Report No. 27005-BD Bangladesh Private Sector Assessment for Health, Nutrition and Population (HNP) in Bangladesh November 18, 2003 South Asia Human Development Sector Unit & HD Network Health, Nutrition and Population Team Document of the World Bank CURRENCYEQUIVALENTS CurrencyUnit - Taka US$l - 58.45 BDT GOVERNMENTFISCALYEAR July 1-June30 ACRONYMS AAA Analyticaland AdvisoryAssistance ADB Asian Development Bank AIDS Acquired Immune Deficiency Syndrome ALWARI Acute (Lower)RespiratoryInfection APP AlternativePrivatePractitioner BBS BangladeshBureau of Statistics BCC Behavior Change Communication BCG Tuberculosisvaccine BINP BangladeshIntegratedNutrition Project BMA BangladeshMedical Association BMDC BangladeshMedical and Dental Council BMMS BangladeshMaternal Health Servicesand Mortality Survey BNC BangladeshNursing Council BPCDOA BangladeshPrivate Clinic and DiagnosticOwners' Association BPHC BangladeshPopulation and Health Consortium BPMPA BangladeshPrivateMedicalPractitioners' Association BRAC BRAC (formerly,BangladeshRural Advancement Committee) CBR CrudeBirth Rate CDR CrudeDeath Rate CIDA CanadianInternational DevelopmentAgency CIET CommunityInformation & EpidemiologicalTechnologies CPR Contraceptive PrevalenceRate DFID Department for InternationalDevelopment(UK) DHS Demographic and Health Survey EC European Commission EIS Epidemiological Information System EOC Essential Obstetric Care EPI Expanded Program on Immunization ESP Essential ServicesPackage ,GAVI GlobalAlliance for Vaccination and Immunization GDP GrossDomestic Product GOB Government of Bangladesh HA Health Assistant HDS Health and DemographicSurvey(BBS) HEU Health Economics Unit HDS Health andDemographic Survey (BBS) HEU HealthEconomics Unit HKJ HealthInformation Unit HIV HumanImmunodeficiency Virus HNP Health, Nutrition, andPopulation HPSP Health and Population Sector Program HRDU HumanResources Development Unit ICDDR,B Center for Health and Population Research IDD Iodine Deficiency Disorder IHE Institute o fHealthEconomics (Dhaka University) I M C I IntegratedManagement o f Childhood Illness IMR Infant MortalityRate I-PRSP Interim Poverty Reduction Strategy Paper MBBS Bachelor ofMedicine and Bachelor o f Surgery M C H Maternal and ChildHealth MDG MillenniumDevelopment Goal(s) MMR Maternal Mortality Ratio (No. of maternal deaths/l00,000 live births) MOHFW MinistryofHealth& FamilyWelfare NGO NonGovernmental Organization ORS Oral Rehydration Solution ORT Oral Rehydration Therapy PHC Primary Health Care PPP Public-Private Partnership PSA Private Sector Assessment PSB Pharmaceutical Society o f Bangladesh QA Quality Assurance R T I Reproductive Tract Infection SD Standard Deviation SDS Service Delivery Survey Sida Swedish International Development Cooperation Agency SMC Social Marketing Company STD Sexually Transmitted Disease(s) TB Tuberculosis TBA Traditional BirthAttendant TFR Total Fertility Rate THC ThanaLJpazilaHealth Complex Tk Taka UNICEF UnitedNations Children's Fund USAID United States Agency for InternationalDevelopment WHO World Health Organization Vice President Praful Pate1 Country Director Christine Wallich Sector Manager Anabela Abreu Task Team Leader Sundararajan Gopalan Table of Contents Chapter 1. TheProvisionOfHNPServices in thePrivate Sector........................................................................ i 1.1 Introduction..................................................................................................................... 1 1.2 The Actors....................................................................................................................... 2 1.3 Physicians ....................................................................................................................... 4 1.4 Nurses ............................................................................................................................. 6 1.5 Trained Paramedicals...................................................................................................... 8 1.6 Alternative Private Practitioners ..................................................................................... 9 1.7 HealthFacilities ............................................................................................................ 12 1.8 Other Private HNP InputMarkets................................................................................. 15 1.9 Conclusions................................................................................................................... 15 Chapter2. Demandfor Consumption$Pkuate SectorHNP Services / ......................................................... 17 2.1 Introduction................................................................................................................... 17 2.2 Health Service Consumption ........................................................................................ 17 2.3 Public/Private Shares o f Services ................................................................................. 19 2.4 Determinants of Consumption / Health Care Seeking Behavior .................................. 22 2.5 Conclusions................................................................................................................... 31 Chapter3. Interaction between the Publicand the Private Sectorsin HNP ...................................................... 33 3.1 Introduction................................................................................................................... 33 3.2 Current Situation........................................................................................................... 33 3.3 Perceptionso fthe Public Sector towards Working with the Private Sector.................40 3.4 Perceptionso fthe Private Sector.................................................................................. 41 3.5 Conclusions................................................................................................................... 42 Chapter4. Main messages andpolip options ................................................................................................... 44 4.1 Findings......................................................................................................................... 44 4.2 PolicyFonnulation: A newparadigm........................................................................... 45 4.3 Policy Options............................................................................................................... 46 4.4 The Way Forward......................................................................................................... 55 Executive Summary Bangladeshhas seen remarkable improvements in health indicators over the last 30 years. The infant mortality rate (IMR) stood at 153 deaths per 1000 live births in the mid-1970s while the latest data suggest an IMR o f 62. In a similar way under-fivemortality has declined by two thirds inthirty years andnow stands at 83. Life expectancy at birthis now estimated at 61 years. Two-thirds o f the deaths in children under five are due to acute respiratory infections (ARI, 27%), perinatal causes (24%) and diarrhea (16%). Neonatal mortality (Le., deaths in the first month of life) currently accounts for about two thirds o f infant deaths and almost half o f under- five deaths. More than halfthe neonatal deaths occur inthe first week, many on the first day o f life. This points clearly to the vital importance of safe delivery and post-natal care to improve child survival. The data conceal significant socioeconomic disparities. For instance, in 1997, children in the poorest households suffered 83% higher mortality than children in the richest households. The matemal mortality rate (MMR) is still high by all standards, with recent estimates rangingfrom 320-400 deaths per 100,000 live births. Malnutrition i s an underlying cause of many childhood deaths. Though nutritional status has improved over the years, still an estimated 35% of children are moderately and 13% severely underweight. The data show a wide economic differential in malnutrition, with children in the poorest households beingtwice as likely to be moderately malnourished, and four times as likely to be severely malnourished as children in the richest homes. Malnutrition has declined in all economic quintiles, but faster amongthe richest quintile. The number o f women dying from causes related to pregnancy and delivery i s around 17,000 each year. Thus roughly 45 young Bangladeshi women and nearly 1,000 under-five children die every day; almost all o f these deaths in women and children could be prevented if appropriate care services could be provided. Beyond childhood, a major cause o f morbidity and mortality i s tuberculosis. The number o f deaths due to the disease was estimated at about 70,000 in 2000-01. HIV/AIDS is an emerging challenge, still largely limited to certain high risk groups, but with significant factors which threaten its spread to the general population. Among non-communicable diseases, cancer and cardiovascular diseases are the leading causes o f morbidity and mortality. Projections show that as early as 2010, non-communicable diseases will increase their share as cause o f mortality to 59% from 40% in 1990. Injuries are expected to increase their share from 9% to 11%. Communicable diseases are expected to decrease their contribution to mortality inthe same time period fi-om 51% to 30%. However, set-backs in the control o f communicable diseases may occur iffull vigilance i s not maintained. Thus far, the government's focus inthe health sector has largely been on the establishment and operation o f facilities and services in the public sector. A majority of contacts between people seeking health care and providers, however, takes place in the private sector'. It would seem 1 The private sector i s defined to include all actors outsideof the govemment. This broad definition encompasses both the commercial sector and the not-for-profit sector, non-governmental organizations (YGOs),care providers with or without formal qualifications, practicing allopathy, homeopathy, ayurveda, or other systems of medxine, fachties o f various sizes, hospitals, chits, pharmacists and drug vendors, and suppliers o f health-sector related goods and services. 11 .. obvious that addressing the sector's problems requires the institution o f appropriate public policies to enhance the effectiveness o f the private sector's contribution to public health goals. Inorderto take full advantage ofthe potential-and address the challenges -ofworkingwiththe private sector, governments need to gain a better understanding o f private actors. Who are the private sector actors? What goods and services do they provide? What are the different kinds o f incentives influencing the behavior o f the private sector? In which areas i s the private sector well placed to complement public sector efforts? What are the most effective strategies for engaging the private sector? Although some evidence is emerging regarding these questions, there is a clear need for further information and knowledge. The objectives of this Private Sector Assessment (PSA) are to gain a better understanding of the private health care markets in Bangladesh and to identify areas for increased collaboration betweenthe government and the private sector. While the study analyzes private health care markets ingeneral, it uses maternal and child health (MCH) as an area o f special focus to illustrate general principles and/or draw lessons for the broader HNP sector. M C H was chosen for this emphasis in view o f its great importance to Bangladesh, and because M C H outcomes constitute a significant part o f the Millennium Development Goals (MDG). Conscious choices had to be made to limit the focus o f the study, as the private HNP sector is too complex to be covered in full detail under one study. Inparticular, the study does not include a comprehensive treatment of pharmaceuticals, medical equipment, hygiene products, infant feed formulations or other such health-related commodities, all o f which constitute important private sector inputs impacting on health. The broader context of the Analytical and Advisory Assistance (AM) This PSA is just one o f several studies initiated under the umbrella o f the HNP Policy Options AAA work. Notable among the other closely related studies are: a Labor Market Assessment, a Study o f Health Care Financing Options, a Decentralization Study, a Governance Study, a comparative analysis of efficiency among private and public providers, and a pilot design for pro-poor targeting. The AAA work i s a collaborative venture, with the active involvement o f the Government and various development partners, who are financing specific studies. Broad-based consultationhas been an integral part o fthe AAA process. Itis hopedthat this PSA would, along with other existing andnew evidence, inform future public policy discussions in the HNP sector o f Bangladesh, in the context o f the Poverty Reduction Strategy, currently under development. It i s also contributing to the preparation o f the HNP Sector Program (HNPSP), to be implemented in the next three years. The AAA benefited from the divisional level consultations conducted for the preparation o fHNPSP. The AAA i s particularly relevant to the new reform agenda being envisioned for the HNP sector by the Government, with a focus on areas such as public-private partnerships, rather than solely onthe improvement o fthe publiclyprovided services. ... 111 Main Findings TheProvision o fHNPServices in thePrivate Sector (Chapter I) 0 Private service delivery sector i s dominant 0 Among the private providers, the major proportion is those with low levels o f formal training (Alternative Private Practitioners), raising serious concems about quality o f care 0 Lack o f quality standards 0 Inequitable distribution o fhealth care 0 Lack o f competition between public andprivate providers The PSA analysis confirmed the findings from other studies that the private sector dominates the provision o f basic care, nursing homes, laboratory and ambulatory diagnostic services in Bangladesh. The public sector, however, remains the main provider o f inpatient care. Private sector providers are a heterogeneous group, differing in their training, legal status, system o f medicine used, type o f organization and on whether or not they held a public sector employment as well, Altemative private practitioners (APPs) are by far the largest group o fproviders. These include partially qualified or unqualified allopathic practitioners, drug vendors, and practitioners of non-allopathic or mixed systems o fmedicine. Interms o fhumanresources, the private sector predominates (innearly every category o fhealth professionals, a greater proportion o f them work in the private sector). However, Bangladesh has one o f the lowest nurse to population ratios in the world: 11nurses per 100 000 population, compared to 132 in low income countries, and 750 inhigh income countries. The shortages are accompanied paradoxically by a significant problem o f nurse unemployment because many private facilities make do with unqualified and unregistered nurses, which i s a cause for concern interms ofquality. Gender issues are very relevant when analyzing the human resources capacity in the private sector in Bangladesh. Other than traditional birth attendants and nurses, male private health providers far outnumber females: by about 4 to 1 among qualified doctors and by about 9 to 1 among APPs. This has a deleterious effect on women's access to care. Findings indicate important deficiencies in the technical quality of care delivered even by formally trained practitioners in both private and public sectors. For example only 10% o f private providers in hospitals used medical protocols to treat tuberculosis patients. In the absence o f appropriate mechanisms and institutions to promote quality o f care in a systematic manner it i s impossible to monitor and assure quality o f care; in such a scenario, the quality i s unlikely to be high. These problems are even greater inthe case o f APPs, the largest and least measured group o f providers in the country. Traditional providers' knowledge was particularly poor on M C H issues, for example, the management o f a newborn with pneumonia and the complications o f delivery. Consumption ofprivate HNPServices The following bullet points capture the main demand-side issues, raised inchapter 2: 0 Overall low consumption o f essential care 0 Poor populations demand / consume basic ambulatory care services from private providers; private services are not merely for the rich iv 0 Gender disparities inaccess 0 Financial barriers to access 0 Information and others barriers Overall health care consumption in Bangladesh in both public and private sectors is low compared with other countries and relative to need. The private sector is used for the overwhelming majority o f outpatient curative care, while the public sector i s used for a larger proportion o f hospital deliveries and preventive care. The higher proportion o f institutional deliveries in the public sector should be understood in the backdrop o f the fact that overall proportion o f institutional deliveries i s only 8%. About 90% o f medical care for children with acute respiratory infection (ARI)or diarrhea i s obtained from the private sector. This indicates the importance o f the private sector in terms o f access and signals the need for effective quality o f care measures. The dependence on the private sector for curative care i s also true for the poor in Bangladesh. The poorest 20 percent o f Bangladesh children have a higher dependence on the private sector for the management o f ARI and diarrhea than the richest quintile. The largest differences between the rich and the poor are for medically trained deliveries, antenatal care, treatment for ARI, and immunizations. In contrast, the use o f modem contraceptives and oral rehydration therapy for diarrhea, two commodities where there has been extensive social marketing, do not show such disparities between the rich and the poor. This suggests that social marketing may help inreducing some o f the inequities inthe consumption o f certain health-related commodities across income quintiles. Women and girls tend to receive less medical care than their male counterparts, with gender bias resulting from cultural norms that require women to obtain permissionprior to seeking medical care, andneeding to find someone to accompany them when they do. The situation is made worse bythe lack o f female healthproviders. Perceptions o f provider's experience and familiarity with the provider are important reasons for selecting private health providers. Further studies are needed to examine what specific factors influence care-seeking behaviors inBangladesh. While expectedly the richest quintile spends more than the poorest quintile (by a factor of 6) on health care, the proportion o f the spending that goes to the private sector is higher among the poor than among the rich. The private providers are often closer to the clients and more conveniently located than public facilities. coverage - public or private - appear to be major constraints to access to care for the poor. Financial barriers and lack o f basic insurance Serious efforts need to be made to address the financial, physical and social barriers to access, especially for the women and the poorer population groups. Operations research is needed to see how consumers can influence quality o f care -by being empoweredto demand better quality. Interaction betweenthe public and private sector in HNP Chapter 3 brings out the following findings: 0 Weak regulatory framework; ineffective enforcement 0 Several pilot initiatives include government-NGO partnerships, but most are donor- financed and have not been scaled up 0 MOHFWneeds the fiscal space andgreater and different kinds o fcapacity to enhance engagement with the private providers. V 0 Misperception o f size and scope o f private sector by government and lack o f capacity to play any role other than service provision. The range and magnitude o f government engagement with private providers i s not congruent with their importance. The bulk o f interaction takes place interms o fregulation, andwith regard to private clinics and hospitals. Less formal, less organized providers, such as non-allopathic practitioners, including traditional birth attendants, and drug vendors and retail pharmacists have very little interaction with government. Thus whatever little public-private engagement has occurred inBangladesh has mostly excludedthe providers o f greatest importance to the poor. There are, however, positive experiences in the area o f public-private engagement, including a number of pilot initiatives to work with private, mostly non-profit, service providers. Notably, the very successfil NGO-contracting experiences onnutrition andurbanprimary health care, and other forms o f partnership in areas o f family planning, TB control and immunization, bear important lessons for the rest o f the HNP sector. In addition, involvement of, and collaboration with, some private sector actors has occurred sometimes inpolicy discussions and formulation, though this has not been a consistent feature. Recently the government has been consideringthe possibility o f contracting NGOs to better manage several hundred public facilities at the union2 level with a view to expanding the coverage and improving the quality o f essential 3" services. Such an initiative i s a very welcome step inthe right direction and should be supported actively by the development partners, so that it quickly matures into a well-designed large-scale pilot with the potential for scaling up iffound successful. Secondly, the fulfillment o f government's stewardship responsibilities in the HNP sector could be enhanced. Health services regulation currently appears to be a fairly low priority issue. There i s little collaboration with professional and providers organization, nor support for self regulation. Currently professional and provider organizations are primarily playing the role o f trade unions. Neither consumer nor patients' organizations have yet emerged to play an advocacy role, nor to engage in monitoring o f service quality and outcomes. Instruments to engage private actors require government officials to perform tasks very distinct from their traditional activities. There is currently very little capacity to implement such instrumentsinthe MOHFW or inlocal government bodies. Thirdly, misperception and low capacity underlie weak public-private engagement. The policy- makers' interviews reveal that there is limitedunderstanding o fthe private sector size androle in provision o f care, especially for M C H services in rural areas. Most policy makers - especially those at the national level - believe that private providers mainly cater to tertiary care needs o f the rich inthe capital and other urban areas; while in fact, it i s the poor who are more dependent on the private providers, especiallythe APPs. A union i s the lowest administrative division with fixed public facilities for the provision of health care and consists o f around 20 villages, with around 20,000 population on average. 8-10 unions generally make a sub-district (Upazila or Thana), which has a population o f around 200,000 on average and 6-8 sub-districts make a district. There are about 4,700 unions, 470 Upazilas and 64 districts inBangladesh. vi Key issues Chapter 4 highlightsthe following conclusions: Public sector is not strategically usingthe scarce resources that are available inthe private sector The low level ofpublic expenditure on health care and the fact that all public spendingon health goes to public providers leaves little head room for contractingwith private providers Low level care provided byAPPs andpersistent shortages o f formally trained staff Uneven quality andproblems with access Poor needbetter capacity to make informed decisions about the quality o f care provided byprivate practitioners The poor are more likely to forego medical treatment due to financial constraints Otherbarriers prevent appropriate health seeking behavior Lack o f competition between public andprivate providers leads to inefficient use o f resources Poor coordination betweenpublic and private sector, and lack o f complementarity, which contributes to gaps incoverage PolicyImplications The central policy implication from these conclusions i s to revisit the role o f government in H",given the realities ofresource andcapacity constraints inthe public sector, the already dominant place held by private actors in the financing and delivery o f HNP services and the serious concerns about quality, access, accountability and governance with regard to both private and public services. A policy shift from an approach o f fixing the public sector problems to one of greater engagement with the private sector appears to be warranted. In particular the following three broad areas would appear to deserve priority ingovernment actions: -Under-consamption0fsemz;CeS thepoor and women -Service quality and outcomes -The knowledgebase PolicyOptions The following policy options were discussed during stakeholder consultations in early M a y 2003, and at the policy retreat and dissemination workshop in July 2003. These consultations included government officials, private sector actors, civil society, academia, and development partners. While there was broad agreement on the need to increase the engagement with the private sector and on the value o f the options presented here, it was felt that further debate, consultations, pilot tests and studies are neededbefore policy decisions are taken. a Develop a clear public policy towards the private sector that harness the valuable resources that are available inthis sector. 0 The government needs to create "head room" in its public expenditure envelope so that some public resources will become available for influencing the behavior o f private providers through contracting with private providers and subsidizing care for the poor. a BringAPPs into the service provider systembyworking with them instrengthening skills and increase the number o f formally trained staff through training. vii Increase quality benchmarking, performance based competitive pressures and incentives to attract private practitioners to work in low coverage areas in addition to traditional regulatory and quality assurancetechniques. Make information about the quality and price o f private providers readily available to consumers, especially for the poor. Introduce targeted subsidies and community level insurance for the poor and social insurance mechanisms for civil servants and formal sector workers. Use financial incentives (Le., fees for vaccinations) and social marketing techniques to overcome other barriersto appropriate health seekingbehavior. Increase competition between public and private sector through competitive and selective contracting andperformance benchmarking. Introduce internal markets (make public providers compete for public funding on a performance basis) and new public sector management techniques (ie. contracting out, contracting in, management contracts etc). Redefine the role o f the M O H and strengthen its core stewardship capacity in areas such as strategic planning, monitoring and evaluation, coordination, regulation, quality control and enforcement. It is clear that the PSA has not coveredthe whole ground on the subject and further studies and analytical work are needed. It must be stressed however, these suggestions for further studies are not to be misconstrued as a reason to delay policy actions for which considerable evidential basis already exists. A distinction must be made between operations research to pilot-test the policy options and the other research activities aiming to generate new knowledge. The Way Forward The authors do not wish to be prescriptive about the solutions for the issues emerging from their study. Rather, public policy should evolve through a participatory process in Bangladesh, with the active and broad-based involvement o f all stakeholders. Therefore, the policy options presented here - boththe "what" and the "how" - are merely a startingpoint for national debate. As the Government i s preparing its new Health, Nutrition and Population Sector Program (HNPSP), and has outlined a new reform agenda for the future, this study and the other related studies under the AAA work could meaningfully inform policy dialogue, taking a fresh perspective on sector reforms needed to achieve HNP outcomes as part o f the MDGs. In order to advance such national policy debate and enable the government of Bangladesh to better harness the potential o f the private sector for the achievement o f health outcomes, the following are possible next steps: Set up a Public-Private Task Force inthe MOHFW. 0 Create the necessary fiscal space or "head room" inthe public resource envelope. Capacity development inthe MOHFW to enhance its engagement with the private health sector. Participatory policy-making and more inclusive planning andprogramming. 0 Pilot activities to test the selected policy options. ... Vlll Completion o f other related studies (demand-side financing, May 2004 governance, pro-poor targeting, comparative study of efficiency) Disseminationo f existing evidence, multi-pronged communication Julyto December 2003 exercise, consultations across the country Development o fbroad-based HNP Policy Options July 2003 to June 2004 Initiation o fPilot Interventions (e.g., vouchers, micro-insurance, January 2004 results-based contracting with private sector, demand-side subsidies, pro-poor targeting) Chapter 1. The Provision of HNP Services in the Private Sector By far the main suppliers of clinical care to Bangladeshis are the alternative private practitioners (APPs), who includepartially qualiped or unqualiJied allopathic and non- allopathic practitioners, and village pharmacists. The best estimate is that APPs outnumber all qualiJied allopathic physicians by about 12:l. Bangladesh has one of the lowest nurse topopulation ratios in the world. The challenge is not only to increase their numbers, but to enhance the role and quality of nursing care. The widespread use of unqualijied workers as "nurses" is a serious causefor concern. Although physicians are also in short supply, they outnumber nurses by 1.7 to I, compared to an average ratio of 0.6 to I among low-income countries. The government invests relatively heavily in the education of physicians compared to other providers. Current plans to expand the numbers of physicians ought first to address the existing problems in theproduction and retention of physicians. Newfindings in this study show a signipcant "brain drain" of doctors migrating to high-income countries. Other than traditional birth attendants and nurses, male health providers far outnumber females: by 4 to 1 among qualijied doctors and by 9 to I among APPs, which results in signijicant gender disparities in access. New findings indicate that there are important deficiencies in the technical quality of care, especially though not only by APPs. Institutions and mechanisms topromote quality in the health sector are lacking, in the public as well as private sectors. There is little experience in Bangladesh with strategies to influence where or how private providers practice medicine. Not surprisingly, the largest group of providers, the APPs, is also the least influenced by currentpublic policies. Thegovernment could make the biggest gains in ensuring theprovision of health services by vigorously and systematically tackling quality issues, and by beginning large experiments to learn how to monitor and influence the most signiJcant group of health providers, the APPs. 1.1 Introduction This chapter deals with the issues related to the supply side o f private health services, specifically examining the health labor market3 and private health infrastructure. We begin by examining the different types o f private sector providers o f health services. Inthe last section o f the chapter, we outline what i s known about the types o f private health facilities, particularly the different types o finpatient andoutpatient facilities, pharmacies, andlaboratory services. This chapter does not fully deal with the supply o f health goods, notably the commodities such as contraceptives, bed-nets, oral re-hydration fluids, and infant formula and feeding supplements. Also, we do not examine other private key input markets such as those for manufacturers and distributors o fpharmaceuticals, medical equipment, and construction o fbuildings. Greater details on the labor market o f HNP sector inBangladesh can be found inthe related study on the health labor market, which was fiianced by CIDA. This chapter's treatment o f the labor market discusses key aspects about the supply o f services. 2 1.2 The Actors In this report, we categorize the health providers as qualified allopathic providers who are formally recognized through qualifications that allow them to be registered with a public agency (e.g. doctors, nurses, and trained paramedicals who practice "westem" medicine), and providers without formal training and recognized qualifications in allopathic medicine, whom we have labeled as alternative private practitioners (APP); dais OY traditional birthattendants (TBA) form a specialized type o f APP, but we have treated them as a separate group o f providers inview o f the largenumbers andthe very limitedscope o fservices they provide. The main sub-categories o fAPPs include: Non-qualified allopathic providers, who have not received formal education to practice allopathic medicine. Pharmacists who supply allopathic or other medicines, but who do not have formal qualifications to diagnose illness or prescribe medicines. Kabivaj, who practice an ayurvedic system o f medicine, based on ancient Hindi systems of medicine and commonly involving diet, herbs, and exercise. Some o f these providers have been formally trained inayurvedic colleges. Totka, who combine ayurvedic, unani (a traditional Muslim system o f medicine), and shamanistic systems. They may use allopathic medicines as well. Spiritual healers, who often rely on chants or sacred readings intheir treatment. Homeopaths, who follow the homeopathic system o f medicine, which involves treatment through minute quantities o f the presumed cause o f disease. In some cases, these providers have been formally trained and recognizedinhomeopathy. The distinction between the public and private sector providers is not always clear. A considerable proportion o f qualified allopathic doctors and other formal sector providers work entirely in the private sector (see below), but many o f the public sector doctors also practice privately -- either by working at private clinics and hospitals after public hours, or by charging private fees while practicing at public facilities. In this chapter, we consider this type of practitioner as first a public sector health worker, even if their participation inthe private sector dominates their time and income. Private providers are clearly a heterogeneous group, differing in their training, legal status, nature o f service, mix o f public and private practice, and type o f organization (Table 1.1). The consequence o f this is that some types o f providers are less visible for study or regulation, making it difficult to ascertain the true size or nature o f the private sector. This is particularly the case for those that are not formally trained and arepracticing illegally. - 3 Table 1.1-Visibility of Private Health Providers According to Selected Characteristics Characteristic More Visible Less Visible Legal status and Formally trained and Informal -untrained and illegal training operating legally (e.g. (e.g. shopkeepers, itinerant vendors) doctors, nurses pharmacists) Organization IIncorporated for-profit or Unincorporated solo practitioners non-profit organization Size o f facilities ILarge hospitals, networks Solo practitioners o f clinics Nature o f service Comprehensive clinical Single product or service (e.g. drugs) care Public-private mix Fulltime public practice or Illegal dual practice legal dual practice Alternative private 284,000 231 100% (Ali,77 473 2001) (Sarder& Chen, 1981) Allopathic providers 110,000 90 100% 38 145 (Claquin, 1981) (ORQ-MargQuest, 2000) ~ Traditional providers 173,000 50 239 Dais (TraditionalBirth Attendants) 119,000 Sources: Midrange estimates derived from weighting provideripopulationratios from Peters et al, 2003; Ali et al, 2001; ORQ-Marg Quest Ltd, 2000ab; BBS, 1998; Sarder and Chen, 1981; Claquin, 1981 Doctors and nurses are here considered as public sector providers ifthey are employed by the public sector, even if they also practice privately. This leads to an under-estimate of the proportion o f these categories in the private sector. The more appropriate approach would be to count dual practitioners under both the public and the private sectors. While that would be double-counting those providers, it would lead to a more accurate reflection o f the respective share o f the market held by the public and private sectors. 4 International comparisons on health personnel are fraught with difficulties, because o f the differences indefinitions, and the poor quality o f data. Compared with other countries, there are fewer qualified physicians and nurses per capita in Bangladesh (Table 1.3). Another striking finding is that the physician to nurse ratio is very highinBangladesh. The implications are that Bangladesh needs to consider ways o f dealing with the shortage o f physicians, but even more urgentlythe shortage ofpracticingnurses. Table 1.3 - International Comparisons of Physician and Nurse to Population Ratios (Around 1998) Country Physiciansper 100,000 population Nursesper 100,000 population Physician/Nurse Ratio Bangladesh 19 11 1.8 India 106 94 1.1 Pakistan 57 34 1.7 Sri Lanka 37 103 0.4 Global Average 146 334 0.4 Global Median 114 233 0.5 Low Income Countries 73 132 0.6 Middle Income Countries 142 278 0.5 High Income Countries 286 750 0.4 1.3 Physicians There i s no precise count o f the physicians actively practicing inBangladeshtoday. As of March 2003, the Bangladesh Medical and Dental Council (BMDC) reported a cumulative total of 34,541 physicians registered. But according to BMDC, until 2002 Bangladesh had 28,537 doctors (HRDU, 2003). However, a survey of their membership registered through the year 2000, returnedonly 9,988 responses (BMDC, 2003). In 1998 an HEUreport made reference to a study, based upon a 1993 census, reporting there to be approximately 13,200 practicing physicians with an additional 2,800 abroad and 2,000 unemployed. The same report also cited a Bureau o f Statistics Survey o f Professional and Miscellaneous Services Personnel, which said that there should have been22,356 doctors in 1993-1994based upon the same 1993 census. Registereddoctors work inboth the public and private sectors, but estimates o f those working only in the public sector are difficult to come by. In the B M D C survey, 27% of respondents reported they were only in the public sector (BMDC, 2003). However 39% reported mixed public employment with private practice. Another study reported that 69% o f practicing physicians were in the public sector, but the percentage of those who also had a private component was not given (HRDU,2003). The estimated percent o f those working exclusively in the private sector varies from 22% (BBS, 1998) to 31% (Health Economics Unit, 1998b) to 34% (BMDC, 2003). "Brain drain" i s a serious problem for Bangladesh, with many physicians leaving the country to work elsewhere (See Box 1.1). 5 Box 1.1-The BrainDraininthe HealthSector inBangladesh Ina background study examining graduatingMBBS classes from three Bangladeshmedical colleges, it was found that more than 20% of the 1975 graduating class had emigrated, comparednearly 28% of the 1985 cohort, with an annual loss of nearly 1.5% of doctors graduating between 1975 and 1995. The USA, Saudi Arabia, and the UK are the most common countries of emigration. Further details about the physiciansandtheir characteristicsmay be found inthe Labor Market Assessment (Peters et al, 2003). The total cost of losing a doctor cannot be easily measured. However, the costs of medical educationcan be reasonablydetermined. There are currently 13 medical collegesinthe public sector (of which five are new) and 20 nongovernmental medical colleges (HRDU, 2003). Estimates for the cost of medical education for an MBBS degree varies between establishedpublic schools, newly opened public schools and private schools. Based upon a weighted average of five establishedpublic medical colleges, it i s estimatedthat Tk 250,000 (US$4,000) i s spent over the average of 5.8 years it takes to produce anMBBS graduate (Health Economics Unit, 1998a). A comparable figure for a five-year program at Dinajpur Medical College, a new public school (including costs of establishing the school and annual recurring costs) i s Tk 491,000 (USSl0,OOO). Private schools are the most expensive with total costs per graduate estimatedto be between Tk 800,000 and 1,000,000 (US$l6,000-20,000) (HRDU,2003). 1.3.1 IncomeLncentives/Disincerztivesfor Physicians Estimates o f income of physicians vary by type o f practice and level o f experience. Ina survey o f private health service establishments, monthly income varied from Tk 7,500 to Tk 150,000 (US$ 150 to US$ 3,000) with the majority o f the practitioners earning incomes at the lower end. However, 95% o f those sampled also had some public component to their income (BBS, 1998). In a separate study o f public employees who also had a component o f private practice, 79% reported a monthly government salary between Tk 5,000-10,000 (US$ 100-200), 16% had salaries up to Tk 15,000 per month, and 5% earned Tk 5,000 or less. As for their private salary component, 19% earned less than their government salary, 21% earned an amount similar to their government salary, and 56% earned more. The overall average total income was Tk 27,500 (US$ 550) per month. In the absence of third party payments and reliable income tax information, these numbers are likely to be underestimates even though efforts were made to verify them (Gruen et al, 2002). It i s estimated that only 12% o f doctors are listed by the National Board o fRevenue, which i s the agency responsible for tax collection (UNB, 2002). Disincentives for private practice in the rural areas include lower purchasing power o f patients, competition with alternative private providers, and a weak infrastructure. Another concern i s harassment from local mafia-like structures (Gruen et al, 2002). Another difficulty with attracting doctors for rural areas i s that most physicians come from urban backgrounds. They are reluctant to give up that life and do not want their children to be deprived o f the opportunities available incities (Chaudhury and Hammer, 2002). Inthis study o f absenteeism, determinants o f government physicians' likelihood o f attending their rural postings included living in the same locality as that o fthe clinic, access to roads, and electrification. Inan extensive surveydone for the MOHFW on incentive schemes for public sector doctors and other health workers, the most frequently mentioned source o f dissatisfaction was "lack o f promotion", followed by "low salaries". Figure 1.3 shows reasons for discontent among medical practitioners: non-recognition o f good work was also a cause o f discontent, along with 6 inadequate residential and clinical facilities, lack o f access to quality health care for the physicians and their families, and physical and social insecurity (SRGB, 2002). 1.3.2 QualiQ of CareProvided by Physicians There i s little assessment or reporting on the quality o f physician care in Bangladesh, in the public as well as the private sector. A background study on private practitioners found that about 90% o f private hospitals maintained patient records, and nearly 60% used standard treatment protocols (HEU/IHE/NIPORT, 2003) (Figure 1.1). Nearly all facilities claimed to use safe disposable syringes. Ninety-eight percent o f private physicians appropriately recommended ORS for treatment o f acute diarrhea, but only two-thirds correctly identified the need to do a sputum test to assess a suspected case o f tuberculosis, and only 61% would correctly do an examination for a woman with post-partum bleeding. Though limited inscope, these findings show that there i s significant room to improve the technical quality o f care among private practitioners. In Chapter 2, quality o f care i s examined from the patient's perspective. Figure 1.1- Percentage of Private Hospitals using MedicalProtocols to Treat Patients I I C h i l d h o o d d i s e a s e s I I I I I I I I I I II M a t e r n i t y c a r e I I II I I S e x u a l l y t r a n s m i t t e d II I I I I d i s e a s e s I H I V I A I D S ,I I I I T u b e r c u lo si5 I I 1 I 0 % 5 % 1 0 % 1 5 % 2 0 % 2 5 % 3 0 % 3 5 % 4 0 % 4 5 % 5 0 ` Source: HEU/IHE/NIPORT, 2003 [Study commissioned specifically for this PSA] and authors' calculations 1.4 Nurses Eventhough the estimates on the number ofnurses working inBangladeshare not consistent, by any standard the number o f qualified nurses i s very low and the number actually employed i s even lower. As o f February 2003, the Bangladesh Nursing Council reported a total o f 19,066 nurses (93% female) in their registry. Distribution by year o f registration and sex i s shown in Figure 1.2. But based on a national census o f professional services personnel, there were just over 12,000 nurses working in 1993-94, a ratio o f about 11nurses per 100,000 population (CSIP 1995). Ninety-five percent o f the nurses work inhospitals and clinics inurban settings. 7 Figure 1.2 - Registration of Nurses by Year and Sex Registered Nurses by Year and Sex I2O0 I 1000 11 I Fl 1 .v, I 2cn 600 5 400 I Z I 200 I I 0 Source: BangladeshNursing Council, 2003 Data on immigration o f nurses from Bangladesh into other countries are not available, and a survey o f Bangladesh nursinggraduates did not produce meaningful results on the magnitude o f the problem. Giventhe longstanding global shortage o f nurses, it i s likely that many nurses that are qualified to work inother countries will emigrate if employment opportunities remainpoor in Bangladesh. However, nursing training facilities are inpoor condition, and the vast majority o f students and teachers are not proficient in English, which is a requirement for work in many other countries (Peters et al, 2003). The nursing market in Bangladesh is further undermined by a lack o f standards and regulation. To obtain work in private clinics, qualified nurses must compete with unregistered individuals who act as nurses, even though they may have minimal private training to give injections and provide unskilledcare to patients. The average monthly salary o f registered nurses i s Tk 8,700- 10,000 (US$200-230) inthe public sector, though unqualified nurses may work for much lower wages inprivate clinics (Begum, 1998). Low salaries and lack o f promotion appear as some o f the major concerns o f public sector nurses, who share many o f the same concerns about incentives as physicians (Figure 1.3). The challenges o f enhancing the role o f nursing, improving the quality o f nursing education, and better regulating the nursingmarket are at least as important as increasing the number of nurses in Bangladesh. In order to deal with a growing population and maintain the ratio with physicians, it i s estimated that by 2020 an additional 20,567 nurses would need to be trained. If one wanted to improve the doctor:nurse ratio to 1:1, an additional 45,649 nurses would be required (HRDU, 2003). Before accounting for losses due to retirement, emigration, and leave, this would requiremore than doubling the current output o fnurses. 8 Figure1.3 - Causes of DissatisfactionAmong DoctorsandNursesin the Public Sector Percent Ranking Either First or Second as Cause of , Dissatisfaction 40% 5 30% 2 20% I I I 10% I 0% I Cause of Dissatisfaction I I Source: SGRB, 2002 1.5 TrainedParamedicals There are many types o f paramedical professionals practicing in Bangladesh, though there i s little reliable information on their numbers or types o f practices. Most o f the paramedical schools require a grade 10 or secondary school certificate (SSC) to enter their three-year training programs. As can be seen from Table 1.4, the public sector employs some types o f specialized paramedical personnel trained specifically for public health functions, such as the Health or Family Planning Inspectors. On the other hand, the private sector has the dominant number o f laboratory technicians (94%), medical assistants (89%), and pharmacists (69%); though in each case, it i s doubtful that the private sector practitioners actually have the full educational qualifications as those employed in the public sector. The vast numbers o f private sector personnel in these categories suggest that there i s a considerable private market for the types o f services they can provide. Inthe case o f the medical assistants and pharmacists, they are likely also acting as physician providers, i.e. making diagnosis and prescribing treatment to patients. There i s very little information on paramedics. Basic questions concerning how paramedical professionals are trained or supervised, how they practice their professions, what their concerns are, or what contributions they are making to the health system have not been seriously addressed. Given the lower cost o f training, and the relative ease o f selecting paramedical trainees from rural populations, the main strategic issue for Bangladesh i s to consider whether paramedicals could take up more o f the responsibilities in the health care system. This question i s particularly relevant in considering options in remote areas where it i s difficult to get MBBS doctors and registered nurses to be stationed in either public or private sectors. 9 Table 1.4 -Estimatesof HealthWorkers inBangladeshOther than PhysiciansandNurses Provider Public Sector Private Sector I Medical Assistant I 2002 1996-97 5.598 I 45.603' I Pharmacists 7,622 1,789 Licensedpharmacists (without universityor technology 0 15,477 Laboratory Technicians 1,840 29,085 Radiographers 1,054 ? Health Inspectors 1,401 0 Family PlanningInspectors 4,110 0 Dentists (& Dental Surgeons) 1,740 1,247 Other trained paramedical6 3,574 ? Total 28.941 1.247 Sources: Public sector FromHRDData Sheet 2002 (HRDU, 2003). Figures do not include other para-professionals, including Family Welfare Assistants (22,350), Health Assistants (2 1,016), Assistant Health Inspectors (4,202), and Family Welfare Visitors (5,248). Private sector Most categories are from BBS (1998), except for pharmacists (ORQ, 2000), which are likely under-estimates. The estimates do not include another 45,820 health related workers estimated to be working in the private health sector, and i s intended to exclude those working as village doctors (BBS, 1998). There are also a large number o f auxiliary health workers who have shorter periods o f training and lower entry requirements than the paramedical workers considered above. Many o f these workers have been specifically trained for the public health workforce, including some 50,000 Family Welfare Assistants, Health Assistants, and Family Welfare Visitors, and an estimated 25,000 traditional birth attendants who have been involved in public sector programs (PRU, 2002). NGOs and government have also trained various types o f community health volunteers, such as BRAC's Shasthyo Shebika (Khan et al, 1998). While the drop-out rate of community volunteers can be quite high, their success seems to be dependent on carehl selection o f volunteers, involvement o f communities, supportive supervision, and good training (Islam et al, 2002; Khan et al, 1998; Amhold, 1979). In the case o f tuberculosis treatment, BRAC's use o f illiterate community volunteers turned out to be 50% more cost-effective than the comparison government program (Islam et al, 2002). Around the world, the recurrent lesson has been that it i s easier to initiate these programs than to sustain them (Walt, 1988). Further opportunities for usingcommunity volunteers inBangladesh should be considered only when it is clear that proper attention can be paid to the conditions that make community volunteers successful beyond an initialperiod, and where adequate monitoring can be sustained. 1.6 AlternativePrivate Practitioners It has long been recognized that APPs provide the majority o f health care in Bangladesh, particularly inrural areas (Claquin, 1981;Sarder and Chen, 1981). These providers are very well embedded into the culture and society o f villages (Bhuiya, 1992; Feldman, 1983; Ashraf et al, 1982; Leslie, 1976). APPs are also becoming more organized, forming their own professional 'Probably includes many healthproviders who do not have a three-year Medical Assistant training Other trained paramedicals includes Physiotherapists, Family Planning Officers andrelated professionals 10 associations in local areas. Despite the dominance of these types o f providers, relatively little is known about the actual number of the different types o f alternative private providers, the types o f practices they have, or how their behavior can be influenced. Other than a brief period when government sponsored the palli chikitshak training program in the early 1 9 8 0 ~policymakers ~ have largely ignoredthe informal sector. 1.61 Typesof Qualifications andPractices Formal training in traditional systems o f medicine now exists in Bangladesh, with govemment recognizing nine unanicolleges and six ayurvedic colleges with each havinga four-year diploma course (BBS, 2002ab). However, the vast majority o f APPs practicing in Bangladesh have not received formal education in their system of medicine, though a substantial proportion have received some semi-fonnal training. For example, in a study in Brahmanpura in 2000, 61% o f APPs had some kind o f certificate o f health training (ORG-Marg Quest Ltd, 2000b). Most traditional providers have hadtrainingthrough apprenticeship (Feldman, 1983; Sarder and Chen, 1981). The general education levels among the APPs tend to be higher in the allopathic providers than in the traditional practitioners, with the majority o f allopathic providers having completedjunior secondary school, and many having completed high school (ORQ-Marg Quest Ltd, 2000ab; Bhuiya, 1992; Sarder and Chen 1981). The gender distribution o f private providers, which has important consequences for women's access to health care, i s described in Box 1.2. Box 1.2 - Gender Distributionof Private Providers With the exception o fthe traditional midwives, who are female, the APPs are largely male, particularly the allopathic practitioners. In 1976-77, a nationwide survey found that 99% o f alternative providers were male (excluding traditional midwives). Ina census taken inMatlab thana in 1978, similar proportions o funqualified allopathic and homeopathic providers were male, whereas the kabiruj and totkas were more evenly distributedbetweenmale and female (Sarder and Chen, 1981). Inthe Brahmanpara study mentioned above, 98% o f village doctors were male (ORG-Marg Quest Ltd, 2000b). The obvious implication o f such a male dominance o fproviders ina traditional m a l society i s that it i s more difficult for women to access health care even through the APPs. This issue i s discussed further in Chapter 2 where the demand for and use o f health services are considered. Most o f what i s known about the practice patterns o f APPs comes from small area studies, making it difficult to generalize results to the entire country. APPs are known to provide services for a wide range o f health conditions, though almost exclusively on an outpatient basis. Village doctors nearly always sell medicines (94% o f village doctors in Brahmanpara; 95% o f allopaths in Matlab), and the majority provides dressings. Kabivaj and totka providers are less likely to sell medicines, and many o f the spiritual healers provide a much narrower range o f services for a more limited set o f conditions. Some providers, such as bonesetters, provide a very specific set o f services. Similarly, most traditional birthattendants tend to provide services only for childbirth. Most APPs provide services close to their population base. The allopaths and homeopaths tend to have a small building for their business in a local market or in the village (ORG-Marg Quest Ltd, 2000b; Bhuiya, 1992), as do those who are pharmacists (ORG-Marg Quest, 2000a). Traditional healers tend to operate fi-om their homes, and are more likely to make home visits (Bhuiya 1992). Allopaths are more likely to work on a full-time basis than other providers 11 (ORG-Marg Quest Ltd, 2000b; Sarder and Chen, 1983). InClaquin's (1981) study, only 35 YOof APPs worked on a fulltime basis, seeing an average o f 17 to 52 patients per week. In Brahmanpara, the village doctors nearly all worked fulltime, and averaged 18 patients a day (ORQ 2000). Incontrast to these providers, traditional midwives averaged only two patients a week (Claquin, 1981). 1.62 Incentive and Disincentives of APPs There are a few reports on the incomes and fees charged by APPs, though both appear to be generally modest. Claquin (1981) reported that APPs charged between two to four Taka for consultation fees on average, which was equivalent to $0.12 to $0.25 at the time o f the survey. However, payments from medicines could average 10 times this amount, with unqualified allopaths charging more than ayurvedics, followed by homeopaths and spiritual healers. In Brahmanpara, village doctors claimed to eam about Tk. 1,600 (US$ 32) per month from prescribing medicines to their patients, and another Tk. 2,200 (US$ 44) per month through other pharmacy sales (ORG-Marg Quest Ltd, 2000b). There i s little other information on the motivations o f APPs, nor about their aspirations or practice constraints. The Brahmanpara study reported a near unanimous interest among village doctors in working in partnership with the public sector (ORG-Marg Quest Ltd, 2000b). Although there have been a number o f projects that attempted to train APPs, these are not well documented, and little i s known about what strategies would be effective in improving their skills, preventing them from practicing illegally, or reducing the potential harm caused by poor quality services. 1.63 Quality of ServicesProvided by Alternative Private Practitioners Poor quality o f services is one o f the major concems regarding informal sector providers. In a detailed anthropological study on village practitioners in three villages, Ashraf and colleagues (1983) emphasized that a large variety o f drugs are being prescribed, usually inappropriately. Because o f a fatalistic attitude o f the villagers, the health providers are rarely blamed or held accountable for poor practice. Bhuiya (1992) pointed out that understanding o f the causes o f diarrhea was quite limited among providers, and that only 60% o f the providers used oral rehydration solution (ORS) for treatment, despite working in an area where ORS had been actively promoted for decades. H e also pointed out that the allopathic providers were more likely to provide appropriate care for diarrhea than the traditional ones. Ina similar vein, Ali and colleagues (2001) found that access to allopathic providers (both qualified and unqualified) was related to lower childhood deaths due to pneumonia, whereas access to traditional providers was related to higher mortality. In a background survey for this study, different types of private health providers were asked about how they would handle specific medical conditions (Figure 1.4) (HEU/IHE/NIPORT, 2003). With one exception (a newbom with pneumonia should be treated with antibiotics at a hospital, may not require a referral from a doctor who practices at a hospital), the correct medical care would include an affirmative answer to each o f the responses shown at the bottom o f the figure. The results indicate a very low level o f quality of care by the APPs in absolute levels, and in comparison to private hospital doctors. Among the APPs, the allopathic drug vendors performed better than the homeopaths and traditional providers in cases of childhood diarrhea 12 and newborn pneumonia, but similarly badly for the appropriate investigations for suspected TB or management of a women with post-partum bleeding. Figure 1.4 - Percent of Different Types of Providers Offering Appropriate Medical Care for SpecificMedicalConditions 120% 100% 80% 60% 40% 20% 0% ORS for childhood Referralfor newborn Sputum test for PhysicaVpelvic exam Ergometrinefor post- diarrhea pneumonia suspected TB for post-partum partum bleeding bleeding 1 Hospitaldoctors Allopathic drug vendors 0 Homeopaths 0Traditional practitioners 1 Source: HEU/IHE/NIPORT, 2003 and authors' calculations 1.7 HealthFacilities Information on the numbers and types o fprivate health facilities inBangladesh i s quite weak. In the public sector, there are precise counts on the locations and types o f health infrastructure, though little i s reported on how functional they are (Le. staffed, equipped, and seeing a full complement o f patients). Because most private practitioners operate out o f small clinics or their homes, these types o f doctor's chambers are hard to enumerate, monitor, or regulate. The last census o f private health facilities was undertaken in 1997-98 (BBS, 1998). It focused only on facilities that had inpatient beds or provided laboratory services (Table 1.5). The data suggest that government inpatient facilities comprise about half (51 %) o f the total inpatient facilities, but a much larger majority o f the hospital beds (72%). Just examining the number o f registered private facilities would also seriously underestimate the total number o f private facilities: only about 70% o f the private inpatient clinics that were enumerated had been registered with the government. The majority o f the other facilities had no type o f registration, though some had obtained various types o f licenses from local authorities. Inanother analysis o f 13 private clinics and hospitals in Bangladesh, the Health Economics Unit (1998a) found that only 27% o fthe 252 clinics sampled hadbeen registered. Table 1.5 -Estimated Number of Government and Private Health Facilities in 1997-98 Source: BBS, 1998. The Health Economics Unit extrapolated from their sample to estimate that there were 584 private hospitals nationally (HEU, 1998a), compared to 613 reported here. NA not available - laboratories inthe - public sector are nearly always part ofpublic hospitals The largest gap concerning private health facilities is the absence o f estimates on private outpatient clinics (or doctor's chambers). As noted above, APPs tend to work out o f a building, often their own home or place in the market. In many cases, it may be difficult to distinguish their places o f practice from a pharmacy. But many public sector physicians and private MBBS doctors also have outpatient clinics, about which very little i s known. Giventhe limited information on even the number of health facilities inBangladesh, particularly for outpatient clinics, it i s obvious that there will be even less information available about their quality, efficiency, or pricing. The HEU study (1998a) on 252 private medical hospitals provides a rare insight into their operation. Bedoccupancy rates averaged only 56%, rangingfrom 12.5% to 97.5%. The average length o f stay was 5 days, compared to about 7.2 days in the comparison government owned hospital. However, information on the case-mix o f patients i s not known, so the value o f comparison i s limited. The study estimated that average return to capital was 38%, suggesting that private hospitals were making profits much larger than would be available in other sectors. The HEU study concluded that because facilities with a bed size o f 11-20 beds had the lowest average costs, this was resultingina large number o fprivate facilities o fthis size. A background study o f private hospital and clinic managers reported on some o f their main constraints (HEU/IHE/NIPORT, 2003). As shown in Figure 1.5, their main concerns are with motivating and recruiting qualified staff, problems with infrastructure, ability to purchase supplies, and uncertainty over government policies and regulations. Problems with electricity supply were universally reported as the main infrastructure problem. Incontrast, issues related to obtaining credit, or with having a low demand are far less o f a concern to them. It may be that barriers to entry related to obtaining high quality personnel and supplies, and the difficulty o f managing infrastructure problems, allows the existing private hospitals to make substantial 'Includes facilities that have obtained licenses from local authorities, comprising 8% o fthe total Nursing homes have inpatient services but no outpatient services 14 profits in spite o f the relatively high number o f beds unoccupied. Caution i s needed in interpreting these findings, as the surveywas conducted on existing facilities. Ifit were possible to study cases inwhich private hospitals were sought to be established by investors but failed to materialize, or had started but failed thereafter, the list o fbarriers to entry mightbe different. Figure1.5 -PrivateHospitalManagers' Opinionson the BusinessEnvironment 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Difficult to motivate staff to do well Problems of infrastruct Difficult getting qualified s Availability of suppl Regulations (non-bank Uncertainty about govt poli Availability or cost of spare p Getting space for hospital-cl Taxes and tax admini Too little demand for s Problems with getting Level of interest rates to Too little financing from Gove Problems with money transfer Source: HEU/IHE/NIPORT, 2003 and authors' calculations Turning to laboratory facilities, it i s clear from the number o f people working as laboratory technicians in the private sector (over 29,000 in 1996-97) that the number o f registered laboratories under-estimates the number o f facilities providing laboratory services by a large margin. Given that the average laboratory on the average employs somewhat less than two laboratory technicians, there would be about 16,000 laboratories inBangladesh, compared to the 1,042 identified inthe 1998 survey (BBS, 1998). Data from the private health service establishment survey found that on average, laboratories performed about 28 tests per day, and had an annual operating profit (total income - total expenditure) o f Tk. 404,000 (US$ 8,600) (BBS, 1998). The market i s currently allowing for highprofit margins on nearly all types o ftests. The operatingprofits onnineteen laboratory tests examined in the same study was 175%, ranging from a low o f 45% for plain x-rays o f the abdomen, to 400% for mammography. Overall, the largest profits were being derived from blood sugar tests and hemoglobin examinations, and the least on x-ray procedures. It i s possible that equipment costs and technical ability to operate the machines are creating the conditions where laboratories can charge highprices relative to their costs. 15 In contrast to provider and hospital medical services, the quality of laboratory services is relatively more measurable, using standard sample selection techniques andreference laboratory standards. Quality assurance procedures should be easier to establish to assess the quality o f laboratory test results. However, systems to monitor the quality o f laboratory services do not exist inBangladesh for all but a few laboratories that do so on their own, and largely for research purposes. The result i s that the validity o f most laboratory test results i s unknown. 1.8 Other PrivateHNPInputMarkets Whereas a detailed assessment o f the other input markets i s beyond the scope o f this study, it i s important to recognize the scale and influence o f these components o f the private health sector. For example, inthe pharmaceuticals market, there were 767 licensed drugmanufacturingunits in 1997, 1,353 drug wholesale trading firms, and 33,975 retail trading firms (BBS, 2002a). Locally produced drugs were valued at US$ 312 million in 1997, accounting for 1.0% o f GDP. About 40% o f non-clinical family planning methods were distributed through private pharmacies in 1989 (BBS, 2002a). Spurious drugs are a serious problem. The Social Marketing Company (SMC) is the dominant player insocial marketing inBangladesh and distributes health related products to 200,000 retail outlets. SMC distributes 170 million condoms (70% o f estimated use in Bangladesh) and 37 million cycles o f oral pills (30% o f estimated use) annually. In addition to reproductive health products, it distributes 105 million sachets of oral rehydration solution (ORs), which accounts for about 60% o f estimated use (SMC, 2003). ORS was first marketed inBangladeshin 1986, andhas been an importantpart o f health initiatives to reduce child mortality from diarrhea. The ORS market in Bangladesh has grown substantially through consumer education and advertising, and now encompasses 15 brands. In addition to providing ORs, SMC also provides training to ApPs on diarrhea management. SMC i s considering to market zinc as a supplementary dose to persons with diarrhea, andbelieves that there is social marketingpotential for safe deliverykits (SMC, 2003). Currently SMC i s subsidizedby the government, which provides the commodities free o f charge for distribution, and SMC charges the consumer a nominal price. SMC i s 40% self-financed. Bed-nets, another important health commodity, are imported and sold in the private sector. However, the market for bed-nets i s currently not very large. Soap i s currently too expensive for the poor population, who mainly uses ash for hand washing (SMC, 2003). More detailed studies are requiredto understand the private markets for goods and commodities relevant to HNP, e.g., bed-nets, infant food formulas, hygiene products, pharmaceuticals, vaccines, and medical equipment. 1.9 Conclusions There are clearly large information gaps concerning the supply o f private health services in Bangladesh. However, enough i s already known to conclude that the biggest portion o f health providers i s also the most neglectedbypublic policy-makers. For policy-makers, there are a number o f key recurring themes that run through this analysis o f the private health sector. One o f these themes is the challenge to improve the quality o f care. There i s no single best way to address this in Bangladesh, but what i s known from experience elsewhere is that high level leadership and commitment i s needed. Also critical i s active 16 participation from key stakeholders, including provider groups, government agencies, community and consumer representatives, and often some independent monitoring agencies. In the short term, some o f the key steps may include pulling together leaders, testing quality improvement tools, and assessing performance and developing standards and benchmarks for different types o f providers. Inthe medium to long term, credible institutions will be needed to promote professional self-regulation, consumer protection, targeted regulation, and use o f payment mechanisms and information disclosure techniques to continually improve health services performance. Another urgent issue i s to correct the imbalances in the production o f health personnel. There are too few nurses and doctors, and an over-abundance o f unqualified providers, over whom the government has little influence. There are also too few female health care providers. Ambitious plans are proposed to increase the number o f physicians, but given the resource constraints, alternative sources for the production o f qualified medical care need to be considered. A number o f strategies could be pursuedto try to retainhealthprofessionals inthe country. One relatively simple measure to reduce losses to emigration i s to demand reimbursement or a bond covering the costs o f medical expenses from physicians who obtain visas to work overseas. Developing more nurses andparamedicals that could take on more o f the primarycare diagnostic and therapeutic responsibilities that physicians now assume i s likely to be a more cost-effective strategy than focusing on training more physicians. This could be coupled with admission strategies that target rural communities. Engaging with the existing APPs also has potential to reduce the needs for new formally trained healthpersonnel. Finally, the question o f private practice among public practitioners i s a persistent and pervasive issue that affects the credibility o f government health services. Partial approaches to this issue, such as formalizing private fees in government hospitals are likely to exacerbate the problem. To help the government in taking steps further work i s needed to test the feasibility o f different options to improve the governance o f public hospitals, change labor relations, develop contracting o fprivate physicians, and strengtheningmonitoring andreporting o f hospitals. 17 Chapter 2. Demandfor / Consumptionof PrivateSector HNP Services Overall health service consumption in Bangladesh is low in comparison to other countries and to levels of need. The use of maternity services is particularly low: e.g. only 8% of deliveries occur in a healthfacility. The private sector is used for the overwhelming majority of outpatient curative care, while the public sector is used for a larger proportion of hospital deliveries and preventive care. The dependence on the private sector for curative care is even more true for the poor in Bangladesh; the poorest quintile of Bangladeshi children have a higher dependenceon theprivate sectorfor acute respiratory infection and diarrhea care than the richest quintile. Poverty is a signijkant constraint to health care access and hence, use. The largest differences between the rich and thepoor arefor medically trained deliveries, antenatal care, treatmentfor acute respiratory infection, and immunization. Women and girls tend to receive less medical care than males, with gender bias resultingfrom cultural factors and the relative lack of female health providers. New analyses show that nearly all private health spending is at privatefacilities (88%). I n absolute terms, the richest quintile spends about six times as much as the poorest quintile on health care, presumably purchasing a higher quality of health service. The cost of health care often results in foregone medical treatment. The cost of drugs and transport, and distance to the provider are some of the most important barriers to health care. Theprivate providers are generally closer and more conveniently located than public facilities. New studies show that perceptions of provider's experience and familiarity with the provider are also important reasonsfor selecting private providers. Further studies are needed on determinants of care seeking behaviors. More attention is needed to overcome the social,financial, and physical barriers to care. There are indications that social marketing can help to reduce some of the disparities in use of services between the rich andpoor. Further experimentation is needed to see how consumers can influence the quality of care, how health decisions in the home can be improved, and how thefinancial impact of health care costs can be reduced. 2.1 Introduction This chapter addresses the issues related to the demand for health services. The chapter begins by an analysis o f consumption levels for various types of health services, and the differences between public and private sector consumption. We then tum to the question o f barriers to access, andwhy people choose to use certain health providers. 2.2 HealthService Consumption Although data on overall outpatient clinic visits or hospital utilization are not available in Bangladesh, it i s possible to use Demographic and Health Survey (DHS) data to compare selected services with other countries (Table 2.1). These data suggest that consumption o f maternal and child health services in Bangladesh i s quite low. Bangladesh has lower rates of institutional deliveries and use o f medical services for antenatal care and treatment o f childhood 18 diarrhea. Among 45 countries with comparable DHS data, Bangladesh has the highest rate o f home delivery (World Bank, 2003). Hospital deliveries and births attended by a medically trained person are remarkably low in Bangladesh - only 5% o f all deliveries were in a health facility in 1996/97, compared to an average of 49% for all other developing countries. By the time o f the 2000-01 DHS survey, only 8% o f deliveries occurred in a health facility, and a medically trained person attended 22% of all deliveries. In contrast, untrained traditional birth attendants were at 54% o f deliveries, and a relative (or no person) was at 24% o f deliveries (NPORT et al, 2001). Table 2.1 InternationalComparisonsof HealthService Consumption Average for 45 Indicator Bangladesh India Nepal Pakistan developing countries % of children with diarrhea seen 22 61 14 48 39 medically % of childrenreceiving full 54 35 43 35 51 immunization schedule % ofbirths where antenatal care is from 26 49 38 26 71 a medically trained person Of all deliveries, infacility 5 26 8 15 49 Of all deliveries, % attended by 8 34 10 19 53 medically trained person 19 Ensor et al (2002) found that overall levels o f per capita consumption o f the essential service package (ESP) - which i s targetedtoward the poor, women, and young children - would have to increase by 40% in order to achieve desired consumption levels. Consumption o f the child health component o f the ESP would have to increase by 12% to reach desired levels. The same study found that consumption of maternal health services was particularly low relative to desirable levels; consumption would have to increase by 122% to reach the desired per capita consumption level. Two critical questions remain: (1) how can demand for the essential service package be raised to match the healthneeds? (2) can access to quality essential health services be met by relyingentirely on the public sector, or should the much larger private sector be used? 2.3 PublidPrivate Shares of Services Bangladeshis are much more likely to use private providers than public providers for most o f their ambulatory care services. In a background study on household use of health services conducted in2003, the preliminaryresults show that o f those who sought care outside the home for an illness, 87% o f urban residents used private providers, compared to 75% o f rural residents (HEU/IHE/NIPORT, 2003). Private allopathic providers were used as the first source o f care for 32% of the cases, while pharmacies were used 26% of the time. Another 10% of the cases were seen at private hospitals, private non-allopathic providers saw 9%, and the public sector was used 21% of the time. The qualified private providers were used twice as often as unqualified allopathic providers (excluding pharmacies). Since the unqualified allopathic providers vastly outnumber the qualified providers, and since the outpatient workloads between the two appear comparable (Chapter l), it is likely that the public are more likely to view their allopathic provider as qualified to practice medicine, even when they do not have an MBBS degree. Other surveys show that consumption of outpatient curative services are largely in the private sector. The 2000 Service Delivery Survey (SDS) found that the public sector accounted for only 21% o f visits in the last month, whereas NGO and for-profit providers covered 30%, and alternative private practitioners (APPs) had 49% o f the visits (CIET Canada and MOHFW, 2001). The same survey also found that visits to a private provider were more likely to be for curative care (90%), compared to visits to a public provider (71%). Another study reported that 75% of the first point o f contact for care occurs in the private sector (Sen, 2001). Levin et al (2001) also found that people used primarily village doctors and traditional practitioners for health care, and together they accounted for almost two-thirds of the care utilized. A study o f infant mortality and health seeking behavior in a rural area of Bangladesh found that 90% of parents sought treatment for their sick children from private providers (Bhardwaj and Paul, 1986). The same study reported that 53% o f sick children received treatment from ApPs; the majority o fthese were kabiraj, followed by non-qualified allopathic providers. International comparisons o f the distribution o f the use of public and private health services are difficult to come by, but data from comparable DHS surveys provide some insights. These surveys show that Bangladesh stands out for its high level o f dependence on the private sector for treatment o f childhood diarrhea, ARI, and institutional deliveries (Figure 2.1). Although other South Asian countries also have a high dependence on the private sector for childhood curative services, Bangladesh leads the region, and i s second only to Haiti among all 45 countries with comparable data. O f children with diarrhea inBangladesh, 22% were brought to a health facility. O f those children, 92% were seen in a private sector health facility. Of children 20 with ARI, 33% were brought to a health facility. O f those children, a private provider saw 89% (Gwatkin et al, 2000). Bangladesh's position i s even more striking when making intemational comparisons o f obstetric services. The most significant point i s that Bangladesh leads the world indeliveries that occur at home - about 91% of all deliveries inthe three years preceding the1999-2000 DHS (NIPORT et al, 2001). Of those few deliveries that occur in a health facility, Bangladesh still has a substantial share occurring in private facilities (40%). Public facilities provided more treatment for obstetric complications (73%) than private qualified allopathic providers (27%), yet private providers performed a higher proportion o f caesarean sections (56%) than public providers (44%) (ACPR and UNICEF, 2001). The higher proportion o f caesarean sections in the private sector may suggest that there i s a problem with supplier-induced demand: patients using private providers have higher caesarian section rates than would be neededbecause these providers have an incentive to give more services, a common issue where the private sector i s poorly regulated (Peters, 2002). The use of private providers is not restrictedto the rich, as the poor also have a high dependence on private providers for childhood curative care (Figure 2.2). However, as noted above, there may be differences in the type o f private provider used, with the poor more likely to rely on unqualified practitioners, and the better off able to afford qualified physicians. Figure 2.1 - International Comparison of the Share of Medical Care Used in the Private Sector For Selected Services 10 0 Bangladesh India Nepal Pakistan Average for 45 developing countries Source: World Bank, 2003; based on DHS inthe 1990s (1996197 for Bangladesh) 21 Figure 2.2 - Proportionof Richest and Poorest QuintilesUsingPublic and Private Health Services Institutional Deliveries richest 20% Institutional Deliveries poorest 20% Childhood ARI treatment richest 20% Childhood ARI treatment poorest 20% Childhood diarrhea treatment richest 20% Childhood diarrhea treatment poorest 20% I I I I I I II I II I 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Source: World Bank, 2003; based o n Demographic and Health Surveys inthe 1990s (1996197 for Bangladesh) Figure 2.3 - InstitutionalDeliveries in Private and Public Facilities in Bangladesh and 44 OtherDevelopingCountriesAccordingto Level of Wealth 70 I 60 10 0 Poorest 20% 20-40% 40-60% 60-80% Richest 20% 1~ I Source: World Bank, 2003; based on Demographic and Health Surveys inthe 1990s (1996197 for Bangladesh) 22 Figure 2.3 shows that there are enormous differences inthe use o fprivate institutional deliveries between the richest quintile (8.7% o f deliveries) and poorest quintile (0.1% o f deliveries) in Bangladesh; a rkhipoor ratio of 87. O f all countries with comparable data, Bangladesh has the greatest inequality in private institutional deliveries, but also ranks behind only Nepal and Pakistan in terms of inequality of public sector institutional deliveries (World Bank, 2003). In sum, women have very little access to any institutional delivery care inBangladesh, and the rich predominantly consume the little that i s usedinboth public and private sectors. 2.4 Determinantsof Consumption/ HealthCare SeekingBehavior Inthis section, we summarize the data available inBangladesh on determinants ofhealthservice use, focusing on physical, financial, and social barriers, and users' perceptions o fhealth services. In a background household survey about health services use/consumption (HEU/IHE/NIPORT, 2003), the preliminary results on 342 children who received care for an illness showed a familiar dependence on private providers: 43% were seen by a qualified private physician, 30% by an APP, and 27% were seen at a government facility. The main reasons for their choices are illustrated inFigure 2.4. These results suggest that knowledge about the provider's experience i s particularly important in choosing private practitioners, whereas cost and physical proximity are more important reasons for those who chose a public facility. For those who chose an APP, the second most common reason was familiarity with the provider. The provider's behavior, and the availability of medicines were less frequent reasons for selecting a particular provider. Figure 2.4 - Main Reason for Choosing a Specific Provider for Treatment of the Most RecentChildIllness Experience of provioer c o s t Proximity Fam iliantywith provider Behaviour of provider I Availabilityof medicine 0 10 20 30 40 50 60 70 Percent Source: HEUIIHEINIPORT,2003 23 2.4.1 PhysicalAccess Numerous studies have shown that physical access to health services i s an important determinant o f consumption in Bangladesh. Location is one o f the most important factors to determine the access to health services in Bangladesh as documented in the CIET baseline survey (CIET Canada and MOHFW, 1999). Geographic access at least partially explains why consumption rates are higher inurban areas compared to rural areas (NIPORT et al, 2001). Physical access i s a barrier to maternal and child health services inparticular. Inthe 1999-2000 DHS, 79% o f women reported that the lack of a health facility nearby was a constraint to consumption (Streatfield et al, 2001). Inthe same survey, 50% o f women respondedthat getting to the health facility was a problem to them. Levin and colleagues (2001) confirmed the significant negative association betweenboth distance to the provider and travel time and the use o f health services. A child was less likely to be taken to a qualified allopathic provider or a traditional practitioner than a village doctor, ifthe travel time was 40 minute or greater compared with travel time o f 15 minutes or less. Other research has shown that a majority (74%) o f sick children in a rural area o f Bangladesh were taken less than two miles for treatment; and that a majority o f those children were seen by APPs. In contrast, children who were taken more than two miles for treatment received health care from qualified allopathic providers (Bhardwaj and Paul, 1986). These findings strongly suggest that the distance from the household to the health provider i s an important factor in determining whether or not to use a health provider. Because o f the importance o f physical access, the large number and close proximity o f APPs to most rural Bangladeshis give them an advantage over other providers inthe public andprivate sectors. 2.4.2 Financia1Access The cost o f health care can be a strong determinant of health care use, as well as a cause o f poverty. Ability to pay i s a particularly important determinant o f access when a highproportion o f health care is financed privately, and without any type o f financial risk protection from health insurance. In Bangladesh, 60% o f total health expenditure in 2000 was in the form o f out-of- pocket payments by individuals (64% o f total health expenditure was from private sources), so that households' ability to pay for care is important (WHO, 2003a). There i s essentially no social security or private health insurance, although public hospitals are intended to provide a form o f insurance incase o f serious illness. Different types o f cost items can be barriers to the use of health care. Health care costs can be divided among direct medical costs (e.g. medicines and service fees), direct non-medical costs (e.g. transportation costs) and indirect costs (e.g. traveling and waiting time, lost earnings). In the SDS, the cost o f medicines was the most important cost element that prevented people from using health services, followed by transportation costs (CIET Canada and MOHFW, 2001). Even ifthe cost o f medicines could be reduced, such as through non-profit organizations that are able to purchase drugs in bulk and distribute essential drugs more efficiently, other approaches may be required to deal with patient transport costs, and with gender bias inhousehold decision- making. A background study analyzing the BangladeshHousehold Income Expenditure Survey (Peters et al, 2003), not surprisingly, found that 88% of private health expenditures are for services from 24 private providers, and that the rich pay far more for medical care than the poor (the richestlpoorest quintile ratio was 6.2) (Figure 2.5). As a proportion o f total income, health spending by households was found to comprise 6.8% o f their total income overall, with the richer groups spendinga slightly higher proportionthan poorer groups. It is noteworthy that even though the poorer people spend less per episode o f illness, a greater proportion o f that expenditure goes to private providers. Thus the poor are in fact more dependent on private care, contrary to a common beliefthat private care is for the rich. Figure 2.5 - Household Per Capita Payments to Private and Public Health Providers in Bangladeshby IncomeQuintile 1407 Poorest 2040% 40.60% 60-80% Richest 20% Income Quintiles Source: BBS, 2001 and authors' calculations Total household payments to health providers i s estimated according to the results o f the HIES (2000) in 1.5 billion dollars which includes fees, hospitalization expenses, medicines, laboratory and tests, transportation costs, tips and others charges. Gwatkin et a1 (2000) disaggregated the Bangladesh 1996-97 DHS data by wealth quintiles based on assets o f the household, and then reported the use o f health services (Table 2.2). The largest differences between rich and poor were found for medically trained deliveries, antenatal care, treatment for ARI, and absence o f any childhood immunizations. For example, they found that the poorest 20% o f children with ARI are less likely (23%) than richest (51%) to be taken to any medical facility for treatment. There were far smaller differences in the treatment o f childhood diarrhea and use o f contraceptives. This may be due to the fact that in Bangladesh there have been decades o f social marketing and public programs to support family planning and oral rehydration therapy, and to some extent childhood immunizations. 25 Table 2.2 - Wealth Differences in Health Service Consumption Indicator Poorest Richest Population Rich/Poor 20% 20% Average Ratio Proportion o f surviving children (12-23 months) who received all vaccinations (BCG, DPT, oral 47% 67% 54% 1.4 polio, measles) Proportion o f survivingchildren (12-23 months) who receivedno vaccinations 18% 5% 12% 0.3 Proportion o f children (under five) with diarrhea inlasttwo weeks who receivedORT 62% 68% 61% 1.1 Proportion o f children (under five) with diarrhea inlast two weeks who were taken to any medical 22% 24% 22% 1.1 facility for treatment Proportion o f children (under five) with AFU in lasftwo weeks who were taken to any medical 23% 51% 33% 1 2.2 facilitv for treatment II II II Proportion o f birthsfor which a woman received at least one antenatal care consultations by a 14% 59% 26% 4.2 medically trained person (doctor, nurse/ midwife) Proportion o f births for which a woman received two or more antenatal care consultations by a 9% 51% 20% 5.7 medicallv trained person Proportion o f births attended by a medically trainedperson 2Yo 30% 8% 15.0 Proportion o f married women who report use of modem means o f contraceDtion 39% 49% 42% 1.3 Proportion o f married men who report use o f modern means o f contraception 46% 54% 49% 1.2 Source: Gwatkin et al, 2000; Bangladesh data based on DHS 1996197 Figure 2.6 - Proportion o f Household Expenditure Spent on Healthby Income Quintiles 7.4% , I 7.2% 6.6% L $ sf 6.4% - P Y 6.2% 2 .d 6.0% L E 5.8% 0 n 5.6% 5.4% 5.2% Poorest 20% 20.40% 40-60% 60-80% Richest 20% Income Quintlle Source: BBS, 2001 and authors' calculations fry pri~sttcattd public pro 27 Figure2.9. Healthconsumptionof the poorestthe quintileof income. NGOs and others- 6% Governmentproviders Traditionalproviders 3% Private providers 37% Pharniacyidisp 45% Figure2.10. HealthConsumptionof the richestquintile of income. NGOs and others- 4% , Traditional DrOvlderS ~ Government providers Pharmacyidisi 23% 28 Table 2.3 Total householdpaymentsto healthprovidersinBangladeshby incomequintiles Quintiles Fees Hospitalization Medicines Lab. And Transport Tips Other TOTAL test charges Monthly total household payments (millions tks and percentage). Poorest quintile 5.9% 2.5% 9.4% 3.6% 5.2% 4.0% 4.7% 562 (7.5%) Second 10.4% 2.1% 16.0% 5.5% 9.7% 22.2% 8.1% 958 (12.8) Third 13.2% 7.8% 19.3% 14.1% 18.3% 5.3% 14.2% 1,274 (17%) Fourth 23.3% 9.8% 22.2% 16.3% 24.1% 26.8% 17.3% 1,550 (20.7%) Richest quintile 47.2% 77.7% 33.1% 60.5% 42.7% 41.7% 55.6% 3,126 (41.9%) Total Monthly total 680 524 4,712 681 419 50 405 7,476 (100%) household payments (100%) (100%) (100%) (100%) (100%) (100%) (100%) (millions tks and percentage). 29 qualified allopathic providers and homeopaths provided most o f the health services for female children (Bhardwaj andPaul, 1986). The information on gender differences inhealthspending is not as clear-cut. One study o fpublic sector health expenditure found that on average, more money was spent on males compared to females (Ensor et al, 2002). However, analysis o f the Bangladesh HIES (Peters et al, 2003) shows that innearly all income quintiles, spendingon women i s greater than on men or children (Figure 2.11). Figure2.11 -Average AnnualAmount (in Taka) of Private Out Of PocketSpendingon the Health of Women, Men, andChildrenAccordingto IncomeQuintile,Bangladesh,2000 'O 1 60 50 40 Y m I- m 30 20 10 0 Poorest 20% 2040% 40-60% 60-8O% Richest 20% Income Quintile 1 a Women Men 0 Children Source: BBS, 2001 and authors' calculations 2.4.4 Patient Perceptions of Providers Although there is relatively little data on patient or client satisfaction in health care in Bangladesh, what i s available suggests that clients are more satisfied with private than public providers. In the 2000 service delivery survey, 41% of households reported that they think public sector health and family planning services are "bad", whereas only 10% of households thought the same about private providers. Conversely, only 10% o f households thought that public health services were "good", while 25% thought the same about private providers (CIET and MOHFW, 2001). Table 2.4 summarizes the main reasons for dissatisfaction with public or private health and family planning services (CIET Canada and MOHFW, 2001). Poor staff attitudes and problems with the lack of medicines or their quality were the biggest problems with the public sector. while costly medicines, and lack o f doctorshurses were cited inthe private sector. 30 Table 2.4 -CommonlyIdentifiedProblemsin Public and PrivateServices Proportionof Respondents Problem PublicFrivate Public Private Ratio Lack of, and poor quality of, medicines 58% 18% 3.2 Bad service 40% 33% 1.2 Bad staff attitude 25% 6% 4.2 Have to pay for medicines, expensive medicines 17% 41% 0.4 'Lack Lack o f doctors, nurses, specialists 14% 32% 0.4 o f different services 14% 24% 0.6 Dirtyandpoor equipmentand facility 13% 8% 1.6 Doctors not available 13% No data -- Extrapayments to doctors and other workers 12% 11% 1.1 Too few beds, lack o f facilities 7% 7% 1.o Source: CIET Canada and MOHFW, 2001 In the background study 'on household use of health services, satisfaction with child health services was also assessed for 342 children who were treated for an illness outside the home (HEU/IHE/NIPORT, 2003). As showninFigure2.12, there was little difference inoverall levels o f satisfaction between users o f different types o f providers. Figure2.12 Levels of Satisfactionwith ChildHealthServices, by Type of Provider . . . .. . . , . .. . . . . . 1" " " P u b l i c Facility Private Qualified D o c t o r Source: HEUIIHEMIPORT, 2003 31 Figure2.13 Satisfactionwith ChildHealthServices: MostImportantFactor Accordingto Type of Provider 1' . . . , . .. . . . , , . .. ....... .... ...I......,,.. .. .... . .. . . ... . I I Courtesy of provider Perce veo nealtn 0-tcom e Explanation about treatment Cost for services I I I Cleanliness II I I I I Consultation time I I I I I 0 5 10 15 20 25 30 35 Percent Source: HEUIIHEINIPORT,2003 Other studies also provide insights into how satisfaction with health services affects consumption. For example, 54% o f female respondents in the 1999-2000 DHS reported lack o f confidence inhealth services as a reason for not using those services (Streatfield et al, 2001). In the SDS there were differences in levels o f satisfaction between types of providers. The proportion o f users satisfied with overall services was 62% for public sector services, 88% for qualified allopathic providers (for-profit and NGO), and 88% for APPs (CIET Canada and MOHFW, 2001). 2.5 Conclusions This chapter confirmed that Bangladeshis overwhelmingly use the private sector for their health care, though much o f this care i s obtained from unqualified providers and other APPs. In the case o f maternal health services, there are remarkably low levels o f access, particularly for delivery and antenatal care. The data also show that there are considerable gaps in information concerning the demand for health care that policy makers need to be aware for the design and implementation of policy options. There i s little experience in how to influence demand for higher quality o f health services. Family planning and use o f oral rehydration therapy, which have benefitedfrom long public health education campaigns, are associated with relatively small disparities between rich and poor in Bangladesh. But it i s not known how behavior change communications can be used to influence the type o f care offered by health providers. Interventions are needed to test how to empower people to demand better quality o f health services from both private and public providers. The biggest need i s in matemity services and general health services for girls and women. 32 The dependence on out-of-pocket financing o f health care puts Bangladeshis at risk o f poverty and foregone medical treatment. This is more so for the poor. Interventions to reduce the financial impact o f drug and transport costs have not been tested in Bangladesh, and the effectiveness o f interventions such as subsidies, demand-side financing, or health insurance on financial protection i s a challenge ina large scale. These ought to be tested. Taken together, the findings in this chapter emphasize the importance o f people's reliance on different aspects o f the private sector for their health care. More than ever, there is a need in Bangladesh to implement actions that take advantage o f the existing private sector to address basic public health goals of improving quality, access, and affordability o f health services. The identification o f activities where the private sector has comparative advantage for the delivery o f health services is crucial to scale-up the process o fpurchasingprivate services with public funds. 33 Chapter 3. Interactionbetweenthe Publicandthe Private Sectors inHNP Reviewing the interaction between government and theprivate sector in health generates several key conclusions. First, the range and magnitude of government engagement with private providers is not congruent with their presence serving the poor. The bulk of interaction takes place in terms of regulation and with regard to private clinics and hospitals. APPs have very little interaction with government. Thus, public-private engagement has largely excluded service providers of greatest importance to the poor. There are, however, positive experiences, including a number of pilot initiatives, where thepublic sector haspartnered withprivate, mostly non-profit, health serviceproviders - - e.g. in nutrition, TBAeprosy, family planning, urban primary health care and immunizations. I n addition, the involvement of private actors has occurred sometimes in policy discussions andformulation, though this has not been a consistentfeature. Second, the government stewardship responsibilities to the health sector could be enhanced. There is little collaboration withprofessional and provider organizations, or support for self-regulation. Currently professional and provider organizations are primarily playing the role of trade unions. Neither consumer nor patients ' organizations have yet emerged strongly to play an advocacy role, nor to engage in monitoring of service quality and outcomes. Instruments to engageprivate actors require government officials toperform tasks very differentfrom their traditional activities. Third, misperception and low capacity underlie weak public-private engagement. The policy-makers ' interviews reveal that there is a limited understanding of the private sector size and role in provision of care, especiallyfor maternal and child health services in rural areas. Most policy makers believe that the bulk of private provision is tertiary carefor the rich. Broad-based efforts are needed to improve quality ofprivate health services and specijk initiativesfocusing on APPs. I n addition, there is a need for an increased capacity of the MOHFW and strong engagement of the key stakeholders, through the setting up of a Public-Private TaskForce in the MOHFW 3.1 Introduction The previous chapters described the size and characteristics o f the private heath sector supply and the demand and consumption of private health services by the population. This chapter will focus on the interactionbetween the government o f Bangladesh and the private health sector. It will attempt to shed some light on the current situation and the existing experiences in the country, but also draw on some o f the perspectives that governmental actors and private actors have on the issue of collaboration. It aims at providing and informing the debate on the two sides and it will conclude by suggesting some options to improve the interaction between public andprivate sector inH" inBangladesh. 3.2 CurrentSituation As described inthe previous chapters, it is clear that the private sector, inits many, varied forms, is the predominant source of health services for the people o f Bangladesh. Health policy in Bangladesh, as inmany developing countries, has traditionally focused on the public sector, and, in particular, on administration of public facilities. Notwithstanding the public-sector focus, 34 however, there are a number o f areas in which public and private actors interact. Below, this interaction i s presented according to the following categories: governmental regulation; self- and non-governmental regulation; service and facilities planning; information dissemination; disease surveillance; contracting; and, grants and subsidies. 3.2.1 GovernmentalRegulation Regulation o f health services i s a central role for governments in health systems with private delivery. While virtually all such health systems have the basic components o f health service regulation in place, in many developing countries, there i s a huge gap between the legal provisions and implementation (Afifi et al, 2003). This i s also the case in Bangladesh. Regulations are in place relating to most key inputs for health services, including: premises, equipment, and education and licensing o f medical and health workers. Regulations also exist with regardto pricing o f goods and services. The Directorate o f Health Services, headed by the Director-General o f Health Services (DGHS) i s responsible for implementation o f most regulations. An autonomous government agency, the Bangladesh Medical and Dental Council (BMDC), is responsible for undertaking and enforcing registration, as well as approval o f curriculum for medical education programs. Another autonomous agency, the Bangladesh Nursing Council (BNC) recognizes nursing and allied health traininginstitutions. The standards specified in the laws are significantly out o f date, and therefore not applicable to the current status and state o f development in the health sector. For example the provisions for maximum fees for surgical operation, normal deliveries and diagnostic tests were established in 1984, andhave neitherbeenupdated nor revoked since then. The effectiveness o f regulation is very limited. At the most fimdamental level, major constraints appear to be insufficient capacity and attention to regulatory issues. There are too few human and financial resources devoted to regulation. It is evident that the capacity o f the govemment bodies to implement this legislation i s weak, or non-existent insome cases. While registration i s formally required to practice, the vast majority o f providers practice without being registered (Health Economics Unit, 1998ab). Inspections for monitoring o f service quality are not possible with available resources. There are nearly 85,000 healthpersonnel under the Director General o f Health Services, hence it i s not an absolute shortage o f humanresources that explains the weak capacity. Rather too few staff are assigned to these activities, and there i s a shortage o f personnel with relevant training. Even a substantial portion (30%) o f inpatient facilities, which are fewer innumber and easier to identify, operate without registration (BBS, 1998). Inspection o f health facilities occurs only when a license i s initially given. Subsequent re-licensing i s done based on self-reported information. Out-of-date standards and rates established in regulation seriously undermine compliance. The fact that the standards or values set inthe law are so incongruent with reality inprivate practices andclinics, diminishes the leverage andpotential influence o fthe regulations. 35 3.2.2 Sew- and Non-governmental Regulation The role o f professional and provider groups is another critical element o f a well-hctioning regulatory framework. Invirtually all well-performing health systems, these non-governmental bodies undertake extensive quality assurance activities, with varying degrees o f oversight by government (Afifi et al, 2003). The technical expertise and credibility o f these bodies makes them invaluable partners indeveloping standards for education andpractice. They often play a substantial role inimplementationo fregulation as well. There i s a substantial number o f professional and provider organizations inthe health care field in Bangladesh (see Box 3.1). Currently, these bodies undertake few self-regulation activities. To date, the organizations function primarily as trade bodies, that is, they focus their activities on protecting the interests o f their provider-members as businesses. The government does involve representatives from these organizations in some policy initiatives, for instance consultation on proposed legislation. However, it generally does not involve professional or provider organizations in implementation o f policies. The degree o f governmental support and oversight for self-regulationbyprofessional or provider bodies i s likewise low. Consumer or patient organizations are not developed, nor i s a role established for them in regulation or government activities. Victims o f malpractice or mistreatment can only make an appeal to the Director General o f Health Services, an event that i s fairly uncommon. When asked, private providers confirmed that no other non-governmental groups strongly influence their environment (HEU/ME/NIPORT,2003). 36 Box 3.1 -PotentialSelf-Regulatory Bodies in Bangladesh The Bangladesh Medical Association (BMA) is the main representative organization o f the medical profession. Its formal objectives include improvement o f medical sciences and enhancing the status and honor of physicians. Inpractice, its main concerns relate to protecting the interests, rights and privileges of its members. Physicians working both in the public and private sectors are members o f the BMA. Many other professional societies of special medicallsurgical disciplines are affiliated with the BMA. Theprofessionaljournal ofBMAi s publishedirregularly. Private medical practitioners have an association of their own, namely the Bangladesh Private Medical Practitioners 'Association (BPMPA). Many of its membersare also members ofthe BMA. Theprofessional association o fregisterednurses is theBangladeshNursing Association. Similarly, the Pharmaceutical Society of Bangladesh (PSB) is the apex professional association of graduatepharmacists. Diplomapharmacists have a separateprofessional society. Private pharmaceutical manufacturers (mainly the larger ones) have their own organization, the BangladeshAushad Shilpa Samity. The owners of retail pharmacies have a society of their own, the Bangladesh Chemists and Druggists Samity. The owners o f private clinics and laboratories have recently formed their own association inthe name of Bangladesh Private Clinic and Diagnostic Owners' Association (BPCDOA). It has a nine-member committee to fix more or less uniform charges and to address quality of services issues. Like the other professional associations mentionedhere, this entity functions more or less as a trade organization. Source: Hve. 2003 3.2.3 Planning In mixed-delivery health systems, government service and facility planning must take into account the existing private sector capacity, as well as their plans for development. Otherwise, the impact o f public sector construction and service development is reduced through duplication and"crowding out". Service and facilityplanning inBangladeshis undertakenby the MOHFW. Currently, these planning activities are based only on public facilities and services. Some private hospitals and clinics evidently receive information from the government as to planned development (see Table 3.1). However, information about private capacity appears not to be taken into account ingovernment facility andinvestment plans. 37 Table 3.1 -Public/PrivateSector Interactionby Type of Provider Source: HEU/ME/NIPORT, 2003 and authors' cal~ulations.~ 3.2.4 Policy Making Inmixed-delivery health systems, private providers are typically involved, through a range of mechanisms, inthe policy makingprocess. Most often this involvement i s formal, and provider representative bodies have an official "seat at the table" in deliberations regarding policy development. InBangladesh, representatives from the more formal and organized professional and provider groups (allopathic qualified physicians; private clinics/ hospitals) are sometimes involved inpolicy development and implementation in the health sector, especially where such policies affect the private sector. Typically, these representatives serve as members o f formal government appointed bodies or committees. Many components o f the service delivery sub- sector, such as traditional healers, homeopaths, pharmacists, are less organized and more atomistic. Not surprisingly, neither they nor their representatives are involved in the policy making process. Uncertainty about govemment policies towards the private sector was mentioned as an important constraint to their operations by a large majority o f the private providers interviewed(see Table 3.1). 3.2.5 Information Dissemination Information dissemination activities are a critical part o f a govemment's stewardship in the health sector. Such activities constitute a substantial part o f government activities in the health sector in successful mixed-delivery health systems. Governments usually partner with self- regulatory and consumer or patients' organizations inimplementing these activities. Information disseminated to consumers usually relates to appropriate health-seeking behavior and what constitutes high-quality health care. This knowledge guidespatient consumption andpreferences toward higher quality care and providers. Better informed patients are also less likely to pressure providers for unneeded, even harmhl interventions - pressure to which private providers are known to be responsive. Information dissemination to providers about policies, regulatory and planning activities promotes compliance and reduces uncertainty in private investment and development. Information dissemination to providers increases knowledge, and improves coordination and clinical practice. The percentages were calculated among the private actors who provided an answer to the question. 38 InBangladesh, government information dissemination activities appearto reachalimitedportion o f health providers. From the provider survey, 40.6% o f hospital or clinic managers receive some information about health sector planning for example. A smaller proportion o f clinicians (26.8%) indicated that they receive similar information. Beyond the formal provider group however, virtually no information is received from government (see Table 3.1 above). With regard to information dissemination activities to ,consumers, though very little information is available it appears that such efforts are limited to special programs (family planning; child health; TB) and, with the exception o f family planning, are associated with donor-driven initiatives. 3.2.6 DiseaseSurveillance Disease surveillance i s a central responsibility o f government in the health sector. For surveillance systems to be effective, it i s essential that all providers o f health care actively contribute information. It is equally important that all providers, both public and private, participate in programs to appropriately identify, treat and/or refer patients. Currently in Bangladesh there i s no systeminplace for reporting o f infectious diseases from private sector in the main surveillance system. Private providers are not legally required to report disease patterns nor the number o f cases treated. The exception to this situation is the reporting o f tuberculosis cases inthe context o f the government collaboration with NGOs (see Box 3.2). 39 Box 3.2 -Government-NGOCollaborationto Implement the NationalTB Program In 1994, the Government signed two Memoranda of Understanding separately with BRAC and the Leprosy Coordinating Committee, the latter being an umbrella of NGOs working on leprosy. Both agreements, renewed on annual basis, state the principles o f collaboration in implementing the National Tuberculosis Program. NGOs' resources are channeled to the provision of standardized TB services, to rural and less accessible areas, through the use o f community health workers. The results are very encouraging. The BRACprogram achieved highrates of case detection and treatment compliance, with a cure rate of at least 85% (Chowdury et al, 1997). Inaddition, when compared to the government, the use of CHWs was found to be more cost-effective in rural Bangladesh. With the same budget the BRAC programcould cure three TB patients for every two cured inthe government program (Islam et al, 2002). The successful collaboration in TB control in Bangladesh has been achowledged internationally (Kumaresan et al, 2000) and could serve as a model for other programs in the country. Crucial to the success seems to be the clear identification of roles and responsibilities o f the different actors. The division o f responsibilities i s presented in the following table, and reporting o f tuberculosis cases i s clearly mentionedunder the NGOs' responsibilities. Area ofCoIIaboration Government 1NGOs Source: Barkat et al, 2003 3.2.7 Contracting Engaging in long-tenn purchasing arrangements with private providers i s the most influential instrument used to guide independent service providers in mixed-delivery health systems. The interest o f private providers in a predictable revenue flow yields the government purchaser strong leverage over their operation. Health services contracts are typically longer-tenn (relative to other public services), and subject to on-going consultation and coordination. Therefore, health services contracting both requires and generates frequent communication and interaction with private providers, and, as such constitutes an important component of a government's overall interaction with the private sector (Taylor, 2003). InBangladesh, a number ofpilots have beeninitiated to engage NGO providers via contracting to provide health care services. These experiences constitute a rich resource for evaluating and improving the interaction between the GOB and private service providers. To date, most of these public-private partnership initiatives have been implemented in conjunction with donor- supported initiatives. Recently, the MOHFW has declared its intention to contract out NGOs in 350 unions to provide essential service delivery. 40 3.2.8 Grants and Subsidies In some health systems, governments support private service providers via grants or subsidies. Since the hnds are not formally linked to any service or output, such arrangements are usually applied to NGO service providers, whose activities are linked to government's sector objectives. InBangladesh, there are some instances of such support to NGO providers (see Box 3.3). As with contracting however, these experiences are taking place mainly in connection with donor supported programs and pilot activities. Indicators Numbers Percent National Women who received at least two antenatal care (ANC) consultations out o f 28959 74.0 30.3 total deliveries Deliveries conducted by qualified person out of total deliveries 14439 36.9 NIA Women receiving at least one postnatal care (PNC) consultations within 42 days 19886 62.1 NIA o f delivery 1 Contraceptive Acceptance Rate (CAR) N / A 63.O 53.8 Neonatal death I I I 1 I 968 1 25.8/1000 I 50.4/1000 I death I 1.9/1000 4.34000 Source: Barkat et al. 2003 3.3 Perceptions of the Public Sector towards Working with the Private Sector The information described below was collected during the semi-structured interviews conducted with a total of 21 policy makers and managers inthe government system at central level to elicit their views o fpolicy making inthe health sector, especially withregard to public/private mix o f services and the overall role o f the private health care sector in achieving national health goals. 41 Mid-level government officials -- those directly in charge o f implementation o f health sector policies -- were interviewed to obtain an understanding o f their views on private provision (Forsberg andAxelsson, 2003). Many o f the persons interviewed felt that there i s not a clear andwell-communicated government policy towards the private health sector. However, it was said in some o f the interviews that the government has recently decided to contract out some health services to private actors. More active supervision o f the private for-profit providers in Dhaka district has also taken place recently following an initiative taken by the new government to improve services in the private sector. Senior officials tended to describe the private health sector as providing tertiary care to well-off people. The widespread use o f private providers at other levels and by all socioeconomic groups, as documented in surveys, was not put forward by most respondents. The fact that publicly employed doctors often work in the private sector and therefore lose focus on their duties inthe public sector was often described as a problembythe persons interviewed. 3.4 Perceptionsof the Private Sector As part o f the policy-makers interviews, four privatehealth care providers were also interviewed to provide a private sector perspective on government health sector policies and the interaction between the private and the public sector. In addition, during the provider survey several questions elicited their perspective o f the interaction with the government (HEU/IHE/NIPORT, 2003). Private providers in general perceived that there was little interaction between themselves and government programs and activities. Again, the informal andnon-allopathic providers perceived the least interaction. Nevertheless, all groups of providers expressed strong willingness to work with the government and the MOHFW. The majority expressed an interest inworking with the government to expand their involvement in health promotion activities - an area where private providers are typically perceived as weak. Inaddition, the private clinic managers indicated that training o fprivate staff would be a usefil support from the government (see Table 3.2). While there was a surprisingly large amount o f interest inworking with the Government, private sector representatives expressed some reservations, which they attributed to corruption and lack of accountability infinancial management. "Support by donors to health care provision could go directly to private actors. When finds go through the government there is corruption and too much diversiodleaking o f funds." Or "One major reason why the public sector is not a good choice for provision o f care i s that the accountability o f the government is poor. The quality o f public services therefore becomes very poor." 42 Table 3.2 -Public/PrivateSector Interaction:the Perceptionby Type of Provider wouldbe usefbl support from NIA NIA NIA (87.5% govemment lo The percentages were calculated among the private actors who provided an answer to the question. 43 access to quality services for these population groups. Much has been learned from these experiences about the opportunities and challenges o f "hamessing" the private sector to meet health goals in the Bangladeshi context. For example the collaboration between the national tuberculosis program and large NGOs could serve as an example for other programmatic areas. To date, however, few o f these pilot initiatives have been expandedor "scaled up". Buildingon these initiatives, to apply "lessons learned" and to reach greater numbers o f people remains an important unused opportunity. Encouragingly, recent statements by the GOB indicate increasing awareness o f and openness to working with the private sector and in particular with NGOs to provide service delivery at the unionlevel. 3.52 Public-Private Engagement Excludes Providers of GreatestImportance to thePoor Ingeneral, the bulkofinteractionbetweenthe government andprivateproviders takes placewith regard to private clinics and hospitals. Less formal, less organized providers, such as non- allopathic practitioners, including traditional birth attendants, and drug vendors and retail pharmacists on the other hand have very little interactionwith government. This i s unfortunate, as the poorer and more rural populations tend to utilize the latter. While there are multiple factors that contribute to this situation, a stakeholder analysis o f the relevant political economy may be helpful to identifyresistance points anddevelop appropriate solutions. 3.53 Misperception and Low Capacity Underlie WeakPublic-Private Engagement A number o f factors undoubtedly contribute to the current poor state o fpublic policy toward the private health sector. 0 There appears to be limited knowledge o f the role (instruments) o f the government in guidingthe private sector towards contributing more towards social objectives related to access and service quality, inparticular. 0 The policy-makers' interviews revealed that there is a limited understanding, indeed, even misunderstanding o f the private sector size and role in provision o f health care, in particular to the poor in rural areas. Most policy makers believe the bulk o f private provision i s concentrated in tertiary care in the capital, overlooking the significant role altemative private providers play in serving the poor and especially for maternal and child health services inrural areas. Instruments to engage private actors (providers and representative organizations) are inherently complex to operate. They also require government officials to perform tasks very different from their traditional ones related to administration o f public facilities. There i s currently very little capacity to operate such instrumentsinthe MOHFW or inlocal government bodies. The bulk o fhealth services are provided by altemative private practitioners. This segment o fthe health services sector is consistently the most difficult to monitor, engage and influence. Experience from many countries, both developed and developing, reveal that initiatives to influence service provision inthis segment i s particularly challenging. 44 Chapter 4. Main messagesandpolicyoptions 4.1 Findings (a) Overall health service consumption inBangladesh (from any source, public or private) i s low compared to levels o freported illnesses andto levels inother countries. (b) The poor are far more likely to forego medical treatment (30% o f poor sick people are not treated), with differences between rich and poor households being the largest for medically trained deliveries, antenatal care, treatment for ARI, and immunizations. The differences nearly disappear when comparing the use o f modern contraceptives or oral re-hydration therapy for diarrhea, two services where there has been extensive social marketing. (c) Women and girls tend to receive less medical care than men and boys despite the fact that females show a greater disease burden than males. HIES 2000 indicates that 58% of curative services are utilized bymen. (d) Theprivate sector is used for the overwhelming majority o f outpatient curative care, while the public sector is used for a larger proportion o f hospital and preventive care. This broad division o f roles cuts across economic strata of the consumers, contrary to a common perception that private sector caters mainly to the tertiary care needs o fricher populations. (e) The poor also use private curative health care services and pay for those services with higher proportion o f their income than the richest households. However, both the poor and the rich choose private providers insimilar proportion. Inthe poorest households, 81% o f the health care services are provided by the private sector, and 88% inthe richest households. Also 40% of total out-of-pocket expenditures are made by the 40% bottom poorest households. (f) There are major gaps inknowledge concerning the private health sector - the actual numbers o f providers, the services they provide, the conditions under which they practice their trade, their incentives and disincentives, etc. (g) One very well established fact is that a majority o f private providers o f health services in Bangladesh (referred to in this report as alternative private practitioners, APPs) do not have formal training and recognized qualifications inallopathic medicine. The poor, especially, make heavy use o f these APPs. Implementing training programs to drugs shop workers and birth attendants i s needed to improve quality o f services to the poor. (h) The bulk ofprivate health service providers are males (the exception beingtraditional birth attendants), which poses a major problem o f access to their services bywomen. (i)Although there isrelativelylittleknownabout the quality ofcare ofindividualprivate providers, or the health facilities they work at, the available information suggests that assuring technical quality is a signzjkantproblem, particularly among alternative private practitioners. ('j)Amajorityofconsumers, however, reporttobesatisfiedwiththeprivateservicestheyhave received, and rate them superior to government-provided services. Alternative private practitioners are given preference over qualified doctors mostly on account o f their easy access. Availability o f drugs i s also reported to be a key factor in choosing private practitioners (qualified or not) over government facilities. 45 (k) Inspite ofthe obvious importance o ftheprivate sector, healthpolicy inBangladesh thus far has focused on the public sector and, inparticular, on administration o f public facilities. There has been insufficient attention paid to the potential o f using the know how and resources o f the private sector more systematically to achieve societal goals in health. There have been some instances such as the collaboration with NGOs (e.g., National Tuberculosis Program, Integrated Nutrition Project, social marketing o f contraceptives), and a number o f pilot initiatives as in urban primary health care. Such initiatives have yet to be scaled up and lessons from these experiences yet to be evaluated andbe applied to HNP program. (1) Government regulations are in place for many aspects o f health service provision inprivate facilities but enforcement has been uneven. The main constraints appear to be insufficient capacity and attention. There has been little collaboration between the government and professional and provider organizations in ensuring adequate standards, and there is no wide support for self-regulation. Neither consumer nor patient organizations have yet emerged to play anadvocacy role nor to engage inmonitoring o f service quality andoutcomes. Inmanyways, boththe public andthe private sectors have failedto meetthe essentialhealthcare needs o f the people in terms of both access and quality. Near half o f the users of the publicly financed essential service package are non-poor people. The relevant question i s whether the performance o f the private sector can be improved to meet the needs o fthe people and to provide the value for the money they spend for their health care. 4.2 Policy Formulation: A new paradigm The above messages suggest that there are several areas where a reformulation o f government policies would be desirable, with a view to helping the attainment o f the social goals in health included in the government's December 2002 interim PRSP. Given the lack o f resources and capacity constraints inthe public sector, there is a need to shift the role o f the Government from provision o f health care services to the purchase o f health services. This new paradigm is justified by the current dominant place held by the private sector inthe financing and delivery o f HNP services, and the serious concerns about quality, access, accountability and govemance with regard to both private and public services. Inparticular, the following policy areas would appear to deserve priority in government thinking and action to strength the implementation o f the next Health, Nutrition andPopulation Program. 4.2.1 Under-ConsumptionofHeaJth Careby thePoor and Women The fact that many patients have a preference for private providers suggests that the observed under-consumption o f certain essential health services, especially by poor households and by women, cannot be remedied without increasing the access o f under-served populations to private providers. The problem is particularly important for matemal health services, especially assisted delivery by a skilled attendant. The more traditional government approach o f expanding the supplyo f services provided by government employees out o fpublic facilities is unlikelyto reach the poor. The gender distributiono fthe providers poses an additional disincentive for women. 4.2.2 Service Quality and Outcomes Private health services (mostly clinical services) appear to be o f good quality in the eyes o f consumers. Thisjudgment i s likelybased on those characteristics o f private services that can be 46 easily assessed by patients, such as ease o f access, degree o f courtesyhespect, and means to obtain both advice and medicines in one place. Much more problematic for consumers i s the ability to assess correctly the technical quality o f private treatment received or to relate such treatment to outcomes o f illness episodes (good or bad). To make up for this important deficiency, and ensure that in most cases consumers derive good value from their purchases o f privately provided services, deliberate and well-conceived collective action will be required. Issues o f quality need to be looked at separately for the formal and informal sectors as public policy interventions would be different for these two groups o f providers. The multitude o f provider types inthe private sector warrants a mix of differentpolicy options. 4.2.3 ThekhowledgeBase While various public interventions could be conceived based on what i s now known about the private health sector in Bangladesh, the large knowledge gaps that exist would magnify the uncertainty always associated with new policies and courses o f public action. The knowledge base about private health services needs to be widened to enable the progressive refinement o f policies and programs, and also to help users to make informed decisions related to the utilization o fhealthproviders. 4.3 Policy Options The followingpolicy options were discussed during stakeholder consultations inearly May 2003 with government officials, private sector actors, civil society, the academia, and development partners. While there was broad agreement on the need to increase the engagement with the private sector and on the value o f the options presented here, it was agreed that firther debate, consultations, pilot tests and studies are neededbefore policy decisions are taken. Provision Develop a clear public policy towards the private sector that harness the valuable resources that are available inthis sector The government needs to create "head room" in its public expenditure envelope so that some public resources will become available for influencing the behavior o f private providers through different measures such as contracting with private providers and subsidizing care for the poor. BringAlternative PrivatePractitioners (APPs) into to service provider system byworking with them in strengthening skills and increase the number o f formally trained staff throughtraining. In addition to traditional regulatory and quality assurance techniques, it is necessary to promote health services standards, performance based competitive pressures, and incentives to attract private practitioners to work inlow coverage areas Consumption 0 Make information about the quality and price o f private providers readily available to consumers, especially for the poor 0 Introduce targeted subsidies and community level insurance for the poor and social insurance mechanisms for civil servants and formal sector workers 47 Publicprivate interaction 0 Increase competition between public andprivate sector through competitive and selective contracting andperformance benchmarking and service standards. 0 Introduce intemal markets (make public providers compete for public fbnding on a performance basis) and new public sector management techniques (ie contracting out, contracting in,management contracts etc). 0 Redefine the role o f the MOH and strengthenits core stewardship capacity in areas such as strategic planning, monitoring and evaluation, coordination, regulation, quality control andenforcement The options described below are initial ideas to address the three main sets o f issues raised, Le., under-consumption, quality concems and knowledge base. These options need to be pilot-tested and evaluated first, before being scaled up nationally, but such pilots should be large enough to yield meaningful lessons. 4.3.1 ToAddress Under-ConsumptionofServices by thePoor and Women Pilot contracting private providers with government finding: the government has already had some experience in contracting out some of its services out to non-profit organizations, but most o f these models have been through donor-financed projects and have yet to be scaled up. Serious consideration needs to be given to the option o f larger scale contracting o f HNP service provision to the private providers, financed by public finds. The recent government commitment to contract out the management o f 350 community clinics and Union Health and Family Welfare Centers is a positive step. Such contracting should, however, follow a transparent and fair process for the selection o f firms/NGOs and should include rigorous monitoring and accountability procedures, to ensure boththe quality of care and the efficient use o fpublic resources. The contracting and the monitoring should be performed by an independent body such as an NGO or private company. Health services should initially be focused on child andmatemal services. Contracting arrangements should be performance-based. Contracts fees should be linked to agreed outputs and health outcomes. Box 4.1 provides some pointers on "contracting in" and "contracting out", based on experiences inother countries. Expand social marketing: Bangladesh has a positive record o f social marketing in contraceptives and oral rehydration therapy. Social marketing has been shown to reduce inequalities inaccess to such commodities. This experience could be expanded to other essential health-promoting commodities, such as bed nets and soap, which i s known to have a positive impact on the health o f the poor. 48 Box 4.1:Contracting-Inand Contracting-Outto ImproveHealthServices Utilizationand QualityinCambodia Inresponseto poor healthoutcomesandapublic infrastructure ravagedbya quarter-century of conflict, Cambodia has recentlyundertaken innovative approaches to collaborating with the private sector for health services delivery. The Ministryo f Health i s currently piloting health financing reforms inselected districts, through its Accelerated District Development (ADD)program. With the assistance o f the Asian Development Bank (ADB), Cambodiahas piloted two models o f contracting for primary healthcare services inan eight-district interventioncovering a total o f one millionpeople: 1) contracting-out, where contractors (typically NGOs) have full responsibility for service delivery and directly employ their staffi and 2) contracting-in, where contractors provide only management support to civil service health staff. The contracting model was intendedto address one o f the primary causes o fpoor health system performance: poor staff morale and inadequate management. Preliminary results show that those districts with contractingmodels outperformed control districts on several health service coverage indicators, with the contracting-out model yielding the greatest gains. For example, districts with contracting out increased annual per-capita health care contacts among the populationto 1.7, compared with 1.2 for contracting-in districts and 0.8 for control districts, with lower recurrent costs (Bhushan et al., 2002). Contracting-out also proved equitable, as utilization o f government-financed services increased disproportionately among the poor, whose out-of-pocket payments were also reduced, up to 70% incontracting-out districts and 40% in contracting-in districts. Several factors o fthe Cambodian contractingexperience have contributedtoward its success, including: establishment o f contract management capacity inthe Project CoordinatingUnit (PCU) o f the MOH; a transparent competitive biddingprocess; and an agreed-upon monitoring process. With assistance from ABD, the Ministryof Health specified objectives and solicitedcontract bids from for-profit and not-for-profit internationalagencies, and subsequently monitored awarded contracts through the PCU. For more details on the contracting-inmodel, please refer to an in-depth case study o f the Pereang District in: Soeters, R, and Griffiths, F. 2003. "Improving Government Health Services Through ContractManagement: a Case from Cambodia." Health Policy andPlanning, 18: 74-83. Sources: Bhushan, I., S., and Schwartz, B.March 2002. "Achieving the Twin Objectives o f Efficiency Keller, and Equity: Contracting Health Services in Cambodia." Asian Development Bank. ERD Policy Brief No. 8; Loevinsohn, B. "Contracting for the Delivery o f Primary Health Care in Cambodia: Design and Initial Experience o f a Large Pilot-Test." The World Bank; Soeters, R, and Griffiths, F. 2003. "Improving Government Health Services Through Contract Management: a Case from Cambodia." Health Policy and Planning, 18: 74-83. Explore insurance / risk-poolinn and mepayment mechanisms: Since people are paying for private sector HNP services from their pockets, well-designed community insurance schemes could provide a feasible option for better delivering o f a benefit package with acceptable quality. Such schemes would not only pool risks and resources for curative services for minor illnesses, but also provide for catastrophic coverage (if public subsidy for such coverage is considered appropriate) and could include incentives for seeking preventive services by buildingdifferential 49 co-payments or deductibles. The Bangladesh experience such as Grameen Bank's schemes for their members may prove to b e h e l p f d in designing locally appropriate community insurance mechanisms. Box 4.2 provides a brief account o f the Grameen Insurance system. The inclusion of maternal and child health services in benefit packages is crucial, particularly for assisted delivery and complications requiring hospital care. Insurance schemes could b e o f different types in terms of the benefits package, beneficiary pool, and other aspects, ranging from micro- insurance schemes at the community level to the social insurance programs covering large populations employed inthe organized sector. Box 4.2 GrameenKallyan(a communityhealthinsurancescheme) Grameen Kallyan has been operating since the year 1996. The major actors in this organisation are: Grameen Kallyan, Grameen Bank members, and non-members o f Grameen Bank. The target beneficiaries are the rural poor families (within / outside Grameen Bank members). It i s Grameen Bank's long-term experience flowing from its involvement in micro credit operations that the rural poor in Bangladesh are prone to serious setbacks emerging from natural calamities as well as from personal or family-based misfortunes from time to time. This type o f vulnerability deprives them of the opportunity to continue their thrift savings sustainably. Inaddition, the level and quality o f social security, especially in the area o f medical care being one o f the lowest, Grameen Bank thought of introducingmicrofinancing coupled with quality health care services toward improvedhealth status of the rural poor. The goal o f Grameen Kallyan i s to provide sustainable quality primary health care services. The existing interventions o f Grameen Kallyan spread through the districts o f Tangail, Dhaka, and Comilla. This organisation does not receive any donor aid. In this sense, it i s limited to being a partnership o f the nature o f Private (NG0)-Community Venture. The major activities included in Grameen Kallyan are: collecting premium, consultation services, selling essential medicines, selling pathological services, referral o f pathological cases to outside laboratories, referral o f patient cases for prescriptionto outside doctors, payment o f certain fees to a patient in case o f admission to an outside hospital on referral,. Grameen Kallyan's services are two-fold: (a) management o f microfinancing through health insurance, and (b) provision o f health care services. Grameen Kallyan, from time to time, also arranges for health care camps especially inthe area o f care for cataract. angladesh. Expand direct information Campaigns: Such campaigns should address under-served households to enhance their appreciation o f the importance of the health services they are not seeking - and the risk linked to the care obtained from unqualified practitioners. Strategic planningo fthe Government would guide the private sector inthe provisiono fpublic information campaigns on individual and household behaviors, with attention to b e paid to changing health- care seeking behaviors, especially of mothers with children. Bangladesh does have some successfil experience in modifying health-seeking behaviors through information campaigns, e.g., family planning, immunizations and oral rehydration therapy. Areas where information campaigns could have been more effective include the use o f iodized salt, maternal health care, and tuberculosis control programs. Clearly thought out communication strategies, based on formative research which study not only health-seeking behaviors but also their socio-cultural 50 determinants are a key pre-requisite for successful information campaigns. Other conditions for success would be effective leadership and political will to carry out such campaigns, backed up by highquality services that the population can rely on. An example o f information campaigns not backed up by appropriate services might be found in the manner that Bangladesh has been dealing with the problem o f arsenic contamination o f drinking water; while information campaigns warned people not to drink contaminated water, no viable alternatives have yet been made available inan affordable and acceptable manner. Explore demand-side subsidies: Public resources may be usedto provide health coupons or such similar instruments to poorer population groups, giving them the necessary purchasing power to consume essential services from the private sector; this approach puts the choice o f providers in the hands o f the consumers, and empowers them. The success o f demand-side subsidies depends on the capacity for identifying the potential beneficiaries, and the availability o f appropriate services o f acceptable quality; hence, this option needs to be used in conjunction with other policy measures geared towards quality improvement, e.g., mechanisms for identification o f the poor and accreditation systems . 4.3.2 ToImprove Service Quality and Outcomesin thePrivate Sector A recurring theme o f this study is how to improve the quality o f health care services. There are extremely weak and insufficient systems for assuring the quality o f health care in Bangladesh, and as a result, there i s little information about the quality of care. Weak performance o f the civil service andpersistent governance problem underpinthe need to design and implement adequate legal and economic incentives and to establish autonomous regulatory agencies financially and administratively independent o f the MOHFW. Such agencies must ensure, promote and disseminate best practices in quality services. Quality o f services will only be achieved as result o f consistent set o f actions. If Bangladesh's focus on quality becomes merely a program to regulate the health sector, it is unlikely to change the quality o f health care significantly inthe near future. It will be necessary to tackle the quality issues on several fronts, using strategies that may be characterized as "top-down"; "bottom-up", "leading edge", and "driving force" (Table 4.1). The different strategies for addressing quality involve different actors, have complementary objectives, andhave their own limitations and timefiames. 51 Tble 4.1 QualityImprovementStrategiesRelevantto Bangladesh: MultipleFronts Strategy Examples Objectives Limitations Top-down Licensing, egistration Set minimum Limited impact in informal sector; accreditation standards effectiveness linked to good governance Bottom-up (i) Consumer education & (i) Raise demand and (i) Long run solution, but limited advocacy expectations for experience in countries where quality education levels are Idw, modest (ii) Pilot quality improvement successes have been achieve in projects in leading (ii) Strengthen specific areas organizations capacity for quality (ii) Limited scale Leading edge Collaborative professional Improve performance Little experience in low-income learning networks and change patterns countries of practice Driving force Financing incentive Influence provider Medium to long term solution; mechanisms linked to behavior to improve requires group purchasing of health clinical practices quality services through insurance or pre- payment The top-down strategies usually involve government agencies setting standards of care, most often with an objective o f demanding a minimum level o f quality and safety. In Bangladesh, most standards inhealth care concem the qualifications of staff and physical inputs at health facilities, rather than quality assurance processes, such as clinical guidelines, standard diagnosis, treatment guidelines and continuing medical education linked to certification and accreditation . Most top-down approaches are mandatory, though accreditation by professional bodies i s more often a voluntary process. In many other countries accreditation i s mandatory because large payers (e.g. governments and insurance companies) will require accreditation for an organization to be eligible for finding. Top-down approaches work best when the criteria can be easily measured and enforcement is straightforward; this is far from being the case in Bangladesh. While improvements and additions could be made to top-down strategies in Bangladesh, concentrating efforts on these approaches i s not likely to make substantial improvements in the quality o f care in the short and medium term, since most providers work outside the formal sector, and the govemance environment i s weak. Bottom-up approaches involve both demand and supply sides o f the equation. On the demand side, such strategies involve educating and empowering consumers and consumer organizations. The main purpose i s to enable the public to expect and demand better health care, or insome cases to redress harm caused to patients. Consumer ratings o fproviders, facilities, and products (e.g. pharmaceuticals) can lead to changes in the behaviors of providers or the quality of products. Although there i s a tremendous theoretical appeal o f approaches that reduce information asymmetries between patients and providers, in practice, they have had limited success. Where successful, as inthe case o f reducingthe demand for antibiotics for sore throats, the campaigns were quite targeted, and occurred in conditions where levels o f education and access to health services are much higher. Supply side strategies involve individual pilot projects by innovative and leading health care institutions. The limitation with these approaches i s that they often depend on the individual circumstances o f the hospital or organization 52 undergoing the changes, and are not taken to a larger scale. Yet both types o f bottom-up strategies are lacking in Bangladesh, and are worth initiating soon, even if their impact i s more likelyto befelt inthe mediumto longterm. Probably the most gains inthe short and mediumterm would be made through so-called leading edge approaches, characterized by collaborative leaming networks. These strategies involve pullingtogether networks o f providers and facilities to establish priorities for services, in a framework that uses up-to-date quality improvement methods and information sharing. In contrast to top-down strategies that try to set minimum standards, these approaches seek to improve care, and to continually develop best practices, while buildinghuman and institutional capacity. They are particularly effective inconditions where health systems are fragmented, as i s clearly the case inBangladesh. Finally, financing mechanisms can be considered as a driving force for quality improvement. Financing tools can be linked to the demonstration o f good practice or good outcomes though subsidies, quality-linked payment rates, or contracting that is partly based on quality provisions. Their purpose i s to influence provider behavior in a way that improves the quality o f care and increases accountability. One limitation with these strategies in Bangladesh i s that people make most health care payments individually from out-of-pocket, so that the ability to use the power o f group purchasing to influence provider behavior i s lacking. These payments are essentially in the nature o f "fee for service", and there i s little scope to use other mechanisms o f compensating the provider (e.g., capitation payments for general practice), as levers to influence provider behavior. Risk-pooling mechanisms, e.g., community health insurance schemes, might help develop different kinds o f incentive mechanisms for providers and consumers alike and could help improve quality o f health services in Bangladesh. Inthe short to medium term, pursuing public sector contracting for non-public health services would help to build up experience with financing strategies. Whatever sets o f strategies are chosen to improve quality, leadership and consensus building will form important components o f the first steps. The strategies and technical components will need to be seen as being endorsed by influential decision-makers in the political, professional, and civic arenas, and the clinical aspects to be developedendorsed by the top clinicians. Government should play a leading role in catalyzing these initial steps. For example, govemment may establish a task force or commission to develop and oversee its quality improvement strategies. Another way govemment can immediately support a movement to improve quality inhealth care i s to support research and assessments o f quality o f health care. Providing information on the quality o f current practices i s important to raise awareness and support for improving quality, and i s also needed to provide a basis for setting standards and benchmarking, and to get healthworkers involved inquality improvement activities. Based on the above considerations, the following policy options aimed at improving quality are targeted at both the formal sector, Le., the qualified allopathic providers and facilities andthe informal sector (APPs): Institute specific initiatives to improve aualitv o f services provided by altemative private practitioners: Encourage the development o f APP representative organizations, both local and national, through formal involvement in consultation and policy dialogue. Expand existing initiatives (contracting, training, information dissemination) to include APPs where viable; of 53 particular relevance are the informal providers o f maternal and child health services, including the traditional birthattendants. Creation o f a National accreditation agency: Both public and private providers and facilities should be accredited, by a established independent body with a reliable system o f regular monitoring and maintenance o f standards, and legal capacity to provide incentives and enforce sanctions. This agency would provide information to consumers for judging the quality of services. The work o f this agency should be linked to direct information campaigns to households. Upgrade the regulatory framework: This should include the revision o f outdated regulations and identification o f appropriate mechanisms and resources for enforcement o f existing, appropriate regulations. Considering the existing, limited capacity to monitor and enforce regulations, this may be a long-term goal. Professional/provider associations need to be identified andtheir capacity to play the role o f a self-regulatory bodyneeds to be strengthened. To promote the dissemination o f franchising: This involves brandname development o f health services, to give the consumers o f health care a way to choose providers with an assured standardo f services. Franchising could also provide credit andmanagement support to providers andstaff and a sense o fbelonging andpride Direct information campaigns to households: A direct information campaign, as mentioned above, can increase the consumption o f services. It can also influence the demand for increased quality o f services, resulting in pressure on private providers to improve their practices. Such strategy could initially be targeted to the poorest districts and the poor areas o f selected urbancities. Promote the formation o f one or more consumer organizations in health: This would supplement and reinforce the ability o f individual consumers to demand better quality services and to represent their concerns and even negotiate more competitive prices. Consumer organizations can also play a role in the monitoring and improving o f quality o f care in the private sector. Promoting training of private providers: This option i s particularly usefbl for UPS, whose qualifications and skills are variable and inadequate in most cases. Recognizing the ineffective enforcement o f the laws that ban their illegal practice, an alternative or complementary public policy might be to equip them with acceptable levels o f knowledge and skills, both to improve their effectiveness and reduce the potential harmthey cause to the public. Special focus should be given to the training o f local midwives, birth attendants and drug shop workers through the provision o f scholarships to be trained and further monitoring o f performance. A.3.3. To Improve the Knowledge Basefor Policy-Making Subsidize further research, including operational research and pilot initiatives: This was emphasized in the stakeholder consultations as a pre-requisite for making policy decisions on public-private interaction. Policy decisions must have a solid evidence-base. Filling information gaps and testing new approaches was recognized as essential for enhancing the private sector contribution to health outcomes. 54 To enrich our understanding o f the dynamics o f the private sector, further work should include a more in-depth analysis of: (a) factors market (pharmaceuticals, medical equipment, consumables, etc); (b) labor market dynamics; (c) capital markets; and (d) potential for insurance markets. A broadeningo fthe knowledge base is also necessary inother related subject areas. It must be stressed however, these suggestions for further studies are not to be misconstrued as a reason to delay policy actions for which considerable evidential basis already exists. A distinction must be made between operations research to pilot-test the policy options and the other research activities aiming to generate new knowledge. While the specific areas o f research would need to be determined through future consultations, some suggestions are listedhere: e Conduct a more comprehensive, nationally representative survey o f private sector providers, their consumers (geographical and socio-economic strata), types o f services provided, prices o f these services, quality, client satisfaction, level o f training o f providers, types o f medicine practiced. Such a follow-on survey and training should include all types o f alternative providers as well, particularly traditional birth attendants, and drug store managers; and should include a comparative analysis o f strengths and weaknesses o f private providers, including NGOs. e Carry out econometric analysis o f the private health care market, including an analysis o f the supply-demand curves, price-elasticity, income elasticity, unit-costs and such other aspects that could enlighten the policy-makers more about the economic drivers o f supply of, and demand for, private services. e Conduct a further analysis o f HNP-related commodities (pharmaceuticals, vaccines, baby food formula, hygiene products, bed nets, etc.), to study both supply and demand side factors relevant to these markets and look at ways o f buildingpublic- private partnerships inthese areas. e Conduct a labor market assessment, including the market dynamics o f supply and demand in relation to human resources, and the incentive mechanisms influencing providers inthe private sector. e Conduct a full-fledged study o f governance issues as they apply to private as well as public sector, including the issue o f dual practice. e Good research and development programs (such as those developed by HMOs in the US)might contribute significantly to policy options aimed at improvingquality o f services. Based on a thorough analysis o f the various markets relevant to HNP services, it i s essential to develop a strategy for creating a more competitive environment among the private providers and also between the public and private providers, using public policy as a facilitating tool. Develop better information systems: Expand the health information system to collect more reliable data from the private sector, especially with regard to disease burden but also in private sector resources (practitioners, infrastructure, equipment etc.). Information sharing: Such efforts would aim at disseminating information to policy- makers, managers, and other key actors inthe public sector to enhance their understanding o f the 55 private sector and vice versa. Key data to be shared are: public budgets and resource allocation across districts, health indicators, consumer preferences on health, market analysis on health, best management practices on healthprovision etc. 4.3.3 Promotingprivateinvestmentin heaIth The scaling-up o f private health care providers requires inthe long term o f additional resources from local and foreign investors. Investment will provide financial and physical resources, "known how" and best practices that can be rapidly spread across existing private and public providers with a positive impact in quality and unit cost. The Government should initiate an aggressive policy to encourage private investment in the curative health care sector. Some specific actions are suggested as follows: To install a Suecial Task-Force Unit in the Office of the Primer Minister to help local and foreign investors to seek potential business inthe health sector in Bangladesh. This unit would have to define goals interms o fthe value o fthe investment, andwill coordinate all neededpublic actions to accomplish those objectives. The Unit will have to develop the appropriate legal conditions and to monitor the dissemination o f lease management, concessions as options to contract agreements and also to evaluate the transfer o f ownership o f public health facilities to the private sector. Bonuses to NGOs or private providers that serve inunderserved areas. To encourage alliances between the MOHFW and the private sector to achieve specific public health goals such as immunization, tuberculosis and malaria control etc. Creation o f Health Management Organizations Implementing social insurance for civil servants and private employees. This implies a mandatory earnings contribution (about 4% o f total earnings) that could go either to the Ministry o f Health and Family Welfare providing a right to receive medical attention without fees or to a health maintenance organizations (HMO). This strategy may allow to obtainnot less than 200 milliondollars a year into the health sector to strengthenthe developmentof a competitive marketwith "incentive schemes" and "minimum standards" in the search of better quality and coverage. This option requires the design of a new legislation that promotes the formation o fHMO. The accreditation o f activities and health plans offered by these HMO will be under the responsibility o f the NAAH. The Health observatory/Consumer alliance will inform consumers about the best choices and disseminate user's rights. 4.4 The Way Forward This section provides some options on how the above policy options might be realized into action. The original intention o f the authors was not to be prescriptive about the solutions for the issues emerging from their study. However since the Government is preparing its new Health, Nutrition and Population Sector Program (HNPSP), the findings o f this study and the suggested policy options presented here - both the "what" and the "how"- could serve as a useful vehicle for policy dialogue inBangladeshand also as starting point for the design and implementation o f the sector reforms required to achieve the desired HNP outcomes as specified in the MDGs, in the context o fa choice competitive model. The following are possiblenextsteps: 56 0 Set up a Public-Private Task Force in the MOHFW to provide a focal point for developing processes and activities towards a fruitful relationship between public and private actors. This task force would promote, institutionalize, and coordinate public- private interactions. This would be a temporary measure until durable processes, systems, financing mechanisms, and regulations relating to public/private partnershipin the health sector are well-established and absorbed by the relevant agencies. 0 Create the necessary fiscal space or "head room" inthe public resource envelope, so as to ensure the availability o f the substantial additional resources needed to finance increased engagement with the private sector and the required pilots on a large enough scale to makethem replicable (over and above the current health sector budget, most o f which is already committed to public sector provision). 0 Capacity development inthe MOHFW to enhance its engagement with the private health sector. Such capacity requires the development o f new types o f skills, in areas that have not traditionally been among the functions o f MOHFW such as negotiation, social marketing, monitoring performance based on results etc. The capacity o f other relevant actors, such as civil society, development partners, and private sector actors should also be considered. This would include substantial education for public sector staff on the size and distribution o f private HNP sector in Bangladesh, and ways o f building partnerships with private actors; similarly, training and confidence-building activities targeted at the private sector would also be critical. 0 Participatory policy-making and more inclusive planning and programming: The findings o f this and other related studies should be actively and widely disseminated, as input for the for national policy debate, involving stakeholders at all levels. Private sector actors should be included in such process and a full consideration given to strengthen private sector capacity in service and facility planning. Participatory policy debate should take advantage o f the preparation o f the full PRSP from the I-PRSP. 0 Establishment o f anAccreditation System An Accreditation body would work with consumers, health care purchasers, legislators and the providers in developing standards for both public and private providers in Bangladesh. Participation in accreditation and certification programs would be voluntary. This would create incentives to good providers to distinguishthemselves from badproviders. The NAAHshould evaluate health care inthree different ways; a. Through accreditation (a rigorous on -site assessment o f key clinical and administrative processes) b. Through healthplanemployer data ifthey exist. c. Througha comprehensive user's andprovider's surveys. Since regulation i s usually costly and often ineffective, accreditation is recommended to "protect a minimal set of standards on health service provision". Non-regulatory interventions or incentives would also be essential for improving private sector activity. 57 The NAAHmay use several factors for the elaboration o fthe rankingo fproviders such as: a. Access and service -Are there enough doctors and specialists to serve the number o fpatients? -Which are the services o f the health facility? -What i s the access to training o f the healthworkers and practitioners inthe health facility? -Sufficient and adequate resources: human (health personnel), equipment, and infrastructure o f health facilities. -Availability o f diagnostic tests and drugs Qualifications o fproviders -Ensuring each doctor is licensed and trained to practice medicine and that the users are satisfied with the services received bypractitioners. -How do the providers rate their own doctors? -How do the providers deal with bad practices and complaints against their practitioners or healthworkers? -Which i s the level o f cleanliness o f the health facility? -Which i s the level o f access to water, electricity and other services inthe health facility? -What i s the quality o f the food providedby the health facility? -What i s the ratio o fpractitioners/administrativeworkers inthe health facility? -What i s the status o f financial procedures? -Does the hospital have clear and written standards and protocols for treatment and diagnosis? -What i s the average waiting time o f a patient before he/she receives treatment? b. Healthoutcomes -Are the health activities o f the provider helping people to stay healthy? -Are preventivehealth activities appropriate? -Evaluate whether people recover effectively from illness. c. User's satisfaction Measure by the level o f comfort, promptness of service, waiting time to receive the medical attention, treatment by the doctors and health workers, privacy, availability o f services and food receivedby the patient. To be eligibly for accreditation, a health provider would need to be in operation for a minimumo fthree years. The rankingo f providers should be simple and easy to understand for the public such as: excellent, good, satisfactory, below average, fail. The NAAHwill use two options for accrediting healthcare providers: a. D o the inspections directly with own personnel. b. Contractinginspectors from other NGOsandprivate institutions. Both options will demand special training o fpersonnel. - Creation of a Health observatory/Consumer Coalition The mainactivities ofthe HealthObservatory/Consumer Coalitionwould be the following: 58 a. Advocacy Help to educate, assist and protect the rights o f individuals through consumer information, consumer participation, consumer advocacy programs, data collection and independent quality oversight. Draft model policies or legislation on the areas o f interest. Help consumers know about options o f coverage, provision and treatment. b. Grassroots organizing Broadencoalitions by creating worker/consumers partnerships at the local andnational level. Implement a quality watch-line (toll free number) which collects individual experiences o f poor quality o f care services from consumers or health workers in the country. These would be real life stories very useful for the design ofpro-consumer strategies inthe health sector. c. Participation inquality measurement Ensure that the consumer voice is listened to inall forums andwork groups relatedto legislation or decisions on the health sector. d. Accountability The observatory should be independent of providers and financers o f health care and free of conflict o f interests. The agency should collect and disseminate information to strengthen the national data system. A National Contracting Agency The formulation o f a strategy framework for purchasing services from NGOs and other non- public providersshould be defined clearly. The themes for further definition andwork are: a. The services to be purchased. b. Criteria for choosing the geographical areas o f interventions. The proposal is to start in the unions were there are already NGOs delivering health services. c. The procedures to monitor andto supervise the performance o f the NGOs. d. The establishment o fthe contract agreements andbiddingprocess. e. Types o f partnership between the GOB andNGOs. f. The payment mechanisms to providers. One option is to initiate the process with a simpleper capita allocation. g. To give the contracted NGOs the possibility ofretention o fuser fees. h. Accreditation shouldbemanagedbyanindependent agency. A private institution (The National Contracting Agency) would be contracted by the GOB to manage the bidding process, and the supervision and monitoring o f performance, together with the training o fhealthworkers inthe NGOs. A permanent GOB-Civil Society committee lead by the MOHFW would guide and provide inputs to the National Contracting Agency on the national goals o f targeting the poor. The National Contracting Agency would beresponsible for: 59 1. The decision o fwhat to contract out. 2. The decision of from whom to purchase services. 3. The definition o fthe contract payment option. 4. To negotiate the terms o f the contract with the selected provider. 5. To supervise andto monitor the performance of the contracted provider. 6. To modify contracts base on performance. 7. To promote the formation o fpublic andprivate healthnetworks at the unionlevel. The NCA will need also to develop indicators to measure contracting outcomes such as: Improvements in average health indicators with significant reduction in disparities between the rich and the poor, betweenmale and female. Infant Mortality Rates reduced to a target rate with a significant decrease inrich/ poor , male/female ratios. Under-5 Mortality Rates reducedto a target rate with a significant decrease in rich/poor, male/female ratios. Maternal Mortality Ratio (or suitable proxy) reduced to a target rate with a significant decrease inricldpoor ratio. Percentage o f fblly immunized children against 6 diseases within the first year o f life increased to x %. Proportion o fwomen with obstetric complications treated at facilities increased to x %. Use o f curative essential services by women, children andthe poor increased to x %. Discontinuation rate o f contraceptionreduced to x %. Proportion o fwomen who receive antenatal care. Proportion o fwomen who receive post-natal care. Fertility rates. Pilot activities to test the selected policy options. Selected policy options need to be tested through operations research to determine their feasibility and measure their impact in the Bangladeshi context. The design o f such pilots should build on previous experience, and evidence available about public-private partnerships such as demand-side financing, micro-insurance schemes, and contracting with results-based financing. A summary o fissues and suggested policy options is given intable 4.2. 60 A preliminary timeline for the next steps is suggested subject to previous agreement with the Government: Completion o f other related studies: BetweenAugust - Governance issues inthe health sector. 2003 and May 2004 - Pro-poor targeting mechanisms. - NGO contracting evaluation. - Comparative study of cost-effectiveness among public andprivate health care providers. Dissemination o f existing evidence, multi-pronged communication From July to exercise, consultations across the country December 2003 Development o f broad-based HNP Policy Options dialogue May 2004 ~ ~ ~ Initiation of Pilot Interventions (e.g., vouchers, micro-insurance, methods Starting January for contracting-out non-public providers, demand-side subsidies such as 2004 cash transfers to the poor ) "1 i r c fi .I x 4 4 a cl s 75 e, 2 0 0 N r- N e 29 .3 F F B2 References ACPR and UNICEF (2001) Review of availability and use of emergency obstetric care (EmOC) services in Bangladesh. Dhaka: ACPR andUNICEF. Afifi, N.H,, R Busse, Harding, A.L. (2003) Regulation of Health Services, inHarding, A.L. and Preker, A.S. (eds.) Private Participation in Health Services. Washington, DC: World Bank. Ahmed S. (2002) The Politics of Reforms in South Asia: Bangladesh and Pakistan. South Asia Region Internal Discussion Paper. Washington, DC: World Bank. Ali M., Emch, M.,Tofail, F. and Baqui, A.H. (2001) Implications of health care provisionon acute lower respiratory infectionmortalityinBangladeshi children. Social Science and Medicine, 52: 267-277. Ali, Q.L.(2001) Calculation oftotal unit cost for diarrhoeal management at district hospital and Thanahealthcomplex. Dhaka: HealthEconomics Unit,Ministry o fHealthandFamily Welfare. Alisjahbana A., Williams, C., Dharmayanti, R., Hermawan, D., Kwast, B.E., and Koblinsky, M. (1995) An Integrated Village Matemity Service to Improve Referral Pattems in a Rural area in West-Java. InternationalJournal of Gynecology & Obstetrics, 48(Suppl): 83-94. Amin, R., Chowdhury, S.A., Kamal, G.M. and Chowdhury, J. (1989) Community Health Services and Health Care Utilization inRural Bangladesh. Social Science and Medicine, 29(12): 1343-1349. Amhold, R.G.(1979) Paramedicalprograms inruralBangladesh. SouthMed J, 72(8): 992-996. Ashraf, A., Chowdhury, S. and Streefland, P. (1982) Health, Disease and Health-Care inRural Bangladesh. Social Scienceand Medicine, 16: 2041-2054. Axelsson, H.,Bustreo, F. and Harding, A. (2003) Private sectorparticipation in child health: A review of WorldBankprojects 1993-2002. Washington, DC: World Bank.Inpress. BangladeshBureau o f Statistics (1999a) Statistical Yearbook of Bangladesh. Twentieth Edition. Dhaka: PlanningDivision, MinistryofPlanning. Bangladesh Bureau o f Statistics (1999b) Health situation and health care expenditures in Bangladesh: evidencesfrom nationally representative surveys. BangladeshBureau of Statistics (2001) Preliminary report of household income and expenditure survey -2000. BangladeshBureau of Statistics (2002) 2000 Statistical Yearbook of Bangladesh, 2IsfEdition. Dhaka: PlanningDivision, MinistryofPlanning. BangladeshBureau of Statistics (2002) Report of the Labour Force Survey Bangladesh 1999- 2000. Dhaka: Planning Division, Ministryof Planning. BangladeshBureau of Statistics andUNICEF (2000) Progotir Pathey: On the road toprogress - Achieving the goals for children in Bangladesh. Report on Multiple Indicator Cluster Survey (MICS) 2000. BangladeshBureau o f Statistics (1998). Report on survey of private health service establishment 1997-1998.Dhaka: Statistics Division, Ministry of Planning. Bangladesh Medical Association (2002) Proposed policy framework for "Health Sector Program (HSP: 2004-2006) ". Draft submittedto The MinistryofHealth and FamilyWelfare. BangladeshMedical andDental Council (2003) Personalcommunication. BangladeshNursing Council (2003) Personalcommunication. Baqui, A.H., Sabir, A.A., Begum, N.,Arifeen, S.E., Mitra, S.N. and Black, R.E. (2001) Causes of childhood deaths inBangladesh: an update.Acta Paediatrica, 90:682-690. Barkat, A., Ara, R., Huque, M. and Syeduzzaman, K.M. (2003) Case Studies in privatehublie partnership in the HNP sector in Bangladesh: Preliminary report. Background study to the Private Sector Assessment. Unpublished. Begum, H. (1998). Health care, ethics and nursing in Bangladesh: a personal perspective. Nursing Ethics, 5(6), 535-541. Bhardwaj, S.M. andPaul, B.K. (1986) Medical Pluralismand Infant Mortality ina Rural Area of Bangladesh. Social Science and Medicine, 23(10): 1003-1010. Bhuiya, A. (1992) Village Health Care Providers in Matlab, Bangladesh: A Study of Their Knowledge inthe Management o f Childhood Diarrhoea. JDiarrhoeal Dis Res, 1O( 1): 10-15. Bhuiya, A. and Streatfield, K. (1995) Feeding, home-remedy practices, and consultation with health care providers during childhood illness in rural Bangladesh. Journal of Diarrhoeal Dis Res, 13(2): 106-112. Bhushan, I. et al(2002) Achieving the twin objectives of efjciency and equity: contracting health services in Cambodia. Manila: Asian Development Bank, Economics and ResearchDepartment PolicyBriefNumber6. Buchan, J. (2000). Nursingbrain drain. Nurse Stand, 15(4), 22-23. Cash, K.et al(2001). Telling them their own stories: legitimizing sexual andreproductive health educationinruralBangladesh.Sex Education, 1:43-57. Centre for Software and Info-processing (1995) Report of the Survey on Professional and Miscellaneous ServicesPersonnel. BangladeshBureauo f Statistics. Chakraborty, S. et al (2000) Improving private practitioner care o f sick children: testing new approachesinruralBihar.Health Policy and Planning, 15(4): 400-407. Chanda, R. (2002) Trade in health services. Bulletin of the World Health Organization, 80(2), 158-163. Chaudhury, N. and Hammer, J.S. (2002) Ghost Doctors: Absenteeism in Bangladeshi Health Facilities. Washington: World Bank, draft mimeo. CIET Canada and MOHFW (1999) Health and Population Sector Program 1998-2003 Bangladesh, Baseline Service Delivery Survey: FinalReport. Claquin, P. (1981) Private Health Care Providers in Rural Bangladesh. Social Science and Medicine, 15: 153-157. Data International (1998) Bangladesh National Health Accounts 1996/97:Final Report. Dhaka: Prepared for the HealthEconomics Unit,Ministryof Health and FamilyWelfare. DKT Intemational(2003) Accessed at www.dktintemational.org. Dmytraczenko, T. et a1 (1999a) Bolivia: a mid course assessment of the National Mother Child Health InsuranceProgram. Health Reform Prior Sen, Summer-Fall: 19-20. Dmytraczenko, T., Scribner, S., Leighton, C., Novak, K. (1999b) Reducing maternal and child mortality in Bolivia. Bethesda, MD:Partnershipsfor Health Reform. Ensor, T. (2001) Projecting the cost of the Essential Service Package. Health Economics Unit, Dhaka. ResearchPaper26. Ensor, T., Dave-Sen, P., Ali, L., Hossain, A., Begum, S.A. and Moral, H. (2002) Do essential service packages benefit the poor? Preliminary evidence from Bangladesh. Health Policy and Planning, 17(3): 247-256. Feldman, S. (1983) The Use of Private HealthCare Providers inRuralBangladesh: A Response to Claquin. Social Scienceand Medicine, 17(23): 1887-1896. Forsberg, B.C. and Axelsson, H. (2003). Interviews with Senior Managers in the Public and Private Health Sector on Public Policies versus Private Care and Experience on Private Care. Background study to the Private Sector Assessment. Government of Bangladesh and UNICEF (1999) Situation assessment and analysis of children and women in Bangladesh. Greenwood, A.M., Bradley, A.K., Bypass, P., Greenwood, R.W. et a1 (1990) Evaluation o f a primaryhealth care program inthe Gambia. The impact o ftrained traditional birthattendantson the outcome ofpregnancy.Journal of TropicalMedicine and Hygiene, 93: 58-66. Gruen, R., Anwar, R., Begum, T., Killingsworth, J.R. and Normand, C. (2002) Dual Job holdingpractitioners inBangladesh: an exploration. Social Scienceand Medicine, 54: 267-279. Gwatkin, D.R., Rustein, S., Johnson, K., Pande, R.P. and Wagstaff, A. (2000) Socio-economic differences in Health, Nutrition, and Population in Bangladesh. Washington, DC: World Bank. Harding, A.L. and Preker, A.S. (2003) Private Participation in Health Services. Washington, DC: World Bank. Hay R., et al. (2002) Health Futures in Bangladesh: Some Key Issues and Options. Report to The World Bank (Unpublished). Health Economics Unit (HEU) (1998a) Private Sector Medical Clinics and Hospitals Survey. Research PaperNo. 15. Dhaka: MinistryofHealth and FamilyWelfare. HEU(1998b). Economic aspects of human resource development in health andfamily planning in Bangladesh: Costs of education and training of health workers in Bangladesh. ResearchPaper No. 14.Dhaka :Ministryof Health andFamilyWelfare. HEU, Institute o f Health Economics (ME), National Institute for Population Research and Training (NIPORT) (2003) Private HNP sector assessment: Jindings of surveys on private providers and consumers. Background study to Private Sector Assessment. Unpublished. Hossain, S.M.,Bhuiya, A., and S. Rasheed(2001) Correlates of perceived malarial episodes and treatment-seeking behavior in a malaria-endemic rural area in Bangladesh. Southeast Asian Journal of TropicalMedicine and Public Health, 32(4): 707-719. Human Resources Development Unit. (2003). Task forces reports on human resources development in health, nutrition and population sector. Dhaka: Ministry of Health and Family Welfare. Huang, P (1995) "An overview of hospital accreditation in Taiwan, Republic of China", The international Journal o fHealthPlanning andManagement. 10(3), 183-192. Hye, H.K.M.A. (2003) Health Regulation Review. Background study to Private Sector Assessment. Unpublished. Islam, M.A., Wakai, S,Ishikawa N., Chowdhury, A.M., Vaughan, J.P. (2002) Cost-effectiveness of community healthworkers intuberculosis control inBangladesh.Bulletin of the World Health Organization, SO(6):445-50. Khan, S.H., Chowdhury, A.M., Karim, F., and Barua, M.K. (1998) Training and retaining Shasthyo Shebika: reasons for turnover of community health workers in Bangladesh. Health Care Superv, 17(1):37-47 Kumaresan, J.A., de Colombani, P., and Karim, E. (2000) Tuberculosis and health sector reform inBangladesh.IntJTuberc LungDis, 4(7):615-21. Leslie, C. (ed.) (1976) Asian Medical Systems. University o f CaliforniaPress. Levin, A., Rahman, M.A., Quayyum, Z., Routh, S. and Barkat-e-Khuda (2001) The demand for child curative care in two rural thanas of Bangladesh: effect of income and women's employment. International Journal of Health Planning and Management, 16: 179-194. Loevinsohn, B. (2000) Contractingfor the delivery of primay health care in Cambodia: design and initial experience of a large pilot-test. Washington, DC: World Bank Institute Flagship Program Online Journal. Luby, S. et al (2002) Improving private practitioner sick-child case management in two urban communities inPakistan. TropMed Int Health, 7(3): 210-219. Newbrander William (1999). `Accreditation of providers for the National Health InsuranceFund of Tanzania". Management Sciences for Health, 2000. Marek, T. et a1 (1999) Successfbl contracting o f prevention services: fighting malnutrition in SenegalandMadagascar.Health Policy and Planning, 14(4): 382-389. Miller, L.C., Jami-Imam, F., Timouri, M., and Wijnker, J. (1995) Trained traditional birth attendants as educators ofrefbgeemothers. WorldHealth Forum, 16(2): 151-156. Ministry of Finance, Government of Bangladesh (2002) A national strategy for economic growth, poverty seduction and social development. Ministry of Health and Family Welfare, Government of Bangladesh(2001) Public Expenditure Review (2000/01) of the Health and Population Sector Program. Ministry of Health and Family Welfare, Government of Bangladesh (2002) Conceptual Framework on Health, Nutrition and Population Sector Program (2003-2006): Interim Draft. Planning Wing. MinistryofHealthandFamilyWelfare, Government ofBangladesh(2001)Bangladesh National Strategyfor Maternal Health. National Institute of Population Research and Training (NIPORT) (2002) Bangladesh Maternal Health Services and Maternal Mortality Survey 2001: Preliminay Report. Nieves, I., Forgia, G.M., and Ribera, J. (2000) Large-scale government contracting of NGOs La to extend basic health services topoor populations in Guatemala. Washington, DC: IESE, Inter- American Development Bank, World Bank. NIPORT, Mitra and Associates, and ORC Macro (2001) Bangladesh Demographic and Health Survey 1999-2000. Dhaka, Bangladeshand Calverton, USA. ORC Macro (2003) Demographic and Health Surveys - STAT Compiler. Available at www.measuredhs.com. Accessed April 13,2003. ORG-Marg Quest Ltd. (2000a) Report on Medical Practitioners and Pharmacists in Bangladesh. Dhaka: Report for BritishCouncil andNICARE. ORG-Marg Quest Ltd. (2000b) Survey on Village Doctors Practicing in Brahmanpara. Dhaka: Report for BritishCouncil andNICARE. O'Rourke (1994) An Evaluation of a TBA Training Program: Its Effect on TBA Practices and Perinatal Mortality. PhD Thesis. Amherst, MA: University o f Massachusetts, School o f Public Health. Pang, T., Lansang, M.A., and Haines, A. (2002). Brain drain and health professionals: a global problem needs global solutions. British Medical Journal, 324: 499-500. Peters, D.H. (2002) Oversight o f Health Sector Policy and Performance in Low-Income Countries: The Example of Sexual and Reproductive Health Services in India. Reproductive Health Matters, lO(20): 82-94. Peters, D.H., Yazbeck, A.S., Sharma, R.R., Ramana, G.N.V., Pritchett, L.H. and Wagstaff, A. (2002) Better health systemsfos India 's poor: findings, analysis, and options. Health, Nutrition, andPopulation Series. Washington, DC: HumanDevelopment Network, World Bank. Peters, D.H. et a1 (2003) Labor market assessment for the HNP sector in Bangladesh. Unpublished. Policy andResearchUnit (2002) HRDData Sheet 2001. Ministryof Healthand Family Welfare Rosen, James "Contrating for Reproductive Health Care: A Guide". The World Bank, December 2000. Saade, C. et a1 (2001) The stoy of a successful public-privatepartnership in Central America: handwashing for diarrheal disease prevention. Arlington, VA: BASICS 11, EHP, UNICEF, USAID, WorldBank. Sarder, A.M. and Chen, L.C. (1981) Distributionand Characteristicso fNon-Government Health Practitioners is a RuralArea of Bangladesh. Social Scienceand Medicine, 15: 543-550. Schellenberg, J.R. et a1 (1999) KINET: a social marketing program o f treated nets and net treatment for malaria control inTanzania, with evaluation of childhealth and long-term survival. TransR Soc TropMed Hyg, 93(3): 225-231. Schellenberg, J.R. et a1 (2001) Effect of large-scale social marketing o f insecticide-treated nets on child survival inrural Tanzania. Lancet, 357(9264): 1241-1247. Schuler, S.R., Bates, L.M. and Islam, M.K. (2002) Paying for reproductive health services in Bangladesh: intersections between cost, quality and culture. Health Policy and Planning, 17(3): 273-280. Sen, B. (2001) The health divide: analysis of inequalities in health in Bangladesh. Dhaka: BangladeshInstitute for Development Studies. Smith, E., Brugha, R. and Zwi, A. (2001). Working with Private Sector Providers for Better Health Care, An Introductoy Guide. London: Options and London School o f Tropical Hygiene andMedicine. Smith, E. (2002) Social Franchising Reproductive Health Services: Can It Work - A Review of the Experience. Working PaperNo. 5. London: Marie Stopes International. Social Marketing Company (2003) Personalcommunication. Soeters, R. and Griffiths, F. (2003) Improving government health services through contract management: a case from Cambodia. Health Policy and Planning, 18(1):74-83. Stenson, B. et a1 (2001a) Real world pharmacy: assessing the quality of private pharmacy practice inthe Lao People'sDemocratic Republic. Social Science and Medicine, 52(3): 393-404. Stenson, B. et a1 (2001b) Private pharmacy practice and regulation. A randomized trial in Lao P.D.R. Int JTechnol Assess Health Care, 17(4): 579-589. Streatfield, P.K., Hadley, M. and Chakraborty, N. (2001) Status of program indicators: annual program review 2001 of Health and Population Sector Program 1998-2003. Dhaka: World Bank. Survey Research Group o f Bangladesh. (2003) Research for Development of Alternative Incentive Schemefor HRD in Health & Population Sector. Dhaka: Draft Report to the Policy and ResearchUnit, Ministry of Health & Family Welfare. SustainableHealthcare Enterprise Foundation(2003) Accessed at http://www. shefoundation.org/projects/index.html Taylor, R. (2003) Contracting for Health Services, in Harding, A.L. and Preker, AS. (eds.) Private Participation in Health Services. Washington, DC: World Bank. UNB(2002) NBRplansfresh huntfor new taxpayers.The Independent,December12. UNICEF (1999) Situation Assessment and Analysis of the Children and Women in Bangladesh. Dhaka: UNICEF. Walt, G. (1988) Community Health Workers: Are National Programs in Crisis? Health Policy and Planning, 3(a): 1-21. Winch, P. (2003) Personalcommunication. World Bank and Asian Development Bank (2002) Poverty in Bangladesh: Building on Progress, joint report by the WorldBank and Asian Development Bank. World Bank (1998) Health and Population Program Project. Project Appraisal Document. World Bank (2000a) Bangladesh: Financial Sector Distress and Lost Economic Growth. Washington, DC: World Bank. World Bank (2000b) Corruption in Bangladesh, Costs and Cures. Dhaka: WorldBank World Bank (2001) BangladeshCountry Assistance Strategy. Dhaka: World Bank. World Bank (2003) HNP Poverty Net. World Health Organization (WHO) and John Snow Inc.(2002) County ProJile on Child Health and Development in Bangladesh. Dhaka: WHO. WHO (2003a) SelectedIndicatorsfor Bangladesh. Accessed at http://www3 .who.int/whosis/country WHO (2003b) Report of county visit to Bangladesh, B. Daelmans. WHO (2003~)WHO EstimatesofHealthPersonnel. Accessed at: http://www3.who.int/whosis/health Dersonnel/health personnel.cfm Young, A. (1983) The relevance of traditional medical cultures to modern primary health care. Social Scienceand Medicine, 17: 1205-1211.