. Doctor 2 (including CHO) - Medical Officer 1 ) - Community Health 1 ) Officer 2. Staff Nurse - Staff Nurse Paramedical ) 5 - Centre - Phannacist ) or ) - Compounder 1 ) - Health Officer 1 ) Feemale/ANM - Health Assistant 1 ) (Male) ) - Health Assistant 1 ) (Female) or LHV ) - Lab. Technician 1 ) - Field - Health Educator 1 ) (1 + 2N) 4. Administration - General ) 2 - LDC 1 ) - UDC 1 ) - Others 5 - Driver 1 (subject to availability - Class IV of vehicle) Note: Voluntary workers paid Rs. 50/- per month are not treated as part of regular staff at the Sub-Centre. N = Number of Sub-Centres in a PHC.

189 163 Annex 2 Page 10 of 28 Table 2-7: Staffing Norm for Non-Hospital Services 9.3: CHCs Staff CateNorvo. 1. Doctor 1/ 4 2. Paramedical 2/ Non-medical 3/ 8 (Administrative & others) Total 25 /f Either justified or specially trained w work as surgeon, obstetrician, physician and pediatrician. One of the existing Medical Officers similarly should be either qualified or specially trained in Public Health. 2/ Paramedical includes nurse mid-wives (7), dresser (1), pharmacist (1), lab. technician (1), radiographer (1) and ward boys (2). Non-medical includes washerman (1), sweepers (3), gardener (1). watchman (1), aya (1). and peon (1).

190 Public Provisions in Sample Districts: Facilites. FACILITIES ospilal Non-Hospital CHC's spiic's Sub-Centres Block Level or Additional All PIIC's M ain _- No Beds Beds per No. Beds No. of No. of No. No. per No. No. per No. of No. No. per No per 1000 beds per CHC's per 1000 rural 1000 rural PHC's per 1000 rural disirict 1000 rural populalion population populalion populatio population populaiion population p. ulation_ Gatjarat _.. Pancbt Mahal_ BIhavnagar Valsad Tamil Nadu _ Dharmapuri South Arcot Tlanjavur Pradesh Azamgarh la AImora , Meerul i West Bengal Ptiruliya ;a Medinipur Barddhaman Average = =_=_=_.22 * For hilly and tribal districts, the norms are IPHC per population and ISC per 3000 population. = 1.04 I

191 165 Annex 2 Page 12 of 28 Public Provision in Sample Municipalities: Facilities and Staff BEDS AND UNITS Hospuai Non-Hosp:azi Sume o. lbeds Bds per Unis Unit per (MuCicipalky) 1000 Popuazion 1000 Populaton Gujarat Rajkot (MC) NA 856 I.S4 NA NA Dhmngadhn (M) NA NA NA Tamii Nadu Coibatomre (M Corp.) 3 1,S Neg. Kovilpa (M) Uttar Pradesh. Allahabad (MC) 22 1, Nawabganj(MB) West Bengal Hson (MC) I Ng Chakdaha (M) NA NA NA

192 Pubic Sector Healt Staff In Sample Dbitrcts Number Tobal Per 'M pulaltiom = lospitsi oae-ilospkal llospital on-hlospital Culnr Statc Doc NHon-Docaor Support Doctor Noe- suppot Doctr Hoa-Docior Suppon Doctor Non-Doctor (Distriets) Support (Purued wdocor 1 Staff Staff (Paamedi Staff (Pauud. I Staff Including (Non- (P&f&jnel (Non- Including (HoR- including (Nonnuse) medical or kackadaig medkcal or nwsc) Iedical or nuxse) medical Adinb.) nmm) Admin.) Admin.) of. I- Admin.) Pancltmhals ibhavnagar Valsad Tamil Nadu -_.- Dharmapuri South Arcot S Thanjavur Utlar Pradesh Azamgarh I 1tS Almorah Meerut W est Bien gal Puruliya I1I _- Medinipur Buddhaman

193 Public Sector Health Staff in Sample Municipalities Tolal Number lhospilal Per 'o00 oppslatlion Non-llospital I ospilal Non-I lospitai (Municipality) Doctor Non- Support Doctor Non- Support Doclor Doctor per StafTNon- Non-Doctor Support Doctor Staff Doctor Staff per Non- Non-Doclor 1000 Support (Parumed. Staff Non-medical or 1000 (Paramed. 1 (Paramed t medical or (Puamred t medical or Popultion Non-medical including or Including Admin. Adminisiration Including Population Admin. including Adminisiration nuse) nure) (Ocoeral (General and nurse) (General nurse) (General and pcr and 1000 Others) Per and per 1000 Popuation Others) Others) 1000 Populalion Others) Population Per 1000 Population Cujarat. Rajkot (MC) Dhrangadhra NA Neg NA 0.01 NA 001 (M) NA NA NA -a Tamil Nadu Coimbalore 159 I (M Corp.) Kov.ipatti (M) I Uilar Prailesh Allahabad (MC) _ Nawabg( j (MD) I W est Blengal Ilaora (MC) Chakdaha (M) NA NA NA NA NA NA. b I Paramedical includes Centre as well m field stalf.

194 168 Annex 2 Page 15 of 28 Staffrng Gaps in Public Health Facilities CHC's Main + Additona Main PHC's Additional PHC's PHC 's DismCis Acu SiCUlowd AuAJ SaC-gAd AcnM Samnco AnAaU Saifm Norm amfnwdl Nonm we IId/ Norm u Norm nn m fil Nom_ Norm Noam Norm Gujarat Panchmaha2l _ Bbavnawar Vatsad [ Tamil Nadu Dharrnapurm South Arcot Tbanjavur NA NA Uttar Pradesh Azamgarh AUmora Meerut West Bengal Pubuliya M e d i r p u r NA NA I BRsdrb:harnan Average i

195 Staffilng Levels (Actual and Sanctioned) Relative to Nornm for Different Types of Staff and Faciities CIIC_ Main PIIC _ Additional Doecons PlCs hnmeds - crarc Dociors paramds - ceae hrameds - Iield Docort Pararneds - certie AMN Parameds (S-A) - ficid AMN (S-A) A/N (S-A) AIN (S-A) AIN (S-A) AIN (S-A) A/N (S-A) IN AIN _ (S-A) IN IN _ IN IN _ IN I N.. ~ ~ ~ ~~~~IN Gujurat _ Panchmahals O.( Bhavnagar Valsad Tanlii Nadu _. Dharmapuri South Arcol l,0( Thanjavur _, Utlor Pradesh Azamgarh AImorals Meetit WYest llengal Puruliya Medinipur Blarddliaman Average A/N - actual slaff In place dividd by the norm for hat type of staff and facility (S-A)IN sa-etloned bul unmiled posts divided by th norm for llua typc of staff and faclityc Adding togher (AIN) a (S-A)/N gives the total funde sltff podsion relalive to norm.

196 170 Annex 2 Page 17 of 28 Adequacy Score for Drugs and Supplies*: All PHC's (Main + Additional) DRUGISUPPLY CATEGORY Sate (Disuic) Geneml Emercency Vaccmca and Contraceptives Totai Scores Rciauve w Medicines Medicines Supplements Maximum ~~~~~~Score Gujarat _ PawAcmahais Bhavnagar Valsad Tamil Nadu Dharmapun Soutb Arcot Thanjavur Paas Ute Uft, Pmdhs_ Azgarh Almorab NA - - Meerut West Beagai Pur_lya Medinipur Barddhianan Maximum Scarm (Per PHC ) The drug adequacy score was based on PHC medical officers' opinons obtained during interviews at a sample of facilities visited in each distict. For each class of medicines, officers were asked about several indicator items. Thme were ranked according to whether the item was unavailable or out of date (0), inadequate in supply (1), or adequate in supply (2). _ Notes 1. Since the number of items listed under each drug/supply category is different, the maximum score varies across categories. Ibe number of items listed under each of the four above categories are: 10, 4, 7, and 3 respectively. The maximum score for a siaglc item is Since the number of sample PIIC's varies considerably across disrict due to population size and other factors, the maximum score (last row) is standadized for it.

197 171 Scores on the Adequacy of Drugs & Supplies: Districts Gujama Annex 2 Page 18 of 28_ Tamil Nadu Category of Drugs PM 8a,sVMk0h5 5a hs~ General Medicines A&nalgesticsiAntipyratirs ntimalari2is Anttbiotic is Antilistafflies A.nddiarnhocal i5 I I I1 12 Sedative A.ndasthmztics Andhypemstev/Dliuretics Antispasmodics ORS Total (16.88)(16.38) (17.75)(16.30) (15.06) (15.2) Emergency Medicines Corticosteroids Oxygen oxytocia ~~ I.V. Fluids Tots.! (3.25) (6.50) (3.63) (3.30) (4.18) (4.60) Vaccines & Supplies BCG Polio DPT TT ~~~~~15 is Measles Vitamin A F.A. Tablets Total (12.38) (12.63) (13.13) (11.00) (13.40) (12.80) Contraceptives i.u.d oral Conzceptives (O.C.) Nirodhi s Total (5.15)(5.38) (5.25)(5.00)(3.30)(5.80) GRAND TOTAL (37.75) (40.89) (31.75) (35.70) (35.90) (38.40)

198 172 Annex 2 Page 19 of 28 Uttar Ptadesh West Bengi Category of Dugs /m;un Akmn Mm Pwsy Mpw aij General Medicines An&lgesics/Antpyretics Antimalarifls Antibiotic Arntistanies Antidiarrhoeua S 11 Sedatve A&niasthmnlcsr Antihypertnsives/Diureics Antispasnodics OFLS TotI (8.80) (14.40) (10.40) (8.20) (7.10) Emergency Supplies Corticosteoids oxygen S Oxytocin I.V. Fluids Total (2.70) (3.10) (3.14) (1.80) (2.10) Vaccines & Supples BCG Polio DP-T TT Measles Vitamin A IS I.FA Tablet Total (9.40) (13.10) (11.90) (12.10) (12.30) Contraceptives I.U.D omrlcontaceptives(o.c.) Nuxnvh Total (4.00) (5.90) (5.90) (5.70) (5.40) GRAND TOTAL (24.90) (36.50) (31.30) (27.90) (26.90)

199 173 Annex 2 Page 20 of 28 Notes: 1. Scoing paum for adequacy for drugs and supplies. Abash or nutofdate0; ina u1 l andtaequaim-2 2. Scor aggan an individuitem Ls the toalscore received from WI sazple PHCL 3. Numibe in parenthesis are the score per PHCs, i.c Toml Scorc divided by the No. of PECs GCanz 8 in each distric. Tamil Nadu.: Dhamapuri-6; South Arcot-7 and Thmnjavur-8. Uttar Pradesh Azmgarh-9; Almorah (Cmfomon not obrnned); and Meau-8. West Bengal: Purliya-7; Medinpur-14; and Barddhaman-lZ

200 Estinated Financial Cost of Mising Health Facilities in Sample Districts State CHCs PHCs SCs Annual Annual Total Current Facility Facility (Districts) needed needed needed capital recurrent facility total finance finance cost cost finance annual gap as gap as Rs. Rs. gap recurrent percent of percent of Millions Millions cost current current Rs. spending spending Millions level (rull level capital and (annual recurrent Gujarat recurrent cost) cost only) Gujarat Panchmahals % 45% Bhaynagar Is Valsad Tamil Nadu i Dharrnapuri South Arcot Thanjavur NA Utlar Pradesh _ Azamgarh Almora Meerult West Bengal I Puruliya Medinipur NA Barddliaman s19 ' Facilities needed estimated as gap betwecn current level and completing facility norm for CHCs, PHCs, and SCs (sec Table 4.7). Estimates for Panchmahals and Almora usc the higher figures per population for hilly and tribal districts. Fraclions of arc facilities rounded needed d&=n to next integer. n Capital cost estimaled as Rs. 7,500,000 per CHC, Rs. 2,500,000 per PIIC, Rs. 500,000 per SC. Recurrent cost estimated as staff full complement according to norms, costed at stat estimate for average salary for each type of staff.. Non-salary costs added in as equail; to 30% of salarv cnsis for each facility

201 175 Annex 2 Page 22 of 28 Staffing Resource Gap in Financial Terms for District Level Non-Hospital Based Services: CHC's and PHC's Combined Gap (Rs. 000) Gap as % of Toai I I (Hospital& Non-Hospital Expendintre) State (Disuic) N-S S-A Tot N-S S-A Tota _ Gujarat Pancbmabals Bhavnagar Valsad L Tamil Nadu Dharmapuri _ ls. l South Arcot Thanjavur Uttar Pradesh Azamgarh ALmorah Meerut West Bengal Punrliya Medinipur Barddhaman _ _S.2 Notes: 1. PHC figures incude missing personnel only for existing sub-centers, not for sub-centers which have not been created yet. 2. N-S = the difference between the rupee cost of the normative stafflng levels and the amount of staff actually sanctioned, valued at the average salary cost for each level of staff. 3. S-A = the difference between the rupee cost of the sanctioned stamng level and the amount of staff actually In place, valued at the average salary cost for each level of staff. 4. Where staff levels exceed norms, the financial value of excess staff has been used to reduce the staff finance gap estimate, i.e. this is a "net" estimate of finance required to meet personnel norms.

202 176 Anner 2 Page 23 of 28 Index of Programme Performance: District Level State (Dismac) Medical Public Healkh Family Welfb Ovouul Mm Scare Panchmahals Ibavnag VaLud Tamil Nadu _ Dhrmapmi South Arcot Thanjavur Unar Pradesh Aznmgarh Almon Meert Ii 1.5 W est Daspi P'zliya M.dinip 1.0I Bwddhman

203 177 Annex 2 Page 24 of 28 Table 2-18: Sources of Data on Inputs and Output - District Level Guiarat Tamil Nadu Urar Pradesh West Benzal HB NHB HB NHB HB NHB HB NHB L Financial Resources MS (G) RDD (CHC) M DDHS MS (H) CMO DD (AAU) DD (H) AD (H) DHO (MS) MS (F) MS (i) DD (M) DD (A) DLO D (MC) D (SB) MS (P) MS(G) CMO DAO LS (MC) JD (FP) DAUO CMO MO (E) DO (TB) P (MC) DO (TB) STBS CHC DLO B (PH) DMO DD (A) MO (PAC) ZLO U. lnputs (Real terms) (i) Staff (No.) -do- -do- -do- DDHS MS (H) -do- MS (H) CMO MS (F) DCMO MO (E7 CHC, (PHC) (fi) Drugs/Supplies -do- -do- -do- -do- MS (M) -do- DD (CMS) CMS In-kind) MS (F) DCMO D (SB) DD () (ii) Logistics -do- -do- -do- -do- MS ) -do- CMO C-MO (Inchlding MS (F) -do- DD (CM) DD (AA) equipment etc.) I Output (i) Medical Inpatient/ -do- -do- -do- DDHS MS (M) -do- MS (H) MO (CH) outpatients (Nos.) MS (F) PHC.B PHC (ii) Public Health DHO DHO DO(TB) DO (TB) D (SB) DO(TB) MS (H) DCMO a. NTEP DO (TB) D (SB) DO (PHC P (MC) CHC,B PHC) b. NCBP -do- -do- JD (MS) MS (I) CMO -do- CMO LS (MS) MS (F) D (SB) DCMO D (SB) c. NMEP -do- -do- DDHS DDHS D (JB) D (1B) -do- DCMO DMO DMO DD (M) DMO MO (B. P (MC) PHC) CHC PHC d. NLEP -do- -do- DLO DLO D (SB) DLO ZLO ZLO DLO CMO P (MC) D (SB)

204 178 Annex 2 Page 25 of 28 Table 2-18: Sources of Data on Inputs and Output - District Level Guiarat Tamil Nadu Uttar Pradesh West Bengal HB NHB HB NHB HB NHB HB NHB (iii) Family Welfare a. MCH -do- DHO DDHS DDHS JD (FF) CMO MS (H) DCMO MS (F) JD (FP) DCMO MO (BP) D (SB) PHC, P (MC) CHC b. UIP -do- -do- DDHS DDHS JD (FP) -do- -do- -do- MS (M) MS (F) P (MC) c. FP -do- -do- DFWB DFWB DFWB -do- -do- -do- Note: See list of abbreviations attached. eih: HB: NHB: Hospital-based Non-Hospital-based

205 179 Annex 2 Page 26 of 28 Table 2-18: Sources of Data on Inputs and Output - Municipality Guiarat Tamil Nadu Uttar Pradesh West Bental HB NHB HB NHB HB NHB HB NHB Financial Resources MS (M) DHO LS MS (ESI) P (MC CMO P (DAUH) CMD MS (F) P (MC) MC (C) MS (M) MC CMO C (M) AD (H) D (ESI) MD MS (F) MS(CG MS (J) D (CM) DAO MC PO (TB) HS) MS (P) MS (E) MO (P) P (DAUH) CMO (R) MS (G) JD (MS) DAUO MS (J) MS (PHC) MO (PAC) MS (EST) DMC II. Inputs (Real terms) (i) Staff (No.) -do- DHO -do- -do- -do- -do- MS (G) -do- MS (DPHC) -do- (ii) Drugs/Supplies -do- DHO -do- -do- -do- -do- -do- -do- (In-kind) (iii) Logistics -do- -do- -do- -do- -do- -do- -do- -do- (Including equipment, etc.) in. Output (i) Medical Inpatients/ -do- -do- -do- -do -do -do- -do- -dooutpatients (Nos.) (ii) Public Health a. NTOP DHO DHO -do- -do- DO (TB) DO (TB) NS (BPHC) C (M) CMO D(CMO) P (MC) b. NBCP DHO DHO -do- -do- MS (Nl)) CMO -do- -do- MS (F) P (MC) P(MC) MS (E) c. NMEP DHO DHO -do- DMO DMO -do- -do- D (IB) P (MC) d. NLEP DHO DHO -do- DLO DLO -do- -do- D (SB) P (MC)

206 180 Annex 2 Page 27 of 28 Table 2-18: Sources of Data on Inputs and Output - Municipality Guiarat Tamil Nadu Uttar Pradesh West Benial HD NHB HB NHB HB NHB HB NHB (iii) Family Welfre a. MCH DHO DHO -do- MS (F) CMO MS (G) C (M) P (MC) MS (E) MS (BPHC) D (CMO) D (SB) JD (FP) b. UIP DHO DHO -do- MS (M) -do- -do- -do- MS (F) P (MC) JD (FD) c. FP DHO DHO -do- -do- -do- -do- -do- Note: See list of abbreviations attached. li: HB NHB Hospital-based Non-Hospial-based

207 181 Annex 2 Page 28 of 28 Table 2-18: List of Abbreviations 1/ MS (M) MS (F) CMO (R) D (HS) D (MC) CMO LS P (MC) DD (H) DHO JD (H) JD (MS) JD (FP) DCMO P (DAUH) DAUD DD (HM) MS (CGHS) MS (J) MS (P) MO (P) DMO DDM DO (TB) STBS DLO ZLO DFWB D (SB) D (JB) AD (H) RDD (CHC) DD (AAU) MS (H) DD (A) DDHS MS (PHC) FWC (U) MC C (M) DAO MS (E) MS (ESI) MO (PAC) DD (CMS) Medical Supdt., Male Hospital Medical Supdc., Female Hospital Chief Medical Officer (Railway Hospital) Dean, Health Services Dean, Medical College Chief Medical Officer Lay Secretary, Medical College Principal, Medical College Deputy Director, (Heath) Distt. Health Officer Joint Director, (Health) Joint Director, (Medical Services) Joint Director, (Family Planning & Social Welfare) Deputy, Chief Medical officer Principal, Distt. Ayurvedic & Unani Hospital Distt. Ayurvedic & Unani Hospital Deputy Director (Homeopathy) Medical Supdt., CGHS Dispensary Medical Supdt., Jail Hospital Medical Supdt., Police Hospital Medical Officer, Police Hospital Distt. Malaria Officer Deputy Director Malaria Distt. T.B. Officer Supdt., T.B. Sanatorium Distt. Leprosy Officer Zilla Leprosy Officer Distt. Family Welfare Bureau Director, Swastha Bhawan, Lucknow Director, Jawahar Bhawan, Lucknow Additional Director, Dept. of Health, Gandhinagar Regional Deputy Director, CHC, PHC, BPHC Deputy Director, Allopathy, Ayurvedic & Unani Medical Supdt., Hospital Deputy Director, Ayurvedic Deputy Director, Health Services Medical Supdt., PHC, BPHC Famnily Welfare Centre, Urban Municipal Commissioner Chairman, Municipality Distt. Ayurvedic Officer Medical Supdt., Eye Hospital Medical Supdt., ESI Hospital Medical Officer, PAC Deputy Director, Comptroller of Medical Supplies j/ Refer to Tables 2A and 2B.

208 182 Annex 3 Page I of 42 Annual Incidence of Morbidity (Per 1000 Populaion) During and Compared with Annual Rate of Prevalence of Morbidity Estimated by NCAER* During May-July 1990 and 1991, by Rural and Urban Residence. Stae Rmal Areas Urban Areas * 1991** Gujarat Mahamshua Tamil Nadu Utar Pradesh West Bengal Madhya Pradesh Gwalior 2052 Datia 2256 Utar Pradesh Mathura 2424 Hardoi 2280 Rajasch Alwar 780 Tonk 1788 * NCAER Survey on Morbidity during May-July 1990 with a reference period of two weeks preceding the dat of suvey. NCAER Survey on rural household health care need, using a 1 month reference period. Data were colited during the three seasons in 1991 to correct for seasonality.

209 183 Annex 3 Page 2 of 42 Illness Rates for Specific Diseases NSS 42nd Round and Recent National Estimates Incidence (Per 100,000 Population) Gujarat Maha- Tamil Uttar West National rashtra Nadu Pradesh Bengal Rural Areas TB Malaria Diarrhoea Urban Areas TB Malaria Diarrhoea Total TB 220 Malaria 785 Diarrhoea Note: Source for national data is the GBD data base.

210 184 Annex 3 Page 3 of 42 Treatment of Illness Out of Hospital Relative Roles of Govt. and Non-Govt. Providers (Percent of all episodes taken to each type of provider) Gujarat Maha- Tamil Uttar West rashtra Nadu Pradesh Bengal Rural Areas Govt Non-Govt Urban Areas Govt Non-Govt

211 185 Treatment of Ilness by Age and Sex (Percent of episodes taken to each type of provider) Annex 3 Page 4 of 42 ua Mahrh Uttar Pradesh Tamil N ati West Bengal G NG G NG G NG G NG G NG Rual Areas < > males Fenmles Urban Areas < > Males Females

212 186 Use Patterns by Household Expenditures Class (Percent of episodes taken to each type of provider) Annex 3 Page S of 42 G;ujarat Mabarsa Uffr Prades Tamil Nadu Wet Rengal G NG G NG G NG G NG G NG Average Overal MPCE Rural Areas I II m IV V Urban Areas I II DI IV V Average by Type of Provider Rural Areas I m IV V All Urban I m IV V AU Areas Note: MCPE = mean per capita expendinre I is lowest expendinmre clss, V is highest.

213 187 Percentage of DJness Treated Outside Hospital by Health Care Provider in the Public and Private Sector by Group of Diseases, Annex 3 Page 6 of 42 Gujanit Mahrashtra Tamil Nadu % Share of % Share of % Sham of No. Public Private No. Public Private No. Public Private Rural Areas I Infections * * I Dysentery * * * 2 Typhoid * * * 3 TB * Malaria * * Restof * II Neoplawm I 00.0Q LUI Endocrine * * v Mental * " VI Nervous * * * vu Circulazory * VIII Respiratory * IX Digestive * * S X GU @ )a Pregnancy * * * xii Skin * IV/XIII Blood, etc * * * All * * *

214 188 Percentage of Ulness Treated Outside Hospital by Health Care Provider in the Public and Private Sector by Group of Diseases, Annex 3 Page 7 of 42 Gujarat Mahazishtra Tanmil Nadu % Share of % Share of % Share of No. Public Private No. Public Private No. Public Private Urban Areas I Infections * * Dysentery * * 2 Typhoid * TB * 4 Malaria g.9* * S Restof * * * II Neoplasm @ @ m Endocrine * * V Mental * * * VI Nervous * * VII Circultory * * * Viii Respiratory * * * IX Digestive * * X GU @ * x Prgzacy XII Skin * * IV/=XI Blood, etc * * * AU * * *

215 189 Percentage of Illness Treated Outside Hospital by Health Care Provider in the Public and Private Sector by Group of Diseases, Annex 3 Page 8 of 42 Uttar Pradesh West Bengal % Share of % Share of No. Public Private No. Public Private Rural Areas I Infections * * I Dysentery * * 2 Typhoid * * 3 TB * Malaria * * 5 Rest of r * * (I Neoplasm * @ IIn Endocrine * V Mental * * VI Nervous * * VII Circulatoly * * VIII Respiratory * * IX Digestive * X GU * * Xa Pregnancy * XII Skin * * IV/XIII Blood, etc * * All * *

216 190 Annex 3 Page 9 of 42 Percentage of Illness Treated Outside Hospital by Health Care Provider In the Public and Private Sector by Group of Diseases, Urban Areas I Infections * * 1 Dysentery * * 2 Typhoid * * 3 TB * * 4 Malaria * * 5 Restof I * * II Neoplasm @ @ III Endocrine * V Mental * * VT Nervous * * VII Circulatory ' ' VIII Respiratory * DC Digestive * * X GU @ )a Prepancy ' * xii Skin ' * IV/XM Blood, etc ' All * Note: Absolute number shows the number of cases of llness in the sample for each state/area of esidence. g Not estimated because of a small number of cases in the sample (less than 10). * Difference statistically significant (p<.05)

217 191 Annex 3 Page 10 of 42 International Classification of Diseases (ICD)-9 Codes 1. Infectious and Parasitic Diseases II. Neoplasms III. Endocrine, Nutritional and Metabolic Diseases, and Imrnunity Disorders V. Mental Disorders VI. Diseases of the Nervous System and Sense Organs VII. Diseases of the Circulatory System VII. Diseases of the Respiratory System IX. Diseases of the Digestive System X Diseases of the Genitouninary System XL. Complications of Pregnancy, Childbirth and the Puerperium XII. Diseases of the Skin and Subcutaneous Tissue IV & XI. Diseases of the Blood and Blood Forming Organs (IV), and Diseases of Musculoskeletal and Connective Tissue (XIII) This classification of diseases relates to tables on pages 6 through 9 (Annex 3)

218 192 Annex 3 Page 11 of 42 Publiclpri,ate Serices, Poverty and Specific Duasmes (Row Prcentage) tigrar wrma-sohrm ltrpr&adbeh Tamii adu WenP4Ssn G NG G NG G NG 0 NG G NG Rural Areas TI AUl II m IV V Al I II m rv V AU I H m rv V Nonc: I is lowest expendiure clss, V is highe.

219 193 Annex 3 Page 12 of 42 Public/Private Services, Poverty and Specific Disease (Row Pecentae) Qij2rat baprba Uk I desh IamLaU WNBanuL G NG G NG G NG G NG G NO Urban Areas TB AUl a m IV V All I TV V Dysentery AUl i H m : IV V Note: I is lowest expeaditre class, V is highe.

220 194 Annex 3 Page 13 of 42 Percent of Total Household Expenditure to Health and to Non-Hospital Treatment by Expenditure Quintiles Gujarat Maharashta Uttar Pradesh Tamil Nadu West Bengal Health Non- Health Non- Health Non- Health Non- Health Nonas % of Hospital as % of Hospital as % of Hospital as % of Hospital as% of Hospita: Total as% of Total as% of Total as % of Total as % of Total as % of Exp. Total Exp. Total Exp. Total Exp. Total Exp. Total Exp. Exp. Exp. Exp. Exp. Rural Areas In IV V Total Urban Areas II III i.s :2 IV V Total

221 195 Health Expenditure for fliness Treatment by Expenditure Quintile (Average Rs. expenditure per episode) Annex 3 Page 14 of 42 sfiarac Mahaha t U nd Tamil Nadau West Bensal G NG G NG G NG G NG G NG Rural Areas Expenditure Quintile fl m IV V AOvg/AI Urban Areas Expendibure Quintle I II if rv V Avg/All Note: I is lowest expendimre class, V is highest.

222 196 Annex 3 Page 15 of 42 Cost per Treated Episode as a percent of Monthly Household Consumption Expenditure (Users Only) C'Ttinat MahrashJ Uar Tamil Nau West Bengal Pradesh G NG G NG G NG G NG G NG Rural Areas I I m lv V All Urban Areas I II[ IV v All Note: MTHE = mean toal household expenditures. I is lowest expenditire class, V is highest.

223 197 Annex 3 Page 16 of 42 Access to Free Care for Illness Treatment Gujarat Maha- Uttar Tamil West rashtra Pradesh Nadu Benal Rural Areas % received free medicine % of those w/fe medicine who went to govt. service Urban Areas % received free medicine % of those w/free medicine who went to govt. service % of free medicine by SES Rmral Areas II EmI IV V Urban Areas I II m Iv V Note: I is lowest expenditure class, V is highest.

224 198 Annex 3 Page 17 of 42 Insurance Coverage and Illness Treated Outside Hospital GujaraE Maha- Tamil Utear West rashtra Nadu Pradesh Bengal Rural Areas Proportion of episodes for which insurance coverage was available Sex Total Males Females MPCE Quintie I Iv v Proportion of people who have insurance coverage wo were treated in the government and private sector Government Private Total Urban Areas Proporrion of episodes for which insurance coverage was available sex Total Males Females PCE Qunde I I m IV V Proportion of people who have insurance coverage wo were treated in the government and private sector Govermment Private TOa Note: MPCE - mean per capita expendimure I is lowest exdendimre class. V is highest.

225 199 Annex 3 Page 18 of 42 Illness Treated Outside of Hospital for Total Population and Insured Population CiUiarat Mahamshtr Tarnil Nadluta Pradesh West Bengal G NG G NG G NG G NG G NG Rural Areas Total population Insured population only Urban Areas Total population Insured population only

226 200 Annex 3 Page 19 of 42 Average Total Payment (in Rupees) for Treatnent Outside Hospital by Public and Private Sector and Group of Diseases, Gujarat Maharashira Tamid Nadu Public Pvuza All Public Private All Public Prvi! AU Rural Areas I Infections * * 50 1 Dysentery * 38 2 Typhoid * * 67 3 TB * * 59 4 Malana 26 41* * ' Restof I II - m Endocrine * 65 V Mental VI Nervous 53 95* VII Circuatory 29 Bs* * * 44 VII Respiratory DC Digestive X GU ( XI Pregnancy * * 26 XII Skin * * 43 IV/XKI Blood,etc * 33 All 60 94* * * 41

227 201 Annex 3 Page 20 of 42 Average Total Payment (in Rupees) for Treatment Outside Hospital by Public and Private Sector and Group of Diseases, Gujarat Maharashtra Tamil Nadu Public Private All Public Privame All Public Private All Urban Areas I Infections 34 59* * 38 1 Dysentery * 33 2 Typhoid * 70 3 TB * * 24 4 Malaria 23 42* * 28 5 Rest of I * 51 @ III Endocrine * 67 V Mental * VI Nervous * * * 88 VII Circulatory * ' 45 VIm Respiratory * IX Digestive * * 28 X * 81 XI Pregnancy * 29 xim Skin ' * IV/XIII Blood, etc * * 33 All * * 46

228 202 Annex 3 Page 21 of 42 Average Total Payment (in Rupees) for Treatment Outside Hospital by Public and Private Sector and Group of Diseases, Utar Pradesh West Benzal Public PrivZe All Public Privaze All Rural Areas I Infections * 51 1 Dysentery * * 36 2 Typhoid * 59 3 TB * Malaria Restof I II Neoplasm 169 @ III Endocrine V Mental VI Nervous VII Circulatory VUI Respiratory IX Digestive * 63 X GU )a Pregnancy * 27 xii Skin IV/Xm Blood, etc * 48 AU * * 52

229 203 Annex 3 Page 22 of 42 Average Total Payment (in Rupees) for Treatment Outside Hospital by Public and Private Sector and Group of Diseases, Uttar Pradesh West Bengal Public Privare All Public Privaze Au Urban Areas I Infections * I Dysentery Typhoid TB Malaria Restof I m Endocrine 6 72* V Mental 72 47* VI Nervous Vii Circulatory * 41 VIII Respiratory ix Digestive X GU 146 @ )a Pregnancy xii Skin IV/XMI Blood, etc * 82 All * * Not estimated because of a small number of cases in the sample (less than 10). * Difference is statistically significant (p<.05)

230 Logistic Regressions of Deleriiiiians of Cloice of Trealnieni for Illness Nol Requiring I lospitalization, (0 public. I - pliiv ac RURAL AREAS Plediclous Gujanal Mabauasbtca Tantit Ha"N Utug Pradksh Wcst Ilengal Coclficient Eap (b CoelEcicika Exps (b) Coefliciknt Exp(b) I cocrnicicnt Exp (b) I COCfTKiCnl Exp (b) I ChOi ff Maic.140 I.ISO " S Age @ " -.044'.9S7 Age Sq. ( ' I N Mmakd " SIS "'.546 hkad ' :4ucation Iksuted 2.401' ' ' I75.I53 * 1.75'.309 IL. Charaeterletls rfthe NonSCST 1.777' ' ' ' 1401 Main Sgugce of Income Casual Labor I Itcgula Labof ' '.601 (thcls " LnMTIIE S ' I o Top 20'/.ofMPCE -.3S IS UIISIZE S Osl3 9 OSS AF Edu " III Cha,adet Lft lid. Male SIS' Age SSI' U ' I IdEdu ' [Y. rype of Ailownm Dysetlay ' TypluiJ S OS rno ' ' ' SS' '.302. H Mialasia S' 9I I S16 Cancec S * * *.514".205 Nervons S I ka.os S6.29S 1144 AuIhawn OS ' Acci.cIu S I9S Consiaa ' 1264 %ofconsctly Piedkad to Observed Outcome

231 URBAN AREAS PlCdiECOrS Gujai MahaYashtta Tuanil Nadu UlIY Piadesh West Ilcngal Cocrncicnt Ep(b)h Coemnckn Exp (b) Coeflkcscni Exp (b Coctifcicni Eup(b) I Cociflcickn 1. haciefiolics Exp lb) of 111 Male Agc -079' ' ".965 * Agc Sq. (10 2 ) " '' * N MuIkd * ".564 * lcad t7-451"* 637 * lducatinu.1336 I lns,ucd ' ' ' A C lhaiesiklca ofhe NonSCST 1.051' Mainu Soee of.554' Income Casual Labr ' RegulaI -.569' labor "* ' OaKIus '' ' S LaMTIIE I t02.422' t" Top % of 1IPCE '' Il1ISIZE K -.069* AF Edu.001 I ' ' Il.fd IkS IId.Male lidi. Age ' lls * IIdIdu t ' IV 'rs1t of Alil nin DysenCAy i Fyploid ' ' ; Tll oq 1.779' ' '.26S MaI:"ia ' ' Caw#Ct s t, " Ncrvons llcmn t AsdIsnsa s Acckde * " ss Ceusitn U %ofcomealyh dcikkdla Obstt'vcd OdCo 302 For lds am suhbsequctd -pc 10 egitcssiitn: ' p(os < * p<.oi

232 206 Annex 3 Page 25 of Type of Health Seryice Expenditure on Non-Hospital Treatment with and without Price Variable rmiamy MahnrnMharara Tpmil1Ipdim Uttr Pr',e,Sh West Ben? Without With Without With Without With Without With Without With Price Price Price Price Price Price Price Price Price Price Rural Areas 1. Pvt. Hosp * ' * Medicine ' * 3. X-Ray.686^ ^ ' 4. Dignos -.547^ * -.789* -.no S. Physioth Surgery ^ [I. Duration of Treanent 7. Visits '.132^ *.032'.035* * Duranon '.076* ^.040' *.082^ Prolong * ' * * -.522' ' ' m. Type of Ahment 10. Dvscnt * Typhoid "* ^' TB *" '.390' Malaria -.709' * -.499' " Polio *" '.782Z S. Cancer ' 1.504' @ ' Nervous ' -.596' " Hean ' ' 488' 18. Astbma S Accident '.209".3800".382-' W. Charcteristics of m Person 20. Male.337*.234*" '.186"^ Age.029".028* '.011"' Age Square -.032*" -.027" ' -.013' Not Married lOS ' * ' -.558^ -.40i 23. Head Whether Completed Schooling Primary -. I Second Iasured @ ' *

233 207 Annex 3 Page 26 of 42 Expenditure on Non-Hospital Treatment with and without Price Variable z l Mlahargthr TarnmiL Nadu Urtar Pradesh West Rencal Without With Without With Without With Withour With Without Wilth Price Price Price Price Price Price Price Price Price Price V. Chamcteristics of the Household Rural Areas 26. SC ST SAEMPL ' CASUAL 309@ La MPCE * ' * HH Size *" -.081O C -.017* ' 30. Highest EducaEion of Adult Female AF PRIM @.185@.213* S4 AF SEC S " VT. Characteristia of the Head of Household 31. HD.MALE HD.AGE HD.WDS.429@ ' ' Education of the Head HD. PRIMARY ' 'HID. SEC Constant 1.751'' ' * R Square i S.390 Adj. R Square F Value No. of Cases

234 208 Annex 3 Page 27 of 42 Expenditure on Non-Hospital Treatment with and without Price Variable Gu rkahamchn M Tamil Nadu Uttar Pfrsdec-h Wesmlealzi I. Type of Health Service Without With Without With Without With Without With Without With Price Price Price Price Price Price Price Price Price Price Urban Areas 1. Pvt. Hosp i Medicine.921' ' * * 3. X-Ray ' 1.023' 1.311' 4. Dipos S. Physioth Surgery ' ' Duration of Treatment 7. Visits.077' ''.051'.022''.033' Duration ' ' '.054*.049' 9. Prolong ' O * ' ImI. Type of Ailment 10. Dysenw * "* Typhoid 1.075* '.363-' '.444' ' 12. TB "* '.512' ' Malaria Polio '' 1.325* 1.272* S. Cancer * "' Nervous * Heart '.403'.360*.368' Asthma "' * ' Accident -.704' S'' Iv. Chamcteristics of IIl Person 20. Male Age -.026" Age Square.036' "' Not Married '' ' -.333' Head Whether Completed Schooling Primary Second Insured * * '

235 209 Annex 3 Page 28 of 42 Expenditure on Non-Hospital Treatment with and without Price Variable Guiarar Maharah amil au Ur Pdh±est Without Witb Without With Wtho With Without With Without With Price Price Price Price Price Price Price Price Price Price 26. SC ST * SAEMPL * CASUAL La MPCE *.391* *29. HH Size * -.036* -.089* -.033* Highest EAuCaron of Adult Female AF PRIM *s AF SEC * VI. Chancterisdcs of the Head of Household 31. HD.MALE HD.AGE ' 33. HD.wDS ' Education of the Head HD. PRIMARY HD.SEC Constant 1.641*' * * R Square Adj. R Square F Value No. of Cases

236 210 Annex 3 Page 29 of 42 Hospital Treatment of Illness: Roles of Government and Non-Government Facilities % of Episodes Treated by: Gujarat Maha- Uttar Tamil West rashtra Pradesh Nadu Bengal Rural Areas Govt Non-Govt Urban Areas Govt Non-Govt

237 211 Annex 3 Page 30 of 42 Hospital Use Patterns by Households Expenditure Class (Percent of episodes to govermment and non-government providers) Tjarat Mahamshrra IIIur P Tamii Nadu Wes Rinl G NG G NG G NG G NG G NG Rural Areas AU I ai m rv V Urban Areas AU U ;2 II rv v Note: I is lowest expenditure class. V is highest.

238 212 Annex 3 Page 31 of 42 Hospital Use and Referral % of Hosp. Cases Consulting Prior to Hospital Gujarat Maharashtra Uttar Pradesh Tamil Nadu West Bengal Rural Areas Urban Areas Location of prior consultation ryujgrat Mahahtr Uttar Pradesh Tamil Nad West Bonal G NG G NG G NG G NG G NO Rural Areas Govt Non-Govt Urban Areas Govt Non-Govt

239 213 Annex 3 Page 32 of 42 Mix of Diseases Reported Treated in Hospital Gujarat Mahamt Utraeshz P Total G NG Total G NG Total G NG Rural Areas I Infections TB Malaria Dysentery Typhoid Rest of Neoplasm III Endocrine v Mental VI Nervous VII Circulatory VIII Respiratory rx Digestive X GU )a Pregnancy XII Skin IV/XIII Blood, etc All I Infections T Malaria Dysenterv Typhoid Rest of [I Neoplasm III Endocrine V Mental VT Nervous VII Circulatory VIII Respiratory IX Digestive X GU X) Pregnancy XII Skin IV/=XiI Blood, etc All

240 214 Mix of Diseases Reported Treated in Hospital Annex 3 Page 33 of 42 Tamil Nadu Wct Bengal Total G NG Total G NG Rural Areas I Infections TB Malaria Dysentery Typhoid Rest of II Neoplasm III Endocrine V Mentai VI Nervous VII Circulatory VIII Respiratory ix Digestive X GU XI Pregnancy XII Skin IV/X= Blood, etc All I Infections TB Malaria Dysentery Typhoid Rest of I [I Neoplasm III Endocrine V Mental VI Nervous VII Circulatory VIII Respiratory IX Digestive X GU Xl Pregnancy XII Skin IV/XIII Blood, etc All

241 215 Annex 3 Page 34 of 42 Utilization by Sector, Expenditure Class, and Disease Category (Percent of episodes to government and non-government providers) Cjuiarat Maharashtr Uttar Pradesh TamiliNadu West Rmenal G NG G NG G NG G NG G NG Rural Areas TB All II lii IV V Malaria All I II III IV v Dysenterv All III IV V

242 216 Annex 3 Page 35 of 42 Utilization by Sector, Expenditure Class, and Disease Category (Percent of episodes to gover.nent and non-government providers) ujart Mahsaratr Uttar Pradesh amil Nad West engai G NG G NG G NG G NG G NG Urban Areas TB All I II UI rv V Malaria All II m IV V Dysentery AUl II III IV V Note: I is lowest expenditure class, V is highesl

243 217 Annex 3 Page 36 of 42 Percent of Total Household Expenditure to Health and to Hospital Treatment by Expenditure Quintiles Gujarat Maharashtra Uttar Pradesh Tamil Nadu West Bengal Health Hospital Health Hospital Health Hospital Health Hospital Health Hospital as%of as%of as%of as%of as%of as%of as%of as%of as % of as%of Total Total Total Total Total Total Total Total Total Total Exp. Exp. Exp. Exp. Exp. Exp. Exp. Exp. Exp. Exp. Rural Areas II III IV V Total Urban Areas II III IV V Total Note: I is lowest expenditure class, V is highest.

244 218 Expenditure Per Episode for Hospitalization by Expenditure Quintile and Type of Provider Annex 3 Page 37 of 42 lj.arak mmabashch UtIrPrades Tamil Nadu Wes Rea G NG G NG G NG G NG G NG Rural Areas Urban Areas cpendirure class, V is highest.

245 219 Annex 3 Page 38 of 42 Expenditure per Episode of Hospitalization and Percent of Expenditure Going to Provider (Indian Rs) Gujarat Maharashtra Uttar Tamil West Pradesh Nadu Bengal Rural Areas Avg. ExplEpisode Govt Non-Govt % of cost to Provider Govt Non-Govt Urban Areas Avg. EspJEpisode Govt Non-Govt % of cost to Provider Govt Non-Govt

246 220 Annex 3 Page 39 of 42 Expenditure per Episode and Hospitalization Gujarat Maharashtra Uttar Tamil West Pradesh Nadu Bengal % Receiving Free Bed Rural Areas Urban Areas % with Free Bed Getting Govt. Health Care Rural Areas Urban Areas % with Free Care by SES Rural Areas I IV v Urban Areas II in IV v Note: I is lowest expenditure class, V is highest.

247 221 Average Total cost (in Rupees) for Treatment in a Hospital by Public and Private Sector and Group of Diseases, Annex 3 Page 40 of 42 Gujarat Maharashua Tamil Nadu Public Private All Public Privue All Public Pnvaze All Rural Areas fections * * ' * * @ * @ Tine ' i * * * 264 s * * * 673 tory * * 293 tory * * 652 re * * * * * ' * * * * ' ' ' 449 Urban Areas * ' * ' ' ' * * * * * * * ' @ * ' 709

248 222 Average Total cost (in Rupees) for Treatment in a Hospital by Public and Private Sector and Group of Diseases, Annex 3 Page 41 of 42 Gujarat Maharashra Tamil Nadu Public Pnvare All Public Private All Public Private All Urban Areas V Mental * * * 684 VI Nervous * 764 VII Circulatory * * 475 VIII Respiratory * * 373 IX Digestive * * 2637 X GU * * 850 XI Pregnancy * 525 XII Skin * * * 484 IV/XIII Blood, etc * * * 689 All * * * 662 Uttar Pradesh West Bengal Public Private All Public Private All Rural Areas I Infections * Dysentery * 82 2 Typhoid TB Malaria 249 @ 5Restof I * 262 II Neoplasm * 3901 III Endocrine V Mental VI Nervous 'I Circulatory III Respiratory * Digestive * * 645 GU * 250 Pregnancy 355 @ Skin * mi Blood, etc * * * * 402

249 223 Annex 3 Page 42 of 42 Average Total cost (in Rupees) for Treatment in a Hospital by Public and Private Sector and Group of Diseases, Uttar Pradesh West Bengal Public Private All Public Private All Urban Areas I Infections * 307 I Dysentery * Typhoid 735 @ 3 TB * Malaria 193 @ 5Restof II Neoplasrn m Endocrine V Mental * 648 VI Nervous Vi' Circulatory * 362 VIII Respiratory * 1841 IX Digestive X GU * 519 )a ( XII Skin IV/XM Blood, etc * * 1333 All * * Not estimated because of a small number of cases in the sample (less than 10). * Difference is statistically significant (p<.05)

250

251

252 .~~~~~~~~~~~~ 0 IMAGING ~~~~~~~~~~~Report No: 1:3042 IN _ l~~~~~~~ye:s