TOWARDBETTER REPRODUCTIVEHEALTH FOR POOR WOMEN IN SOUTH ASIA: SUMMARY FOR POLICYMAKERS This is a synopsis of a study undertaken by the World Bank in Jive countries: Bangladesh, India, Nepal, Pakistan and Sri Lanka. A more comprehensive presentation of the Jindings and conclusions is available in the report, Better Reproductive Health for Poor Women in South Asia. Background reports on each of the Jive countries, on financing issues, global best practices in reproductive health, and decentralized action planning are available on request from mchatteriee@,worl~lbank.or,q. The study was financed by a grant from the Bank-Netherlands Partnership Program (BNPP) Authors MeeraChatterjee Ruth Levine Shreelata Rao-Seshadri Nirmala Murthy Team Meera Chatterjee (Team Leader) Ruth Levine (Adviser) Bina Valaydon(Bangladesh) Farial Mahmud (Bangladesh) Tirtha Rana (Nepal) Shahnaz Kazi (Pakistan) Kumari Navaratne (Sri Lanka) Pranita Achyut (Research Analyst) P.N. Rqjna(Research Analyst) Ruhi Saith (Research Analyst) G. Srihari (Program Assistant) Anabela Abreu: Sector Manager, SASHD Julian Schweitzer: Sector Director, SASHD Praful Patel: Vice President, South Asia Region Table of Contents Acronyms and Abbreviations.................................................................. v Chapter 1. The Opportunity t o Improve Lives and Livelihoods through Better Reproductive Health: A Preview of the Recommendations....................................................................................... 1 Chapter 2. Reproductive Health Services: Inadequate and Unused............................................................................................................ 9 Chapter 3. An Ambitious Agenda For Five Countries I n South Asia: Toward Better Reproductive Health For Poor Women... 35 Chapter 4. A Practical First Step: Decentralizing Aclmionst o Bridge Inequalities.................................................................................... 59 References ................................................................................................... 67 Acronyms and Abbreviations AIDS Acquired Immuno-deficiency Syndrome ANC Antenatal Care ANM Auxiliary Nurse Midwife ARI Acute Respiratory Infection ASHA Accredited Social Health Activist AYUSH Ayurveda, Yoga, Unani, Siddha and Homeopathy (Indian medical systems) BCC Behavior Change Communication BNNP Bangladesh National Nutrition Program CAC Comprehensive Abortion Care CHC Community Health Center CNA Community Needs Assessment DAP Decentralized Action Planning DCPP Disease Control Priorities Project DGFP Directorate General of Family Planning DGHS Directorate General of Health Services DHAP District Health Action Planning DHM District Health Mission DHS Directorate of Health Services EmOC Emergency Obstetric Care EOC Essential Obstetric Care EPI Expanded Program of Immunization ESP Essential Services Package FP Family Planning FWA Family Welfare Assistant GDP Gross Domestic Product GOTN Government of Tamil Nadu HIV Human Immuno-deficiency Virus HLSP Health and Life Sciences Program HPSP Health and Population Sector Program ICDS Integrated Child Development Services' Program ICPD International Conference on Population and Development IEC Information, Education and Communication IFA Iron Folic Acid IUCD Intra-Uterine Contraceptive Device IUD Intra-Uterine Device LBW Low Birth Weight LHW Lady Health Worker MCH Maternal and Child Health MDG Millennium Development Goal MMR Maternal Mortality Ratio M0 Medical Officer MOH Ministry of Health MOHP Ministry of Health and Population MOPW Ministry of Population Welfare NGO Non-Governmental Organization NHA National Health Accounts NIPORT National Institute of Population Research and Training (Bangladesh) NRHM National Rural Health Mission NSS National Sample Survey OC Oral Contraceptives OPD Out Patient Department PAC Post Abortion Care PDHS Provincial Director of Health Services PHC Primary Health Center PHI Public Health Inspector PHM Public Health Midwife PHNS Public Health Nursing Sister PNC Postnatal Care POA Programme of Action (of the ICPD) PPP Public-Private Partnership PRHFPS Pakistan Reproductive Health and Family Planning Survey PRSP Poverty Reduction Strategy Paper PRI Panchayati Raj Institution (Local government institution) PSI Population Services International RCH Reproductive and Child Health RTI Reproductive Tract Infection SMP Social Marketing Pakistan ST1 Sexually-Transmitted Infection TB Tuberculosis TBA Traditional Birth Attendant TFR Total Fertility Rate TNMSC Tamil Nadu Medical Services Corporation (India) -IT Tetanus Toxoid VHW Village Health Worker WHO World Health Organization THE OPPORTUNITYTO IMPROVE LIVES AND LIVELIHOODS THROUGH BETTER REPRODUCTIVE HEALTH: A PREVIEW OF THE RECOMMENDATIONS This summary of a five-countrystudy of reproductive health among poor women in South Asia is intended to raise awareness about the major reproductive health problems faced by poor women; to shed light on their primary determinants and consequences; and to recommend high-priority actions for consideration at the policy and program levels. It argues that, within the framework of the International Conference on Population and Development's Programme of Action, and the Millennium Development Goals, countries in the South Asia region face an urgent imperative to increase the responsiveness of the health system to the reproductive health needs of women who marry and start childbearing at young ages and have compromised nutritional and health status, in a way that is cognizant of the underlying determinantsof ill health, including gender discrimination,poverty and lack of education. This will require program changes, drawing on global good practices, and an increase and reorientation in spending on reproductive health services. While each country faces special challenges, the health sector strengthening processes currently underway in all represent opportunitiesfor greater attentionto reproductive health priorities. 1.1 Poor reproductive health outcomes are a major contributor to ill health among poor women in South Asia. Moreover, they represent significant losses to community and social welfare, and contribute to the transmission of poverty fiom one generation to another. Based on a multi-year project supported by the World Bank, which included a series of significant background analyses, this summary report examines the extent of reproductive health problems in five South Asian countries - Bangladesh, India, Nepal, Pakistan and Sri Lanka. It highlights the justification for an urgent and systematic effort to improve health sector policies and practices in the pursuit of better reproductivehealth. It proposes an ambitious agenda for learning and action that could be taken on by the national and sub-national governments, as well as their development partners. And, as a practical first step, it describes the ways in which a district-level action planning process can contribute to identifying the main strategies through which improvemeints can be made within the health sector. This summary, which is one of several outputs fiom the project, is not intended to be comprehensive; rather, it is intended to raise awareness about the major reproductive health problems faced by poor women, the primary determinants and consequences, and the high-priority actions for consideration at the policy and program levels. 1.2 This represents an auspicious moment for paying renewed attention to the challenge of improving reproductive health. It is now approximately 12 years after the International Conference on Population and Development (ICPD), in which a far- reaching agenda for improved reproductive health was agreed upon by national leaders from virtually all countries of the world. Thus, it is timely to reflect on what the track record has been in this area in the South Asia region. Moreover, the Rdillennium Development Goals (MDGs) provide specific new targets for reproductive health.' Finally, each of the five countriesincluded in this review is engaged in important changes in the financing,organization, management and implementation of health programs, with significant emphasis on reachng poor households and communities with needed services and protection from health-related financial hardships. Specific strategies may be required for these changes to benefit women affected by reproductivehealth problems. 1.3 The remainder of this chapter highlights the main reproductive health challenges, and provides a preview of the recommendations for program and policy changes that would respond to those challenges. It sets the stage for the subsequent chapters. Chapter 2 describes in some detail the current deficits in the supply of and demand for reproductive health services, and the patternsof financingof reproductive health services. Chapter 3 lays out an agenda for technical improvements in reproductive health programs, drawing on global good practices; and the policy changes to support those improvements. Chapter 4 presents an outline of a first, practical step, which was tested on a pilot basis during the conduct of this analytic work- District Action Planning. Poor ReproductiveHealth Affects Women, Families and Communities 1.4 Each year, nearly 185,000 women in South Asia die from causes related to pregnancy (WHO/UNICEF/UNFPA, 2004). Millions more are affected by acute or chronic illnesses or disabilities as a result of childbearing. Many mothers suffer from under-nutrition that predisposes them to problems during pregnancy and affects the starting life chances of their infants. In South Asia, 11 million infants are born every year under the acceptable weight of 2.5 kilograms, accounting for 50 percent of the world's total number of low birth weight babies. About 2.4 million of them die due to this and other preventable causes. Reproductive health deficits also compromise children's health, contributing to high rates of under-5 mortality (see Figure 1, which compares recent levels of maternal and child mortalitywith the MDG targets). 1.5 Reproductive health problems represent a major priority for public sector attention because of their direct and indirect consequences for a broad range of welfare measures. Different population sub-groups are affected differently by distinct health problems. The following sections provide a picture of the reproductive health of adolescent and adult women among poor households in the five countries studied, and highlight the underlyingrisk factors. '. The MDGs include the followingreproductivehealth-related goals: Between 1990 and 2015 it is expected that under-5 mortality rates will be brought down by two-thirds,and maternal mortalityratios will be reduced by three-quarters. The spread of HIVIAIDSis to be stopped and reversed. Progressin achievingthe goals is to be monitored also on the basis of the proportion of deliveries attended by trained health personnel,contraceptiveprevalence,and HIV prevalenceamong adults (15-49 years). While the ICPD set a direction for reproductivehealth, the MDGs provide these specific indicatorsand targets. Figure 1.1 Current Reproductive Health Levels and Goals for the Five Countries of South Asia Maternal Mortality Ratio (Deaths per 100,000) (Deaths per 1,000) Bangladesh India Nepal Pakistan Sri Lanka Note: The graphs above are based on modeled estimatesprovided in WHO (2004) which facilitate comparisonsacross the five countries. The 2006 Nepal DHS has provided new estimates of the MMRs for Nepal in 1995 (539) and 2000 (470) and a revised goal for 2015 (213) based on the estimated 1988 (rather than 1990) level (GON-MOHP,2007). It also reports a U5MR of 65 in 2006. A Quick Look at the Risks 1.6 Gender discrimination in the early years constitutes a central risk factor for poor reproductive health outcomes in South Asia because it translates into a lifetime of social and health vulnerability. Many girls are subordinated and subject to discrimination within their families and communities - a process that begins in childhood. This is manifest in their poorer health and nutritional status compared with boys, lack of education and information, heavier domestic work burdens and constrained mobility- all of which have a bearing on their reproductive health at this stage as well as later in the life cycle. The practices of seclusion and purdah, which confine girls spatially and in terms of social interactions, begin in earnest during puberty, and have a serious impact on schooling. While educational attainments are high in Sri Lanka, and rising in the other four countries, girls and boys still have unequal access to schooling and vastly different levels of achievement. For example, in India 86.7 percent of males it1the 15- to 19-year age-group are literate compared to 62.4 percent of females. 1.7 Coupled with the prevalent practice of dowry, the perceived low economic value of women fuels a vicious cycle of low investment in girls, early marriage and child- bearing, and limited access to resources and power. The disadvantaged position of women also exposes them to gender-based violence, including sexual abuse and violence within and outside the family. As in most matters concerning sex and reproduction, secrecy shrouds this issue, so that the actual extent of harm is largely unknown. 1.8 Marriage and childbearing at young ages, which are also related to Ithe lack of alternative social roles and economic opportunities for women, greatly affect women's prospects of reproductive health. The majority of girls in South Asia begin sexual activity during adolescence, usually within the context of marriage. In general, only a small proportion of girls report premarital sex, e.g., below 10 percent in India, while about one-quarter to one-third of boys do (Jejeebhoy, 2000).~These sexual encounters are often unplanned and unsafe. Young women (and men) in low-income urban areas appear to be particularly vulnerable to sexual coercion, including physical force. While girls tend to report casual encounters or a steady partner, boys often have multiple (usually casual) partners and almost always fail to use a condom as protection against pregnancy and sexually-transmittedinfections. 1.9 Marriage is almost universal throughoutSouth Asia, and early marriage continues to be the norm, especially among girls. This results in early sexual activity and exposure to pregnancy and infections and their attendant consequences. There are signs that this is changing, albeit very slowly. In the past 20 to 25 years, Nepal and Bangladesh have experienced a marginal increase in the age at marriage. Young Bangladeshi women increasinglyregard early marriage as undesirable (Islam, 1995). This trend is also visible in India although the median age at marriage has increased only from 15.8 to 18 years over the past 25 years. At present, in Bangladesh, India and Nepal, one-third to one-half of adolescent girls are married. Among ever-married women in the 25- to 49-year age- group, half were married by the age of 16 years, two years below the legal age of marriage. The median age at marriage is higher in Palustan (18 years) and Sri Lanka (23 years). Despite a recent increase in the age at marriage in Pakistan, adolescent marriages are still prevalent among girls, particularly in rural areas and among relatively uneducated girls (Population Council, 2003). 1.10 The relationship between early marriage and education is a strong one. Among women aged 20 to 24 years, 68 percent of the illiterate were married by the age of 20, compared to 13 percent of those with 10 or more years of education. 1.11 High fertility, with short intervals between births, places a strain on women's health. In South Asia, where great emphasis continues to be put on the fertility of women, desired and achieved fertility are high in many communities. However, evidence suggests substantial fertility decline in many areas. The highest fertility rates among the countries in the region in 200012001 were those of Pakistan (4.1 children per woman) and Nepal (4.1) followed by Bangladesh (3.2), India (2.8) and the lowest in Sri Lanka (1.9). The first three countries experienced fairly significant declines in fertility during the 1990s; Sri Lanka saw a somewhat smaller decline from its already low level, and Bangladesh a very slight decline. The 2006 DHS in Nepal reports a TFR of 3.1, reflecting a remarkable drop in fertility in the five years between 2001 and 2006. In India, a few states (e.g.,Kerala, Goa and Tamil Nadu) have already achieved replacement level fertility or are close to attaining it. However, five states (Bihar, Uttar Pradesh, Madhya Pradesh, Rajasthan and Orissa) account for nearly 40 percent of the country's population and will contribute well over 50 percent of its population increase during the next decade. 1.12 Poor nutritional status and other pre-existing conditions represent major health challenges, particularly among adolescents. Adolescent girls are at high risk of multiple nutritional deficiencies (of iron, vitamins and energy) that adversely affect their 2.Higherrates arereportedin Sri Lanka and Nepal. reproductive health and outcomes. More than half of all 15 to 19 year-old mothers in India, and one-third even in Sri Lanka, are anemic (IIPS and ORC Macro, 2000; USAIDIOMNI, 2000). Adolescents have almost the same prevalence of anemia as all women of reproductive age, suggesting that anemia sets in early in the reprodluctive life cycle as iron intakes fail to keep pace with the increased requirementsafter menarche. 1.13 Systematic information on knowledge of RTIs, STIs or HIVIAIDS is unavailable for the five countries, but some studies suggest that awareness is much lovirer among women than men. In India and Nepal in 2000, only half of women surveyed vvere aware of HIVIAIDS compared to two-thirds of men. Knowledge of RTIsISTIs was even lower (NACO and UNICEF, 2002; GON-MOH et al., 2002). Two-fifths of Nepali women and two-thirds of men believed that AIDS could be avoided. By 2006 in Nepal, awarenessof' HIVIAIDS had increased to 73 percent among women and 89 percent among men (GON- MOPH, 2007). 1.14 Among the majority of women, particularlythose from low-income householdsin rural areas, access to quality health services is very low. The utilization of maternal health services shows great variation - high in Sri Lanka and lowest in Nepal and Bangladesh. Institutional delivery is nearly universal in Sri Lanka. In the other four countries, despite the presence of fieldworkers, considerable proportions of women are not reached effectively during pregnancy. Less than one-fifth deliver in an institution in Bangladesh, Nepal and Pakistan, and there is hardly any provision of postnatal care. Again, adolescents are particularly disadvantaged. Poor uneducated y0un.g mothers living in rural areas, trapped in a cycle of repeated pregnancies, account for the majority of maternal deaths and are most in need of family planning, abortion, and maternal and child health services. 1.15 Several supply- and demand-sideconstraints affect service utilization. Awareness of the availability of and need for services is low. Cultural norms and social stigma hinder the use of services even for recognized problems (e.g., RTIsISTIs) or result in women approaching the wrong provider. In addition to lack of knowledge, cost is an important demand-side constraint. Quality issues that constrain service provision and utilization include: inadequate staff, particularly women providers; inadequate supplies, equipment and basic amenities; irresponsible and rude behavior among service providers; lack of privacy and overcrowding; inadequate and inappropriate information provision; poor technical competence;lack of follow-up;and demands for informal payments. Health System Goals to Achieve Better Reproductive Health 1.16 Improving reproductive health in South Asia requires adoption of an ambitious agenda to tackle several priorities. First, in all five countries (including some areas and services in Sri Lanka) mechanisms to increase the sup~lyof reproductive health services for poor women must be strengthened. This should start with those services, such as birth spacing and safe abortion, for which there is considerable 'unmet demand' among poor women. The chief approaches are to target poor geographic areas for special planning and resource allocations (at the most decentralized level possible), and the poorest villages and households for attention through local outreach mechanisms (e.g., fieldworkers, camps, mobile services, etc.) and demand-side financing. 1.17 Reproductive health services must also target adolescents (married and unmarried) as they are central to the achievement of reproductive health goals. They require information as well as services which can be provided through frontline health workers if these are given a clear mandate and trainingin the social and counselingskills required to access this difficult group. These interpersonal efforts must be bolstered by behavior change communication programs through mass media, schools and community institutions. Innovative approaches have been developed and could be supported through public grants. 1.18 Second, a corollaryof targeting is to enhance demand among the poor for services that are poorly understood and under-utilized, notably ante-natal care and safe delivery. For this, behavior change communication efforts should be made relevant to poor women, and demand-side financing used to reduce cost barriers, particularly for the use of indoor services and purchase of medicines. Supply-side improvements that address the problems listed above would also enhance demand. 1.19 Third, all countries need to deliver the complete essential package of reproductive health services in an intemated manner. (The essential package is described later.) The services that should be provided through single-window primary health facilities and workers are: maternal and child health care, nutritional prophylaxis,family planning, safe abortion, diagnosis and treatment of RTIsISTIs, all relevant information and counseling, and referral to emergency/surgical/specializedcare at secondary fa~ilities.~To start with, it is advisable to integrate separate departments of family planning, health, and nutrition, and develop unified policy and program guidelines. In the field, providing clear guidelines, tools and training would help workers implement a client-centered approach efficiently, and managers should encourage and monitor performance on the basis of a continuum of care. Creating and disseminatingthe know-how for this should be a central hnction while implementation and management are decentralized. Integration will improve demand for and use of services. 1.20 Some neglected aspects of the essential package require special attention4 Throughout South Asia, the availability of safe abortion sewices must be increased through both the public and private health sectors. Public systems could increase medical abortion and vacuum aspiration facilities, and public financing could help increase the availability of private services. Providing capital grants andlor per-service subsidies (especiallyfor poor women) through contracts, and social franchising of clinics are some promising approaches in South Asia. Simultaneously, governments must take their regulatory role in this area seriously, supporting safe abortion and cracking down on unqualified providers to eliminateabortion-related mortality and morbidity. 3.While, for example, Sri Lanka has many of the elements of the essential package in place (including early registration of pregnant women, ANC, delivery care, PNC, child immunization, family planning and nutrition, it needs to complete reproductive health services by adding adolescent care, comprehensive abortion care, and RTVSTI services. RTVSTl services are also important and neglected, but emphasized less here because it is already on the radar screen of South Asia's health systems, mostly under the aegis of HIVIAIDS programs. 1.21 Counseling to improve mother and child nutrition, anemia prophylaxis and care of the under-nourished are currently inadequate everywhere and must be enhanced by training health staff better to preventlmanage under-nutrition. They should start by focusing on poor women who are at greatest risk of bearing low birth weight infants. The health systems must take responsibilityfor this care because it is central to maternal and infant survival. 1.22 While ANC and skilled delivery are receiving attention and efforts to increase the availability of essential obstetric facilities continue, postnatal care needs more emphasis. Improvementsin timing and quality could help to reduce maternal and neonatal mortality and morbidity. Women who deliver at home should be visited within 24 hours of delivery by a qualified female paramedic, and transport vouchers or h d s should be provided to those who need medical attention. 1.23 Fourth, progress in poor women's reproductive health will depend greatly on improvement in the quality of care, particularly through outreach by frontline women workers. Good quality outreach services need to be readily available, calling for larger numbers of workers, more efficiently deployed, more highly skilled, adequately equipped, and supplied with medicines. The critical role played by women providers in ensuring South Asia's health must be fully recognized and rewarded as their status is reflected in their behavior toward clients. Women providers must be the focus of the "health systemfix." The Policy Path 1.24 To bring about the required improvements requires important changes in the health sectors of the five counties. First, particularly to improve the supply ;andquality of services to the poor, outcome-based planning and monitoring can be introduced or expanded. Planning actions and allocating resources should be decentralized at least to the distict level in all countries, requiring higher levels to commit to flexibility in decision-making. Decentralized action planning (DAP), which was pilot tested as part of the study summarized here, identifies what needs to be done and can be done locally, and measures results in repeated cycles using local data. It can improve the effectiveness of available resources by ensuring their application to priority problems and by helping service providers and managers do what works in local experience or promising examples. DAP can enhance the technical and practical knowledge of those involved as information is shared during the planning efforts. In addition to improving the supply of services, it can help integrate services and motivate efforts to create demand. It can involve local government members, private providers, clients and others, in addition to health staff and managers, to ensure that public resources are used efficiently and to mobilize other resources - good results can help attract additional public or private local resources. Decentralized and participatory planning could be the cornerstone of increased ownership and accountabilityin South Asia's health systems. 1.25 Second, the recommendations above call for robust human resource development in the five countries, including attention to strategic human resource planning and to developing staff skills, motivation and perfomance. Some specific measures are: Increasing the numbers of qualified female staff (especially doctors and paramedics) at the frontline; the important strategies to achieve this include additional training and recruitment, contracting in or out, improved allowances and support, and perfomance-based incentives; Enhancing organization and management of integrated reproductive health service delivery through technical and managerial training and by making implementation and monitoring guidelines and good practice information available everywhere; Improving attitudes and behaviors toward poor women through sensitization programs and accountability measures; and Increasing accountability for health outcomes among providers and managers using perfomance incentives in addition to decentralized planning. 1.26 Third, in addition to better spending through decentralized planning and monitoring, more public finance must flow to the reproductive health sub-sector. This is necessary to ensure that pressing needs for staff, equipment, medicines and other inputs in the public system are met, and that the availability of services to the poor is increased by reducing the costs of using private services. From an equity perspective, general revenue financing is desirable as it is a progressive source of health care financing and when combined with low user fees and universal coverage it provides high levels of financial protection against catastrophic ill-health. Promising demand-side financing options include voucher schemes to assist poor women to have institutional deliveries in the private sector, reimbursement of transport and other out-of-pocket expenses when they use public facilities, and conditional cash transfers (e.g., after completion of three antenatal care visits). Private resources for poor women's reproductive health can be enhanced through public-private partnerships, including social marketinglfi-anchising schemes, contracting out, grants or subsidies (e.g., for safe delivery or abortion facilities), and support in cash or kind to services (e.g., for adolescents in schools or communities). 1.27 These improvements fit squarely with the overall agenda for health sector refom in the region, which includes greater responsibility to sub-national and local authorities for the delivery of essential services, improved efficiency in health spending, and the development of financing mechanisms to reduce the burden on the poor of out-of-pocket spending. As health refoms are strengthened, special attention must be paid to reproductive health. 2.1 Interventions to improve reproductive health have been examined closely across the world and this has led to formulation of an essential package of services, which includes family planning, safe abortion, treatment of reproductive tract infections/sexuallytransmitted infections (RTIsISTIs), antenatal, natal and postnatal care, child health services, and treatment of post-menopausal problems. To achieve reproductivehealth goals, this package must be delivered to the majority of those in need.. As part of a broad diagnosis of the challenges faced in improvingreproductivehealth, the first part of this chapter examines the actual use in South Asia of the services for which data are available.' The second part draws upon a new analysis to describe patterns of spending on and financing of reproductivehealth services in the region. Utilization of Services Family Planning 2.2 Sri Lanka has a high overall level of contraceptive use, with about 70 percent of reproductive age women using any method, and 50 percent using a modern contraceptive method in 2000. Bangladesh and India have moderate levels of contraceptive use, and Nepal and Pakistan have low levels (Table 2.1).~ 2.3 The five countries have three distinct patterns of contraceptive use. Permanent methods predominate in India and Nepal, temporary methods in Bangladesh ancl Pakistan, and there is an equal distribution in Sri Lanka. In India women who chose sterilization already have four children on average Pathak et al., 1998),more than their I.Due to data limitations, RTVSTI services and the care of older women are not discussed in this chapter. 2.Between 2001 and 2006 Nepal increased modem contraceptive use to 44.2 percent, including 19.2 percent of temporary methods. desired number, but they delay acceptance because terminal methods are irreversible. In effect, poor availability of temporary methods results in unmet need and excess fertility. Although women who use spacing methods generally have lower parity, discontinuation of contraception among them is an important concern. In Bangladesh temporary methods are discontinued by almost 50 percent of women in the first year of use. This demonstrates the need to manage perceptions of as well as actual side effects among women (NIPORT, 2001). Table 2.1 Use of Contraceptives in Five Countries, Various Years (percent) Method Bangladesh India Nepal Pakistan Sri Lanka 1999-00 2003-04 1998-99 2001 2003 2000 Any Method 53.8 58.1 48.2 39.3 32.1 70.0 Any Modern Method 43.4 47.3 42.85 35.4 25.2 49.5 Pill 23.O 26.2 2.11 1.6 3.1 6.7 Injectables 7.2 9.7 - 8.4 3.4 10.8 9.01.2 rtJD 1.2 0.6 1.64 0.4 4.4 5.1 Norplant 0.5 0.8 0.6 -0.3 0.1 Condom 4.3 4.2 3.06 2.9 6.4 3.7 Temporary 36.2 41.5 6.81 13.9 17.6 26.4 Sterilization Female 6.7 5.2 34.14 15.0 7.5 21.0 Male 0.5 0.6 1.89 6.3 0.2 2.1 Any traditional 10.4 10.8 5.0 3.9 6.9 20.5 method Men - 5.1 Women -3.9 Sources: Bangladesh: NIPORT, 2001; NIPORT and ORC Macro, 2004; India: IIPS and ORC Macro, 2000; Nepal: GON-MOH et al., 2002; Pakistan: NIPS, 2003;Sri Lanka: GOSL-DCS, 2002a. 2.4 The public sector is the major source of contraceptives in India, Bangladesh and Sri Lanka, particularly for the poorest women. The private sector in India contributes to condom and pill use and, in Nepal, mainly to condom use. Condoms are used as contraceptives by only three to five percent of couples in the region. Injectables and implants are hardly used in India and Pakistan. In sum, all countries need to expand the choice of contraceptives available, especially temporary methods in India, Nepal and Pakistan, terminal methods in Pakistan and Bangladesh, and male methods in all countries. 2.5 Determinants of Contraceptive Use. Contraceptive use depends on awareness of different methods, access to them, perceptions of their side effects, and desired fertility. Knowledge about contraception is generally high in all five countries. Almost all 15 to 49 year-old women interviewed in the Demographic and Health Surveys (DHS) conducted around the year 2000 in each country knew of at least one method. However, they knew little about the options available, their correct use, side effects, or sources of to terminate her pregnancy up to 12 weeks, up to 18 weeks if the pregnancy resulted from rape or incest, and at any time if recommendedby a qualified medical practitioner. 2.12 India legalized abortion in 1971 through its Medical Termination of Pregnancy (MTP) Act. Under this, abortion can be performed if a pregnancy carries the risk of grave physical harm to a woman or endangers her mental health, and when conception results fi-om contraceptive failure or rape, or is likely to result in a physically or mentally abnormal child. Bangladesh allows menstrual regulation up to eight weeks of pregnancy, and abortion by qualified physicians in hospitals only, if approved by two physicians and necessary to save a woman's life. Sri Lankan law is the most restrictive as abortion is prohibited except when needed to save a woman's life, while in Pakistan preserving physical and mental health are also valid reasons. Sri Lanka's law results in an important difference between poor and non-poor women although almost all other reproductive health services are equitable. Women from higher-income households are able to obtain abortions at private clinics or government hospitals while lower-income women resort more often to backdoor abortion and suffer its negative consequences. Antenatal Care 2.13 Coverage of pregnant women with even one antenatal care (ANC) visit varies substantially among and within the five South Asian countries - from 83 percent in Sri Lanka to 37 percent in Bangladesh. In Pakistan and Nepal half of all pregnant women are reached, while in India two-thirds receive a check-up. The majority of .women are registered in the second or third trimester. The value of one or two visits is questionable especially as the quality of care also requires vast improvement - for example, blood pressure is often not measured in India and Nepal. In all countries coverage with tetanus toxoid (TT) is higher than that with antenatal check-ups, but again there is wide inter- and intra-country variation. Sri Lanka has a high level of coverage (96 percent) fiollowed by Bangladesh (81 percent), India (75 percent), Nepal (55 percent in 2001 and 72 percent in 2006), and Pakistan (51 percent). Opportunities to provide check-ups at the time of TT immunization are clearly missed. While health workers have a strong mandate and supplies to provide TT, they apparently lack the skills, time, equipment, incentive andfor household 'permission' to cany out good antenatal care. 2.14 Despite the high prevalence of anemia in South Asia and uniform policy to treat this, only one-fourth to half of pregnant women in four countries (excluding Sri Lanka) received iron folic acid (FA) supplements. Inadequate knowledge among health workers of nutrition - even of this simple micro-nutrient supplementation action - is a major contributory factor. For example, in three districts of Nepal only 13 to 44 percent of Maternal and Child Health Workers knew the appropriate dose of FA. 2.15 Antenatal care coverage among illiterate women is one-half to one-fourth that of women with secondary or higher education in four countries (except Sri Lanka). Even in Sri Lanka, only two-thirds of illiterate women received ANC compared with three- fourths or more of literate women. In Bangladesh, rural India, Nepal and Pakistan, only women fi-om the richest 20 percent of households have coverage levels comparable to Sri Lanka's overall level. In urban India, coverage among the richest quintile is universal, and the gap between rich and poor is somewhat lower. Only one out of three or four women from the poorest quintile were checked in Bangladesh, Nepal and Pakistan compared with over three-fourths of the richest quintile. In Sri Lanka less than half of pregnant women in the estate sector receive antenatal check-ups. 2.16 There are wide differences in ANC coverage among women of different religions in India, ranging from 84 percent among Christians to less than 65 percent among Hindus and Muslims. Among Scheduled Tribe women, 43 percent had not received even one antenatal check-up. 2.17 Maternal education, economic status and husband's education play strong positive and independent roles in determining the use of ANC. The exceptionis Sri Lanka, where none of these factors play a role because services reach all women. Young maternal age is a strongly negative factor. Adolescent girls are thus doubly disadvantaged:not only do they face higher biological risks but they also have lower chances of receivingcare. Care During Delivery 2.18 In general, about 85 percent of births are normal and 15 percent have associated complications(WHO, 1996). While 'normal' births may not need obstetric intervention (and hence could occur at home), complicated births require assistance at adequately equipped facilities to increase the likelihood of a positive outcome. Importantly, it is impossible to predict into which group a woman would fall as all women are at risk of complications. Untreated or improperly treated conditions during pregnancy or delivery can lead to maternal or neonatal death. Hence, the current recommendation is to ensure skilled attendance at delivery whether at home or at a health facility and a well- functioning referral system. Skilled attendants include doctors, nurses and midwives trained to conduct normal deliveries and to diagnose and manage or refer obstetric complications. 2.19 Differences in reproductive health care among the countries of the region are widest in care during delivery. Almost all births in Sri Lanka occur in an institution and the use of government facilities is widespread (93 percent). However, a lower proportion of women in the estate sector (87 percent) deliver in institutions. Institutionaldelivery in Nepal and Bangladesh is strikingly low, at less than 15 percent; in Pakistan it is 23 percent and in India, 34 percent. Institutional deliveriesare more common than antenatal check-ups in Sri Lanka, while it is the reverse in the other four countries. 2.20 Rich-poor and education gaps are highest for institutional delivery among all reproductive health services (Figure 2.2). In Bangladesh and Nepal three to four percent of illiterate women deliver in an institution compared to half to two-thirds of those with secondary or higher education. In India and Pakistan about 15 percent of illiterate women deliver in an institution. The situation is similar among the poorest women. In Bangladesh, rural India, Nepal and Pakistan, even among women from the richest 20 percent of households, half or less deliver in institutions. In Sri Lanka most socio- different contraceptives. Although "lack of knowledge" was a reason for not using contraception among 5 percent of women in Nepal, fear of side-effects was reported by 25 percent. Service providers also have misconceptions. Health staff in Nepal, for example, thought that women needed at least one child to be permitted to use condoms and two children to use intrauterine devices (IUDs) (VaRG, 2003). In India low awareness of the ease and benefits of vasectomy, and misconceptions (e.g., that it leads to impotence) are responsible for current low use of this method. 2.6 Social norms strongly influence method use. For example, the key olbstacles to contraceptive use in Pakistan Punjab are perceptions that it is culturally or religiously unacceptable, and wives' beliefs that husbands are opposed to it (Casterline et al., 2001). However, recent research is positive about male support to contraception. In Bangladesh, for instance, almost all men are able to name modem contraceptives and, more importantly, almost all agree with their wives on the desired number of children. 2.7 Contraceptive use increases with education and economic status. The gap between rich and poor is particularly striking in Pakistan: use is four times higher in the richest 20 percent of households compared to the poorest (Figure 2.1). The situation in Sri Lanka is the reverse: women in the poorest fifth of the population have higher use (80.3 percent) than those in the richest (69.5 percent). Use of modem methods is lower, and of traditional methods higher, among the richest women compared to the poorest. As there is no evidence of higher family size or higher desired fertility among the better-off, these differences suggest they are more able to control their fertility through means other than contraception - for example, by "saying 'no' to sex". Abortion, which is more easily accessibleto the better-off,may also play a role in establishingsmaller familysize. Figure 2.1 Met and Unmet Need for Contraceptionin Four Countries,Poorest and RichestIncome Groups (quintiles) 90 80 70 60 percent 50 40 30 20 Poor Rich Poor Rich Poor Rich Poor Rich Poor Rich Bangladesh India Rural - India- Urban Nepal Pakistan I Sources:Bangladesh: NIPORT, 2001; India: IIPS and ORC Macro, 2000; Nepal: GON-MOH et al., 2002; Pakistan: NIPS, 2001; Sri Lanka: GOSL-DCS, 2002a. 2.8 Contraception is higher in urban than rural areas in all countries. In Sri Lanka, use is lowest in the estate sector. While use is lowest among 15 to19 year-olds, variation across countries in this age-group is revealing: it ranges fkom a low of five percent in India to 50 percent in Sri Lanka. Other important social differences occur in India: use is higher among Hindus (44 percent) and Christians (45 percent) than among Scheduled Castes (40 percent),Scheduled Tribes (35 percent) and Muslims (30 percent). 2.9 The effects of various socio-economic and demographc variables in determining contraceptiveuse are similar in four of the five countries (excludingSri Lanka): mothers' education, economic status and urban residence have independent positive influence^.^ In Bangladesh, the impact of educationis significant only among those with secondaryor higher education when compared with illiterate women. Women's age and son preference also play important roles. 2.10 The desire to end childbearing among women in all five countries is influenced greatly by education and economic status, particularly in ~ n d i a . ~Illiterate and poor women are more likely to have a desire for more children. Despite this, the poorest still have the highest unmet need, indicating that even the low demand of illiterate and poor women living in rural areas is not met. Ubiquitously, low use among adolescents and young women arises fkom their poor knowledge of reproduction (e.g., low perceived risk of pregnancy), lack of information about contraception, and/or inability to express their desire to delay conception and to interact with service providers. Among older women, unmet need is related more closely to inadequate access to providers, the methods available, and/or their perceptions of the quality of services or of the side effects of contraceptives. For example, much unmet need in India is due to the paucity of spacing methods in the government family planning program and wrong beliefs about those methods that are sparingly available (such as IUDs and oral pills). Social barriers to contraception (e.g., low spousal, familial or societal acceptance) also are important determinants of use. In Sri Lanka unmet need (8 percent) is concentrated among women in plantations, factories, displaced populations, urban slums and underserved rural areas and, along with contraceptive failure, underlies much abortion because emergency contraception is not readily available. 2.11 Unwanted Pregnancies and Abortion Services. One outcome of unmet need is unwanted births - nearly one-fifth of births in the region are unwanted or mistimed. At the individual and family levels, eliminating unwanted births could enhance the well- being of women and children, and at the societal level it could reduce mortality and fertility substantially. A drastic consequence of unwanted pregnancies is abortion. The legal environment and attitudes of service providers determine how women end unwanted pregnancies and how safe the procedures will be, but have little effect on the incidence of abortion (Rahman et al., 1998; Cook et al., 1999). Abortion laws vary across the five South Asian countries. Nepal's law, enacted in 2002, allows any woman '. The variables are: mother's education, husband's education, economicstatus, woman's employment, place of residence, autonomy, exposure to mass media, maternal age, number and sex of children, socio-religious group, and desire for more children. 4. The recent DHS in Nepal shows a 'U-shaped' relationship between education and contraceptive use. A sizeable proportion of women with little or no education undergo sterilization. economic and demographic variables (except birth order) have no influence on institutional delivery as this is near universal, while in the other countries education and economic status have independenteffects. Figure 2.2 Institutional Delivery in South Asia among Women in the Poorest and Richest Income Quintiles (percent) BD IN-R IN-U NP PK SL Ei! Poorest Richest I Computed using data from the followingsources: Bangladesh:NIF'ORT, 2001; India: IIPS and.ORC Macro, 2000; Nepal: GON-MOH et al., 2002; Pakistan: NIPS, 2001; Sri Lanka: GOSL-DCS, 2002a. 2.21 Antenatal care plays a strong role in increasing institutional del.ivery. In Bangladesh and India the odds of women who had more than three antenatal visits delivering in institutions were 15 and 9, respectively, over those who had no ANC. In Nepal even one ANC visit led to odds of 4.6. The chances of an institutional delivery were lower in the youngest age-group than among older women (odds between 0.3 in Bangladesh and 0.6 in India). 2.22 Many women fear going to hospitals to deliver, believing that they would not experience a normal delivery there. As a poor woman in Bangladesh put it:"For money, they (doctors) are ready with knives for surgery." In Bangladesh, more than maternal education and antenatal care, identification of potential delivery complicatio~nswas the most significant factor determining use of modem health facilities for childbirth (Paul and Rurnsey, 2002). Very few home-based births are attended by skilled personnel: under 5 percent in Bangladesh and 10 percent in India. In Nepal there has been an increase in assisted childbirth after the introduction of a maternity incentive scheme under the Support to Safer Motherhood program in 66 (of 75) districts in the country. 2.23 The determinantsof skilled attendance are broadly similar to those of institutional delivery. Power dynamics within households determine the use of health facilities as well as of skilled providers. South Asian women tend to get low priority in household resource allocations, and decisions about contact with outsiders are often made by husbands or mothers-in-law. Such dependence and deprivation can have serious repercussions. At a large public hospital in Karachi, Pakistan, about one-third of obstetric cases were dead upon arrival. Delays were often due to husbands' absences fiom home and the women's lack of autonomy to approach health providers or facilities on their own (Jafarey and Korejo, 1995). Inadequately skilled service providers can aggravate the situation. An evaluation of the Lady Health Worker (LHW) program in Pakistan found that nearly 40 percent of LHWs failed to recognize the need to refer cases of pre-eclampsia; in three districts of Nepal, less than one-fourth of Village Health Workers (VHWs) recognized convulsions as dangerous (Oxford Policy Management, 2002; Bhattarai, 2000). Improving the reach and quality of skilled birth attendance are clearlyof critical importancefor South Asian women. Postnatal Care 2.24 Different norms exist for postnatal care in the five South Asian counties. In Nepal the Ministry of Health and Population recommends that all women receive a postnatal visit within two days of delivery, while in Sri Lanka the policy is to ensure at least two visits within 10 days to women who had institutional deliveries and three to those who delivered at home. India recommends three visits during the first six weeks, the first of which should be within a week of delivery. Thus, comparisons of postnatal coverage across the five counties must be made cautiously. More importantly, appropriate standards need to be adopted and implemented, recognizingthat most women in South Asia deliver at home. 2.25 Many of the differentials in postnatal care (PNC) across and within countries are similar to those for other services. Use of PNC is highest in Sri Lanka (75.5 percent), low in India (36 percent) and Pakistan (23 percent), and lowest in Nepal (17 percent) and Bangladesh (11 percent). Coverage is two to three times higher among those with secondary or higher education compared to illiterate women in Bangladesh, India and Pakistan. However, even among well-educated women, levels of PNC are substantially lower than those of skilled birth attendance and institutional delivery. For example, in Bangladesh only 28 percent of women with higher secondary education and above received postnatal care compared with 66 percent who delivered in institutions and 77 percent who had skilled assistance. These gaps show that the follow-up even of women who have been in contact with the health system is poor. In four countries (excluding Sri Lanka), while postnatal coverage of women in the richest fifth of the population was very low, it was two to three times higher than that of the poorest quintileof women. Childhood Immunization 2.26 In South Asia, childhood immunization is provided through reproductive or maternal and child health programs - hence its consideration here. Child immunization coverage is an indication of continued contact between mothers and health services, and is less affected by the taboos that keep post-parturnwomen away from service providers. In the region, Sri Lanka has the highest child immunization coverage (94 percent) followed by Pakistan (77 percent), Nepal.(66 percent in 2001; this increased to 83 percent in 2006) and Bangladesh (65 percent). 2.27 Despite continuous and strong emphasis on this service in India, only half of all children are fully immunized (UNICEF, 2004). Data on dropouts indicate a lack of tracking mechanisms and follow-up - for example, in Bangladesh, although 95 percent of infants received the first vaccination(BCG), only 70 percent obtained three doses each of DPT and OPV and only 65 percent received one dose against measles. 2.28 Immunization coverage of children increases substantially with. maternal education and economic status. Although levels among the richest 20 percent of householdsin Nepal and Pakistan are similar to those in Sri Lanka (around 90 percent), in India even in the richest urban quintile of households, coverage is 76 percent. One reason for India's current low coverage is the decline of routine immunization activities under the pressure of 'crash' programs such as the Pulse Polio campaigns. 2.29 On the demand side, resistance continues among families who are not adequately reached by information efforts. Substantial differences are seen between children of women who are illiterate and those who have had at least some primary schooling (e.g., in India coverage in these two groups was 28 percent and 47 percent, respectively,and in Nepal it was 57 and 83 percent in 2001, and 74 and 88 percent in 2006). Although immunization increases further with more education, differences are not as large. In Sri Lanka, more than 80 percent of children of illiterate women are immunized. Reproductive Health Service Systems 2.30 This section first briefly describes the government health systems that provide reproductive health services in the five countries, and then discusses private services. Both the public and private sectors are extensive in all five countries. The section also describes the systemic constraints faced by poor women who wish to obtain reproductive health care. Public Sector 2.31 The broad nature of reproductive health services calls for a look, first, at the structures that make policy decisions, plans and intra-sectoral financial allocations in the five countries. In Nepal, India and Sri Lanka, health and family planning services are under one Ministry at the national level.' Even so, coordination among departments that need to be involved in reproductivehealth may be weak.6 In India, for example, both the Health and Family Welfare departments have roles in reproductive health as well as a These are the Ministryof Health and Population(MOHP) in Nepal, the Ministryof Health and Family Welfare in India, and the Ministry of Healthcare and Nutritionin Sri Lanka. In Nepal, the Population Division of the MOHP is responsiblefor other aspects of populationpolicy and activitiesother:than service delivery. 6. In Sri Lanka, reproductivehealthis solely the responsibilityof the FamilyHealthBureau of the Ministry of Health Care and Nutrition, and is implemented through a 'straight line' of health officialsat the provincial,sub-provinceand district levels. Hence, coordinationis not a critical issue. department in charge of women and child nutrition in a separate ministry. In Bangladesh family planning is under the Director General of Family Planning (DGFP) and health services under the Director General of Health Services (DGHS). In Pakistan a decision was taken in 2003 to integratethe Ministryof Population Welfare (responsiblefor Family Planning) and the Ministry of Health (responsible for Maternal and Child Health). Integration of the outreach workers of the two programs was accomplished successfully but coordination at the facility level and above remains a challenge. 2.32 In all countries, reproductive health care is provided largely through primary and secondary facilities. While the service structures are broadly similar, worker- and facility-population norms and skill levels differ. Two differences are immediately apparent between Sri Lanka and the other countries, which may partly explain Sri Lanka's better performance. At the primary level in Sri Lanka, a well-functioning preventive health network gets most reproductive health services actively out to women, rationalizing their use of curative facilities, while in the other countries primary preventive outreach activities are limited, so that health centers receive a mix of preventive and curative demands. Second, the technical expertise available at each service level in Sri Lanka is generally hgher, and the quality of similar cadres of service providers may be better than elsewhere. For example, the Public Health Midwife in Sri Lanka is better trained, skilled and focused on reproductive health care than her counterparts in the other countries. She also has clear responsibility for counseling and for serving pregnant adolescents (even if they are unmarried), which is not the case elsewhere. 2.33 Both clients and providers perceive the inadequacy of staff to be a key problem in service delivery. In Pakistan, providers mentioned that at outlets of the Population Welfare Department the lack of trained staff results in a concentrationon family planning services although the mandate is to deliver maternal and child health care. In India, Primary Health Centers oAen do not conduct deliveries because of the absence of female doctors or indoor staff (Murthy and Barua, 1998; Tata Institute of Social Sciences, 2003; Kavitha and Audinarayana, 1997; Rana and Johnson, 2003). Both vacancies and absenteeism are caused by unwillingness among trained professionals to work in rural areas with poor infi-astructure, inadequate compensation, and limited opportunities for professional growth. Female doctors and paramedics are reluctant to live in isolated places with inadequate security. The conflicts in Nepal and Sri Lanka have severely depleted staff in the affected areas. 2.34 In addition to staffing-up,greater attention is needed to provider skills, motivation and efficiency. Technical knowledge needs to be improved at all levels. Even simple procedures may be poorly performed. A study of hygiene in Nepal observed that hands were not washed routinely before and after treating patients (Friedman, 1996). Workers are poorly motivated as (in addition to their heavy workloads, poor skills and inadequate supplies) supervision tends to be punitive rather than supportive or capacity-enhancing, and there are few incentives. Many doctors and nurses engage in dual practice (whether permitted or not), often referring patients fi-om public facilities to their private clinics and cutting corners in their public efforts. Among the most important needs of public reproductive health care (which is highly dependent on motivated 'workers to reach out, especially to poor women) are strategiesand plans to improve worker performance. 2.35 Basic amenities and maintenance are often poor at health facilities. Inadequate water and toilets deter clients and hamper service delivery. For example, an assessment of 40 facilities in Pakistan found that only 13 percent of Basic Health Units and 62 percent of Rural Health Centers had "good" basic infrastructure (Aziz, 1998). Problems are often found in the availability or functioning of equipment, from the simple to the sophisticated. The lack of one item can render an entire system dysfunctional: for example, an Upazilla Health Complex in Bangladesh had blood grouping and storage facilities but the absence of a blood-bank nearby led to patients who needed blood transfusions being referred to a higher-level facility. Failure to supply medicines is a major cause of client dissatisfaction. 2.36 Lack of privacy and rude or insensitive behavior among public service providers are among the important reasons why women prefer to approach private practitioners. In India women complained that government doctors dismissed their complaints about the side effects of contraceptives or post-abortion complications (Murthy et al., 2002). Consequently, they either did not seek treatment (30 percent) or sought it from private providers (50 percent) who were often unqualified. In general, patients preferred private facilities as they found staff more attentive to their needs, especially in cases of emergency. Providers at public facilities give a number of explanations for these behaviors, ranging from high patient loads to clients' "ignorance" - particularly among the poor or uneducated. Serious social biases are evident; clients belonging to lower castes in India and Nepal often feel discriminatedagainst. In Sri Lanka, where the use of public health facilities is very high (85 percent of inpatient care is provided by the government), the majority of respondentswere satisfied with way they were treated. Private Services 2.37 Private reproductive health services in South Asia are provided b:y both the commercial and not-for-profit sectors. Commercial providers include individual medical practitioners, clinics, dispensaries, nursing homes and hospitals following allopathic ("Western"), homeopathic or indigenous systems of medicine. There are also registered pharmacies, laboratories and formally-trained paramedics. An informal sector includes "less-than-fully-qualified"practitioners,faith healers and traditionalmidwives (dais). It1 the five countries, the formal private sector is concentrated largely in urban areas and spread thinly in rural areas, while the informal sector is predominantly,though not solely, rural. For example in Pakistan, rural areas with almost 70 percent of the country's population have 30 percent of the country's private health facilities. This pattern is repeated in Bangladesh, India and Nepal. Incompletedata indicate that in these countries there may be three to four times as many formally-qualified practitioners in the private sector compared with the public, and up to twice again as many "less-than-qualified" providers. In Sri Lanka there appears to be a more even distribution between the public and private sectors, the ratio of trained to untrained practitioners is higher, and there is a better spatial distribution of trained providers. 2.38 The private sector provides care to the majority in all countries except Sri Lanka. In Pakistan, for example, it serves 60 to 70 percent of the population (Haq, 2001). Even poor people use private care to a greater extent than public services. Ease of access importantly determines use of private facilities, but women seeking reproductive health care also feel that they receive better quality services and appreciate the greater confidentiality, "better behavior" and flexible payment arrangements of private providers. In India the private sector bears 60 percent of the case load of maternal and child services (Kavitha and Audinarayana, 1997; IIPS and ORC Macro, 2000). In Bangladesh the proportion of households using government health services for treatment was estimated as 10 percent, while 60 percent of clients consulted unqualified practitioners (CIET, 2003). In reproductive health, the poor qualifications of private practitioners are evidenced by the high proportion of abortions that are botched. Using spending as a yardstick to compare the use of private and public services in the health sector as a whole, one-half (in Sri Lanka) to over three-quarters (in the other countries) of health care is provided by the private sector. 2.39 Non-Governmental Organizations. A wide range of non-profit organizations (NGOs) are active in the reproductivehealth sector in South Asia. In Bangladesh, NGOs provide as much as 40 percent of all reproductive health care, in contrast with the other four countries where their coverage is limited to a few percent. They provide both curative and preventive care, the latter including contraceptive services and health education, especially for family planning, STIs (including HIVIAIDS) and safe motherhood. In curative care, their main focus is maternal and child health. While some NGOs are localized, others have extensive networks across their country of operation. For example in Pakistan, the Family Planning Association of Pakistan, Marie Stopes Society, and Aga Khan Health Services, and in Bangladesh, BRAC and the Marie Stopes Society have a wide presence. The Family Planning Association of Sri Lanka and Population Services Lanka Limited have a large network of associated sub-national organizations. The Family Planning Association of Nepal is involved in the full range of reproductive health services, having gradually shifted its focus from family planning to comprehensive sexual and reproductive health. In Sri Lanka in recent years NGOs have increased their care (and legal assistance) to victims of gender violence. On balance, however, the non-profit sector in South Asia leans toward preventive and promotive health, rather than curative and clinicalservices. 2.40 In India, the NGO sector covers an estimated 4 percent of the population and is diverse in capacity and performance. An evaluation of a USAID-supported program that involved 131 small and large NGOs providing primary health care in remote villages in 13 states found, on the whole, that the NGOs performed rather poorly, particularly in ensuring community participation and the sustainability of their efforts wed, 1997). Generally, however, NGOs in the health sector are believed to be pro-poor, client- fiendly and more "rational" in their use of medicine than the public sector, thereby achieving greater efficienciesand reputations for more sensitiveand better managed care. These attributes and their willingness to work in backward or remote areas make them useful partners to meet reproductivehealth goals. Publicprivate Shares of Reproductive Health Services 2.41 Few data are available for the five countries (particularly Nepal and Pakistan) on the private and public shares of reproductivehealth care. In general, and in sornecontrast with "general" health care, public sources tend to outstrip private services for immunization, family planning and antenatal care. On the other hand, private sources (particularly if traditional providers are included) surpass public providers for treatment of RTIsISTIs, abortion and home-based delivery. There are exceptions to these generalizations, For example, the public sector in Bangladesh is expanding the availability of medical abortion; and private provision of birth spacing is increasing through community-based distribution and social marketing. As institutional deliveries in India increase, a greater share is being provided by the private sector; and the private share of condom distribution has increased due to higher over-the-countersales as well as to social marketing, some of which is supported by public-private partnerships. In Sri Lanka, provision of reproductivehealth care is overwhelminglypublic. 2.42 In addition to differences by type of service, use of publiclprivateproviders varies by socio-economic status. In Kerala, the Indian state with the most favorable reproductive health indicators and good services in both the private and public sectors, there are wide variations in the use of services by levels of household wealth. The poorest people obtain a greater share of services from the public sector, while the reverse is true for the richest and middle-income groups (Mahal et al., 2001). In other states which do not have as well-developed services nor as favorable heath indicators, the proportion of hospitalization at public facilities among the poorest is roughly the same (57 percent in Kerala and 55 percent in UP); but among the richest, 29 percent of hospitalizations in Kerala and 39 percent in UP were at public facilities. Furthermore, while women from the richest quintile of households in Kerala had about 1.5 times the hospitalization rate of the poorest, in Uttar Pradesh this ratio was 5. In Orissa, another poor performing state, the private sector accounted for less than 15 percent of hospitalizationsamong the richest quintile, and only 8 percent of those among the poorest quintile. During the 1990s the private share of out-patient and hospitalization services in India grew considerablyin both urban and rural areas and among the poor. 2.43 In Bangladesh 36 percent of institutional deliveries took place in private (including NGO) facilities, and the rest at government health facilities (NIPORT et al., 2003). Of 62 percent of pregnant women who sought care for a life-threatening condition, almost as many went to an unqualiJied private doctor as to a qualified doctor (private or public). Among 42 percent who sought treatment for non-life-threatening conditions, again, almost equal shares went to unqualified and qualified doctors. Among those who sought treatment for life-threateningconditionsoutside the home, almost equal shares approached government facilities and private/NGO facilities. However, for non- life-threatening conditions a far larger share used private facilities. The poorest quintile used more unqualified doctors than the richest, and fewer qualified doctors overall for both types of conditions. When facilities are compared, however, the poorest use government facilities proportionately more than private facilities for life-threatening conditions than for less severe ones. A study in Gujarat suggests that women's education, income and family structure are important predictors of the use of private/government/NGO/traditionalservices, and that women are more sensitive to the social and indirect costs of services (travel time, transport costs) than to direct costs (Vissandjee et al., 1997). Public-Private Partnerships 2.44 Many and various types of public-private partnerships (PPP) have been developed in the reproductive health sector in South Asia. In a unique example fiom Sri Lanka, the services of doctors (especiallyspecialists)fiom the government sector are provided to the private sector through "channeling centers." Doctors officially provide services after office hours and charge consultation fees. This arrangement helps the private sector to cater to the demands of those who are well-off and able to afford paid services. Further, there is co-operation in emergency care. Government specialists who are Visiting Consultants in the private sector have the option to refer patients in emergencies to the governrnent hospital where better facilities exist.7 These arrangements exemplify public to private human resource flows in Sri Lanka and contrast with PPPs in the other four countries, which involve contractingin of private providers or contracting out of services because of severe manpower or infrastructuredeficits in the public systems. 2.45 In Bangladesh, the Government's Health and Population Sector Program (HPSP) includes an initiative to support NGOs to deliver MCH and FP services to poor and under-served communities. Essential health care services have been implemented through this PPP model in two upazillas. While successful, it is unclear how this will be scaled up to the national level (i.e., to about 460 upazillas). In India's Child Survival and Safe Motherhood Program and its successor, the Reproductive and Child Health Program, a large number of NGOs received support for service delivery, and many involved in health communications and training. Although useful to extend coverage, the reach of these efforts has been short of expectations and the hoped-for adoption of innovativeor improved practices by the public sector has been slow. 2.46 In Pakistan, the Ministry of Population Welfare has Reproductive Health Centers located within larger public and private health facilities to provide a range of family planning services. A bold initiative in the mid-1980s was the social marketing of condoms through private channels to expand information and services to a larger clientele. This PPP has been quite successful. Another type of partnership involves public facilities being run by private agenciessuch as the Punjab Rural Support Program. If current examples work well, the governrnent aims to turn over other poorly-functioning health centers to private managers. A similar strategy has been successful in some states of India (e.g.,Maharashtra and Gujarat), but scale-up has been slow. Not only are there a limited number of suitable applicants, but many consider the government's terms and conditions too rigid. In contrast,in Bangladeshand some Indian states, private practice by governmentstaff is permitted after working hours but there are no formal arrangements,while in Nepal, Pakistan and other Indian states, such dual practiceis not permitted. 2.47 A number of other PPPs have also been tried in India in addition. to social marketing and franchising of family planning services and contracting out of government health facilities. Private service providers from anesthetists to laundries and security guards have been contracted in. NGOs have been involved extensively in community mobilization, IEC, HIVIAIDS counselingand care, training and advocacy in reproductive health. Under the RCH program CHCs which have hctioning operation facilities can hire private gynecologists and anesthetists to perform Caesarean sections. However, implementation is scanty due to a lack of interest among private practitioners (who fear losing their practice) as well as government staff (who do not want the additional workload). Costs and Financing of Reproductive Health Services 2.48 South Asian countries face significant challenges in organizing health institutions and financing to ensure sufficient resources to improve access to and equity in health care, and to protect people - especially the poor - against health shocks. Significant improvements in reproductive health outcomes in South Asia will be realized only with larger and more effective investments in health services, oriented to benefit poor households who currently face the greatest constraints to service use. New analyses carried out in this study to estimate the level and allocation of financing of reproductive health care in the five South Asian countries highlight several "points to watch" as decision-makersassess financing options.' Four findings stand out: Across the countries of the region, per capita spending on reproductive health services delivered through the public sector is not correlated with health outcomes. This suggests that in at least some countries there is significant room to increase the efficiencyof spending with respect to outcomes. In most countries a large financing burden falls on households. In the absence of effective risk-pooling mechanisms, the poor are disproportionately affected. As financial barriers limit the use of preventive services and/or lead to delays in seeking medical attention, this contributesto poor health outcomes. '. Important data limitationsaffect our ability to make definitive statementsabout financingpatterns. Comprehensive National Health Accounts (NHA) and breakdowns,including private spending, were available only for Bangladesh,Sri Lanka and the Indian state of Rajasthan, thus limiting the analysis of private and total spending. Estimates of public financing were compiled for all countries, but a detailed comparison of Pakistan with the other countries was not possible because of accounting inconsistencies. Differencesin the data available for each country result in some variations in the methods used to analyze public (including donor) and private expenditures on reproductivehealth. Details of the methods used for each country and assumptionsmade are presented in Annex 2 Section 2.5. For example, Bangladeshand Sri Lankahave NHA, enabling direct computations;for the Indian states of AndhraPradesh and Rajasthan, it was necessary to use a combinationof NHA-based methods and private expendituredata from the National Sample Survey. In Nepal, data on private expenditure were not available, while in Pakistan, go,vernmentdata were incomplete. The details of the householdsurveys that provide out-of-pocket expendituredata also vary across countries. While public expenditurescould be analyzed by service componentsin five locations (excludingPakistan) private spending could be disaggregatedonly in Bangladesh,Sri Lanka and Rajasthan. Hence, comparisonsbetween the countrieslstatesare made cautiously, mindful of these variations. The priorities of donors and governments are poorly aligned, hindering implementation. Both need to focus squarely on supporting the essential package of reproductive health services. In particular, while donors have focused on support to family planning services, the greatest share of the financing burden on households has been in the area of maternal services. This is an important area in which greater spending is required to achieve improved outcomes - donors need to increase their investments and national governments need to increase their investments even more. Creative approaches to using general revenues that are spent on reproductive health, such as demand-side financing, and greater contributions of the private sector to the health of poor women, possibly through partnerships, could help to improve outcomes. 2.49 In South Asian countries, the public health sector provides free or highly subsidized health care funded from general revenues (supplemented to varying degrees by grants from bilateral donors andlor loans from international financial institutions). Whether funding is from national or sub-national sources varies by country. 2.50 In all countries, private out-of-pocket financing plays an important role. User fees are levied on services at many public facilities (except in Sri Lanka), typically with exemptions for low-income patients. However, in several countries, informal payments are the norm, even for the poorest patients. Social insurance mechanisms have not been of major importance (notwithstanding the scheme which covers formal sector workers in India). Thus, private spending occurs primarily as out-of-pocket expenditure, which is least desirable from an equity perspective. 2.51 In all five countries, a variety of changes are underway in health sector financing and organization to support expansions in quantity and quality improvements in health care. In Nepal, for example, government policy has focused on three areas: privatization, use of community schemes particularly with respect to essential drugs, and income generation at public facilities (Agarwal, 1998). In India, financing reforms at the state level have included the introduction of user fees, setting up "autonomous bodies" to manage public hospitals, various types of public-private contracts, and health insurance schemes. 2.52 Although the region has re-conceptualized reproductive health care in recent years, institutional constraints have impeded implementation. Among these, a particularly important constraint to reproductive health improvement is the continued separation of organizational structures for MCH-Family Planning, general health services and nutrition. This three-way split not only causes a separation of service delivery sites and affects integration of services but also results in inefficient allocations of scarce public sector resources and, most likely, increased out-of-pocket spending as well. To date, only limited attention has been devoted to addressing demand-side barriers to improve utilization of key reproductive health services. 2.53 Within these contexts, Sri Lanka's long standing policies for maternal and child health are noteworthy. The country has put a strong emphasis on motivating women to obtain institutional care for delivery as well as for treatment of infections at the same facilities. Investing simultaneouslyin its preventive network, it achieved efficiencies by ensuring that PHMs covered all the relevant target groups in their areas, including adolescents and, as the country's age structure changed, older women could seek care at Well Women Clinics that have been established over the past 12 years. Sri Lanka's investments in health have been proportionately greater than in the other countries and egalitarian, protecting the poor from the costs of inpatient services and keeping levels of private spendingrelatively lower, especiallyfor reproductivehealth. Resource Mobilization for ReproductiveHealth 2.54 A picture of resources currently mobilized for reproductive health is provided by public and private spending on these service^.^ The services covered include maternal health (ANC, PNC and dietary supplementation),childbirth (delivery, care of pregnancy complications, and abortion), infant care (immunization, micronutrients, growth monitoring and health care), family planning (including goods and services, counseling and IEC), prevention and control of RTIsISTIs including HIVIAIDS, and other personal reproductive health services for women (all obstetric and gynecological services).'' In the case of integrated services, the estimates include the relevant shares of program overheads, supportinginfrastructureand other non-servicedelivery expenditures. 2.55 Total Spending. Estimates of total (public and private) expenditures on reproductive health are only presented for Bangladesh, Sri Lanka and the Indian state of Rajasthan. As a percentage of GDP they vary considerably between Sri ILanka (0.3 percent) and Rajasthan (1.3 percent), mainly due to differences in private spending." As a share of total health spending, reproductive health spending varies in a similar way: from 10 percent in Sri Lanka to 21 percent in Rajasthan. 2.56 Public Expenditure. Public sector health expenditures as a share of GDP are high in Sri Lanka and Rajasthan, but relatively small shares are allocated to reproductive health (Table 2.2). Despite its low expenditure, however, Sri Lanka provides near universal access, which supports high utilization of reproductive health care. Another Indian state, Andhra Pradesh, has lower overall health expenditureas a share of GDP, but spends a much larger share on reproductive health. Bangladesh and Nepal allso allocate over a quarter of public sector health resources to reproductivehealth services. 9.New estimates of expenditureson reproductivehealth care were developed for this study. The estimates are full costs (as measured by expenditures)including both the recurrent and capital costs of providingservices, but do not represent full economic costs of service delivery. 10 . Reproductivehealth expenditures do not include child nutritional supplementationother than rnicronutrients. Among child health expendituresbeyond infancy,some such as immunization are included but others - such as general illness consultations - may not be. ". The privatespendingfor Rajasthancould be overestimated,as the estimatesare based solely on household surveydata and have not been reconciled with provider-sidedata as in Bangladeshand Sri Lanka. Table 2.2 Public Sector Expenditures on Health and Reproductive Health in South Asia India: India: Sri Bangladesh Andhra Lanka 2000-01 Pradesh 1999-00 1998-99 199, 2000-01 Total health expenditures 332 266 188 47 252 (constant2000 US $ millions) Total reproductive health expenditures(constant2000 88 1 1 1 28 13 39 US $ millions) Healthexpendituresas a shareofGDP0 0.8 0.9 1.7 0.9 1.6 Reproductivehealth expendituresas a share of 0.2 0.4 0.3 0.2 0.2 GDP (%) Reproductivehealth expendituresas a share of 26 42 15 28 16 total health expenditures(%) 2.57 Private Spending. Household resources spent on both health and reproductive health as a share of national or state income are significantly larger in Bangladesh (2.6 and 0.4 percent) and Rajasthan(4.2 and 1 percent)than in Sri Lanka (1.6 and 0.1 percent) (Table 2.3). Although public and private health expenditures as shares of GDP are roughly the same in Sri Lanka, private reproductive health expenditure is half that of public, reflecting the greater degree of support provided by the state for reproductive health- an outcome of its policy emphasis on maternal and child care that began early in the 20th century. In Rajasthan, private reproductive health expenditure is more than three times public expenditure. Indeed, in most parts of the region, the levels of out-of- pocket spending on essential health care (relative to public investments) are cause for concern, particularly given the barrier that this can create in the health-seekingbehavior of poor households. 2.58 Public-Private Financing Mix. Overall resource mobilization for reproductive health as a share of national resources is inversely related to the level of public financing. For example, the lowest level of reproductive health spending is in Sri Lanka, which has the highest proportion of public funds, and the highest overall spending is in Rajasthan with the largest private fimding mix. Since households pay for both private spending (directly out-of-pocket) and public spending (indirectly through taxes), the financing strategythat would be least burdensome for households is likely to be that which involves the most extensiverisk pooling. Table 2.3 Private Expenditureson Health and ReproductiveHealth in South Asia - - Private health expenditures(constant2000 1,055 462 257 -- Privatereproductivehealthexpenditures 144 111 (constant2000 US$ millions) I Private health expendituresas a share of 2.6 4.2 1.6 GDP (or state GDP) (%) Private reproductivehealth expendituresas 0.4 1.O 0.1 -a share of GDP (or state GDP)(%) Privatereproductivehealth expendituresas 13.6 24.0 a share of total private health expends.(%) 2.59 Inter-country Comparisons of Public Expenditure per Capita. When standardized, the levels of public sector reproductive health expenditure per capita vary somewhat among the five countries.I2 Expenditure in Andhra Pradesh (U-S$ 2.5) is greater than in Sri Lanka (US$ 2) and the other countries (around US$ 1.5). However, these variations could not account for the differences between countries in service delivery levels and reproductive health outcomes. Instead they suggest that efficient use of financial resources can achieve better outcomes. Sri Lanka has better reproductive health than the other countries not because it spends more but because it uses a similar level of finance more efficiently and equitably. 2.60 The standardized private expenditures reinforce the conclusion made above and in the preceding section. The burden on households with reproductive age women in Sri Lanka (US$ 0.6) is substantially lower than that in the other countries (US$ 2.7). It appears that Sri Lanka's stronger public sector effort, including higher efficiency, considerably reduces out-of-pocketcontributions by households for reproductive health. 2.61 Donor Assistance. Donor support to reproductive health varies greatly across the region from 3 percent of public reproductive health expenditure in Sri Lanka to around 30 percent in Bangladesh and Rajasthan and 65 percent in ~ e ~ a 1In ~ . lIndia as a whole, a large proportion of public expenditure on reproductive health is funded by international agencies. For instance, in 2001-02, two-thirds of the estimated US$ 228 million of Family Welfare spending by the Central Government was met from external finds (GOI- MOHFW, 2002). 2.62 In Pakistan donor support to the Population Welfare program has declined over time and over the past decade the program has relied largely on government funding. However, donors continue to finance the two major social marketing agencies which distribute more than two-thirds of all condoms and one-third of oral pills in the country. 12. . Pakistancouldnotbeincludedin thiscomparativeanalysis in theabsenceofrigorousaccountingestimates. 13.. These estimates are based on an analysis of donor project data for Nepal and Sri Lanka, national health accounts for Bangladesh and reproductive health accounts for Rajasthan. comparable estimates could not be produced for A.P. and Pakistan because similar data were unavailable. Foreign assistance as a proportion of public health sector allocations varied between 4 and 16 percent during the period from 1998199 to 2004. 2.63 Who Benefits from Public Reproductive Health Spending? Severe data inadequacies introduce difficulties in estimating reproductivehealth expenditures at sub- national levels. In Bangladesh, India and Sri Lanka household surveys have the necessary data, but the sample sizes of those reporting reproductive health expenditures at the provincial or district levels are too small to permit any meaninghl analyses. Therefore, only distributional analyses of the national samples were attempted. The distribution of use of services is an acceptable proxy to assess how well public resources are targeted at poor users. The data indicate that the richest quintile of women obtain a disproportionate share of institutional deliveries at government health services in most South Asian countries (Figure 2.3). The exceptions are urban India, where the distribution is progressive as richer women have more private sector alternatives,and Sri Lanka where institutional deliveries are near universal. The distributions of other reproductive health services demonstrate a similar pattern, with few exceptions. These data suggest that public resources need to be targeted more efficiently at the poor andlor that governments need to achieve universal coveragewith essential health services. Figure 2.3 Distribution of Deliveries in the Public Sector, Poorest and Richest Economic Groups (quintiles), Various Years 2.64 What Resources are Needed for Reproductive Health? The International Conference on Population and Development Plan of Action contained estimates of the resources required to provide basic reproductive health services, covering family planning, ANC, delivery, PNC and abortion services, STIs, EC, treatment of infertility and some capacity building. However, resources required for emergency obstetric care, child survival programs, broader ST1 programs including HIVIAIDS and strengthening primary health care delivery systems were not included. Subsequently, the UN Millennium Project estimated the resourcesrequired to achieve the health MDGs (among others) at the country level, and the WHO (2005) worked out the requirements to achieve universal coverage with maternal and newborn care. Most recently, Vlassoff and Bernstein (2006) have been able to rework and update the ICPD costs, including health system overheads as well as improvement costs, EmOC, HIVIAIDS preveiition, data collection needs, and other relevant investments. The estimates for one of the focus countries, Bangladesh, show that US $3.6 will be required per capita by 2015, over double the estimate for 2005 (US $1.6). The figures for Cambodia and Ghana were similar- so that an extrapolationto the rest of South Asia would not be ~nreasonable.'~ 2.65 If reproductive health services in Bangladesh called for a total of $229 million in 2005 and US $610 million in 2015, at least these amounts would be needed in Pakistan, and seven times as much in India. Clearly, expenditure in all these countries is far short of what it would take to deliver adequate reproductive health care to serve the MDG goals. Another estimate points out that over the next 20 years increases in populationand changes in the age structure alone in South Asia will result in an increased need of 45 percent in total health spending or 2-3 percent annually (Gottret and Schieber, 2006). The South Asian governmentsmust take the major responsibilityfor meeting the costs of providing services at least to the poor. It is noteworthy that the country that provides the best reproductive health care in the sub-continent, Sri Lanka, also has the lowest proportion of its health budget provided by donors (3 percent). 2.66 Nevertheless, increases in donor assistance would be helpful to achieve the increased spending required. This in turn would increase the proportion of the total reproductive health budgets of the five countries contributed by donors, which is currently quite low in four of the five countries (except Bangladesh). While total donor funding to health globally is estimated to have doubled between 2002 and 2005, much of the increase constituted assistance to HIVIAIDS programs, including treatment and care costs, and South Asia received a relatively low share. Resource Allocation within Reproductive Health 2.67 In concert with mobilizing additional resources for reproductive health care, governments and donors need to pay attention to how these resources are distributed to 14. While there are manystudies on the costs of differentreproductivehealth services in developingcountries, they measuredifferent activities and use different units (e.g., visits, pregnant women, couples, etc.), the costs of individual services range widely,and few have worked out the cost of a total package,correcting for the common infrastructureused in the deliveryof different services. While several studies are useful to estimate the costs of individual services,even in the aggregatethey would not serve well l?owork out the cost of the essential package of RH services for other countries. Information on the costs of individual or 'packaged' RH services is woefully inadequatefor South Asian countries- and thereis considerabledeviationfrom the cost estimatedby Vlassoffand Bemstein (2006). For example, an effort to cost"Essential Health Care Services"in Nepal reported the 2003 cost per woman of 2 ANC visits as US$2.53and the cost of a Mother-Babypackage as US$7.57. Even the latterfigure does not include, for example, the full cost of family planning or ST1services (Alban and Sakya,2004). Considerablywork needs to be done in this area. Cost-effectiveness analysesfocus on deaths averted or DALYs saved, and thus also do not servereproductive health services well enough as the benefits, for example, of family planning are not fullycaptured by these measures. Many studies on the cost-effectivenessof RH interventions are ably summarizedin the DiseaseControl Priorities Project(DCPP) chapterson AdolescentPrograms,Contraception, STIs, HNIAIDS, Matemal and Perinatal Conditions,Quality of Care (for child health), Gender (for "women's conditions") and Overview chapter (World Bank, 2006). achieve greater equity and better health outcomes. This section discusses the current scenario and points to improvements that must be made. 2.68 Total Resource Allocations to Reproductive Health Components. The distribution of total reproductive health resources by different service components is similar in Bangladesh, Rajasthan and Sri Lanka (Table 2.4). Maternal health and childbirth services account for 26 to 33 percent of total spending, and family planning for 7 to 18 percent. One striking differenceamong the three countries is the relative share of resources spent on routine gynecological outpatient and inpatient services. It is less than 3 percent in Bangladesh compared with 35 percent in Sri Lanka and 43 percent in Rajasthan. Table 2.4 Total Expenditureon Reproductive Health by Components(percent) India - ReproductiveHealth Bangladesh Rajasthan Sri Lanka Components 2000-2001 1998-1999 1997 Maternal health 18 11 9 Childbirthand pregnancy complications 14 15 24 Infant care 47 24 23 Familyplanning 18 7 9 Prevention and control of STDs 0 0 0 Other inpatient RH servicesfor women 1 19 Other outpatient RH servicesfor women 2 43 15 Total 100 100 99 2.69 Allocations of Government Resources. Public reproductive health resources in Andhra Pradesh, Bangladesh and Nepal are allocated largely to family planning and infant care (Table 2.5). Maternal health receives moderately large shares in Bangladesh and A.P. Resources to other reproductive health services are relatively small or insignificant in these countrieslstate. 2.70 In Sri Lanka, on the other hand, childbirth and other reproductive health services for women, particularly inpatient services, obtain the largest shares of public resources for reproductive health, thus providing a form of public insurance for inpatient care, which could otherwiseimpose a large financial burden on households. This pattern in Sri Lanka emerged from the very early emphasis on institutional care in the context of addressing maternal and child health needs in the island country. The Indian state of Rajasthan also allocates about 20 percent of reproductive health resources on other services for women (including treatment of RTIs and abortion services which are not captured separatelyby this analysis). Table 2.5 Allocation of Public Sector Resources to Reproductive Health Components (percent) India- I Andhra India- ReproductiveHealth Bangladesh Pradesh Rajasthan Nepal Sri Lanka Components , 2000-01 2000-01 1998-99 1999-2000 1997 Maternal health --15 26 8 11 11 Childbirth and 'pregnancycomplications 8 18 25 Infant care (including immunization) 37 16 26 27 24 Family planning 26 57 16 58 10 Preventionand control 4 of STDs 0 0 0 0 Other inpatient RH 20 Other outpatient RH services 2.71 Trends in Resource Mobilization for Reproductive Health Services. Reproductive health received increasing shares of public health resources during the periods studied in Bangladesh, Andhra Pradesh and Nepal, with significant variations in different ~om~onents.'~In real terms resources increased most in maternal health (45 percent),infant health (30 percent) and family planning (45 percent) in Bangladesh, wlvle in Nepal an increase occurred in family planning. In Andhra Pradesh, expenditures on child birthlpregnancy and family planningincreased by over 40 percent, infant care by 50 percent and maternal health by 62 percent. 2.72 In Pakistan the trends in population welfare expenditures were rather erratic due to the program's heavy dependence on donor funding. Since the late 1990's the population program has relied increasingly on government financing. In Sri Lanka the ratio of public and private reproductive health expenditures to total public and private health expend.itures changed very little during 1990-1997. In real terms, government spending on childbirth and general obgyn services increased by more than 60 percent in this period, while family planning expenditures declined by 20 percent. NGO expenditures on family planningincreased sharply in the same period. 2.73 Donor Support to Various Services. A functional analysis of donor expenditures indicates that an overwhelming proportion of resources to reproductive health goes to family planning in Nepal and Sri Lanka. In Bangladesh, however, infant care expenditures receive the largest share of donor reproductive health resources (37 percent) 15.Expenditureson reproductivehealth were estimated for the followingyears: Bangladesh:1999-00-2001-00; India-Andhra Pradesh: 1996-97-2000-01; India-Rajasthan:1998-99; Nepal: 1995-96-1999-00; Pakistan:1993-94-1998-99; Sri Lanka:actuals for 1990-1997 and estimatesfor 1998-1999. The ratios used to estimatereproductive health expendituresfrom data on total health expenditures were the same across the years for each country. The estimation ratios were generally calculated using household data and other information that werenot available on an annual basis. No trend analysis was possible for Rajasthan. and maternal health and family planning account for 28 percent each. The pattern of donor assistance to Bangladesh stayed roughly the same during 1999-2001. 2.74 In Nepal, 80 to 90 percent of donor support was to family planning during the period from 1995 to 1999. In a country whose maternal mortality ratio is among the highest in the world, less than 20 percent of donor reproductive health assistance was on maternal health and childbirth services. In Sri Lanka, the share of donor resources to infant care has fallen from about 50 percent to 10 to 20 percent in recent years, while the share on family planning (which may include some MCH care) has increased significantly. This is incongruousgiven that the country has had below replacement level fertility since 1993. The low proportions of donor contributions to maternal, childbirth and infant services are surprising, given the importance of safe motherhood to the Millennium Development Goals and donor commitments to achieving these. 2.75 Continued donor emphasis on family planning harks back to historical concerns about population growth in South Asia and demonstrates insufficient agility after ICPD. Sri Lanka's early success in controlling fertility and other positive social indicators kept levels of donor assistance to the health sector low in that country. Support to the other countries has increased over time and expanded from family planning to some MCH activities such as immunization, vitamin A supplementation and child nutrition, but despite recent efforts at sector-wide approaches and budget support, many donors continue to prefer picking up the tabs for specific "vertical programs", components or commodities.16 It is difficult to pick "reproductive health services"because of the broad ways in which these are organized (e.g., institutional deliveries are supported by overall hospital resources, diagnosis and treatment of RTIsISTIs are done by general health services) other than at the field level. 2.76 Unit Costs of Public Services. In South Asia, the unit cost for institutional deliveryat public facilities ranged from over US$100 in Bangladesh and A.P. to less than US$40 in Sri Lanka, Rajasthan and Nepal. The low unit cost in Sri Lanka reflects the high utilization of public facilities for childbirth, while the very low unit cost in Nepal (US$12) reflects low spending. Similarly, very low unit costs for family planning in Sri Lanka can be attributed to high utilization, while high unit costs in A.P. reflect the low use of the services relative to the high level of spending on family planning in the state. Overall, the picture that emerges is that Sri Lanka achieves good reproductive health outcomes with relativelylow per capita and per service spending. Is. Indeed, while the vertical programs (e.g., malaria, TB, EPI, HIVIAIDS)emerged originally from the desire of technical agencies to deliver interventions efficiently to the field level, they have persisted due to separate sources of funding and technical assistance, resulting in waste and inefficiency. Governments, donors (and international partnerships) and technical agencies have been concerned to ensure priority to and mobilize/protect budgets for particular programs, bypass weak systems by having separate delivery 'shutes', and achieve "quick results." As a result, and because resources have been insufficient even in the aggregate, they have been spread thin, systems have remained fragmented and weak, and human resources underdeveloped and uncoordinated -both horizontally and vertically. (In some cases, there may be more than one cadre canying out similar tasks - for example, MCH and FP workers in Bangladesh.) Allocations of Private Resources within ReproductiveHealth Care 2.77 Allocations of private resources to different reproductive health services vary across the five countries (Table 2.6). Other reproductive health services for women take up a substantial share of total private resources in Sri Lanka and Rajasthan. Around one- third of private spending in Sri Lanka is on outpatient services, which are generally low cost and are provided at reasonable quality. In Bangladesh, more than half of private resources are spent on infant care. Table 2.6 PrivateAllocationsto Reproductive Health Components (percent) 1 1 1 1 ReproductiveHealth Components ~ ' 1 p o f i g h India- Sri Lanka Rajasthan 1997 1998-99 I Maternal health 13.2 10 Childbirth and pregnancy -- 18.1 14 Infant care 52.8 23 Family planning 13.2 5 8 Preventionand control of STDs 0 0 0 1 1 1 I Other Inpatient RH services 0 I 50 1 15 IOther Oubatient RH Services I 2.1 I I 31 1 1 Total 99.4 99.4 100 2.78 Private financing of each reproductive health component increases with an increase in overall private expenditure (Figure 2.4). In Sri Lanka, public funding is the predominant source for all components (80 to 90 percent), except routine outpatient obstetric and gynecological services (40 percent), demonstrating a progressive use of public and private resources. A high degree of public financing of maternity (around 55 percent) and inpatient services (80 percent) is seen also in Bangladesh, indicating that public spending responds to some extent to market failure. However, no such pattern is seen in Rajasthan. 2.79 Average Private Costs of Reproductive Health. Average private costs per couple for modem contraceptives are very high in Bangladesh (0.72 percent of per capita GDPQ compared with Sri Lanka (0.05 percent of per capita GDP) and can be attributed almost entirely to NGO spending. 2.80 Survey-based estimates of out-of-pocket payments for reproductive health have been derived for Pakistan and AP. The 1995-96 round of India's National Sample Survey (NSS) provides data on payments for childbirth, which indicate the variation in out-of-pocket expenditure between rural and urban settings. In Andhra Pradesh, institutional deliveries were expensive, more so in rural areas (9 percent of per capita GDP) than urban (7 percent), while home deliveries were cheaper in rural areas (1.2 percent) than in urban ones (2.1 percent). In Pakistan, a visit to a private practitioner for immunization costs four times more than a visit to a government facility (0.4 and 0.11 percent of GDP, respectively). Figure 2.4 Public-PrivateCompositionof Financing of ReproductiveHealth Components,Bangladesh,Rajasthan and Sri Lanka, Various Years (percent) 2.81 Analysis of data on household consumption and health expenditure in Bangladesh indicates that the richest quintile of households accounts for a larger share of household spending on reproductive health care (61 percent) than of household spending on health care in general (34 percent). An important issue is the financial burden imposed on households by reproductive health spending. Analysis of household data for Andhra Pradesh and India indicates that (with the exception of urban areas in AP)the financial burden of facility-based deliveries is somewhat progressively distributed in India, with richer households more likely to incur higher payments as a share of total household consumption. The fact that richer women make use of private hospital services for childbirth may account for much of this difference. However, this distribution of spending also reflects the low overall use of institutional facilities by poor women. AN AMBITIOUSAGENDA FOR FIVE COUNTRIESIN SOUTH ASIA: TOWARDBETTERREPRODUCTIVEHEALTH FORPOORWOMEN A broad agenda for improving reproductive health outcomes among poor women in the countries of the South Asia region has two elements: (a) improving the technical strategies for reproductive health service delivery, within the public sector, with particular attention to the needs of adolescents and women from poor households; and (b) exploring optionsto increase the amount of funding, and improvethe efficiency and equity of resource allocation. Improved technical strategies would involve efforts to expand the reach of services, increase demand, integrate a comprehensive package of services, improve quality, and create an enabling environment across sectors. Financing options include ways to increase efficiency of current spending, mobilize new resources, organize provider reimbursement to reward high performance, and develop financing mechanismsto protect the poor from burdensomeout-of- pocket payments. 3.1 Describing problems is always easier than fixing them. This is particularly the case when so many of the shortcomings in the supply of and demand for reproductive health services, and the poor health outcomes themselves, are a result of a longstanding set of problems that have deep social and economic roots. So, for example, the challenge of overcomingthe barriers between the health system and poor householdsis an enduring one that few countries have tackled in a way that is fully successfil. While there are admittedly no quick fixes, the imperative to take constructive steps forward, within an ambitious and long-tern framework,is urgent. 3.2 This chapter presents a broad agenda to frame programmatic and policy discussions, with two elements: (a) improving the technical strategies for reproductive health service delivery, within the public sector, with particular attention to the needs of adolescents and women from poor households; and (b) exploring options to increase the amount of funding, and improve the efficiency and equity of resource allocation. Technical Strategies toward Better Outcomes 3.3 Five key priorities should dominate the programmaticagenda. These are: Increase the supply of reproductivehealth services for poor women, and particularlypoor adolescent women. Enhance demand among the poor. Deliver the complete essentialpackage of reproductive health services .inan integrated manner. Improve the quality of care Create an enabling environment Unique approachesto addressing these priorities will be required in each country and sub-nationalarea. However, internationalresearch and experiences from the region can shed light on promising approaches. Increase the Supply of Reproductive Health Services for the Poor 3.4 Supply-side improvements encompass the establishment of health centers in underserved areas; improvements in facilities, equipment, drug supplies, etc.; extension, outreach, camp or mobile services; better management and training of workers; community-based mechanisms such as 'lay health workers' and health groups; and several others. These improvements can be brought about through top-down policies, financing and instructions, or through decentralized action planning that concentrates on improving the use of available resources and can help secure additional resources either locally or fi-omhigher authorities. Given the large needs and current limitations in public sector service delivery, it is necessary to mobilize all available resources to enlarge the pie of services available and achieve better health outcomes in poor settings. 3.5 Establishing facilities beyond public hospitals and clinics, such as birthing homes and maternity waiting centers, to improve geographic access to essential obstetric care (EOC) especially for women living in remote areas has worked to reduce maternal mortality. For example, in Malaysia low-risk birth centers with four to six beds are attached to health clinics. They are staffed by doctors, nurses and midwives from the respective clinics. High-risk women are also allowed to deliver at these centers if they do not want to go to a hospital, but are sent to the nearest hospital by ambulance in case of emergencies. Through this strategy both demand and access to good obstetric care have increased and outcomes have improved (Anon, 2003a).' Efforts to improve home-based life-saving skills have been undertaken in India and elsewhere. They improve the ability of pregnant women and family members to recognize danger signs and increase birth preparedness. They have also helped to increase postpartum and post-abortion family planning; develop referral systems with emergency transportation and funds; and establish sustainable networks of community volunteers who manage the entire process and solve problems at the local level (PRIME, 2003). Organizing camps on well- publicized dates and sites has also helped to increase service delivery. 3.6 Other interventions that have success~llyaddressed some key barriers and increased access to quality reproductive health services include those that: encourage communities to take responsibility for their reproductive health needs; make use of private formal and non-formal service providers to increase available options; and increase efficient use of available resources. These also demonstrate that communities and households can take more responsibility in improving their own health when the formal system fails. In a project in Ethiopia that aimed to overcome social and geographic barriers to family planning in rural areas, extension agents lived in villages for six months and worked intensively with the communities to organize them and develop community-based service systems (e.g., to distribute pills and condoms). After three months, the villagers selected and trained local volunteers who then gradually took over the duties of the extension agents including counseling, group education and distribution of contraceptives. This allowed the outside agents to move on to other villages, making efficient use of scarce human resources (Rubardt (2002) cited in Anon, 2003b). In India, the SEARCH Foundation in a tribal area of Maharashtra trains village '. In Sri Lanka, changed social norms, education, health literacy and perceptions that better health services are delivered at centers with medical specialists and sophisticated equipment has led to over-utilization of better equipped hospitals and under-utilization of maternity homes, rural hospitals, peripheral units and district hospitals which have basic facilities and staff for maternal health services. It may be well to redress this imbalance. health workers and traditional birth attendants (TBAs) to diagnose childhood pneumonia and treat it with antibiotics under the supervision of a medical team. This has helped to reduce the case fatality rate and consequently lowered infant and child mortality. The Indian Council of Medical Research is carrying out a multi-center study to assess the feasibility of scaling up this approach through the government health system (Bang et al., 1999; www.searchgadchiro1i.org). In Sri Lanka, for the past ten years trained ANMs have identified children with risk signs for ARI and sent them for medical care. The ARI Control program has already expanded nationwide. 3.7 Considerable scope exists to expand public-private collaboration, including using public funds to purchase health services for the poor fi-omNGOs and other private providers (through a variety of financing mechanisms such as "contracting out" or voucher schemes that will be discussed in Chapter 4). Other strategies to involve the private sector include social marketing and social franchising. At least two excellent examples are available in South Asia itself- the Pakistan Green Star Program and Janani in India (Box 3.1). Assessments point to many positive outcomes of this approach, such as better targeting of services, easier access to new products and procedures, service quality improvements,and utilizationof spare capacity or efficiency improvem.entsin the private sector. The public sector can also collaborate with the private sector in training, communications, organization of emergency transport and so on to improve access to, demand for andlor quality of services. Box 3.1 PromisingApproaches to Increase Access to ReproductiveHealth Services Pakistan: Franchising Family Planning for the Poor. When the social marketing program in Pakistan expanded from condoms to clinical methods in the mid-1990s, it became essentialto expand the pool of skilled family planning providers serving poor communities. The solution was social franchising, recruitingphysicians from Pakistan's highly-developedprivate sector to add family planning to their existing medical practices and market their services under an easily recognized brand name. Population Services International (PSI) and its local affiliate, Social Marketing Pakistan (SMP), designed and launched the Green Star network in 1995. Like commercial franchises, Green Star is governed by a contract between the franchiser, SMP, and the health care provider. Providers agree to deliver services according to protocols and quality standards set by SMP. In return, SMP supplies them with training, support and materials, including Green Star signs to display outside their clinics. By establishing a franchise dues system at the outset, SMP can screen out applicants who are not serious, increase franchisees' commitment to the network, and contribute to cost recovery. The contract clearly spells out the roles and responsibilities of franchisees and outlines enforcement mechanisms to maintain control over the quality of services. SMP also markets the Green Star brand name to attract clients to franchiselocations. The incentive for providers is economic: they hope to attract more clients because of their Green Star affiliation and also to profit on the sale of Green Star contraceptives. Because of cultural restrictions on pelvic examinations, the Green Star network initially recruited licensed female physicians to offer a full range of family planning services including IUD insertions. Later, male physicians were added to the network because of their ability to talk with men about contraceptive use. The next step was to recruit pharmacists who play important roles in disseminating family planning information and influencing client decisions to use family planning. Eventually, Green Star included Lady Health Visitors who deliver health care to the poorest and most underserved urban neighborhoods. Recruitingproviders into the network has become easier since the Green Star brand has achieved recognition and respect. The network can now use more stringentcriteria to select and admit new providers. Experience has shown that the most successful Green Star providers own and operate their own clinics, which makes their employment more stable and follow-up easier. Training is key to a franchise's success as it ensures that providers understand Green Star service delivery protocols and can meet advertised standards of quality. The network has developed different training curricula and reference materials for each type of provider. To ensure that providers meet the quality standards, training staff also have a monitoring function. They make regular site visits, especially to providers who are new to the network, to answer technical questions, assist with procedures and solve problems. In addition, SMP field staff visit Green Star sites quarterly to assess providers' performance, and mystery client surveys are conducted periodically. If a provider's services do not meet Green Star standards, he or she is required to undergo remedial training. Multi-media campaigns, promotions and public relations activities have created awareness of the Green Star brand and continue to generate demand for its services and products. By 1997, 93 percent of respondents in low-income urban areas recognized the Green Star logo and identified it as a symbol of high-quality family planning at affordable prices. In 2003 Green Star provided 30 percent of all couple-year protection in Pakistan, which compares well with the 59 percent provided by the public sector, and a total of 11 percent provided by Key Social Marketing, commercial sales and the NGO sector combined (Ahmed, 2006). India: Janani, a non-profit organization, uses social franchising to provide services to the poor. It is among the largest public-private networks delivering reproductive health care in India. The program covers Bihar and Jharkhand, two of India's poorest states (annual per capita income about $120) and home to 118 million people, and ten districts of Madhya Pradesh, another large and poor state. Janani has developed an extensive distribution network of pharmacies, cigarette shops, grocery stores, general merchants and rural medical practitioners to provide quality family planning services at affordable prices. It has also created a network of general practitioners under the Surya Clinic franchise to provide clinical family planning services. This program covers a number of districts in Bihar and shows that couples are willing to adopt family planning and pay for services if these are easy to access and of good quality. Janani started as a conventional social marketing program in 1996 using shops to sell subsidized products. Subsequently it has refined its strategies to address three major shortcomings of conventional social marketing programs: the inability to address needs beyond urban areas, to deliver clinical services, and to serve the needs of the poorest. This meant that service delivery channels had to extend beyond shops. Janani consequently identified rural practitioners to increase access to products and services for village communities and private doctors to deliver clinical services. By December 2005 the organization had trained over 40,000 providers and established 520 medical clinics, in addition to delivering its products through 31,000 shops. The network will ultimately consist of 40,000 shops, 360 streamlined medical clinics and 57,000 rural centers, one in each village in the areas covered. Janani's clients are predominantly poor, and pay significantly less than commercial health service prices. When Janani began, the program partnered with the Indian government and focused only on the delivery of condoms and oral contraceptives. The integration of clinical services began in 2000. The emphasis has been on enhancing the doctors' surgical skills while using paramedics for routine procedures. The Surya (Sun) Clinics are franchised and quality standards are maintained by a Janani-appointed administrator at each clinic. The network of rural health practitioners, each working in partnership with a woman family member, serves as the link between these clinics and communities. After their training by Janani, the rural practitioners are franchised into Titli (Butterfly) Centers, where they sell non-clinical products and over-the-counterdiagnostic tests. They counsel clients needing clinical services and refer them to the nearest Surya Clinic, thereby earning a commission. This bundling of services has enhanced the viability of franchising. The conventional social marketing infrastructure of shops and stockists serves to sell products in urban and semi-urban areas and to replenish supplies to Titli and Surya centers. In essence, Janani's operational strategy leverages service delivery resources that are already in place. Creating such resources would have been expensive and time-consuming. Janani's administrative strategy is to maintain a small core team with a high level of communication, and to outsource fieldwork in a competitive manner. In the long-term the organization aims to expand service delivery further in Madhya Pradesh and to Chhattisgarh, Uttar Pradesh, Uttaranchal, Rajasthan, Orissa, West Bengal and Assam. In nine years of implementation the program is estimated to have averted 5.52 million births. In 2005 it protected 1.68 million couples, averting almost a million births. Despite the difficulties of taking products and services to interior villages, the cost of protecting a couple during 2005 was Rs. 139 or $3.09. (Internationallythe cost of one couple-year of protection is estimated to be around $20.) About 24 percent of the total cost of the program is met by revenues. Based on material available on the Janani Website(http://www.janani.org). Enhance Demand for Services 3.8 While societal changes that enhance the status of women - including increased education- raise demand for services sustainably, as they have in Sri Lanka, Kerala and Tamil Nadu, a number of shorter-tern approaches have been used to enhance demand among poor women. Information, education, communication and behavior change strategies are used widely and appear to have variable chances of success as they depend heavily not only on content and mode of delivery but on the local social, economic and political situations of poor women. As in the case of adolescents,creative efforts such as the "3W Safe Motherhood Game in Senegal" build on local context, generating interest among clients and providers (Spadacini, 2001). In this example, women's awareness of maternal and child health risks is heightened through association with local cultural images, beliefs and proverbs. Evaluation showed increased knowledge of risk factors among the women who participated, and an increase in the number of women who completed three antenatal visits. The interest created in the communitiesis believed to be spreading the message of safe motherhood even further. 3.9 Also widely practiced are participatory approaches to delivering or managing services which are similarly sensitive to context - unless poor women are targeted specifically, they are often 'crowded out' by the better-off, and unless they perceive the activities in which they are involved as having high returns to their scarce resources of time and energy they are ineffective or unsustainable. Many initiatives have used community-based women's groups to raise awareness of reproductive health needs and rights, mobilize women's participationand enhance use of services. 3.10 Involving men in reproductive health programs has also shown the potential to achieve better outcomes for women and reproductive health (Anon, 2003~).SuccessfuR efforts in India include PRlME I1 and Shramik Bharti s CommunityPartnerships for Safe 7 Motherhood in Uttar Pradesh and Pati Sampark in Gujarat (India). The latter program engaged men in monitoring their wives clinic attendance and consumption of'nutritiona:l supplements during pregnancy. Evaluation found that women whose husbands participated in the program attended clinics six to seven times compared with two or three times among others, and men's awareness of family planning was also increased. Another NGO in Gujarat, Sewa Rural, organizes health workers visits according to 7 men's schedules so that husbands can be involved in discussions and actions for Safe Motherhood. This has led to a 40 percent increase in men seeking out health workers to register their wives early for antenatal care, one-third of men accompanying their wives to hospital, and a significant increase in the number of fathers bringing their infants for immunization. 3.11 In Pakistan Punjab, a project set up demonstration sites for men's reproductive health services, with each site developing its own strategies to promote services. At one, community religious leaders became very involved; at another, rickshaw drivers were trained to answer basic questions about the clinic and men's services. Vasectomies increased four fold as a result of these innovations (EngenderHealth, 2003). Such location-specific action makes programs work on the ground and calls for health managers to be flexible in allowing and financing different approaches rather than prescribingthe "how" of implementationfrom above and afar.2 3.12 In addition to improving awareness, protecting the poor from the adverse financial consequences of illness can increase demand for services. Karnataka's Yeshasvini Scheme is among programs that have devised mechanisms to assist the poor to meet the costs of health care (Box 3.2). Box 3.2 Insuring Poor Health: A Community-basedScheme in India The Yeshasvini scheme is a successful public-private partnership that was designed for the rural poor. It provides health insurance to about 1.87 million farm households, members of some 30,000 cooperative societies in Karnataka. Using the cooperative network has enabled the scheme to reach out and build a large membership in just three years, and the aim is to insure all members of cooperative societies. Recently, coverage has been extended to women's self-help groups that are linked to cooperative banks, and poor urban slum-dwellers who wish to register. The insurer is the Yeshasvini Trust, a registered charitable society. Members originally paid a premium of Rs. 60 per year (now raised to Rs. 120), and the Government of Karnataka provided contributions of about 50 percent of the net income of the program in the first year and 25 percent in the second and third years. Members can have their premiums deducted from cooperative earnings and pay in installments - even of milk if they belong to a dairy cooperative! Membership, authorizations and claims are managed by a third party administrator, a profit-oriented company. More than 150 private hospitals all over the state have enrolled as service providers on a voluntary basis. The good reputation of these hospitals has contributed to the scheme's attractiveness. The benefit package focuses on "high cost-low frequency events" such as surgery and certain accidents and injuries, and includes free outpatient consultations, subsidized medicines and diagnostics for any illness. About 1,600 types of surgery are covered including Cesarean sections. There are reports that Cesarean rates are higher than expected, which may be due to provider moral hazard or self-selection. Future plans include extending benefits to cover normal deliveries and neonatal care. Fixing the costs of the most common surgeries has minimized the risks of cost inflation and escalation. Patients do not need to handle payments as the insurer pays the health care provider directly for pre- approved surgeries. The scheme helps poor clients get surgeries they could not otherwise afford, and the free services provide additional protection. In its first year over 30,000 OPD treatments were covered. Evaluations show that communication with members needs to be improved and authorizations and claims need to be speeded up. Another important issue for the poor is the distance that some need to travel to a participating hospital - some quality may need to be sacrificed for improved access (Radermacher et al., 2005). '. Reproductive Health Outlook (httu://www.rho.org) has many other examples of successful or promising practices related to enhancing demand for reproductive health care among adolescents or families. 40 Deliver an Essential Package of Services in an Integrated Manner 3.13 Self-evidently, the effectiveness of each type of reproductive health service depends on the presence of a comprehensive set of other services, from family planning through post-natal care and beyond. Shown below are the main areas requiring attention in the countries studied. 3.14 Contraception. Expanding the menu of contraceptives (especially to include spacing methods where these are currently limited, as in India), providing informed choice, strengthening follow-up, managing side-effects, and actively promoting male methods and involvement are needed in all countries. A broad method mix would increase overall use and continuation, and reduce failure rates, abortion and coercive sterilization. The difficulties of providing services to adolescents and secluded women and the burden of RTIs/STIs/HIV/AIDS call for condoms and "dual protection" to be promoted actively. Given the aggressive nature of family planning programs in the past, it is particularly challenging to provide contraception in a 'user-friendly' manner and address women's specific needs. Sound programs require strong technical and counseling skills and social and gender sensitivity among providers. Enswring home visits to advise husbands and wives together to discuss their family planning objectives and preferences would be an effective strategy to address unmet need and increase demand. Family planning education through mass media, based on accurate social and cultural understanding (especially of the problems faced by poor young women), could create an enabling environment for services. Public education is also important to dispel misconceptions(such as those that surround injectablesin India). 3.15 Abortion. The large number of abortions, complications and deaths reflect a serious failure of reproductive health policies in the region. Ready access to safe and confidential abortion is a critical need everywhere. Sri Lanka and Pakistan need to liberalize their abortion laws, and all five countries need to enforce safety lavvs, regulate clinics and penalize providers of unsafe abortions. Qualified providers can be encouraged to use the safest methods such as manual or electrical vacuum aspiration. Mifepristone (recently legalized in India) is a safe and effective option if providers are adequately informed. Importantly in South Asia, where women's decisions about abortion are often taken jointly with their husbands, mothers-in-law or other family members, IEC and counseling programs should include these people to erlsure 'safe' abortion decisions, post-abortion care PAC) and contraception. PAC is essential to reduce mortality due to abortion, and includes emergency treatment of complications, family planning counseling and services to prevent repeat abortions, and links to other reproductive health services, such as diagnosis and treatment of RTIs. The experience in Bangladesh of providing menstrual regulation through trained providers such as nurses and midwives could be extended through appropriatetraining to providing PAC. 3.16 Skilled Birth Attendance. Given the positive association between antenatal care and use of skilled birth attendance, expanding ANC could increase women's chances of safe delivery and survival. This is particularly important given the socio-cultural and physical barriers to institutional delivery in South Asia. Demand-side financing mechanisms, such vouchers for pregnant women to use safe delivery facilities in the public or private sectors, andlor reimbursing transportation and other costs, could reduce economic constraints. Although traditional birth attendants (TBAs) are not classified as slulled, they could be given incentives to refer pregnant women to qualified birth attendants or facilities. Safe motherhood demonstration projects in Ghana and elsewhere have shown that a brief course in obstetric life-saving skills given to staff at primary and secondary health facilities can be effective in detecting life-threatening obstetric conditions (Osei et al., 2005). Improving maternal health through ANC and safe delivery will also enhance neonatal and infant survival, as Sri Lanka's experience clearly shows. 3.17 Postnatal Care. Strengthening outreach to ensure that every mother who has delivered at home receives a check-up and care within 24 hours of delivery could help to reduce maternal and neonatal deaths. Financial support (vouchers or refunds) to use medical facilities in case of complications would be helphl at this time. As many home- based births are conducted by TBAs, they must be trained to refer women with postpartum complications. Improve the Qualilyof Care 3.18 Although this is a difficult task in resource-poor environments, some programs in developing countries have been able to demonstrate marked quality improvements with simple enhancements. For example, in Tanzania a package of improvements led to a several-fold increase in reproductive tract infection treatment and family planning, greater patient satisfaction, and increased community participation within two years (Atherton et al., 1999). The package included capacity-building (continuing education, training in service quality concepts, strengthened management systems, and enhanced community links), improving infrastructure, regular provision of supplies and drugs, quality monitoring (through regular management meetings and quality assurance techniques), service integration, community participation, and additional service provision through the private sector. 3.19 Several countries such as Malaysia and South Africa have developed total quality management programs, instituted accreditation to establish national standards, or carry out external assessments to ensure quality. At the Castle Street Women's Hospital in Sri Lanka "work teams" use participatory decision-making processes to identify and analyze problems and devise strategies to overcome deficiencies in the system. A monitoring system was set up and monthly meetings of sectional heads are conducted to review progress and sort out cross-cutting issues. The obstetricians and gynecologists in the system play an advisory role to these groups (Withanachchi et al., 2004). In Egypt, clinical standards and protocols, training courses, a three-tiered supervision system, and a clinic certification plan were introduced to "push" quality into every level of the service delivery system. At the same time, public media campaigns were designed to "pull" quality into health facilities by raising community expectations of the quality of family planning services and prompting clients to demand good quality care (Anon, 2003b). Maternal death audits have also been effective in improving the quality of maternal care in Sri Lanka. It is mandatory for all relevant functionaries to promptly notifLa maternal death, and this is investigated. In addition, District Maternal Death Reviews are carried out every three months to identify managerial and technical problems that may have led to a maternal death, and measures are taken to correct them. The findings are also used to review relevant policies and devise strategies to prevent such deaths in the future (UNICEF, 2004). 3.20 Most critical to enhance quality in reproductive health care in the public health sectors of the five countries are human resource development and improvementactivities. There are many examples that indicate that initiatives in skilled-based training and incentives have been successful in improving the availability and competence of service providers. The importance of having well-skilled frontline workers is demonstrated by the following experience. In Andhra Pradesh (India) an experiment set up private midwifery practice by trained auxiliary nurse midwives (ANMs) in poor remote villages. Women's associations were involved in appointing the ANMs and deciding on the payment for their services. One village appointed a local married woman who was trained to provide basic health care, while the other village appointed a qualified ANM. The qualified ANM was more effective in providing services as she could treat women's illnesses and conduct deliveries. Within a year, she was earning more than she would have earned working at a private nursing home. The basic worker, on the other hand, was not effective as she did not conduct deliveries and could not treat minor ailments. 3.21 Elsewhere, twinning arrangements between providers and training agencies have been effective in increasing women's confidence in and utilization of services. Integration of services has also strengthened service delivery through efficient utilization of available human and other resources, including in the Pakistan Lady Health Worker Program in which this single cadre of community-based workers provides home-based. reproductivehealth services. The program has had a significant impact on the delivery of services and helped to achieve PRSP goals for immunization and contraception. For doctors, the government of Andhra Pradesh established a Tribal Medical Service which provides incentives such as priority in admission to post-graduate medical courses to those who work in tribal areas for three years after receiving their basic medical degree (Andhra Pradesh Vaidya Vidhana Parishad, 1998). This approach contrasts with that of Karnataka and Orissa, where all medical graduates are compulsorily posted to rural areas for 2-3 years. In some states, to fill vacancies and reduce absenteeism, doctors are consulted on their preferencefor posting. 3.22 Finally, using a combination of strategies could produce lasting results. For example, in Bangladesh, a "multi-angle approach" has been adopted to improve the quality of family planning services (Landovitz(1997) cited in Anon, 2003b). Under this project, health officials are oriented in local planning using COPE (Client-Oriented Provider-Efficient services) exercises. They identify factors hindering service quality and develop and implement action plans, which are reviewed monthly. At the same time, service providers are trained in technical, counseling and supervisory skills to make them more aware of and responsiveto clients' needs and rights. Create an Enabling Environment 3.23 The concepts of the ICPD Programme of Action (POA) have been incorporated into the policies of all five countries. However, despite this, several key aspects of reproductive health such as a "client-centered women-friendly approach," delivering an integrated essential package of services, a focus on adolescents and sexuality, and reproductive rights have hardly been implemented. There is little evidence also of broader strategies needed to achieve reproductive health goals, such as pro-poor actions, rights-based approaches, preventive health activities, and reductions in the financial burdens on the poor. Specific activities to implement these approaches are needed to provide an enabling environment for reproductive health care. 3.24 The reproductive health policies in place are most likely to succeed if they are supported by a wide range of stakeholders - from policymakers to service providers, women's advocacy groups to grassroots organizers and client representatives. Bolivia provides an example of robust political and social commitment to improving the quality of maternal and child health care. The Ministry of Health has been implementing the Making Pregnancy Safer Initiative with WHO support, focusing on increasing comprehensive obstetric care and capacity at the first referral level. This has been supported by the adoption of national standards of care, revitalization of the epidemiological surveillance system, and development of a uniform death certificate especially for maternal deaths. The initiative plans to continue to focus on improving the quality of care and to ensure that services are culturally sensitive (Anon, 2003a). This set of reinforcing actions was possible due to strong political support for reforms in Bolivia. 3.25 Many developing countries including Sri Lanka have established a strong enabling environment through health sector reforms, women's education and employment, political and human rights, and successfully improved reproductive health (Box 3.3). In India, Tamil Nadu has made rapid progress in strengthening its health system, improving service utilization and reducingmortality and fertility (Box 3.4). Box 3.3 Sri Lanka's Historical Effortsto Improve Maternal Health Sri Lanka started maternal and child health services through doctors, nurses and midwives at the turn of the 2othcentury when birth and death registration were also made compulsory. In 1927 a Medical Ordinance required all midwives to be registered, the government focused on reorganizing midwifery services and introduced a scheme to train midwives in order to modernize birthing practices in rural areas. Although suitable candidates were initially in short supply (they were required to be 'respectable', which conflicted with the low status of midwives), the scheme gradually succeeded, and the number of midwives trained and employed by government increased almost three-fold between 1931 and 1938. Infant and maternal care services were developed alongside. In 1920 a scheme was formulated for the provision of financial and other assistance to expectant mothers from poorer groups. It focused on two stages of children's lives - antenatal and infancy. The state took responsibility for the health of mothers by ensuring that every woman had the services of a qualified midwife during childbirth. Maternity centers and antenatal clinics staffed with qualified medical practitioners and midwives were set up. After the 1960s,the expansion of family planning services also helped to reduce births and deaths. Sri Lanka is among the few developing countries that have successfully reduced maternal mortality to levels comparable to those of developed countries - its maternal mortality ratio was 23 per 100,000 live births in 2000. In the 1930s, the maternal mortality ratio (IVIMR) was estimated to be over 2,000 per 100,000 live births; by the 1950s it had declined to below 500. Successful efforts to combat malaria and the introduction of modern medical practices for infant and maternal care, includingmidwifery services, deserve much of the credit for this achievement. A functioning referral system for obstetric care has been largely responsible for recent reductions, and strong civil registration and maternal death audit systems have provided the information necessary to assess and accelerate progress in maternal health. In addition, public access to education expanded rapidly during the first half of the 2othcentury, resulting in a remarkable increase in female literacy from 8.5 percent in 1901 to 44 percent in 1946 and 71 percent in 1971. Increased gender equality and effective public investments in improving living standards also contributed. To improve service delivery, the government introduced Health Units in 1926 focused on expanding free services in rural areas, an important step in the improvement of health in the country. Attention was given to preventive and promotive activities at the community level, and especially to control of the major communicable diseases. Each unit had a Medical Officer and a team of field workers who were responsible for serving the population in the area and provided services at the domiciliary level and at health centers. Public health midwives provided care to pregnant women within populations of 3,000 to 4,000. These units remain the main providers of reproductive health services today. Until 1940, skilled attendants assisted only about 30 percent of births. By 1950, after the implementation of policies to introduce and expand the cadre of public health midwives, this percentage doubled and, currently, skilled practitioners attend 97 percent of births, the majority within institutions. While field services were being expanded, institutionalfacilities - from General Hospitals at the higher levels to Rural Hospitals at the lowest - were also improved. Municipalities set up Maternity Homes where trained midwives attended to deliveries. The higher level facilities had specialist services and served as referral centers. Geographical access to these was facilitated by good road networks, and economic access was enabledby free services for all. A well-functioning referral system for pregnancy-related and other health problems was also established early. The number of health facilities was expanded rapidly from 112 government hospitals in 1930 (about 182 beds per 100,000 people) to 247 hospitals in 1948 (close to 250 beds per 100,000). Between 1948 and 1950 the national ambulance fleet was increased from 12 to 67 ambulances. These improvements continued through the second half of the century. A program to reach women on tea estates is another example of a targeted effort to ensure good quality services to all. In Sri Lanka now, maternitycare including surgery and blood transfusionis availablefree to any woman in the country. A strong civil registration system in Sri Lanka provided valuable information to plan and monitor progress in maternal health. The government used death inquiries to identify problems in the delivery of care. Since 1970, all maternal deaths have been required to be notified to the local Medical Officer (MCH) within 72 hours. A Medical Officer of Health then visits the hospital and home and files a detailed report. A full inquiry and a meeting at the institution where the death occurred are completed with two weeks. There are also regional reviews to discuss all deaths and plan remedial measures, and finally a national review at the level of the Director of Maternal Health Services and representativesof the College of Obstetricians. In 1987 management of the state health care system was devolved to the Provincial Councils. The Central Ministry of Health (MOH) is responsible for establishing policy, training health personnel (except Medical Officers who are the responsibility of the Department of Higher Education), managing teaching and specialized medical institutions and bulk purchasing of medical equipment and supplies. Responsibilityfor policy, planning and monitoring reproductive health lies with the Family Health Bureau (FHB), a unit of the MOH. The eight ProvincialDirectors of Health Services (PDHS) are responsible for implementation of RH programs. Below this level there are 23 Deputy DHS, each responsible for a district, and 265 Divisional DHS (which are the current equivalent of the earlier Health Units). These directorates are responsiblefor provision of comprehensive health services (with emphasis on preventive and promotive care) to the population of their defined administrative unit. At the district level there is a Medical Officer for MCH (MO-MCH), some of whom are specialists. The divisional units are staffed by a Medical Officer of Health and several field health persons, including Public Health Nursing Sisters (PHNS), Public Health Inspectors (PHI) and Public Health Midwives (PHM) who focus mainly on preventive and promotive health. They provide antenatal, natal, postnatal and neonatal care, immunization and other health activities aimed at infants and preschool children, family planning and Well Women Clinic services. It is interesting to note that in 2000 there were more Medical Officers (includingwomen MOs)per 100,000 people in this system than PHMs- 41 and 38, respectively. The PHM is the main grassroots worker, still responsible for providing family health services to a population of 3000-4000 through field visits and a field-level clinic. Above her are Supervising PHMs and a PHNS. The PHM maintains an eligible women register and is expected to register all pregnant women and advise them to attend antenatal clinics. She carries out physical exams of mothers and educates them about pregnancy care, especially nutrition. Although her role is to advise institutional delivery, she assists the small percentage of deliveries which occur at home and is responsible for the care of the newborn. She is expected to visit postnatal mothers who have been discharged from institutions three times during the first ten days after delivery - but there are often delays in these visits. Each newborn is given a Child Health Development Record. The PHM also counsels women about family planning and refers them to an appropriate service center. They sell condoms and oral pills at subsidized rates. The PHMs also register all women between 35 and 60 years of age and motivates them to attend the Well Women Clinics which screen women for diabetes, hypertension, cervical and breast cancer. Curative services are provided through Rural Hospitals and Maternity Homes at the primary level to Teaching Hospitals, Provincial Hospitals, Base Hospitals and the National Hospital of Sri Lanka at the tertiary level. The primary facilities carry out ANC clinics and uncomplicated deliveries, and secondary care is provided at District Hospitals, Peripheral Units, and specialist Maternity Hospitals. However, the lack of a well-functioning referral system results in some 'irrational' use, and over-crowding of the higher level facilities. Only a few per cent of births occur in homes and at the primary level - the majority of about 94 percent are distributed almost equally among secondary and tertiary (specialist)hospitals. The decentralization of health management in Sri Lanka presents a mixed picture. Unfortunately, the decentralizationof responsibility has not been accompanied by a commensurate provision of financial resources, and even human resources under the control of the provinces and lower levels are inadequate, especially at the managerial level for planning, management and monitoring. The central MOH continues to manage the posting and career development of doctors, and practices are somewhat outdated. Informationsystems also require strengthening. Sri Lanka has achieved a dramatic fall in maternal mortality since about 1948. The first decline from 2000 maternal deaths per 100,000 live births in 1948 to about 500 in the early 1960s is attributed to improvements in public health, particularly malaria control, and to the expansion of MCH services. Subsequently, the wider availability of antibiotics and blood transfusion services, and fertility control are believed to be responsible for the continued decline. Recent activities such as a two-way referral system for pregnant women, maternal death audits, advanced training for providers, and involvement of professional ob-gyn body in standards of care and medical education have further improved maternal survival. Since about 1992, Sri Lanka's maternal mortality rate has been about 25 per 100,000 live births. However, there are large district variations, with the conflict affected and plantation districts having MMRs of over 50 per 100,000 live births. The main causes of death that declined during the 1960-1996 period were sepsis and pregnancy-induced hypertension, possibly as a consequence of improved antenatal care and skilled attendance at delivery. Though ergometrine administration and blood transfusion services are more readily available, hemorrhage continues to be the major cause of maternal death, followed by septic abortion. Sri Lanka's health and demographic successes are widely attributed to the combination of health and education policies and programs adopted by the country, and to the status of women in society. Education was free and available to women. The education and empowerment of women have had positive effects on women's use of health services and (along with food subsidies and supplementation programs) on their nutrition. Family planning has also made an important contribution to women's health. In the 1950s, the Family Planning Association of Sri Lanka spearheaded efforts to make barrier methods available through clinics and to train doctors in family planning. Family welfare activities were carried out along with antenatal and postnatal care, contraceptives were distributed free, and mass media were used for education purposes. In addition to birth spacing and family limitation services, sub-fertility services were also provided. From the 1960s on, family planning services were integrated with MCH programs and widened to include oral contraceptives, condoms, and IUCDs. Male and female sterilization were introduced in the 1970s. Family planning services provided by the government are complemented by the efforts of four major NGOs, and some provision by the private sector. A social marketing program- provides oral contraceptives and condoms through 8000 outlets in the country. Counseling and choice are hallmarks of the program. Currently, injectables are the most popular modern method, followed by oral pills, IUCDs,condoms and sterilization. Incentives are paid to medical teams and clients for sterilization. Education has also contributed to the increase in contraceptive prevalence and in a rise in the age at marriage both of which have helped Sri Lanka to reduce fertility to1 replacement level by 2001, and contributed to the declines in maternal and infant mortality. Sri Lanka achieved its health improvements with relatively low spending. Analysis of public health expenditures shows that total government spending on health care was a modest 1.8 percent of GDP, with maternal health accounting for 0.23 percent. This has been decreasing - during the 1990s, government expenditure fell to about 1.5 percent of GDP, and after 1999 it fell further to about 1.2 percent. Sri Lanka spent about 4-5 percent of total government expenditure on health during the period from 1996-2001. About 50 percent of total expenditure on health in the country is public, and an equal amount is private expenditure so that total health expenditure (including out-of-pocket and insurance payments) has been about 3.0-3.5 percent of GDP since 1990. Some of the private out-of-pocket expenses are incurred on use of public facilities (e.g., for transport, prescribed drugs, and patient attendants). The remainder is largely fees for out-patient services provided by the private sector -private practitioners contribute about 50 percent of 'first- contact' curative care, including RH services. Government doctors are allowed to practice privately outside working hours. Private hospitals with beds are largely in urban areas - about 70 percent in Colombo alone. Only 6 percent of deliveries take place in private hospitals. The private sector provides some sophisticated services (such as in vitro fertilization) which are not available in the public sector. Box 3.4 Health Sector Reforms in Tamil Nadu, India, and Bangladesh India: Tamil Nadu has made rapid progress in different aspect of health care - policies, facilities and logistics, human resources, information, legislation, quality assurance, financing and outcomes. The government has introduced several innovative practiceslreforms to strengthen the state's health infrastructure and human resources, improve quality of care, and expand public-private partnerships. For example, nurses are hired on contracts to provide round-the- clock delivery care, and private anesthetists and obstetricians are engaged to bridge the gap in the public sector. Improved accommodation for health staff and continuous monitoring have helped to reduce absenteeism, a major hurdle in service delivery. Transfers and postings are rationalized to ensure that specialists are available within the public system. Selected Primary Health Centers (PHCs) are provided additional facilities and personnel to run on a 24-hour basis. Members of Parliament and the State Legislative Assembly are encouraged to utilize their development funds to improve PHC facilities in their constituencies. In order to increase the mobility of village health nurses posted at sub-centers, loans are provided to help them purchase two-wheelers. A visitor's fee of Rs. 5 has been introduced for those visitingGovernment Medical College Hospitals and District Hospital during non-visiting hours, and the funds are utilized to maintain the facilities. Hospitals at the taluka level and below are exempt from this. A pioneering effort has been made to streamline drug supplies to all health facilities - the state has had an efficient centralized procurement and distribution system in place since 1994. The Tamil Nadu Medical Services Corporation (TNMSC) was set up to procure, store, distribute and control the quality of drugs. Through its centralized procurement of drugs, the Corporation has been able to achieve price advantages and optimize public resources. The TNMSC maintains 23 drug warehouses, each responsible for meeting the drug requirements of all the health facilities in its service area. The facilities indent for drugs on a predetermined schedule, and the entire operation is computerized and monitored on a daily basis. To strengthen the Health Management Information System (HMIS) and its utilization, two types outreach and program-relatedservices. These are operatedin integrated manner. Software has been developed to capture data, analyze and prepare reports from the block level upward. Recording and reporting tools have been simplified - importantly, the number of registers maintained by sub-health centers has been reduced from 21 to 7! Recently, the GOTN has decided to handover ambulances to NGOs or communities to improve emergency transport. Steps have been taken to train doctors in the rationale use of drugs, and comprehensive norms have been adopted to standardize the distribution of human resources and services in all health institutions. In its ongoing health sector reforms, the GOTN is addressing issues related to resource allocations, integration of primary and secondary care, equitable access to services, referral, public-privatepartnerships,and the utilization of available information. Bangladesh. The Government is making an Essential Services Package (ESP) including reproductive health, child health, communicable disease control, limited curative care, and behavior change communication available at the community level. The key monitoring indicators for this effort include reproductive health indices such as maternal mortality, infant mortality, under-five mortality, age of women at first birth, unwanted fertility, and total fertility. An important aim is to increase the proportion of deliveries conducted by trained personnel through a network of Emergency Obstetric Care facilities. The availability of goods and equipment has been improved, and timely availability of health and family planning materials has increased substantially. In order to achieve cost-effective and client-focused delivery of the ESP at the Upazilla level and below, the program is delivered through health and family planning workers under a single manager. At the community level, the services are provided from a fixed Community Clinic. The construction of community clinics involves the communities themselves who own the structures and take some responsibility for clinic operation. These strategies have helped to increase access, but there remains a need to cover the poorest and most vulnerable groups through targeting and initiatives to reduce out-of-pocket expenditures. The on-going health sector program includes activities in these directions. 3.26 Collaborations with and efforts within other sectors are important to achieving reproductive health goals. An example of a strategic collaborationwith another sector (in this case Women's Development or Education) is provided by efforts to increase health workers' awareness of gender differentials and violence against women that in turn helps to increase their sensitivityto female clients. In South Africa, a four-daytraining module on gender violence was introduced into the curriculum of primary care nurses. It started by exploring the nurses' own attitudes, beliefs and personal histories of violence. Popular sayings and wedding songs were deconstructed to help them understand gender stereotypes and conditioning. Finally, it focused on their responsibilities as health professionals. They brainstormed about their role in addressing domestic violence, and the practical skills involved in doing so. The program is expected to counter the prevailing gender violencein rural areas of the country(Kim and Motsei, 2002). 3.27 More broad.ly, programs to improve girls' education, empowerment and nutritional status can have profound impacts on their health. Box 3.5 discusses Bangladesh's fertility decline and the importance of health services and other factors, such as the program to enhance girls' secondary schooling. There are many examples in South Asia of integrated health, nutrition and 'women and child development programs' (read: pre-school education and women's awareness and income-generation), such as India's Integrated Child Development Services (ICDS) Program, and the Bangladesh National Nutrition Program (BNNP). Examples of other types of multi-sectoral efforts are community nutrition programs in several countries that improve women's reproductive health directly or indirectly. A program in Senegal is similar to the ICDS and BNNP, and like the latter is also a partnership program involving private administrationand public financing, but it includes water supply provision (Ndure et al., 1999). A program in Swaziland focuses on vegetable gardening, helping to improve the nutrition and health status of HIV-affected households by increasing their vitamin consumption,food security and incomes (Shumba, 2003). Box 3.5 Bangladesh's Fertility Decline and the Role of Education Fertility Decline. Between 1971-75 and 1994-96, Bangladesh almost halved its total fertility rate from 6.3 to 3.3 births per woman. Contrary to experience in most other parts of the world, this occurred at fairly low per capita income levels (US$270 in 1997), high infant mortality (82 per 1000 live births in 1996-97) and low life expectancy (58 years in 1997). While family planning programs were given a large part of the credit for this change because it was made possible by increased contraceptive use (50 percent of all couples in 1994-96), some analysts have argued that socio-economic developments explain a large part of the decline (Cleland et al., 1994; Caldwell et al., 1999). During the same period urbanization was rapid and schooling of both boys and girls increased substantially. Girls' primary school enrolment rose from below 50 percent in the mid-'80s to 100 percent in the early '90s. Subsequently, secondary school enrolment also increased due to motivation and incentives provided by the government (see below). Changes also occurred in land and agricultural patterns, and farm and off-farm employment of men, women and children, freeing the latter for school and providing additional sources of income for poor households, particularly the landless. Rural development (e.g., electrification and water supply) and availability of doctors and nurses also improved. The adoption of family planning at the household level was found to be related to declining expectations of "support from the land" and increasedeconomic choices and desire to invest in health and education (which c;ould not be met if a family was large and thereby costly). While the climate within which family planning acceptance takes place is surely critical, health service factors also play important roles. Between 1978 and 1997, family planning services and counseling were provided largely through Family Welfare Assistants throughout Bangladesh. This "doorstep delivery" increased FP awareness, use and continuity considerably. It also reduced dependence on sterilization. Although overall use of oral contraceptives increased, use of the pill declined in areas where injectables were made available. Where contraceptives are readily available, the incidence of abortion is also believed to be lower than in poorly served areas. Some shortcomings noted in the program, however, included the lack of other reproductive health care (health and family planning were under separate departments in the government ministry), and training, supervision and management affected the quality of the program. Some Family Welfare Assistants (FWAs)tended to neglect the poor and focused on better-off women, and men were also neglected by this approach. Further, some observers argued that doorstep services reinforced women's seclusion, whereas traveling to clinics could increase women's mobility and autonomy. For these and other reasons, and due to cutbacks in donor funding for the large force of FWAs, the Government of Bangladesh (GOB) stopped doorstep services in 1997. The FWAs were placed in "satellite clinics"at the sub-district level to provide basic health and family planning services. By 2003, however, doorstep delivery was reinstated. Between 1993 and 2004, the total fertility rate (TFR) in Bangladesh declined very slowly from about 3.4 to 3.0. Some suggest that this is directly related to deteriorationin the effectiveness of the family planning program (Islam et al., 2002). In effect, not only was the advocacy and availability of contraceptives 'proximal' to Bangladesh's early and rapid fertility decline, but later changes in deployment and management of the main contraceptiveproviders, the Family Welfare Assistants, may have resulted in a slowing down of the program. Although a slowing in the TFR occurred in 1994 before the shift in 1998 from door-to-door contraceptive distribution and motivation to a system based on fixed facilities, satellite clinics and supply depots may have led to discontinuationand reduced contraceptiveprevalence and quality of care. While a reversal of this situation was expected if women were sure that they would get services of acceptablequality if they approached a clinic, and that they could simultaneously access other health services, this did not occur. Finally, the full impact of girls' secondary schooling on the fertility behavior of the schooled cohort has yet to be felt. The Role of Female Secondary School Assistance. The educational attainments of women in Bangladesh are among the lowest in the world. In 1991 only 20 percent could read and write. During the early 1990s, gender disparity was most significant in secondary education only one- - third of students enrolled in secondary schools were girls, and the number of girls completing secondary school was less than half the number of boys. Unlike primary school, secondary education requires payment of tuition fees in Bangladesh. Other costs such as transportation, books, uniforms, school supplies and examination fees have also to be met by families. In a culture where daughters are considered economic liabilities, many parents are not willing to make these investments in their education. In its Fourth Five-year Plan, the Government of Bangladesh set a goal to raise the female literacy rate from 16 to 25 percent. The Plan included a new initiative to provide stipends to girls attending secondary schools, aiming to increase their numbers and to encourage them to go on to participate in other socio-economic development efforts. In 1994, the World Bank began supporting the government's stipend program under the Female Secondary School Assistance Project in 118 thanas of the country where income, female literacy and girls' school attendance were low. The Bank contributed US68 million in IDA funds to the total project cost of US$88.4 million. The program was subsequently expanded nationally to 460 rural thanas with assistance from the Asian Development Bank and the Government of Norway. Under the program, stipends are available to girls as they progress from Grades 6 to 10. They cover tuition, examination fees, and an increasing proportion of other fees, textbooks, school supplies, uniforms, shoes, transport and kerosene for lamps, offsetting the increasing educational costs. This responds to the need for a higher incentive to reduce dropout in upper grades. The project is reaching out to the poorest families and provides them the incentive to keep daughters in school. Schools are also given bonuses to attract girl students. To reduce the supply constraint (i.e., the limited number of school places) the government also funds the program in private schools. The project is also increasing the number of secondary school teachers, particularly women; providing vocational skills to girls who are about to graduate; making healthier and safer for girls and more attractive; and strengthening government institutions for secondary education. As a result of the program, enrolments have increased markedly and there are now more girls in schools than boys. In 2004, almost 49 percent of 15-19 year-olds had achieved secondary schooling compared to 32 percent in the 20-24 year cohort and 20 percent in the 25-34 age group. Although the program has yet to achieve quality goals and to address the problems in urban areas, it is starting to have an impact on malnutrition and fertility. While the total fertility rate is over 3.5 among women with no education in all income quintiles and between 2.5 and 3.2 among those with primary education, it is between 1.5 and 2.3 among those with secondary schooling. The two richest quintiles having a higher TFR than the two poorest, suggesting that those families that can afford children have them while poor households limit their numbers strictly when they are educated. This is because of the greater opportunity cost to mothers who are educated. Use of contraception is seen to increase about 10 percentage points with primary and 18 points with secondary education, compared with uneducated women. Stunting among children declines by 17 percentage points with secondary education of mothers. Based on the changes seen thus far, it is estimated that by 2015 about 60 percent of girls will have attained secondary schooling, and 30 percent primary. This would in turn reduce the TFR to 2.4 from the current 2.9, and stunting from 47 percent to 33 percent, close to the MDG targets for 2015. International experience shows that raising female literacy and school enrolment helps to reduce infant mortality and fertility, and increases income-earning potential and economic productivity. Improvements in education increase the opportunity costs of child-bearing and change the costlbenefit ratio, leading to lower desired family size. The path between increased female education and improved reproductive and child health requires the availability of good reproductive health services in order that the changing aspirations in family size and survival and "quality" of children can be implemented. Beyond program and wealth effects, secondary education interacts with health care provision to make it more efficient and effective. ~ d p ~ ~ i a;Worldl y , BZ(2002)and el Zayed et a/., 2005. I 3.28 In the context of South Asia's tenacious patriarchy, addressing gender issues is essential to improve both supply and demand for reproductive health care. Unfortunately, gender equality is still poorly understood and practiced by the health sector in most areas of the Subcontinent, but some bias could be reduced through better design of facilities and programs (e.g., by ensuring privacy, involving men in maternal care, nutrition, RTI/STI treatment, etc.), availability of services (e.g., male contraceptive methods), and provider status and behavior (e.g., proper remuneration and support of women workers, sensitive treatment of women clients). Indeed, male involvement in reproductivehealth is a promising area of action to promote demand for and ~ti~lizationof services by women. FinancingOptions to SupportBetter Outcomes 3.29 Improving reproductive health in South Asia requires both the mobilization of more resources and better spending through more efficient organization and resource allocation, including payrnent/purchasing systems. Whether financing interventions are optimal and effective depends on how they address demand- and supply-side constraints, and improve the use and quality of reproductive health services. Different financing options address demand and supply issues differently as discussed below. All countries will need to use a mix of financing options. The optimal choice for each country (or sub- national unit, where these are at different reproductive health service and achievement levels) would be based on its particular situation in terms of public resource availability, private sector capacity and willingness, and client demand and ability to 'negotiate' the health system. Options for Resource Mobilization 3.30 General revenue financing has two attractive features from an equity perspective.First, it has proven to be a progressivesource of health care financingin both developed and developing countries (Van Doorslaer et al., 1993; Institute of Policy Studies, 2004). Second, when combined with universalcoverage and low or no user fees, it provides high levels of financial protection against catastrophic ill-health. In South Asia, Sri Lanka has shown the way to achieving these goals. The scale of the other countries' reproductive health problems, the need to keep prices low while substantially increasing the supply of services, and the challenges of increasing demand for and utilization of services among the poor call for greatlyincreased general revenue financing of reproductive health care in South Asia. Additional public and private efforts are needed to cover the poor, improve quality and performance, and provide adequatechoice to clients. Among promising approaches in the region are on-going efforts to provide matching grants to autonomous bodies managing (and partly self-financing) public hospitals in India, grants to local governments in India and Pakistan, and contracting private (usually non-profit) organizations to manage public facilities, provide childbirth or abortion services, or distribute contraceptivesin several countries. 3.31 Unfortunately,the provision of public resources for health care in South Asia has been subject to political pressures. In most countries, health and reproductivehealth have been accorded low priority compared to sectors such as defense, infrastructure, power, water, education and so on. However, direct spending on health is needed to reduce maternal and child mortality and fertility, all of which have implications for resource needs in other sectors. The determinants of reproductive health point especially to the importance of girls' education, women's social and economic empowerment, and nutrition to achieving good outcomes. With the exception of Sri Lanka and Kerala in India, South Asian countries have underinvested in these historically. While efforts are being made more broadly now in the areas of girls' education and women's empowerment, there are still "miles to go" for the majority of South Asian women. Reaching the MDGs will require investments in health as well as sectors such as food, water supply, sanitation, education, and transport which could improve the efficiency of health spending and further improve health outcomes. 3.32 Most official development assistance to the health sector in South Asia either goes through government health budgets or, if off-budget, supports identifiable public efforts. Bilateral and multilateral institutions need to increase their support, ensuring that finds are additional and absorbable, and that they foster much-needed integration and rationalizationof reproductive health services. In addition to maternal and neonatal care, other reproductive health services such as RTIfSTI treatment are under-emphasized by donors relative to family planning and, increasingly in South Asia, HIVIAIDS. Private financiers (once confined to supporting non-governmental efforts albeit significantly in, say, Bangladesh) are increasingly pooling their funds to support public efforts (e.g., national HIVIAIDS control programs). 3.33 Improving allocative and technical efficiency is an important mechanism to finance increased coverage and access to services (Hensher, 2001). Information gathered in this study indicates large variations in cost levels and technical efficiency in the delivery of similar reproductive health services in the public sector. Although Sri Lanka now enjoys social and economic indicators that set it apart from most of the rest of South Asia, its high levels of technical efficiency and low unit costs were the result of public policy and decades of cost-reducing productivity gains (rather than of high levels of education as widely believed) (Pathrnanathan et al., 2003). Low unit cost levels in Sri Lanka contribute to its high levels of reproductive health access, and are not the consequence of it. Although not enough is known about the determinants of productivity improvement to make very specific recommendations, achieving universal access to reproductive services is unlikely to be attained in the rest of South Asia without some contribution from cost-reducing productivity gains. Among the improvements that could be achieved are: increasing the productivity of individual workers by providing motivation/incentivesand reducing absenteeism; improving outputs fiom investments by supplying 'missing ingredients' such as essential staff at health centers, medicines where staff are available, etc.; and increasing the effectiveness of outreach efforts by ensuring better instructions, training, and reducing 'overload' with extraneous tasks. In addition, some of the broader mechanisms discussed such as targeting and integrating services would also enhance productivity. 3.34 Privatefinancing demands attention. In South Asia, the disproportionately large share of reproductive health services financed by households has consequences not only for equity and financial protection against the costs of illness as discussed above, but also for 'protection' through contraception,assisted delivery, identificationof pregnancy risks through ANC, timely diagnosis and prevention of infections, and so on. Households may not purchase reproductive health services - many of which are "preventive" or "promotive"-becausethey perceive them to be low needhenefit-high cost; or they may purchase 'more affordable' poorer quality services. Private financing is desirable for low-cost items for which there are no significant market failures and which would be delivered regardless of public financing. One example is the provision of routine outpatient reproductivehealth care services. The empirical findings of this study indicate that these services are more likely to be privately financed than other types of services. In this situation, public financing is not necessarily best targeted at funding ambulatory services for all at the expense of reduced delivery of inpatient services (Filrner, et al., 2002). It is better to use household resources to finance a large proportion of those services, leaving the public sector to finance more expensive and unaffordable in-patient care. However, this should be determined by the extent to which market failures are present in the supply of different types of services and distributional concerns. For example, the lack of information about appropriate health care is a critical form of market failure in reproductive health. This relates to the lack of awareness of signs/symptoms and clinical quality, and inability to identify appropriate providers. Private financing in such a setting carries the risk of women increasingly seeking poor quality care from unqualified providers or not seeking care at all. For many of the poorest households, even outpatient services are unaffordable with private financing. The optimal resource allocation pattern is therefore to ensure that all or most households are publicly financed for inpatient services, and that outpatient services are funded primarily for the benefit of poorer households. This type of funding pattern is seen in Sri Lanka. 3.35 Scope exists for greater engagement of private sector providers - both commercial and non-profit - in financing provision of reproductive health services for poor women. Governments could encourage private providers to deliver a proportion of their services free or at low cost to the poor through incentive or award programs, contracts (discussed below) or even regulation. For example, in India, land is made available for private hospital construction at subsidized rates provided 25 percent of beds are used to treat poor patients. While results have been mixed (due in part to poor monitoring)such strategies can enlargethe pie of private services for the poor. Options for Resource Allocation 3.36 Resource allocations affect access, cost, quality, satisfaction and outcomes. Other than the risk pooling achieved through national health systems, insurance efforts could optimize resource allocations,but these face some difficultiesin the South Asi,mcontext. 3.37 The national health systems in South Asia have networks of facilities and providers, largely funded from general revenues. As they cover entire popillations the ''risk pool" is broad. However, as discussed earlier, there is a need to increase the efficiency and equity of public spending on reproductive health. The national health systems need to operate more efficiently, reach out to the poor and target resources to them, and actually achieve universal coverage. The wide range of problems that are encountered in health care (health problems as well as difficulties in service delivery) calls for better allocation and rationalization of resources in tandem with productivity improvements. This includes allocations by services- in this case to reproductivehealth and within reproductive health to the package of services that are essential to achieve the MDGs. Low priority expenditurescould make room for more necessary ones, especially interventions that have the greatest marginal impact on the poor. 3.38 Resource allocations must also be improved by level (e.g., to in-patient rural health facilities because the transaction costs to the rural poor of using local health centers are lower .thanthose of using more distant urban facilities) and by "input." For example, imbalances in staff, equipment, drugs and supplies may need to be rectified through appropriate budget allocations. Insufficiencies in staff of certain types may require contracting in (e.g., anesthetists and ob-gyn specialists in India) to provide the necessaryrange of services (in our example, EmOC) and use available resources (such as operation theaters) more efficiently. Allocative and technical efficiencyprovide value for money and hence improve financing of health care. 3.39 In addition to addressing the inadequacy and allocation of resources, and improving the efficiency of national health systems, South Asian governments also need to improve accountability. Poor experience with internal staff or management may call for services (such as cleaning, conservancy, watch or transport) to be contracted out; low motivation of staff may call for incentives (promise or performance-based) to be provided; and so on. Such detailed allocation decisions are best made at the local level through decentralized action planning within an administrative framework that supports the principleof subsidiarity. 3.40 Turning to private spending, South Asia's high out-of-pocket expenditure could be channeled to public or private pooling arrangements to increase financial protection through public sector health systems or various forms of insuranceincluding social health insurance, community health insurance, or privatelvoluntaryhealth insurance. However, private health insurance is less likely to produce more and more equitable health care in most South Asian contexts. 3.41 Social health insurance has been developed successfully for mothers and children in Bolivia. The government allocates 20 percent of national revenues to municipalities on a per capita basis, and they in turn reimburse facilities for services provided. The scheme led to increased use of services by adolescents and poor women, and service improvements (e.g., availability of medicines). As services were provided free, some users shifted from private to public care, and providers were dissatisfied becausetheir remuneration was not enhanced despite increases in workload and clients. 3.42 There is little experience in South Asia with social health insurance- it accounts for only 8 percent of public spending on health (or less than 2 percent of total spending) (Gottret and Schieber, 2006). It would be a challenge to introduceat a national level in at least four of the five countries because of low incomes, predominance of informal sector employment, large rural populations, inadequate administrative capacities, governance issues, poor quality of peripheral health services, etc. These constraints may produce negative consequencesfor equity in delivering and financing reproductivehealth care. It could be argued that social insurance could alleviate the need for public financing of services for middle-income groups, thus releasing resources for the poor. However, international experience indicates that political economy factors result in two-tier systems in which it is difficult to divert resourcesfrom richer groups to health care for the poor. 3.43 Community insurance involves risk pooling at the community level, albeit to a limited extent. It faces the problems of raising sufficient resources from the limited incomes and small populations on which they draw, limited management skills, and dominance of providers over prices and quality. A promising scheme that was begun in 2001 in a few districts of Indonesia was extended country-wide in 2005. The JPK-Gakin Scheme is administered by local governments, which has facilitated responsiveness to local conditions, such as the problems of remote areas and seasonal or geographic vulnerability to epidemics Although a review suggests that many improvements are needed in the scheme, it appears to have successfully increased health care coverage among the poor and deserves to be on a watch-list (Arifianto et al., 2005). In South Asia it is unlikely to work alone, but attention to several aspects could enhance its usehlness as a complementarymeans of financing reproductivehealth care. First, most schemes in South Asia currently do not cover the more expensive in-patient reproductive services including childbirth. Coverage would need to extend beyond the narrow range of services to insure households against the financial consequences of illness. Second, community insurance schemes rarely enroll the richest groups in the population, nor are they able to enroll the poorest groups, as underlined by BRAC and Grameen Bank in Bangladesh(Rannan-Eliyaand Hannan, 1997).~This needs to be addressed to exploit the progressive nature of this mode of financing. Finally, approaches to scaling up and sustaining community insurance schemes need to be developed in order to ensure lasting reproductivehealth benefits in South Asia. Options for PurchasingPayment Systems 3.44 Although the five national health systems provide services through staff who are salaried public employees, the bulk of purchasing in reproductive health care (except in Sri Lanka) is by individuals from private service providers. The goals of more and more equitable reproductive health care can also be furthered through purchasing mechanisms. In the public sector, for example, governments can get better value for the rnoney they spend through a variety of purchasing mechanisms, including decentralized planning and management which were discussed in Chapter 3, and contracting in or out for services and efficiency-basedprovider payments or incentives, discussed above. Other promising approachesfor reproductive health care in South Asia include demand-sidefinancingand social marketing. 3.45 The main advantages of demand-side financing are the potential to target benefits, provide a choice of providers, increase utilization of services and compliance with service or treatment regimens, and improve quality through competition (within a fixed price limit). Subsidizing specific health services for the poor, such as institutional delivery, abortion or contraception, could enhance demand and use, and increase the quantity of services flowing to them. Competitive voucher schemes are considered to be better than those in which there is a single provider, but non-competitive schemes can hlfill the purposes of targeting, providing an incentive to change a health behavior, or '.Privatelvoluntaryhealth insurance imposes financialbamers to access becauseof the low affordability of premia, and so is not considered here as our concern is primarily with poor women. referral. Several countries have implemented successful schemes. A Safe Motherhood scheme in Indonesia distributed a booklet of coupons for MCH and FP services to poor women, who could redeem them through contracted midwives (who otherwise charged for their services). In addition to reaching poor women, the scheme increased utilization of these services, and increased the services of the providers (Gorter et al., 2003). A scheme in Nicaragua provided vouchers to high-risk groups for ST1 treatment and reported a fall in the prevalence of infections as well as a lower treatment cost per patient. Bangladesh has initiated a health voucher scheme to increase demand for maternal and neonatal services and insure poor women against the costs of a normal delivery by a skilled provider or of emergency obstetric services. The vouchers enable pregnant women to purchase antenatal, delivery and postnatal services from a qualified private provider of their choice. Providers are reimbursed fi-oma special fund. 3.46 Public monies can also be used for direct cash payments ("conditional cash transfers" or "incentives") to poor households that achieve certain goals, such as "three ANC visits" or "delivering in an institution". This would protect the poor, allow process monitoring, and improve health outcomes. The PROGRESA program in Mexico gives families below the poverty line a subsidy of about 25 percent of their annual income if they utilize a full package of mother and young child health and nutrition services. In addition to increased utilization of services, improvements were reported in health (Gertler, 2000). India has instituted reimbursement of transport and attendant costs for the use of public sector EmOC facilities. Although this has had variable success it could be improved and expanded, and increased in scope to cover other out-of-pocket expenses for reproductive health care. 3.47 Social marketing is an intervention that has effectively mobilized household resources in the reproductive health sector and simultaneously improved delivery. It involves selling government- or donor-subsidized products through commercial channels, NGOs or public retail outlets. The main advantages are that the products, which are partially subsidized become more accessible to adolescents, poor and high-risk populations, and create demand. However, social marketing is not without drawbacks. Amongst the poorest households, where there are competing demands for scarce resources, having to pay, for example, for contraceptives may deter use even when prices are subsidized. Nevertheless, social marketing has been found to be a viable method for delivering contraceptive supplies in South Asia and is to be recommended, provided that publicly-funded zero-priced supplies are still available to the poorest. 3.48 Social franchising of reproductive health services can involve the private health sector in reproductive health goals, taking advantage of its contacts with the poor and improving quality and accountability. Clients benefit through improved access to information, facilities, services and discounts, and private providers gain as a result of bulk purchases, branding, incentives, mass marketing and referrals. The public sector benefits from the organization of private providers and bulk handling, and relative ease with which quality standards can be maintained and the possibility of performance-based contracts. The Greenstar experience in Pakistan and Janani in India, described earlier, provide useful financing and service delivery models to expand commodity and service distribution. Choosing among Options 3.49 Improvements in reproductive health financing must be developed in the context of broader reforms in the health sector. Depending on starting conditions and underlying values and preferences, the different countries may arrive at distinct "solutions" to their major health financing challenges. In assessing their financing options, several factors should be considered to ensure that adequate priority is given to improving reproductive health outcomes for the poor. All the five national governments need to take responsibility for financing health care strongly and subscribe to the urgency of improving reproductive health to reduce maternal and infant mortality, allow families to control their fertility, reduce RTIsISTIs/HIVIAIDSin the population are large, and enhance well-being among women and adolescents and their contributions to national development. The urgency leads to the conclusion that a broad mix of solutions, achievingwide coverage,needs to be adopted. Reliance on out-of-pocket payments contributes to demand-side barriers, and efforts should be made to introduce andlor strengthen collective funding of reproductive (and particularly maternal) health services for the poor. On-going monitoring of out-of-pocket payments and the benefit incidence of public spendingon reproductive health should accompanyfinancingreforms. Potential exists for demand-side financing to be used to encourage very poor households to seek appropriate reproductive health services. Various approaches are being introduced in South Asian countries,and should be considered for scale- up especiallyin areas where significantdemand-sidebarriers are identified. Even at current low levels of spending,there is room for greater efficienciesin the use of funds. Greater efficiency could be achieved by increasing the productivity of workers, integrating services, and promoting higher utilization, for example, through improved quality and outreach services. Some of these approaches may require increased spending, but would produce proportionately greater impacts if properlyimplemented. It is clear that current levels of investment would not achieve the MDG or ICPD goals so both donors and national governments need to increase their investments many fold. While South Asia's people and governments need to hlly "own" the efforts to improve reproductive health, donors could work more actively with them to identify how adequate resources could be mobilized Erom national and international sources and how these could best be programmed to achieve the necessary outcomes. 4.1 Although the five South Asian countries studied have had programs in place for several decades to address aspects of reproductive health, outcomes have not been commensurate with expectations. Many socio-cultural, political, institutional and technical factors have caused the inequalities in service utilization and reproductive health status discussed in the previouschapters. The inequalitiescan be attributed also to nonnative planning and supply-driven implementation of health programs, without adequate attention to local needs, conditions or demand. In all five countries the most backward areas and worst affected people require urgent attention if disparities are to be reduced and the MDGs achieved. This can be done through decentralized planning, with appropriateresource allocationsto correct inadequacies. 4.2 The chapter recommends a practical and simple method to plan actions to address local needs and inequalities and use available resources optimally. The method involves managers, providers, clients, local government representatives, and other important stakeholders. The planning can be done at the district level or below. Actions are developed on the basis of local analysis of needs and promising approaches, and a cyclical process of planning, implementation and monitoring builds the capacities of the health services and communities to improve health care. Good use of resources and demonstrated ability to get results could help decentralized units get the additional resourcesthey need. 4.3 Decentralized Action Planning can be used to increase the supply of services to poor women by targeting geographic areas, villages and households for attention, including outreach. It can also enhance demand where needed by targeting behavior change communications (BCC) and demand-side financing; identify supply-side improvements that would address local constraints; and ensure that programs for sensitive groups such as adolescents or tribal people are appropriately delivered. It can foster the integration of services, strengthening those in the essential package that are weak locally, providing the necessary tools and training to implement a client-centered approach, and measuring performanceaccordingly. 4.4 To energize preparation of effective action plans for reproductive health in South Asia, the chapter identifies and presents some promising practices. These focus on "how" more and better reproductive health services can be delivered. They include demand- and supply-side efforts and address access, quality of care and program management. Some focus on creating enabling environments to improve women's health. The practices were selected after a thorough search for field-tested efforts that have had measurable impacts. To bring a fillip to local practice, particular attention was paid to practices that are innovative and could be implemented,scaled-up and sustained in South Asia. 4.5 Decentralized health planning has been and is being carried out in various ways in each of the five countries. In 1994 Bangladesh devolved the planning and financial management of family planning services to the sub-district level. It allowed local governments to allocate locally-generatedfunds to implement the program. In 1998 the Local Initiative Program started 'bottom-up' input planning to promote community-based action and ensure more equitable, pro-poor allocations of health resources. However, the approach is considered complex and time-consuming as it involves formulation of logistic and procurement plans among other things (Cooper-Stephens,2001). 4.6 In India, decentralized planning began in the mid-'60s with block-level planning for Community Development, and has continued in the form of sectoral plans and multi- sectoral district and state plans which have established the bases for programs, activities and resource allocations. Districtlstatehealth plans have been funded largely on the basis of 'past actual expenditure plus' per capita or per facility, and have not been prepared in a participatory manner. In 1993 the adoption of the 73rd and 74th Constitutional Amendments provided for elected local governments to manage some aspects of health services, providing a new framework for planning and implementation. The introduction of a Community Needs Assessment (CNA) approach in the Family Welfare program in 1998 was intended to produce quality improvements from the sub-center level upward, but 'top-down' introduction of formats and poor preparation stymied its success (Policy Project, 2000). This experience resulted in recognition of the importance of decentralized service delivery and integration of reproductive health services, and various approaches to decentralizedplanning are being tested in different programs or parts of the country. 4.7 In Pakistan, devolution was introduced through a Presidential Ordinance in 2001, and the roles of district and provincial governments in planning, personnel management and procurement are still evolving. Nepal's Local Self Government Act of 1999 devolved management of health care from the central and regional directorates to lower levels. The first major step of transferring the property, equipment and staff of all Health Posts and Sub-Health Posts to Village Development Committees is unfolding slowly. Beginning in 2004, Pro-Poor District Health Investment Plans are being prepared as part of the Second Long-Term Health Plan (1997-2017), which is expected to accelerate the decentralization process. In Sri Lanka, health administration was decentralized in 1987 with the passage of the 13'~Amendment to the Constitution and planning is decentralized. The Central Government, however, retains responsibility for formulating and administeringthe NationalHealth Policy. 4.8 In essence, although decentralized planning is recognized in all five countries as a way to address local variations in health needs, the appropriatemethods for such planning are still evolving. Current approaches fall mainly into three categories: (i) need-based planning, (ii) resource-based planning and (iii) norm-based planning. Need-based planning involves assessing local health needs using data on the disease burden, demographic information, and people's responses to questions about health care. The approaches in the GOI's RCH Program and Sri Lanka's District Health Planning initiative are examples of need-based planning. Eliciting people's health needs is important but not fully reliable because respondents usually describe their needs on the basis of recent efforts to obtain curative care, and they rarely perceive a need for preventivecare which encompassesmany reproductivehealth services. 4.9 Resource-basedplanning accepts general health objectives as a given and focuses on allocating uncommitted resources to activities that are under-funded or flagging. It serves to improve the utilization of resources but usually does not address major gaps or inequalities. Norm-based planning assumes certain input and activity norms and estimates the resources required. Its weakness lies in the poor fit between norms and local situations, particularly in the most backward areas. India's family planning program was an example of norm-based planning that went wrong - considerable resources were allocated over a long period of time to implement activities based on nonns, but poor outcomes and inequalities were not addressed adequately. Bangladesh's Local Level Planning is another example. 4.10 Althoughdecentralizinghealth servicesis national policy in all five countries, and large units such as provinces and states enjoy some autonomy, below this level "decentralization" is largely restricted to the devolution of certain tasks. Generally, district health organizations plan and finance activities within norms, targets and budget envelops specified by higher-level authorities, manage and supervise the delivery of services, oversee in-service training and coordinate with other government departments. They have little choice between priorities or even activities, and can usually modify resource allocations only on the margin. Existing approaches to planning focus largely on inputs and on the allocation of resources to specific activities. Local action plans are prepared rarely and usually do not addressinequalities. 4.11 Our review of the quality of reproductive health care in the five South Asian countries identified many weaknesses in program implementation. All the countries are interested in accelerating effective implementation through local-level planning in keeping with their decentralization policies. As capacities in the field are limited, they need a simple and practical approach that can be implemented with available knowledge and skills. Consequently, during this study, local managers and researchers developed a decentralized action planning (DAP) method, keeping local limitations in mind and aiming to build capacities over time. The participatory nature of the method harnesses; local human resources and can lead to the garneringof other local resources also. 4.12 Because of size and the availability of a range of human and material resources, the district was deemed the most appropriate unit for planning in all countries. The DAP method was developed simultaneously in 12 districts in South Asia: four in India (two in Karnataka and two in Rajasthan) and two each in Bangladesh, Nepal, Pakistan and Sri Lanka. The method combines concepts from Total Quality Management and the Problem-Solving Approach to address three key questions: (i) How well is the district doing in terms of coverage and equity in reproductive health service delivery? (ii) Which essential services are under-utilized, especially by the poor, and why? (iii) How can the utilization of these services and their quality be improved? The planning initiates a cyclical process including implementation and monitoring, as shown in Figure 6, and described briefly below. Figure 6. DecentralizedAction Planning and Implementation:A Cyclical Process 4.13 Assessing the Local Situation. To assess the gaps between actual and expected levels of reproductive health service coverage and equity in the district, informationwas compiled on the district's demographic profile,health infrastructure,staff, reproductive health indices, government-sponsored reproductive health schemes, and private and NGO resources. Information on private/NGO providers included the types of services they provide and their areas of operation. All the information was presented simply in the local language in consultativeworkshops at the district level (which were also conducted primarily in the local language). Information on good practices in reproductivehealth was also made availableto those involved in the planning effort. 4.I4 ParticipatoryAction Planning. Workshops were organized in each district to develop Action Plans. The district teams typically consisted of 20 to 25 persons, about equal numbers from each of three key stakeholder groups: health program administrators (District Health Officers and those dealing with family planning,reproductivehealth, health education, and mother and child nutrition),service providers(doctors, nurses, supervisors and workers), and private sector, NGO, local government and community representatives. Members of the planning teams were selected using two criteria: (i) they had a role in the delivery of reproductivehealth services, and (ii) they were capable of analyzingprogram constraints and suggestingsolutions. 4.15 Each workshop consisted of five sessions, following the first five steps in the process. In the first session, the information on the district was discussed and validated. Team members provided additional or alternative information. Next, reproductivehealth services were broadly categorized into three groups - Maternal and Neonatal Health, Reproductive Health of Sexually-Active Adults, and Adolescent Health - and participants discussed service provision in their district. They identified one service in each of the three categories that required priority attention using "inadequacy" and "inequality"indicators: Inadequacy Ratio = Performance of Best District in the State/Province/Reglon Performanceof the District InequalityRatio = Coverage among the Non-poor Coverage among the Poor 4.16 This exercise attracted significant discussion. For example, a District Health Officer in India was in favor of giving priority to Safe Delivery because it was the most inequitable service. Service providers,however, felt that attention to Full Antenatal Care would ensure Safe Delivery. After a lengthy discussion, the team selected Full ANC and Safe Delivery for action planning. Similar discussionsoccurredin other districts. In Pakistan, in the absence of district level data, the teams used a method developed by the Ministry of Health and the Multi-DonorSupport Unit to calculate inadequacy and inequality. They scored services on the: Magnitude of the problem addressed by the service, Severity of consequencesif the problem were unattended, Availabilityof feasible interventions, Cost-effectivenessof the intervention,and Public and political demand for the service. 4.17 In the third session, participantswere divided into the three stakeholdergroups to discuss why the services they had selectedwere under-utilized,inadequateand inequitable in their district. Both demand- and supply-sideconstraintswere identified. For example,in Mysore district of Kamataka, India, participantsidentified a lack of awareness,cost and poor quality as the reasons for low RTIISTItreatment. Using ranking or voting techniques,one or two of the most important problems that are amenableto action for each of the services were selectedto be addressedin the action plans. This exercise showed that different problemswere perceived to affect tlhe utilizationof the same service in different districts. For example, the prime obstacle identified for under-utilizationof maternal care in Kamataka was cost, while in Rajasthan (India) it was poor access and in Nepal it was lack of awareness. 4.18 For each problem selected, the planning team identified actions to improve utilization and equity and identified some potential opportunities for and constraints to implementation. In the next session, the team worked on converting their ideas into action plans using a logical framework. Each Action Plan contains the actions, opportunities and constraints identified. The plans were completed by assigning responsibilities, goals and time-frames, and discussed to check whether any aspects had been ignored or unrealistic assumptions made, to ensure final practical Action Plans. Finally, the members of each stakeholder group assessed the Action Plan on the basis of several criteria: innovation, potential effectiveness, practicality, sustainability, effective resource utilization, and community participation. This final step gave all a sense of the challenges that lay ahead and the impetus to make their "own" plan work. 4.19 The decentralized Action Plans had several important features including meaningful roles for communities in their implementation, and re-deploymentof existing human resources for more effective service delivery. The activities proposed were practical and within the district's control, features which were particularly encouraging to those involved. The participants also realized that over time they could tackle other problems because planning and implementation would be repeated in cycles, and they could take on more difficult issues as they gained in experience and confidence. Although it was not possible to do so in this first round of planning, participatory local- level planning has the potential to identify and utilize a range of local public and private resources (e.g., service providers, facilities, donations in cash or kind) and develop meaningful local partnerships. 4.20 Key Lessons from the Pilots. The pilots in the five countries suggest that the following are important to support a decentralized planning process: Concise and clear information on the district's reproductive health status, service utilization, health resources and policy mandate; A manageable number of participants with different backgrounds who can make positive contributions; A skilled facilitator who appreciates community participation, understands the local health system and management practices, knows national reproductive health programs and strategies, and has critical thinking faculties; A set of locally-appropriate 'promising practices', including information on their implementation process; Community views on how to make plans more pro-poor; Methods to assess the services for which data are not available; Proper monitoring and supervision of implementation, with inclusion of communities in monitoring, in order to improve the next cycle of planning; An efficient Management Information System and use of the data for decision making; Functionalinter-sectoralplanning and management teams in the districts. 4.21 The DAP process brings about change through understanding (the situation), agreeing (on goals), adapting (practices), improving skills, expanding efforts, and empowering providers and clients. It is similar to the SEED-SCALE model (lde-Taylor and Taylor, 2002) as it involves"Self-Evaluationfor Effective Decision-making (SEED) and is a "System for the Community (of practitioners) to Adapt, Learn and Expand" (SCALE) as it begins with health providers and those most closely related. The health system catalyzes change within itself first and then in its community of clients. Several of the principles of the SEED-SCALE model are integral to DAP. It builds on the existing capacities and strengths of those involved in health care, and can increase confidence with each round. It focuses on available resources and assets and plans actions to address needs based on what is possible. Decisions are made and actions planned on the basis of objective local data (evidence). The health managers and providers may need guidance to interpret and use the data, which can be obtained from outside experts. Local expertise can also be used to introduce appropriate up-to-date ideas and build capacity, and help to gather better data for the next round of planning, helping the process develop from simple to sophisticated. However, DAP uses fewer resources from outside than working directly with communities, which would require a lot of facilitation. Such resources cannot always be obtained where they are needed or because of the expense involved. Scaling up using NGOs or private professionals would simply take too long given the scale of South Asian countries. In effect DAP eliminates intermediaries except as facilitators and as decided by the local planning/implementation teams. 4.22 Recently, India has begun a widespread process of decentralized action planning under the Reproductive and Child Health Program and National Rural Health Mission (Box 4.1). It aims to improve equity in access to health services through this local level planning and resource allocation process, which is also focused on the key reproductive health goals of reduced maternal and child mortality. Thus, it will be on the watch-list of all those concerned with improving poor women's reproductive health in South Asia. Box 4.1 India's Rural Health Mission and Reproductive and Child Health Program In April 2005 India launched a National Rural Health Mission (NRHM) to provide health care more effectively to rural people throughout the country, especially in 18 (of 35) states with poor public health indicators and inadequatehealth infrastructure. Improving access to, utilization and quality of health services is an avowed aim of the mission, which also emphasizes the equity and gender dimensions of health care. The main objectives of the program are to reduce child and maternal mortality, prevent and control communicable and non-communicable diseases, and stabilize the population, achieving gender and demographic "balance". The second phase of the Reproductive and Child Health Program was launched at the same time and subsumed into the NRHM. It focuses on reducing the IMR, under-five mortality rate, MMR and TFR, with national, state and district plans directed at achievinglocal goals in these areas. Strengthening the Health Structure. The NHRM aims to "undertake architectural correction"of the health system to improve public health management and service delivery. The vertical health and family welfare programs will be integrated. "Comprehensive" primary health care is to be provided. Women health activists called ASHAs are expected to "organize the demand side" and promote use of health care. There is also a focus on strengtheningSub-centers by increasing the number of ANMs, human resource development, quality standards, community support and an "untied fund" to support local actions. PHCs will be similarly improved, and 30-50 bedded CHCs brought up to Indian Public Health Standards to provide better curative care. NRHM is placing long overdue emphasis on human resource improvements including locally resident workers, contractualpositions, multi-skilling,career development and transparentpolicies. Promoting Institutional Delivery. An important interventionunder NRHM is the Janani Suraska Yojana scheme to promote institutional delivery and thereby reduce maternal and neonatal mortality. Women who complete three antenatal visits, two tetanus toxoid injections, and deliver in an institution receive cash payments for these. ASHAs who accompany women to institutions for deliveries also receive cash incentives. Traditional Health Systems. One of the cornerstones of the program is the promotion of local health traditions including Ayurveda, Yoga, Unani, Siddha and Homeopathy (AYUSH). The hlRHMaims to mainstream these systems into public health care. Decentralization. Decentralized management of public services is a core strategy of the mission. Each district has a District Health Mission (DHM), and the states have a State Health Mission. District health action planning (DHAP), which is to be realized through participatory and "bottom-up" processes to make them more responsive to local needs, has been started. It includes a situation analysis of the district, formulation of objectives, identificationof interventions, a work plan and budget, and an M&E plan. The DHAPs will be appraised and approved at the state level. Following this, each district will develop its own plans for Training,BCC, and Logistics in keeping with its health plan, and each health facility (sub-center,PHC and CHC) is expected to develop its own service delivery plan on the basis of the approved DHAP. The DHMs are responsible for overseeing implementation of the DHAPs. Village health plans are also to be prepared by the Village Health Committeeof the Panchayat. "Funds, functions and functionaries" are to be transferred to local government (Panchayati Raj) institutions (PRls), and facility-based health management committees (Rogi Kalyan Samitis) are also expected to play a greater role. Capacity enhancementsare planned for the PRls to control and manage public health services. These mechanisms are expected to address the problems of inter-districtand inter-statedisparities. Attention to Determinants of Health. District planning and management are expected to foster greater attention to the determinants of health and coordination across government departments concerned with nutrition, drinking water, sanitation, and efforts to improve the status of women. The program aims to universalize access to public food, nutrition, sanitation, immunization and health services. The program also aims to activateadolescent girls' and women's groups Health Spending. The NRHM aims to increase public spending on health from India's current 0.9 percent of GDP to 2-3 percent, including communityfinancing and risk pooling. Health care is expected to become more affordable to the poor. The non-profit health sector is to be promoted especially in underserved areas. Allocations of RCHINRHM funds to the states are done on the basis of population with additional weightage given to more needy states. States have the option of allocating equal shares to districts, or providing them funds on the basis of socio-demographic criteria to improve equity, or on the basis of need. Accountability. The goals of IVRHlVl are "time bound" and its processes transparent. Data collection will be strengthened and data used for planning as well as for progress monitoring against standards and goals. The program is expected to report publicly on its progress. References Agarwal, K. (1998) ReproductiveHealth Case Study: Nepal, Policy Project, Kathmandu. Alban, A. and G.M. Sakya (2004) "Costing of the Nepal Health Sector Program Implementation Plan with Unit Cost of Essential Health Care Services(EHCS),"EASE InternationalConsultantReport to the World Bank, Kathmandu, Nepal, processed. Anon (2003a) "Safe Motherhood: Program Examples, Malaysia: Increasing Access to Maternity Care," ReproductiveHealth Outlook http://www.rho.orrz;/html/s ro~examples.htm Anon (2003b) "Family Planning Program Issues: Program Examples," Reproductive Health Outlook h~:i/www.rho.orrr/html/fppprogexamples.htm Anon (2003~)"Men and Reproductive Health: Program Examples," Reproductive Health Outlook http:/iwww.rho.or~htmlimenrh progexam~les.htm Arifianto, A., R. Marianti, S. Budiyati and E. Tan (2005) "Making Services Work for the Poor in Indonesia: A Report on Health Financing Mechanisms (JPK Gakin Scheme) in Kabupaten Purbalingga, East Surnba and Tabanan", SMERU ResearchInstitute,Jakarta, processed. Atherton F., G. Mbekem and I. Nyalusi (1999) "Improving Service Quality: Experience from the Tanzania Family Health Project,"Znternational Journal for Quality in Health Care 11. Aziz, M. (1998) Effects of Quality on Utilization of MCH Service, Ph.D. Dissertation, University name and place Bang, A. T., R. A. Bang, S. B. Baitule, M. H. Reddy and M. D. Deshmukh(1999) "Effect of Home-based Neonatal Care and Managementof Sepsis on Neonatal Mortality: Field Trial in Rural India," Lancet 354, 1955-1961. Bhattarai, S.K. (2000) "A Study on Knowledge, Attitudes and Skills of Health Workers on Reproductive Health,"Processed, DHSFHD-GTZ, Kathmandu. Caldwell, John C., B-e-Khuda, B. Caldwell, I. Pieris and P. Caldwell (1999) "The Bangladesh Fertility Decline," Population and Development Review 25 (1) 67-84. Casterline, J. B., Z. A Sathar and Minhaj ul Haque (2001) "Obstacles to ContraceptiveUse in Pakistan: A Study in Punjab," Policy ResearchDivision,PopulationCouncil, Islamabad. CIET (2003) Third HPSP Service Delivery Survey (July 2003) Preliminary Key Findings Report, Draft., Ministry of Health and Family Welfare, Government of Bangladesh,Dhaka. Cleland, John, J.F. Phillips, S. Amin and G.M. Kamal(1994) The Determinants of Reproductive Change in Bangladesh: Success in a ChallengingEnvironment,The World Bank, Washington,D.C. Cook, R.J., B.M. Dickens and L.E. Bliss (1999) "International Developments in Abortion Laws from 1988 to 1998," AmericanJournal of Public Health 89,579-586. Cooper-Stephens, T. (2001) Review of the Phase 1 of Local-levelPlanning, Processed, HLSP Consulting, Dhaka. El Zayed, S.R., Y. Stopnitzkyand Q. Khan (2005) "Reaping the Benefits of Girl's Secondary Educationin Bangladesh- Impact on Fertility and Malnutrition," South Asia Human Development Sector, The Wor1.d Bank, Washington,D.C., presentation. EngenderHealth (2003) "Country by Country: Pakistan" and related material on website: http://www.enaenderhealth.org/ia/cbci Exterkate, M., M. De Jager and M. Hayat (1999) "Financial Resource Flows for Population Activities: Report of a Case Study in Pakistan," Processed, Netherlands Interdisciplinary Demographic Institute and UNFPA, Islamabad. Filmer, D., J. S. Hammer and L. H. Pritchett(2002) "Weak Links in the Chain 11:A Prescription for Health Policy in Poor Countries,"World Bank Research Observer 17,47-66. Friedman, R. (1996) "Quality of Care, Infection Prevention and Other Aspects in Public Sector Family Planning Facilities in Nepal: An Action-oriented Assessment,"Journal of the Nepal Medical Association 34, 118-125. Gertler, P. (2000) The Impact of PROGRESA on Health: Final Report, Processed, International Food Policy Research Institute, Washington,D.C. Gorter, A., P. Sandiford, Z. Rojas and M. Salvetto (2003) Competitive Voucher Schemes for Health, Background Paper prepared for the Private Sector Advisory Unit of the World Bank, Washington, D.C., processed. Gottret, P. and G. Schieber (2006) Health Financing Revisited - A Practitioners' Guide, The World Bank, Washington, D.C. Governmentof India, Ministry of Health and Family Welfare (2002) Annual Report 2001-02, Ministry of Health and Family Welfare, Government of India, New Delhi. Government of Nepal, Central Bureau of Statistics (2004) Nepal Living Standards Survey, 2004, National Planning Commission,Kathmandu. Government of Nepal, Ministry of Health, New ERA and ORC Macro (2002) Nepal Demographic and Health Survey 2001, Family Health Division, Ministry of Health and New ERA, Kathmandu, Nepal; and ORC Macro, Calverton,Maryland, U.S.A. Government of Nepal, Ministry of Health and Population(2005) Public Expenditure Review in Health- 2003, Ministryof Health and Population,Kathmandu. Government of Nepal, Ministry of Health and Population(2007)'Nepal Demographicand Health Survey: 2006, PreliminaryFindings', Kathmandu. Governmentof Pakistan,FederalBureau of Statistics(2005) Pakistan Social and Living StandardsMeasurement (PSLM) Survey 2004-05, Federal Bureau of Statistics,Islamabad. Government of Sri Lanka, Department of Census and Statistics (2002a) Sri Lanka Demographic and Health Survey 2000, DCS, GOSL, Colombo. Hensher, M. (2001) "Financing Health Systems through Efficiency Gains," Commission on Macroeconomicsand Health, Working Group 3, Paper No. 2, New York. International Institute of Population Sciences and ORC Macro (2000) National Family Health Survey-2, 1998-99,InternationalInstitute of PopulationSciences,Mumbai, India. Islam, M.A., M.M. Islam and N. Chakraborty (2002) "Plateauing of Fertility Level in Bangladesh: Exploring the Reality,"Dhaka, Bangladesh,mimeo. Jafarey, S.N. and R. Korejo (1995) "Social and Cultural factors Leading to Mothers Being Brought Dead to Hospital in Karachi, Pakistan,"International Journal of Gynecology and Obstetrics 50 Supplement 2S, 97-99. Jejeebhoy, S. (2000) "Adolescent Sexual and Reproductive Behaviour: A review of the Evidence from India," in Women's Reproductive Health in India (R. Ramasubban and S. Jejeebhoy, eds.) Rawat Publications,Jaipur, India. Kavitha, N. and N. Audinarayana (1997) "Utilisation and Determinantsof Selected MCH Care Services in Rural Areas of Tamil Nadu," Health and Population Perspectives and Issues (volume and page numbers). Kim, J. and M. Motsei (2002) "Women Enjoy Punishment: Attitudes and Experiences of Gender-based Violence among PHC Nurses in Rural South Africa,"Social Science and Medicine 54, 1243-1254. Landovitz , K. (1997) "Increasing access to contraception" AVSC News 35 (4) cited in Anon (2003b) "Family Planning Program Issues: Program Examples," Reproductive Health Outlook http://www.rho.org/html/~p progexamples.htm Mahal, A., S. A. Yazbeck, D.H. Peters and G.N.V. Rarnana (2001) The Poor and Health Service Use in India, HNPDiscussionPaper Series,The World Bank, Washington, D.C. Murthy, N. and A. Bama (1998) "Integrating Reproductive Health in Health Programs in India," in Gender, Population and Development (M. Krishnaraj and others, eds.), Oxford University Press, New Delhi, pp. 293-309. Murthy, N., L. Rarnachandran, P. Pelto and A. Vasan (2002) "Dismantling India's Contraceptive Target System: An Overview and Three Case Studies," in Responding to Cairo: Case Studies of Changing Practice in Reproductive Health and Family Planning (N. Haberland and D. Measham, eds.) Population Council, New York, pp 25-57. National AIDS Control Organizationand UNICEF (2002) Knowledge, Attitudes and Practices of Young Adults (15-24 years), NationalAIDS Control Organization,Government of India and UNICEF, New Delhi. National Institute of Population Research and Training (2001) Bangladesh Demographic and Health Survey, 1999-2000,NIPORT, Dhaka. NationalInstitute of PopulationResearch and Training and ORC Macro (2004) Bangladesh District Level Socio-demographicand Health Care UtilizationIndicators, Processed,Dhaka. National Institute of Population Research and Training, ORC Macro and Johns Hopkins University Centre for Health and Population Research (2003) Bangladesh Maternal Health Services and Maternal Mortality Survey 2001, Processed, NIPORT, Dhaka. National Institute of Population Studies (2001) Pakistan Reproductive Health and Family Planning Survey 2000-2001: Preliminary Report, NIPS, Islamabad. NationalInstitute of PopulationStudies(2003) Status of Women,Reproductive Health and Family Planning Survey, NIPS, Islamabad. Ndure, K.S., M. N. Sy, M. Ntiru and S.M. Diene (1999) Best Practices and Lessons Learned for- Sustainable Community Nutrition Programming, ORANA, SANA and USAID, Africa. Osei, I., B. Garshong, G. 0. Banahene, J. Gyapong, P. Tapsoba, I. Askew, C. Ahiadeke, R. Killian, E. Bob, P. Combary and W. Sampson (2005) "Improving the Ghanaian Safe Motherhood Programme: Evaluating the Effectiveness of Alternative Training Models and Other Performance Improvement Factors on the Quality of Maternal Care and Client Outcomes," FRONTIERS Final Report, Population Council, Washington, D.C. Oxford Policy Management (2002) "Evaluation of the National Programme for Family Planning and Primary Health Care in Pakistan,"Processed, Oxford Policy Management, U.K. Pathak, K.B., G. Feeney and N. Luther (1998) Permanent Method, Temporary Methods and Fertility Decline in India, NFHS Subject Reports Number 7, International Institute for Population Sciences, Mumbai, India and East-West Center for Population, Honolulu, U.S.A. Pathrnanathan,I., J. Liljestrand, Jo M. Martins, L. Rajapaksa, C. Lissner, A. de Silva, S. Selvaraju and P. J. Singh(2003) Investing in Maternal Health: Learning from Malaysia and Sri Lanka, Health, Nutrition, and Population Series, World Bank, Washington, D.C. Paul, B.K. and D.J. Rumsey (2002) "Utilization of Health Facilities and Trained Birth Attendants for Childbirth in Rural Bangladesh: An Empirical Study,"Social Science and Medicine 54(12), 1755-1765. Population Council (2003) dole scents and Youth in Pakistan 2001-02: A Nationally Representative Survey, Population Council, Islamabad. PRIME I1 Project "India: Involving Men in Partnerships for Safe Motherhood," PRIME Voices 20, accessed 6/27/03. Radennacher, R., N. Wig, 0 . van Putten-Radernaker,V. Muller and D. Dror (2005) The Yeshasvini Trust, Karnataka, India," CGAP Working Group on Microinsurance Good and Bad Practices Study No. 20., The World Bank, Washington, D.C. Rahrnan, A., L. Katzive and S.K. Henshaw (1998) "A Global Review of Laws on Induced Abortion: 1985- 1997," International Family Planning Perspectives 24 (2), 56-64. Rana, B. K and L. T. Johnson (2003) "Hot-tempered Cold Fever: Illness and Impoverishment among Adivasis in Jharkhand," Economic and Political Weekly 38(15), 1451-1455. Rannan-Eliya, Ravi P. and F. Hannan (1997) "Experiences with Resource Mobilisation in Bangladesh: Issues and Options," Review paper commissioned by World Bank, Dhaka. Rubardt, M. (2002) "CARE International's Community-based Distribution Program in Eastern Ethopia Increases Contraceptive Use," At a Glance 8, NGO Networks for Health, cited in Anon (2003b) "Family Planning Program Issues: Program Examples," Reproductive Health Outlook h~:llwww.rho.org/html/~p progexamples.htm Shumba, Paul S.S. (2003) "HIV/AIDS and Food Security, Southern African Development Community - Food, Agriculture and Natural Resources - Regional Vulnerability Assessments, Final Draft Analysis Report,"Consultancy report for UNAIDS, New York. Spadacini, Beatrice M. (2001) "When Culture Helps Save Women's Lives," Plan International, New York. http://~~~.~hild~llllri~al.c0dfeature~fbo0kmarks/3 WGame.cfm Tata Institute of Social Sciences (2003) "Social Assessment of Reproductive and Child Health Programme: A Study in 5 Indian States,"Study done for the preparation of the Second Reproductive and Child Health Project for the MOHFW, GOI, New Delhi. Taylor-Ide, D. and C.E. Taylor (2002) Just and Lasting Change, Johns Hopkins University Press, Baltimore, Maryland. United Nations' Children's Fund (2004) "Sri Lanka: The Use of Maternal Death Review in Problem Analysis and Decision Malung," in A Book of Case Studies: Women's Right to Life and Health, pp. 91- 112, UNICEF Regional Office for South Asia, Kathmandu. United Nations' Children's Fund and World Health Organisation (2004) Low Birth Weight: Country, Regional and Global Estimates, UNICEF, New York. USAIDIOMNI (2000) "Controlling Anemia in Sri Lanka, Issues and Options," USAIDIOMNI and Nutrition CoordinationDivision, Ministry of Plan Implementationand Parliamentary Affairs, C8010mbo. van Doorslaer, E., A. Wagstaff and F. Rutten, eds. (1993) Equity in the Finance and Delivery of Health Care: An InternationalPerspective, Oxford UniversityPress, Oxford. Valley Research Group (2003) "Nepal Family Health Program Behavior Change Communication Formative Research and Evaluation Study: Findings on FCHV Survey," Processed, Valley Research Group, Kathmandu. Ved, R. (1997) "Private Voluntary Organizations in Health 11: An Overview," Indian Journal of Community Health, (volume andpage numbers). Vissandjee, B., R. Barlow and D.W. Fraser (1997) "Utilization of health services among nual women in Gujarat,"Public Health111, 135-148. Vlassoff, M. and S. Bernstein (2006) "Resource Requirements for a Basic Package of Sexual and Reproductive Health Care and Population Data in Developing Countries: ICPD Costing Revisted," Background paper for the LIN MillenniumProject, www.unrni1lenniuin~roiect.org Withanachchi, N., W. Karandagoda and Y. Handa (2004) "A Performance Improvement Programme at a Public Hospital in Sri Lanka: An Introduction,"Journal of Health Organization and Management 18, 361-369. World Bank (2002) Female Secondary School Assistance Project - 11, Project Appraisal Document, Processed, The World Bank, Washington,D.C. World Health Organization (1996) Antenatal Care, Report of a Technical Working Group, (No. WHOIFRHIMSMI96.8), World Health Organization,Geneva. World Health Organization (2004) Maternal Mortality in 2000: Estimates Developed by WHO, UNICEF and UNFPA, World Health Organization,Geneva. World Health Organization (2005) Estimating the Cost of Scaling-Up Maternal and Newborn Health Interventions to Reach Universal Coverage: methodology and assumptions,'' Technical WorkingPaper prepared by the Departments of Making Pregnancy Safer and Health Systems Financing for the World Health Report 2005, WHO, Geneva.