33646 Africa Region Human Development Working Paper Series Trends and Opportunities in Public-private Partnerships to Improve Health Service Delivery in Africa By Tonia Marek, Catherine O'Farrell, Chiaki Yamamoto, Ilyse Zable* Contributors: World Bank (Africa Region, HDNHE, INFPI) IFC Sara Project/USAID Copyright ©August 2005 Human Development Sector Africa Region The World Bank The views, findings, interpretations, and conclusions expressed herein are those of the authors and do not necessarily reflect the opinions or policies of the World Bank or anyof its affiliated organizations. Cover design: The Word Express, Inc. Typography: The Word Express, Inc. Cover photo: Ophelie Marek, picture of a boat on the Niger river in Mopti, Mali, 2004 ii Contents Acknowledgements .......................................................................................................................v Foreword ................................................................................................................................... ix List of Acronyms ........................................................................................................................vii Executive Summary .................................................................................................................... xi I. Why Focus on Public-Private Partnerships in Africa? ....................................................... 1 The facts which destroy the myths ................................................................................... 1 a. Who uses the private sector and who uses the public sector? ...................................... 1 b. Who finances health in Africa? ................................................................................... 2 c. How extensive is private sector coverage?................................................................... 3 d. What are the issues that need to be tackled to avoid "business as usual"? .................. 5 e. Other reasons why we can't continue `business as usual' ............................................ 6 Purpose and scope of this paper ....................................................................................... 7 II. Who Delivers What Kind of Health Services in Africa? .................................................... 9 a. Where do people actually go for services? ................................................................... 9 b. Do people get good quality care in public and private sectors? ................................. 11 c. Are private services more expensive than public services? ......................................... 11 d. Is the private sector draining the public sector from its human resources? ................ 13 III. Experiences and Trends in Public-Private Partnerships in Africa .................................... 17 a. Public financing of the private sector ........................................................................ 17 i. Health Insurance ................................................................................................. 17 ii. Demand-driven community schemes ................................................................... 17 iii. Vouchers ............................................................................................................. 20 b. Regulating the private sector..................................................................................... 21 c. Formal public-private partnerships for service delivery ............................................. 24 i. Contracting the provision of health services ........................................................ 24 ii. Leasing of equipment and facilities ..................................................................... 28 iii. Concessions ........................................................................................................ 28 iv. Divestitures ......................................................................................................... 30 v. Franchising ......................................................................................................... 31 d. Facilitation................................................................................................................ 32 iii IV TRENDS AND OPPORTUNITIES IN PUBLIC-PRIVATE PARTNERSHIPS TO MPROVE I HEALTH SERVICE DELIVERY IN AFRICA e. Shifts in paradigm, trends and opportunities to grasp ............................................... 33 i. Public-public performance-based contracting ...................................................... 33 ii. Decentralization .................................................................................................. 33 iii. Private sector human resources ........................................................................... 33 iv. Private Internet communication technology helps service delivery....................... 35 v. Globalization of public-private partnerships for service delivery ......................... 36 IV. Private Sector and Governments' Challenges and Possible Responses ............................ 37 a. Private sector challenges ........................................................................................... 37 b. African Governments' challenges .............................................................................. 38 V. The World Bank's and IFC's Challenges and Possible Responses.................................... 41 a. Portfolio review of what the World Bank is doing on PPPs in Africa......................... 41 b. What IFC is doing in PPPs ........................................................................................ 42 c. The World Bank's challenges to contribute to PPPs ................................................... 43 VI. Conclusion ..................................................................................................................... 45 Annexes 1. Where Do People Go for Services? ............................................................................ 47 2. Case Studies of Franchises in Africa .......................................................................... 73 3. Case Studies of Vouchers in Africa ............................................................................ 79 4. Case Studies of Contracting for Health Service Provision in Africa ........................... 83 References.................................................................................................................................. 87 iv Acknowledgements F inancial support from a Dutch Trust Solomon Orero, KMET, Kenya Fund, an Irish Trust Fund, and from J. J. Wirima, Mwaiwathu, Private Hospital, USAID through the Sara Project was Malawi provided to carry out this study and K. Jameson Makemele, SUKA Healthcare, South is gratefully acknowledged. Africa The following people contributed greatly to Jubi C. Nxumalo, SMM Healthcare, South this report as participants to the "Workshop on Africa public-private partnership for health in Africa," P. Rugarabamu, HKMU Hubert Kariuki Memo- which took place June 28­29, 2004, in Nairobi, rial University, Tanzania Kenya, with the intention to provide inputs to S. M. A. Hashim, Association of Private Hospi- the first draft of the present paper, and the au- tals in Tanzania thors are most grateful to them: Harold Bisase, Uganda Private Medical Practi- tioners Association, Uganda From the NGO sector: Nina Shalita, Sarah Nabembezi Uganda Private Philibert Kankye, Christian Health Association Midwives Association, Uganda of Ghana (CCHAG), Ghana Ibrahim Aliou Sall, Reseau Sante Sida Popula- From the public sector: tion/Conseil Appui Developement, Senegal James Nyikal, Director of Medical Services Magatte Wade, Agetip, Senegal Nairobi, Ministry of Health, Kenya Peter Lochoro, Uganda Catholic Medical Bu- Bande Karim, Hien Thadde, Zoungranne reau, Uganda Agathe, Ministry of Health, Burkina Faso Michelle Folsam, Christy Hanson, PATH Inter- Cecilia Bentsi, Ministry of Health, Ghana national L. Ofosu, SPMDP, Ghana Mame Cor Ndour, CAS/PNDS, Ministry of From the private sector: Health, Senegal Edwin Muinga, Sam Thenya , Kenya Associa- R. Eddie Mhlanga, University Kwa-Zulu tion for Private Hospitals, Kenya Natal, South Africa S. Ochiel, Kenya Medical Association, Kenya George Bagambisa, Ministry of Health, Jagi Gakunju, AAR, Kenya Uganda v VI TRENDSAND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA The authors would also like to thank Kim views of the World Bank Group, its Executive Beer, Moncef Bouhafa, François Diop/Abt As- Directors, or the countries that they represent sociates, Mohsen Farza, Elisabeth Sommerfelt and should not be attributed to them. and the following colleagues from the World Bank: Eric de Roodenbeke, Gilles Dussault, April Harding, Kathy Khu, Benjamin * Tonia Marek is Lead Public Health Specialist in Loevinsohn, Elsie Lauretta Maka, Oscar Picazo, the Africa Region of the World Bank, contact: Alexander Preker and Khama Rogo for their tmarek@worldbank.org constructive comments and contributions. Catherine O'Farrell is Investment Officer in the Advisory Services department of IFC Chiaki Yamamoto is Private Sector Development The findings, interpretations, and conclusions Specialist in the INFPI Unit of the World Bank expressed in this paper are entirely those of the Ilyse Zable is Economist in the Health and Edu- authors. They do not necessarily represent the cation Department of IFC vi Foreword T here is a growing awareness concern- It has been recently recognized in the World ing the importance of the private Development Report 2004 and the paper "Im- sector in the provision of health ser- proving Health, Nutrition and Population vices in sub-Saharan Africa. Private Outcomes in sub-Saharan Africa: The Role of sector providers can range from traditional heal- the World Bank", that when incorporated judi- ers, informal drug vendors and private for-profit ciously into national health plans, the private providers, to non-governmental organizations sector can be an efficient and effective comple- and community groups. Whether it is through ment to the existing public health services. an offer of lower cost, the provision of higher Interventions such as national health insurance quality care, or a greater sensitivity to patients schemes, vouchers, tax exemptions, accredita- mistrustful of public institutions, private pro- tion, franchising, contracting, and concessions viders are often chosen by Africans of all are just a few of the possibilities for public-pri- socio-economic backgrounds in need of health vate partnerships in health. Although generally care above government health services. This recently introduced and evaluated in various sub- strong patronage, in combination with the ex- Saharan African countries, there have been penditure of the private sector for health service numerous successes and lessons learned upon provision, constitutes more than 50% of total which to build. There is much work to be done health expenditures for many sub-Saharan Af- to improve progress toward attainment of the rican countries and shows that there may be health-related MDGs, and it will thus be im- more money flowing within the health sector perative to fully utilize all the means at our than has been traditionally accounted for. As this disposal. paper illustrates, without greater cognizance as to the extent of private health services and the ability of African patients to choose them, any analysis and design of health systems will be Ok Pannenborg incomplete. Senior Health Advisor and Sector Leader for Health, Nutrition and Population Human Development Africa Region vii VIII TRENDSANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA List of Acronyms AED Academy for Educational IT Information Technology Development MD Medical Doctor BOO Build-Own-Operate MDG Millennium Development BOOT Build-Own-Operate-Transfer Goals BOT Build-Operate-Transfer MHO Mutual Health Insurance CBO Community Based Organization Organization MOH Ministry of Health DHS Demographic and Health NGO Non-Governmental Survey Organization GIE Groupement d'Intéret OBA Output-Based Aid Economique (legally formed OECD Organization for Economic entity of people who can be Cooperation and Development contracted for services) OOP Out-of-pocket expenditures HIPC Highly Indebted Poor PNFP Private Not for profit Countries PPP Public-private partnerships for HIV/AIDS Human Immuno-Deficiency/ health services Acquired Immuno-Deficiency SSA Sub-Saharan Africa Syndrome STD Sexually Transmitted Disease IDA International Development USAID United Status Agency for Association International Development IFC International Finance WHO World Health Organization Corporation ZHAC Zambia Health Accreditation ILO International Labor Council Organization viii Executive Summary T he report, in its first part, destroys So, how to engage the private sector effec- three common myths regarding tively? The second part of this report tries to the private health care sector in answer with examples of some successful pub- Africa: lic-private partnerships (PPPs), using the broad definition of PPPs in health as any formal ar- · First, that the private sector is for the rich rangement between government and a private and the public sector for the poor. As dem- entity established for the purpose of providing onstrated here, the poorer segments of the health services. The report also highlights some population do use the private sector exten- of the new trends in public-private partnerships sively, and the public sector does substantially and how to make use of opportunities which subsidize richer people who use its services. present themselves. Finally, it identifies what · Second, that health is mainly financed by the governments, the private sector, and the World public sector. In fact, the public sector finances Bank Group could do better. less than half of total health expenditures. The It's clear today that unless one considers the rest is being financed from out-of-pocket, health system in its entirety, the full spectrum of which goes primarily to buying services from options for improving health outcomes will not the private sector. In most African countries, be utilized. Considering the limitations on pub- the private sector plays a more significant role lic health budgets and the reality of than government, especially when compared out-of-pocket spending flowing toward the pri- to OECD countries where public financing vate sector, it is time to bring the private sector provides the majority of resources. into the fold as an ally in the struggle to provide · Third, that the private sector is not very de- higher quality services to a greater number of veloped in most African countries. In people. PPPs are a way to optimize the use of fact, in most countries the private sector available resources. provides a third or more of all health services. It is expected that the information in this pa- per will be updated every two years or so, as Yet, in Africa most governments and aid or- more evaluation data become available and as ganizations focus on public delivery of health more public-private partnerships are imple- services. mented. ix CHAPTER 1 Why Focus on Public-private Partnerships in Africa? I n this report, private sector providers some countries the private sector provides are understood as any service provid- half of all health services, and in most coun- ers who are not from the public sec- tries it provides about one third of services. tor. In particular, they include private for-profit providers, traditional healers, NGOs, Yet, in Africa most governments and aid or- community groups, and informal drug vendors. ganizations focus on public delivery of health There are three myths regarding the private services. health sector in Africa: · First, that the private sector is for the rich The Facts which Destroy the Myths and the public sector for the poor. As dem- onstrated in this report, the poorer segments a. Who uses the private sector and who of the population do use the private sector uses the public sector? extensively, and the public sector does sub- stantially subsidize richer people who use its Most people in Africa spend their health care services. money on private services, as Graph 1 shows · Second, that health is mainly financed by the for two countries. In addition, the rich seem to public sector. In fact, the public sector fi- often benefit more from the public sector than nances less than half of total health the poor. This was shown in a benefit-incidence expenditures. The rest is being financed from analysis undertaken in Guinea, where it was out-of-pocket, which goes primarily to buy- found that only 20% of those defined as poor ing services from the private sector. In most benefit from MOH spending on health care, in African countries, the private sector plays a comparison to 35% of those defined as rich. One more significant role than government, espe- of the main reasons public spending is pro-rich cially when compared to OECD countries, is that the rich mostly benefit from MOH spend- where public financing provides the major- ing on equipment-intensive hospitals and tertiary ity of resources. care facilities which cater to the urban elite. The · Third, that the private sector is not very de- Guinea data indicate that 40% of MOH spend- veloped in most African countries. In fact, in ing on hospitals entirely benefits 20% of the rich, 1 2 TRENDSAND OPPORTUNITIES INPUBLIC-PRIVATE PARTNERSHIPS TO MPROVE I HEALTH SERVICE DELIVERYINAFRICA while 40% of the poor benefit from only 14% proportion of the two poorest quintiles also use of MOH spending on hospitalsI. private providers, around 17% in Zambia and Similarly, in Mauritania, benefit-incidence 40% in S. Africa. analysis showed that 40% of the richest people consume 72% of public subsidies given to hos- b. Who finances health in Africa? pitals, while the 20% poorest people only benefit from 2% of those subsidies. The situa- About half of health expenditures in Africa are tion is more equitable for subsidies to primary privateIIIas Graph 2 shows. One should note that care, where 53% of public subsidies for primary the Africa part of this graph is based on data care are consumed by the richest 40%, and 11% from only 10 Eastern and Southern African of public subsidies for primary care are con- countries. Data from other countries has not sumed by the poorest 20%II. This is not a been analyzed. However, data from Guinea situation unique to those two countries, and it shows that donors contribute 47% of public can be found in many other African countries expenditures on health. Since public expendi- when such benefit-incidence analysis is under- tures constitute 9% of all health expenditures taken. (Table 1), donors provide around 4% of all Graph 1 clearly shows that richer population health expenditures, which is much lower than groups disproportionately use private providers. what we see in the graph for Eastern and South- However, more importantly, it shows that a large ern Africa. Graph 2 is thus a conservative Graph 1 On use of private services for different symptoms 90 75 60 % 45 30 15 0 Q1 Q2 Q3 Q4 Q5 By income quintile (poorer to richer) South Africa Zambia Source: M. Makinen et al, "Inequalities in health care use and expenditures: empirical data from eight developing countries and countries in transition," Bulletin of the WHO, 78 (n.b., data was analyzed for only two African countries), 2000. Why Focus on Public-private Partnerships in Africa? 3 Graph 2 Sources of total health expenditures in 24 OECD and 10 Eastern & Southern African countries 100% 90% 27% 80% 59% 70% 60% 30% 50% 7% 40% 25% 30% 36% 20% 16% 10% OECD AFRICA Donors Public Other Private Private (Out of Pocket) Sources: for OECD: World Health Organization. The World Health Report 2002, Annex 5: Selected National Health Ac- counts indicators for all Member States, estimates for 1995 to 2000. Geneva; and World Bank, 2002. World Development Indicators 2002. World Bank, Washington DC. For Africa: National Health Accounts in Eastern and Southern Africa: a comparative analysis, ESA NHA Country Teams, 2000, estimated for 1997­98. estimate of private spending, which is probably This shows there's more money in the health higher in other African countries. sector than is usually thought, and there may be If one considers both private and public room for improving the efficiency of such ex- spending, it becomes obvious that there are lots penditures, most of which are spent on private of resources spent on health care, much more health care providers. than the per capita minimum health service pack- age cost of US$13 per year, as Table 1 shows. c. How extensive is private sector coverage? Although there's no systematic data available Table 1 across all countries, evidence shows that a lot Yearly health expenditures per source, of the services in many African countries are pro- for three countries vided by the private sector. For example, in US$/inhabitant (%) Kenya the private sector delivers 49% of health services. Half of this is given by religious and Government Private NGO facilities and the other half by small- and Country and Donors (OOP) Total medium-size commercial health enterprisesVII. MauritaniaIV $8 (26%) $23 (74%) $31 Another example is Ghana, where 37% of in- patient admissions (Accra excluded) are GuineaV $3 (9%) $31 (91%) $34 provided by missions (Graph 3). This varies by Burkina FasoVI $9 (43%) $12 (57%) $21 region as the graph shows. 4 TRENDSANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Graph 3 Coverage by the private not for profit hospital sector and by the public sector in Ghana, by regionVIII Inpatient Admissions By Region And Ownership (Ghana 2002) 90000 80000 70000 60000 50000 Admissions 40000 of No 30000 20000 10000 0 WR CR GAR VR ER ASH BAR NR UER UWR MISSION 21122 16363 0 26186 29721 34037 41780 16742 18788 13594 GHS 43421 36786 83330 37532 54738 80994 32156 53059 20122 18252 Region GHS = public network (Ghana Health Service); Mission network = CHAG (Christian Health Association of Ghana) Table 2 Tanzania: existing types of health services, 2000XI Facility Government Religious or Voluntary NGO Private for profit Specialized hospitals 6 2 0 Regional hospitals 17 0 0 District hospitals 55 13 0 Other hospitals 6 56 20 Health centers 505 48 16 Dispensaries 2652 612 663 Specialized clinics 75 4 22 Nursing homes 0 0 6 Laboratories 21 3 184 X-ray units 8 3 16 Why Focus on Public-private Partnerships in Africa? 5 Finally, in Mauritania, the Ministry of Health · In addition, the quality of services by private looked at the contribution of the private sector providers could often be improved. For ex- to the immunization program and, based on this ample, data from Uganda shows that only analysis, made some interesting recommenda- 19% of private health facilities correctly man- tions (see Box 1). aged simple malaria, a mere 6% of them did Usually the private sector does not provide so for simple diarrhea without blood, and many of the services that the public sector offers. 36% did so for pneumoniaXII. Often key public health services such as immuni- · Resources should follow the patient rather zation and treatment for tuberculosis are not than the type of provider. This means that available at the private clinics, which refer those public resource allocation should be based patients to the public sector, as was found in S. on health service delivery rather than on the AfricaX. The same study found that private clinics type of provider (public, NGO, for profit, were used mainly for curative care. On the other informal). This encourages the development hand, the public sector was often used for treat- of an integrated health system with different ment of chronic conditions. A look at Tanzania's actors rather than the continued pursuit of private sector (Table 2) shows that partnerships parallel systemsXIII with the private sector to provide laboratory · Organizational adaptations will be needed. and X-ray services should be investigated. Indeed, governments and aid organizations will have to make a special effort to work d. What are the issues that need to be with different groups such as private sector tackled to avoid "business as usual"? associations; this will involve designing focal units to have an interface with the private · Despite all the evidence showing the signifi- sector, for example. In the same way, private cant amount of money being spent by providers will have to get organized so that populations on services from the private sec- they can be heard as one voice by their part- tor, most governments and aid organizations ners, since it is not possible for governments still focus quasi-exclusively on public deliv- to deal with each individual, nor with each ery of health services. By doing so they did organization. not account for a large portion of the avail- · Finally, governments should avoid develop- able health expenditures and service ing two parallel types of care: one for the providers. Only recently has the public health rich and one for the poor. It's important that community started to think in terms of health those who pay, especially the poor, be able systems, and not just in terms of ministries to control the type of care they receive, and of health. With the arrival of HIV/AIDS, rec- that they be able to put pressure on the health ognition grew that all potential service care provider. This is not the case now in providers must be mobilized and harmonized Africa for the majority of people who are to cope with the epidemic. Now, the concerns poor. They are hostage to both the public are more on deciding what services can best sector, which often feeds on them through be provided, by whom, and how, so that pub- corrupt practices, and the private sector, lic health goals are reached. which often requires extremely high pay- · Because the money spent on private provid- ments. In both instances, high fees, whether ers by individual people is unpooled, it raises formal or informal payments, put health ser- the issues of efficiency and equity when so vices out of reach of many people. The much money from the poorest quintiles of availability of insurance programs, which the population is spent on the private sec- could cover formal fees, is limited in Africa tor. and subscription rates are low. The result is 6 TRENDSANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Box 1 Coverage by the private health sector in Mauritania A study done in the capital cityIX showed that: · The private sector provides at least 25% of consultations for infants and prenatal care, and 9% of deliveries. · All private clinics vaccinate kids and women but very few collect the statistics since they mainly note the information on the patients' health card. · The national immunization program has no statistics on activities done by the private sector. · Only two of 15 private structures visited had an adequate refrigeration system but none had all the adequate tools to monitor the cold chain. · Private structures buy their vaccines from private pharmacies in monodoses, which increases the cost to the patient. The recommendations made in that study are relevant for any other type of PPP, namely: · To the Ministry of Health: - Ensure that a representative of the private medical association is included on the inter- agency coordination committee - Establish a temporary six-month collaboration committee with the main public and private actors who have something to do with pediatric/gynecology service delivery to see how such a mode of collaboration works. · To the partners: ensure financial and technical support. · To the Extended Program of Immunization: - Provide cold chain material to the private sector - Provide private structures with vaccines and management tools - Train private sector providers on immunizations - Assess the pilot PPP in the capital city after six months. · To the city's health officials: - Monitor progress with the EPI structure - Sustain collaboration with the private sector while keeping in mind the specificities of the private providers. · To the private sector: - Ensure that the associations of private doctors function well to provide an adequate inter- locutor to government - Transmit monthly the data on EPI activities. that both rich and poor patients can be at e. Other reasons why can't we continue the financial mercy of both sectors, which "business as usual" do not always provide quality health ser- vices. It will be difficult for Africa to come close to reaching the MDGs if it continues "business- as-usual" in the health care sector, as Graph 4 Why Focus on Public-private Partnerships in Africa? 7 Graph 4 Reaching the MDGs in health: Trends in under-five mortality by region 250 200 150 thousand per 100 Deaths Sub-Saharan Africa 50 South Asia Middle East and North Africa 0 1970 1985 2000 2015 Source: World Development Indicators database (2003). shows: the full line represents what is happen- Purpose and Scope of this Paper ing, and the dotted line what should happen to reach the MDGs. Although there are no current large-scale part- Although it's important to continue to nerships between the public and the private strengthen the capacity of MOHs, it's equally sector for delivery of health care services in Af- important to start paying attention to where rica, there are some successes as well as some people actually go for services and to ensure that trends. This paper outlines some of these suc- their money, as well as public money, is used cesses and trends to provide lessons learned as efficiently for quality care. Thus, lately, there of today. has been more attention paid by development It is hoped that this paper will help decision organizations such as the World Bank Group to makers in governments, the private sector, and exploring the role of the private sector in reach- development organizations, build on the lessons ing public health goals. learned to find a better way for public and pri- The World Development Report 2004 state vate entities to work together to achieve public clearly that there is a need to consider different health goals and to empower consumers of types of service delivery arrangements, depend- health care services. ing on the country. Four out of the eight This paper also reinforces the role of the state suggested service delivery arrangements involve as a regulator, policy setter, and strategic pur- contracting with the private sector. chaser. It provides insights into the different types The World Bank's Africa Strategy also ac- of policies and interventions a government can knowledges that Ministries of Health will not choose to influence the private sector to reach be able to reach public health goals unless they public health goals, examples of which are pro- harness the potential of existing private sector vided in Table 3: providersXIV. 8 TRENDS AND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Table 3 Types of policies and interventions available to Governments to influence the private sector Type of policy a Government can use Available PPP interventions or schemes to influence the private sector to be used by Government Demand side policies Financing · National Health Insurance system · Community risk pooling schemes · Vouchers Supply side policies · Tax exemptions Regulation · Accreditation · Licensing Formal PPP · Franchising · Contracting · Leasing · Concessions · Divestitures Facilitation · Give voice and information to the private sector Source: Author's own design. CHAPTER 2 Who Delivers What Kind of Health Services in Africa? a. Where do people actually go macies; in Burkina, Guinea, and Mozambique, for services? the poor mainly use traditional healers; while in Niger the poor usually go to shops. T he analysis of DHS data from 26 SSA The richest quintile tends to make more use countries with 42 data sets for differ- of pharmacies, doctors, and private facilities. ent years (Annex 1) shows that almost The rural population is much less likely to half of the parents of a child who had seek care outside home. When they do, they are diarrhea or a respiratory infection in the past more likely than the urban population to use two weeks didn't seek care. Another 28% the private sector in the form of traditional heal- brought the child to a public facility; and 22% ers and shops. However, in some countries such took him to a private provider. However, an as Benin, Comoros, Namibia, and Senegal, the analysis by income level reveals a slightly differ- rural population uses more public facilities than ent picture. There is no surprise or news in the the urban population. fact that the wealthier groups use the private Trends over time for some typical countries sector. What is news is that the poor also use where two sets of data were available: the private sector extensively, since about half of the poorest who sought care brought their · In Malawi, the use of the public sector di- child to the private sector (Table 4). It also con- minished by about half between 1992 and firms that the public sector often subsidizes the 2000 (going from 33 to 16%); this is true of rich since 52% of the people in the richest in- the rich as well as of the poor. At the same come quintile who sought care brought their time, the use of the private sector went from child to a public facility. 27% to 39% for the poorest quintile and In general, the poorest quintiles are somewhat from 31% to 49% for the richest quintile. less likely to seek care outside the home than This might indicate a dynamism of the pri- the rich: about half of the poorest do, while vate sector and a problem in the public sector. about 60% of the richest do. · In Benin, between 1996 and 2001, in There are differences on where the poorest Cameroon between 1991 and 1998, and in quintile goes to seek care. In some countries such Ghana between 1993 and 1998, the poor as Ghana, the poor primarily go to private phar- sought more care outside the home, favoring 9 10 TRENDSAND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Table 4 Use of private and public facilities among the poorest quintile when a child is sick Country and Among those who sought care outside the home, % who went to: year of DHS Private sector Public sector Other Total % seeking care outside home Malawi 00 74% 24% 2% 100% 52.7 Mali 96 69% 24% 7% 100% 30.5 Uganda 95 69% 29% 3% 100% 73.6 Uganda 01 68% 27% 5% 100% 77.7 Ghana 93 65% 25% 10% 100% 57.7 Niger 98 59% 36% 5% 100% 35.4 Cameroon 91 55% 45% 0% 100% 24.6 Benin 96 53% 47% 0% 100% 37.3 Kenya 98 47% 47% 6% 100% 63.9 Malawi 92 46% 53% 0% 100% 58.8 Nigeria 90 46% 47% 7% 100% 50.9 Comoros 96 45% 47% 8% 100% 56.2 Madagascar 97 44% 56% 1% 100% 44.5 Cameroon 98 44% 52% 4% 100% 37.6 Benin 01 41% 55% 3% 100% 40.3 Guinea 99 38% 54% 8% 100% 44.4 Burkina 99 35% 59% 6% 100% 20.1 Mozambique 97 32% 63% 5% 100% 36.6 Tanzania 96 29% 68% 3% 100% 58.1 Zambia 96 24% 68% 8% 100% 64.0 CAR 95 19% 80% 1% 100% 27.7 S. Africa 98 14% 84% 2% 100% 64.5 Namibia 92 7% 90% 3% 100% 67.2 Gross Average: 45% 51% 4% 100% 48.9 Note: countries with more than 10% "other" were not included in the table. Source: Sara Project, Academy for Educational Development, Washington, D.C., March 2004. Please see Annex 1. public sector providers. The richest portion Cameroon, Ghana, Malawi, Uganda), except for of the population in these countries favors Benin and Cameroon where the trends are very private sector providers. Those trends are very positive, the tendency is for people to seek less good, as they indicate a move towards eq- care outside the home. This can be explained by uity. an increase in poverty, or a decrease in the per- ceived accessibility to health services (due to From the five countries where there are two under-the-table payments for example), a de- sets of data with income quintiles details (Benin, crease in the quality of care. Who Delivers What Kind of Health Services in Africa? 11 The data highlight an equity problem in health found to contain a higher dose than what was services availability, with the public sector be- listed on the labelXVI. Both under- and over-dos- ing used by the rich more than by the poor in ing can be very detrimental and point out a several countries. They also show that some problem of drug quality surveillance by govern- countries have a problem of low utilization of ment. Finding ways to improve the procurement public services in general: people are not seek- of good quality drugs by the private sector is ing care anymore since they probably cannot one of many actions that were recommended. afford it. In many countries, people sometimes Another study by WHO provided similar results have to pay up to 10 times the official rate to be in other countriesXVII (Graph 5), where the pri- taken care of in a public facility. Finally, the data vate sector, represented by vendors, shops, and show that the private sector is growing and is pharmacies, did not provide better quality chlo- being used by all income groups. roquine tablets than the public sector. In Uganda, a study undertaken in five districts and in the capital city showed that 81% of the b. Do people get good quality care in simple malaria cases and 64% of pneumonia public and private sectors? cases were not managed correctly by the 164 private health facilities treating those casesXVIII. There's room for improvement in both sectors. Kenya started to solve this problem when the That is why it's so important for governments Kenya Medical Research Institute worked with to adequately perform their role of regulator of the MOH to train private drug retailers in a ru- the health system, including the private sector. ral area in Kilifi district. About 500 shopkeepers However, there is a perception by many popu- were trained in 2­4 day workshops, with the lations that private providers give service with result being an increase from 7 to 65% of chil- better attention. dren given the right dose of antimalarialsXIX. In South Africa a study of nine clinics showed People often have a better perception of the that sexually transmitted infections had been care provided by the private sector. For example diagnosed using the correct approach in 85% in South Africa people liked the private sector of the private clinics, compared with 68% in because they were being treated promptly and public clinics. In the private sector, 97% of pa- with respect, which, with some exceptions, was tients had received treatment in line with the not echoed in public clinics. The waiting time at Department of Health's guidelines, compared to the private clinics was 10­40 minutes, compared 80% in public clinicsXV. with 50 minutes to 3 hours in the public clinicsXX. In Senegal, the national malaria program asked the U.S. Pharmacopeia Drug Quality and Information Program to provide an assessment c. Are private services more of Senegal's antimalarial drug quality. This was expensive than public services? done performing a random sampling of antima- larial drugs and testing them in the USA. It was Data show no conclusive evidence one way or found that 55% of the sampled sulfadoxine- the other; it depends on the facility and the coun- pyrimethamine tablets contained less than the try. There are great variations in both sectors. A claimed amount; that drugs from the private comparison of the mean cost per visit between sector failed more often than those from the in- private and public care providers in South Af- formal market; and that the public sector had rica and in Zimbabwe provided the results the least failings. shown in Table 5, which did reveal some large Tests on chloroquine found a similar pattern difference in Zimbabwe, but no significant dif- except that all the chloroquine that failed was ference in South Africa. 12 TRENDS ANDOPPORTUNITIES INPUBLIC-PRIVATE PARTNERSHIPS TO MPROVE I HEALTH SERVICE DELIVERY INAFRICA Table 5 Mean cost per visit to the public sector and to the private sector in two countries Public Sector Private Sector Small clinics without Large clinics with Private Private general full-time doctors full-time doctors clinics practices South Africa1: 33.20 65.78 42.14 (part of a 89.44 Total cost/visit chain of clinics) (in Rands) Zimbabwe2: Outpatient consultation US$24­52 (depending on level of infrastructure) US$257 US$125 1 Source: Palmer Natasha et al, Bulletin of the WHO, 81 (4), 2003. 2 Source: Mudyarabikwa Oliver, Madhina Denford, "An assessment of incentive setting for participation of private for- profit health care providers in Zimbabwe," PHR Small Applied Research No. 15, November 2000. Graph 5 Percentage failure of chloroquine tablets content in 7 African countries, by source of distribution 100 80 60 40 20 0 Teaching hospital District hospital District medical Health centre Pharmacy Vendor shop Household store % Failure The private clinics in South Africa kept per- computer system. However, private clinics had sonnel costs low by using nurse practitioners as higher external administrative costs, reflecting the main service providers. Drug costs were con- strong management support from their head of- tained by using a basic company formulary, as fice. well as by strict control of prescription practices However, in some countries, low official prices via regular audit of dispensing patterns, using a do not provide the real pictures. A probably Who Delivers What Kind of Health Services in Africa? 13 Table 6 Ethiopia: costs in private and public clinics and hospitals, 1999­2000 (in Birr) NGO-managed Public-owned Enterprise- Private Public clinics and managed clinics owned hospitals hospitals hospitals Recurrent cost per 9.78 (varied 12.55 (varied patient treated from 6 to 67) from 11 to 14) Recurrent cost per patient 42.03 38.35 Recurrent cost per inpatient 2680 556 460 extreme example is in Guinea, where although century, while a freeze on hiring was imposed most public infrastructures display a poster with on governments to contain their recurrent costs. the price list of services, the price paid by pa- Today, mounting empirical evidence of poor tients in the two main hospitals was 4.4 times public health service delivery is associated in part the official price for outpatients, and 9.4 times with lack of trained human resources. Ghana, for hospitalized patients. Out of those amounts, for example, is suffering from a brain drain of 53% went to buy drugs, 39% went under-the- medical and paramedical personnel to the West. table, 5% was for complementary exams, and Of the more than 2,000 physicians who were only 3% went to the hospital registerXXI. registered in Ghana in 1999, about 700 are en- The study did not investigate private struc- gaged in the public sector while around 300 are tures, and cost comparisons are thus not in the private sector; the remainder are believed available. However, considering that one of the to work outside the countryXXIII. A few other major causes of those under-the-table payments countries suffer from a similar problem. Other was identified as the presence of 400 interns areas of the continent suffer from insufficient ("stagiaires") and volunteers who are not paid, personnel to meet their needs. One example is it is unlikely that this practice is found to the Senegal, where there are 11,000 health work- same extent in the private sector. It should also ers, but there's still a deficit of 3,500 agentsXXIV. be noted that three-fourths of the 53% of the It's unfortunate that personnel from the private- cost went for drugs purchased from the private for-profit sector are not included in that count, sector, confirming that the public sector is a con- especially considering Table 7, which shows that sumer of private sector drugs. the private sector employs a lot of specialized In Ethiopia (Table 6), the recurrent cost per personnel, and Government could use those re- patient is slightly lower in public clinics; however, sources. the cost per treated patient is lower, on average, A contentious issue is the competition for in NGO clinics. Large differences are observed scarce human resources between the rapidly among hospitals, and it appears that enterprise- emerging private sector and the public sector. owned hospitals could be more efficientXXII. However, this could also be seen as an opportu- nity for better public-private partnerships, by: d. Is the private sector draining the · Increasing the number of medical and para- public sector of its human resources? medical students who graduate from private training institutions. This was done in It is interesting to note that the private-for-profit Senegal, where the number of paramedical sector boomed during the latter part of the 20th graduates was doubled by contracting three 14 TRENDS ANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Table 7 Number of medical and paramedical personnel in the public and private sectors in Senegal in 2000XXV MDs Dentists Pharmacists Nurses Midwives Sector Public Private Public Private Public Private Public Private Public Private Number 358 354 53 84 66 306 1871 267 562 45 local private training institutionsXXVI. Qual- remote areas where public employees have ity is standardized through the public service been unwilling to live and work. HIPC re- entrance test they take, which is the same for sources are used to pay them. the public school graduates. In 1997, Tanza- · Encouraging the development of good qual- nia saw the creation of the Hubert Kairuki ity private providers, with centers of Memorial University (info@hkmu.ac.tz), excellence, to retain as many professionals which trains medical professionals and as possible in the country. Mali is developing nurses. Kenya has private training schools, a top laboratory with the help of the Bio- but because of poor regulation, there's an Merieux Foundation, which will become the uneven quality of graduates from those reference laboratory and train other public schoolsXXVII. There's a need to encourage the or private labs throughout the country and development of private training schools and possibly the whole sub-region. In Kenya, the ensure their accreditation or that they are up Aga Khan Foundation hospital is one of the to standards. best in the world. These types of highly so- · Including the private-for-profit sector in the phisticated providers can be used for training health system to increase service coverage. and they provide a quality standard in the This means private providers will be able to country, while hiring local personnel who benefit from on-the-job training from the might have otherwise emigrated. public sector, which is too rare an event to- · Expanding and improving government con- day. For example, considering that in Uganda tracting with private providers for public there is approximately one traditional healer services, thereby improving the long-term op- per 100 people, while this ratio for medical portunities of health professionals, while doctors ranges from 1/10,000 in cities to 1/ simultaneously expanding access to publicly 50,000 in rural communitiesXXVIII, it makes funded services. sense to train those healers in certain medi- cal practices. This also means that certain Africa is not alone in having a large propor- primary care services which are now mainly tion of its physicians in the private sector, as provided by the public sector--such as rou- Table 8 shows: tine immunizations--be made available in the private sector. Table 8 · Contracting private providers to supply pub- lic services. Senegal is doing that by hiring Percent of medical doctors who are private, nurses with a renewable contract for two by continentXXIX years. Those contractors are paid better than Sub-Saharan Latin America their civil service counterparts (twice as much Africa and the Caribbean Asia for medical doctors and about three times as much for nurses), but they fill positions in 46% 46% 60% Who Delivers What Kind of Health Services in Africa? 15 Private practices by public servants: In Africa, A study of dual public-private medical prac- public employees often run the private sector. tice in China, Thailand, Peru, and Zimbabwe Indeed, although most governments have legis- showed that doctors maintain their public sec- lation to prevent public employees from tor jobs despite good income opportunities in practicing privately in public structures, this is private practice. Dual practice occurs even in commonly performed, sometimes to the detri- settings where there are major regulatory restric- ment of the poor and of the public sector. For tions, such as China. Rather than fighting this example, in Guinea, a study in the main two practice, there is a need to make the best use of hospitals in the capital showed that 75% of pa- it, namely to enforce clear guidelines and mecha- tients paid more than the official tariff, and only nisms that will prevent misuse of public sector 14% of what patients paid went back to the resources and provide stable incomes, training hospitals. Patients paid on average 10 times the opportunities, and sometimes other benefits such official tariff. This was due in great part to the as a pensionXXXI. fact that interns are working full-time but are The decision of public health professionals not paid by the government, so they pay them- to engage in private practice depends on sev- selves through informal fees from patients. One eral factors, one of which is how individuals hospital employs 300 public servants, 75 con- are paid. Those paid a flat salary would have a tractors, and more than 400 interns. Those greater incentive to undertake dual practice than interns feel they work in place of public servants those who are paid in an incentive systemXXXII, who are on the government's payroll, but who which advocates for performance-based con- spend their time in their own private practice. tracting. The civil servants themselves sometimes also ask for extra payment from patients or simply refer them to their private practiceXXX. CHAPTER 3 Experiences and Trends in Public-private Partnerships in Africa G overnments can influence the private complement government contributions. Re- sector through different means, in- cently, some governments realized that they cluding: needed to find more efficient ways to finance health. Consequently, a national health insur- · financing the private sector to make it more ance program is being tried in some countries efficient or targeted (financial support for like Ghana. Also, some governments are con- health insurance programs and demand-side tributing to demand-driven programs that pool promotion such as community health mutuals resources, such as community health mutuals, or vouchers) while other governments are trying targeted · legislative/regulatory reforms resulting in en- public subsidies through vouchers. abling environments · formal partnerships with the private sector i. Health Insurance to encourage the delivery of certain types of health services. Health insurance is not yet extensive in the sub- Saharan region, except in South Africa and the This section provides an overview of these planned national health insurance program in means. Many of these practices have been suc- Ghana. There are social health insurance pro- cessful; others lack evaluation data; several grams in Kenya, Tanzania, and Mozambique remain on a small scale; and numerous trends are that are usually mandatory for civil servants and encouraging. Ways to improve those practices and sometimes offered to employees of large corpo- scale up the successful ones are also suggested. rations on a voluntary basis. Medical aid societies exist in South Africa, Botswana, and Zimbabwe, which are usually organized along a. Public financing of the professional lines, for example, sickness funds. private sector ii. Demand-driven community programs From colonial times until recently, governments financed the private sector mainly through grants Community health insurance programs vary in to NGOs. User fees have also been used to ownership, management, benefit offerings, and 17 18 TRENDS ANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Box 2 National Health Insurance Program in Ghana The introduction of a national health insurance program in Ghana has been proposed, fol- lowing passage of the National Health Insurance Act of 2003. Previously, financing for health services came largely through public subsidies, foreign assistance, and user fees--mechanisms that were neither efficient nor always fiscally sustainable over time. The objective of the new health insurance program is to improve the way domestic resources for health are mobilized and obtain better value for scarce resources. Part of the financing burden of the program will be shifted to participants through a contributory health insurance premium. The national health insurance program builds on the experience of the voluntary mutual health insurance organization (MHO) movement, which began in the early 1990s and grew to 159 organizations by 2002. These community-financing programs are spread across the country, and vary widely in design, membership, and management. Although a formal study of their impact has not been conducted, it is thought that the better-run MHOs have contrib- uted significantly to improved access to health services and financial protection from the effects of ill health. The immediate goal of the national insurance reform is to eliminate the "cash and carry" system of co-payments in the public sector, and expand coverage by in- creasing the number of mutual health insurance organizations, their geographic outreach, and their membership. The National Health Insurance Act of 2003 establishes three types of health insurance organizations: district mutual health insurance programs, private commercial health insur- ance programs, and private mutual health insurance programs. Only the district insurance programs will be eligible for a subsidy for the indigent under the National Health Insurance Fund. The initial proposal is to expand insurance coverage to 30% of the population, while universal coverage is the long-term goal. Individuals would be obligated to obtain coverage from one of the programs, but would be free to opt out of the public system in favor of private insurance (and would not pay the public insurance premium). The insured would have a choice as to provider, using both public and private sector facilities, if they are accred- ited (the National Health Insurance Act specifies the scope of regulations governing the program, including accreditation of health care providers). Competition among providers is anticipated in urban areas. The issue of subsidies cur- rently provided to public facilities, which undermine the competitiveness of private providers, is one of a number of issues remaining for the government to address. arrangements with providers, but they share Discontent with the quality, reliability and common features that are a hallmark of health sustainability of publicly funded health services insurance: they involve prepayment and risk has instigated moves for establishing mutually pooling. Graph 6 shows the extent of pre-paid beneficial financing arrangements devoted to plans in sub-Saharan Africa, with the Southern health services. This is clearly the case with the African countries leading the way, followed by "mutuelles" in West Africa, many of which origi- French-speaking countries, and then by Eastern nated from simple community drug-revolving African countries. funds, or complementary programs set up by Experiences and Trends in Public-private Partnerships in Africa 19 Graph 6 Prepaid Plans as % of Private Health Expenditures, 1995 and 2000 S. Africa 76.6 Zimbabwe 46.5 Botswana 21.6 Cote d'Ivoire 12.9 Madagascar 10.3 Senegal 8.7 Kenya 4.5 Tanzania 4.2 Malawi 1.8 Uganda 0.5 Rwanda 0.3 0 10 20 30 40 50 60 70 80 90 1995 2000 Source: World Health Report 2002, World Health Organization. Table 9 Evolution of the number of health mutuals in West Africa Situation in 2003 Situation in Operational Soon to be Being Country 1997 in 2003 designed designed In difficulty Total Bénin 11 42 8 3 0 53 Burkina Faso 6 35 36 14 4 89 Cameroon 18 22 6 8 2 38 Côte d'Ivoire 0 36 1 0 3 40 Guinéa 6 55 17 27 10 109 Mali 7 51 12 4 4 71 Mauritania 0 3 0 4 0 7 Niger 6 9 2 1 1 13 Senegal 19 79 30 18 9 136 Tchad 3 7 0 0 0 7 Togo 0 9 2 6 0 17 Total 76 348 114 85 33 580 Source: Inventaires Concertation des acteurs du développement des MS Afrique 1997, 2003.; www.concertation.org 20 TRENDSANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA civil servants who were not satisfied with cover- ties to provide advisory assistance to govern- age provided by the state. Their number has ments and health insurers for improved multiplied sixfold in 6 years. contracting of health services, as it has done in Health insurance embodies the principles of other regions. public/private partnership, whether the health insurance program is mandatory, a private vol- iii. Vouchers untary program for those in the formal sector, or a small community risk-pooling arrangement. Voucher programs for delivering health services Health insurance encourages purchasing from are a mechanism for subsidizing the provision the managers of the fund (the risk pool), who of particular services to targeted groups through are mandated by the members to accredit pro- the use of a token (voucher) that can be re- viders, negotiate payment terms, and monitor deemed to purchase all or part of a good or the quality and volume of services provided. The service. The general idea is that, under certain contractual arrangement between the health in- circumstances, providing demand-side subsidies surance fund and health facilities provides will be more effective than using resources to incentives for providers to perform better be- provide supply-side subsidiesXXXIV. Demand-side cause there is a risk that their contracts or subsidies in general, and voucher programs in accreditation may not be renewed. particular, have an advantage in creating an ex- Health insurance is one of the best tools to plicit link between the subsidy and the output, empower the consumer of health care, espe- thus providing an incentive to increase the use cially when the program has emerged from the of that output. The objective is to select specific consumers themselves, such as in health health interventions that are cost-effective, di- mutuals. rectly target vulnerable groups, simplify Although the growth of health insurance in administration (reducing the possibility of ir- sub-Saharan Africa has been slow, except in West regularities and false claims), and reduce Africa (see Table 9), there is potential for fur- provider-induced demand. Experience to date, ther expansion and hence for more however, has shown voucher programs to have public-private partnerships. Given this potential, higher transaction costsXXXV. the World Bank is currently researching the im- Voucher programs can be competitive or pact of voluntary health insurance on financial non-competitive. In a competitive voucher pro- protection and on access to health careXXXIII. gram, some form of competition exists among the providers of health services (for the busi- Possible roles for the Bank and IFC. The ness of the voucher holder) and thus creates a Bank should support such community risk-pool- choice of provider for the voucher holder. In a ing programs. This support can be initiated by non-competitive scheme, a designated service financing initial start-ups, evaluations, provider is charged with delivering services. A re-insurances, and supporting dispersed pro- typical competitive voucher program works as grams so they can form federations when they follows: request it. IFC has acquired a minority stake in the Af- · funds are transferred to a voucher agency rica Reinsurance Corporation (Africa Re), · the voucher agency produces the vouchers investing US$10 million in Africa's largest do- and distributes them to the target population mestic re-insurer. This new investment is a big (either by itself, or through a third-party or- part of the organization's strategy to support de- ganization) velopment of well-managed primary insurers in · the recipient of the voucher presents it at the Africa. In addition, IFC is exploring opportuni- service provider of his or her choice in ex- change for specified goods or services Experiences and Trends in Public-private Partnerships in Africa 21 · the service provider returns the voucher to use. Initially, it may be that the least poor are the voucher agency (along with any required able to take advantage of vouchers, with the information) poorest participating only later in the life of · the voucher agency pays the provider an the program. agreed-upon sum for each voucher returned · Minimizing misuse may be more difficult in · the voucher agency reports program outputs larger, nationwide programs, as opposed to and outcomes to the government or donor carefully controlled programs limited to a few providing the subsidiesXXXVI. districts. · The use of third parties to distribute vouch- Data on competitive voucher programs for ers, particularly the social organizations delivering health services is limited, in general, around poor or disadvantaged groups or and particularly in Africa. Lessons from three members of the groups themselves, may programs in Africa--purchasing insecticide- strengthen outreach. Particularly for services treated nets in Tanzania, delivering emergency of a sensitive nature, such as sexual and re- contraception in Zambia, and providing repro- productive health services, the point of ductive health services to young people in contact--both in distributing the voucher and Kenya--are described in Annex 3. Some of the obtaining services--may be critical in ensur- lessons learned are as follows: ing comfort and confidentiality, and thus acceptance of the voucher program. · A substantial amount of time--often several · In some cases, better results were obtained years--is needed for people to understand when the distributor of the voucher/provider and use the voucher program, even if a pro- of information was also equipped to deliver motional campaign is in place. Multiple services. communication channels may be warranted. The targeted population must be made aware Training of providers may be required to en- of the importance/use of the good/service and sure quality of goods and services, and to ensure where to go to obtain it. that providers are sensitive to and welcoming · Careful monitoring may be required, espe- of voucher holders. cially at the beginning of the program, so as to understand why targeted individuals are Possible role for the Bank. World Bank's credits not making use of the program and how bar- and grants could be used to finance such pro- riers can be addressed. grams, their evaluations, and what they need to · A target group that is easy to identify and to go full-scale if the pilot is successful. Operations reach--such as pregnant women--contrib- research could be financed by the Bank to find utes to the success of the program. ways to minimize transaction costs. · Rather than providing unfair competition to the private sector by subsidizing the provi- sion of a good or service through public b. Regulating the private sector facilities, a voucher program can strengthen commercial providers while serving public Regulation, although a powerful tool, is not health goals. generally applied in a consistent manner across · In setting the value of the voucher, there is a the continent, whether with the public or the trade-off between increasing its value (and private sector. In most African countries, basic thus providing greater benefit to the poorest, legislation on private health practice is done who may not be able to afford the service through registration/licensure requirement. even with a lower-valued voucher) and mis- However, there are two major reasons why li- 22 TRENDSANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Box 3 Zambia's experience with accreditation to date* Zambia initiated its comprehensive health sector reform in the early 1990s. Reform was considered particularly urgent in the hospital sector, as nearly two thirds of the country's 79 hospitals were more than 30 years old, and more than half of government hospitals had sanitary systems characterized as poor or worse. With assistance from the USAID's Quality Assurance Project, the government launched the Zambia Hospital Accreditation Program in 1997 to accredit both public and private hospitals. Although there were other mechanisms to evaluate hospital performance, accreditation was chosen as it was considered to be a compre- hensive and transparent program. To design and implement the accreditation program, the Zambia Health Accreditation Council (ZHAC) was set up, which was comprised of 12 members representing government, professional organizations, and the public. ZHAC did not have independent funding or legal mandate, however, and the council members had full-time responsibilities elsewhere. ZHAC selected surveyors and trained them through formal training on accreditation principles and surveyor skills, as well as through a series of practice surveys over a 2-month period. Survey- ors had full-time jobs elsewhere, and were to conduct surveys on an ad hoc basis and receive a nominal payment for their efforts. Only one in five hospitals has completed the accreditation cycle. The question of long-term sustainability has been raised. With change in USAID funding, the program is now stalled. In 2002, an internal examination of the program presented the following observations: · Hospitals appreciated the educational nature of the accreditation process. However, hospital staff expressed the need for technical assistance in meeting accreditation standards. In addi- tion, feedback to the hospitals took a long time, up to a year after completion of a survey. · Although ZHAC's mandate is clearly defined, legislation to make it an independent and non- governmental organization did not go through. Without legal recognition and its own financial resources, ZHAC's capacity to carry out all tasks associated with regular survey and accredi- tation is limited. ZHAC staff perform their tasks on voluntary basis, and the level of participation has been waning due to lack of control over budget. · While training of surveyors was effective, there has been high attrition of surveyors because of low compensation relative to the level of work demanded and opportunity cost. *Source: Bukonda, Ngoyi et al, "Implementing a National Hospital Accreditation Program: the Zambian Experience," International Journal for Quality in Health Care, Volume 14, Supplement 1, 2002. Rooney, Anne, Paul van Ostenberg, "Licensure, Accreditation, and Certification: Approaches to Health Service Quality," Quality Assurance Project, 1999. Montagu, Dominic, "Accreditation and other external quality assessment systems for healthcare," DfID Health systems Resource Cen- ter, 2003. censure (or other legislated regulatory mecha- · Weak enforcement of regulatory control, with nisms) has failed to guarantee quality: limited funding. For example, in Malawi, the Experiences and Trends in Public-private Partnerships in Africa 23 Medical Council is supposed to make initial Appropriate regulation of market practices is inspection of premises for anyone applying lacking in many places. For example, in Tanza- to open private practice, as well as perform- nia and Uganda, both of which have explicitly ing periodic spot checks. In a survey of private promoted PPP, most regulations focus on entry practitioners, however, 73% of practitioner requirements for private providers, yet none ex- were found not to have any refrigerators, and plicitly aim to improve competitive practices. they dispense a wide variety of drugs includ- Tanzania has no explicit protection for the health ing those not on the approved listXXXVII. consumer, and vertical integration is widespread, · Unwillingness to enforce regulation against for example, self-referral of doctors--doctors their own membership. For example, the Zim- owning private facilities such as laboratories babwean Medical Council has not publicized where patients are sent, thereby limiting patient any case of malpractice for fear of damaging choice and often bypassing available public ser- the reputation of the professionXXXVIII. vices. A technical review in Tanzania in 2005 pro- Although many countries are trying to im- vided interesting observations about the prove their regulations, some key gaps remain. regulatory framework in place in that country, For example, regulations in Tanzania and Zim- which are summarized in Table 10 (ref. nbr. xiii). babwe (a) focus on individual inputs rather than Some alternatives to legislative regulation in- health system organization; (b) aim to control clude: entry and quality rather than quantity, price, or distribution, and (c) fail to address the market- · Regulation by contract, where there would level problems of anti-competitive practices and be clear separation between service providers lack of patient rightsXXXIX. Similar problems were (private sector) and purchaser (government). found in a study in South Africa, where it was The contract should contain agreement on the noted that non-state providers tend to be more process to reduce the burden of "independent controlled rather than encouraged or supported regulator" (more on this topic in the next by governmentXL. chapter). Table 10 Results of the Tanzania's Technical Review 2005 regarding PPP regulationsXIII Positive aspects already in place Aspects which need improvement Centralized systems for registration are in place Sometimes system is cumbersome, time consuming Registration of public and private facilities does not take into account health services needs A centralized inspection is in place Not yet decentralized A national quality framework was published No comprehensive national quality assurance system in place in 2004 No national standard for accreditation of health facilities The Pharmacy Act was passed in 2002 and a Regulation of pharmaceuticals is yet to be more effectively Dispensing Manual was published enforced The Tanzanian Food and Drug Authority and the Relationship between government and the pharmaceutical MOH have established a Drug Quality and sector is that of "regulator" and "regulated" and not Assurance Program partnership per se 24 TRENDS ANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA · Existing and emerging franchise programs are c. Formal public-private partnerships performing a role of micro-accreditation for service delivery agency to fill the gap, especially for those ser- vice providers who are not in the scope of i. Contracting the provision formal regulation. of health services · Provide financial incentives such as access to subsidies, government loans, etc., depending Three basic conditions need to be met to con- on the accreditation status (conducted on tract successfullyXLII: voluntary basis) of the private practice. This is a viable option if sufficient capacity to · minimum knowledge of the services to be monitor exists, and might be more applicable contracted (in order to be able to define the for middle-income countries. content of the contract) · Performance-based contracts constitute a way · capacity to manage contracts to regulate the provider by requiring the pro- · sufficient funding to cover the economic cost vider to give monitoring data. This of the service at the projected level of de- necessitates a good evaluation system, which mand1. can be contracted out; a variation is to pay against output only. Contracting is one type of public-private part- · Members of the private sector from Tanza- nership in which the pubic sector purchases nia and from Kenya have suggested that a specific services, be they clinical or non-clinical, "Loan Fund" be put in place to provide low- from a private (for-profit or not-for-profit) pro- interest loans to small-scale health providers, vider.2 The trend toward such partnerships is along with capacity building. In Kenya, the fed by growing evidence of the public sector's formal banking sector's interest rates are failure to deliver high quality and essential ser- above 24%, which makes it difficult for start- vices, including healthcare. Also, there is new up operations to borrow. Such a fund would thinking in public sector management that pri- work closely with professional associations vate sector mechanisms may help to improve that would provide yearly certification for the efficiency, equity, and responsiveness to usersXLIII. quality of care delivered before renewing a This type of partnership is now seen as comple- licence. The Loan Fund would finance train- menting public sector provision in order to reach ing, promotion, and establishment of public health goals. practices in rural areas, and only those prac- The theory is that replacing hierarchical tices that have received their certification management structures typical of public bureau- could graduate to a more substantial loan, cracies with contractual relationships between for example. The idea of this Loan Fund goes purchasers and providers--that is, separating much beyond, and has more potential, than purchaser and provide--will increase account- what exists now on a small scale in Uganda ability and improve efficiency. Partnering with the or Kenya (see Box 4). private sector also has the potential to increase the volume and variety of services provided, and Medium-term alternatives are guided by extend access, including remote areas. what is achievable in the context of most Afri- can countries. In the long-term, regulatory frameworks and appropriate information sys- 1 This last condition was added by the authors. tems for purpose of monitoring should be 2 Contracting may involve two public sector enti- developedXLI. ties; however, such arrangements are not dealt with in this paper. Experiences and Trends in Public-private Partnerships in Africa 25 Box 4 Examples of partnerships with professional associationsXI In Uganda, the Market Day Midwives project, a joint effort between SOMARC (Social Mar- keting for Change) and the Uganda Private Midwives Association, set up midwives in community markets as a distribution system for family planning. SOMARC provided each midwife with a sales booth, training through Service Expansion and Technical Support, and a uniform, and sold products to midwives at wholesale prices (Futures Group International 1995). Then, in January 2001, the Summa Foundation created a US$175,000 revolving fund to provide microcredit to private healthcare providers (nurses, midwives, and doctors) to ex- pand or improve their practices. Commercial Market Strategies Uganda provided training in business skills, marketing, and credit management alongside this, and has produced a busi- ness handbook for private health providers. The 3-year, USAID-funded project is expected to provide training and funds to 280 private healthcare providers (www.cmsproject.com/coun- try/africa/uganda.cfm?view=normal). In Kenya, Futures Group Europe initiated a small network of 38 private sector midwives to provide a range of reproductive healthcare advice and services. In addition to free contracep- tive and vaccine supplies, the Ministry of Health supplied the midwives with free bed nets and malaria treatment, with the sale of the bed nets providing a revolving fund for the mid- wives (www.fgeurpoe.com/site/mdmken.asp). Such projects seem to have potential, but are generally small scale; long-term sustainability, once donor project funding has ended, is uncertain. These types of pilots have tended to focus on the low-technology end of activity, on ambulatory care and normal delivery. In Africa, contracting has been used success- ers would be inclined to partner with the public fully for primary care, training, ancillary services, sector. and nutrition interventions. In addition, con- This chapter reports the characteristics of and tracting has been successfully used in other lessons learned from the review of a number of regions for secondary and tertiary care, espe- contracting programs in Democratic Republic cially for specialized clinical interventions. of Congo, Madagascar, Senegal, South Africa, Contracts are shaped by the nature of the ser- and Zimbabwe (see details in Annex 4). The vice to be provided; the amount of risk that each review is by no means an exhaustive account of party is willing to absorb; and the capacity of contracting activity in Africa, but rather is meant the public entity to create and administer con- to provide an overview of a larger strategy on tracts. Another consideration is the market for public-private partnerships for the delivery of providers--whether there is sufficient private health services. capacity to enable competitive bidding; whether providers are not-for-profit or for-profit, and Types of providers. The review found that gov- thus respond differently to incentives; and ernments contracted with both for-profit whether the market is such that private provid- providers, including institutions and individu- 26 TRENDS ANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA als, and not-for-profit providers, predominantly negotiated agreements. In some cases, it was secular non-governmental organizations seen that there was not sufficient competition (NGOs), community-based organizations within the private sector to support a bidding (CBOs), and religious organizations. NGOs/ process (and thus promote greater efficiency and CBOs and locally based individual contractors cost reduction). The fact that private-for-profit were found to be particularly effective with in- providers supply counterpart funding is appar- terventions that relied on closeness to and ently an excuse for non-competitive bidding. knowledge of communities, as well as long-term Experience in other regions (and sectors) has interactions requiring the changing of attitudes shown that contracts for services are generally and behavior. This was the case with nutrition most efficient and may provide greater public programs in Senegal and Madagascar, where benefits when tendered through a well-designed community participation and ownership were and carefully implemented competitive bidding key to the initiatives' success. Some countries process. still experience problems with the legal status of some private providers who need to have a Government capacity to contract. The costs of status that allows them to enter into a legal formulating, negotiating, and administering a agreement (see the section on Regulatory and contract are high, and the sophistication needed Legal Framework below). to devise and administer a contract, especially for complex services, may be extremely high; in Reasons for contracting. In nearly all of the addition, capacity for such operations may be cases, governments' decisions to contract re- limited. Several of the cases reviewed revealed sulted from a need to reach underserved areas that the government did not have the capacity or to provide services for which the govern- to effectively negotiate and enforce a contract. ment had limited capacity, rather than an This resulted in the government attaining an un- explicit policy to encourage private sector pro- favorable risk-sharing position with respect to vision. the contractor, little monitoring of contractors or enforcement of sanctions, weak incentives to Contract specifications. Contracts were contractors, and few efficiency gains from con- predominantly input-based, with both fee-for- tracting. In the case of a long-standing contract service and lump-sum payment mechanisms. between a private hospital and the government Contracts in many cases were poorly specified of Zimbabwe, the government's lack of capac- in terms of standards of quality, measures of per- ity to appropriately screen patients, as well as formance, and sanctions for non-performance. the fee-for-service nature of the contract led to This was the case of contracts with general prac- excessive use and an enormous cost burden titioners in South Africa, where supervision of (70% of the provincial non-salary recurrent general practitioners was limited and sanctions budget). were rarely, if ever, enforced. Practitioners were In some successful cases, NGOs are taking governed instead by professional ethics, ties to over part (or all) of the job of administering the community, and the goal of serving the pub- contracts, thus relieving the government of a role lic good. for which it is often not well equipped. The suc- cessful Community Nutrition Project in Senegal Tendering process. Competitive bidding has was administered by a delegated contract man- been used in some contracting programs, al- agement agency (Agetip), which managed though more often among NGOs than among contracts for the government and monitored and for-profit providers. The contracts with for- implemented the project. In the DR Congo and profit providers were non-competitive, the HIV/AIDS Disaster Response Project in Experiences and Trends in Public-private Partnerships in Africa 27 Burkina Faso, large, predominantly international or private). Important elements of this frame- NGOs were contracted by the government to work include ensuring the legal status for subcontract many of the health interventions to contracting (both on the government and pro- smaller NGOs and individuals. This also encour- vider sides) and the design and implementation aged small, often local, institutions to participate of contracting policies to minimize the uncer- in health initiatives. tainty of any party to the contract. Some countries, including Ghana, have adopted such Government information. Governments often policies, while others, such as Senegal, are in lack knowledge of the private sector and of costs the design process. of provision, weakening their bargaining posi- tion and reducing any efficiency gains from Payment for services. A critical concern of the contracting. In the case of a contract between private health providers is the rapidity and regu- the South African government and a for-profit larity of payment for services should be (Mills hospital companyXLIV, the government's lack of and Broomberg 1998). In several cases, govern- knowledge about actual costs of provision and ment bureaucracy was such that payments were the extent of competition led to efficiency gains delayed for long periods, seriously handicapping being captured by the company, not the govern- the private provider. Reluctance of governments ment, in the form of higher profits. to cede control of services and funding to the Lack of management information systems that private sector also resulted in delays. Using a cover the size and mix of the private health sec- delegated contract management agency has tor, the regional distribution of private providers, helped solve this problem, at least in the case of the experience and qualification of practitioners, the Senegal Community Nutrition Project, and and the nature of private health infrastructure it's now being tried in other countries. is a serious constraint. To build this capacity, Despite some shortcomings, contracting of some countries recently created PPP units in their services has been quite successful in some in- MOHs to serve as an interface between govern- stances. Contracting is a trend to encourage, and ment and the private sector, and to become governments can be helped to become purchas- specialized in dealing with the private sector; this ers of services rather than direct providers. is the case for Burkina Faso, Senegal, Uganda, However, contracting must be done correctly, and South Africa, among others. with strong regulatory and legal frameworks. A lot of work remains to be done to Service contracts should contain detailed qual- strengthen the capacity of both the public and ity and service standards. Regular monitoring private sector in contracting. WHO and the and adequate evaluation mechanisms are a must. World Bank Institute have launched several courses on the subject in collaboration with lo- Another worldwide review of contracting, by cal training institutions. IFC works directly with the World Bank, looked at 10 well-documented governments--advising and training ministries experiences in Cambodia, Bangladesh, Bolivia, of health and health insurers on contracting and Guatemala, Haiti, India, Madagascar, Senegal, PPP projects through the design, tender, and and PakistanXLV. All 10 studies found that con- implementation process. tracting was successful. Four cases with controlled or before-after groups had improve- Regulatory and legal framework: a major chal- ments in health indicators ranging from 9% to lenge for governments is establishing the legal 26%. Six of the studies compared contractor per- and regulatory framework that protects the formance to government provision of the same rights and responsibilities of all parties: govern- services, and all six found that the contractors ment, patients, and providers (whether public were consistently more effective. 28 TRENDSANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA ii. Leasing of equipment and facilities sector, primarily in South Africa. Under a con- tractual agreement, government defines the Leasing can be used in two ways in the health services to be provided, the standards at which sector: (i) equipment leasing, designed to fund the service will be provided, a pricing frame- necessary equipment to provide services; and (ii) work, and the length of time during which the facility leasing, used to provide new manage- concessionaire will provide the services. Expe- ment to an existing healthcare facility. rience has shown that the most successful Equipment leasing has been widely used in other concessions are output- rather than input-driven, sectors as a cost-effective means to increase ca- leaving the provider to determine the most ef- pacity, but has not been used often in the health fective and efficient way of providing services sector in Africa. IFC has successfully developed within the confines of the contract. leasing projects in other sectors and may extend Concessions require either management of these to the health sector. Leasing arrangements public assets for a defined (usually long-term) expand access to capital and services by fund- period, or require new construction. The ing equipment and transferring usage rights and latter have several structures, the most responsibilities for ownership to the equipment common are BOT (Build-Operate-Transfer), operator, who may then provide services to both BOO (Build-Own-Operate), and BOOT (Build- publicly funded and private-pay patients to gen- Own-Operate-Transfer). These models allow erate income to cover the lease payments. government to shift its role from investor in in- Providers (public and private) can also use frastructure, employer, and provider to a more leasing in a defined service area to pool their concentrated role that defines policy and makes resources and meet requirements for medical strategic purchases of services. Increasingly, gov- equipment that would otherwise be unaffordable ernments are using these tools, with some to individual health facilities. In Guinea, a CD4 adaptations, in the health sector to meet in- counter was bought in 2004 by government and creased demand. leased to a private laboratory that ensured its South Africa has developed the continent's maintenance and tested HIV-positive patients. most advanced hospital concessions as an ex- In South Africa, several public hospitals have tension of the Treasury Department's Public leased surplus bed space (instead of closing them) Private Partnership program (see Box 5). South to the private sector, which needed hospital beds Africa has followed the early concession model but did not have the capital to invest in a stand- developed in the U.K. that provided only for alone facility. support and technical services to be included in Lease arrangements give the private provider the concession. The U.K. has advanced their strong incentives to operate efficiently, because model, following the lead of many European and the provider's profitability depends on how other countries, to include clinical services as much it can reduce costs while still meeting the part of concessions for health, a development quality standards specified in the contractXLVI. that may be soon replicated in Africa. The Pelonomi concession described in Box 5 iii. Concessions includes an interesting feature that is being con- sidered by governments in several other Concessions, widely used for physical infrastruc- countries: the co-location of a private wing ture projects such as transport and utilities, are within an existing public facility. This arrange- a common legal instrument used by governments ment benefits all parties. For the public facility, to manage the private sector provision of public their employees and patients, there may be ac- services. As concessions have become more com- cess to updated equipment, possible facility mon in Africa, their use has spread to the health upgrades, and the potential for limited private Experiences and Trends in Public-private Partnerships in Africa 29 Box 5 Examples of concession in South Africa The Inkosi Albert Luthuli Hospital in Kwazulu-Natal was designed to provide the region (one of the fastest growing in the world) with state-of-the-art tertiary care. The winning consortium, which included Siemens medical systems and equipment and a leading IT sys- tems group, will provide full facility management, medical equipment provision and maintenance, and specified facility upgrades throughout the 15-year concession period. Gov- ernment closed five aging hospitals in Durban neighborhoods to concentrate its resources on Albert Luthuli. Siemens has designated the hospital as a paperless facility and works there in cooperation with South Africa's public medical staff to pioneer IT solutions for healthcare management. It is estimated that this concession will save government 370 million Rands over the 15-year partnership A similar model was used by Treasury to attract investors to refurbish the Pelonomi hospi- tal in Blomfontein. In this model, the Free State Government and South Africa's leading private healthcare provider, Netcare, entered an agreement for Netcare to refurbish and up- date the entire hospital facility. In addition to this capital investment, completed in 2004, Netcare will take over one wing of the hospital to provide services using their own staff and equipment to private pay patients. Netcare will also share access, under strict contractual agreement, to some of the operating and other facilities with the public medical staff. All parties are able to achieve their objec- tives: government increases public access to affordable and updated services, and the private provider makes a cost-effective entry into the middle-income market that is developing in South Africa. practice. A private wing may offer the private forms of contracting--a private wing project and provider the potential to establish hospital-re- concessions in other regions--providing govern- lated services with only a portion of the capital ments with the technical, regulatory, and legal investment that would otherwise be required. framework required to process, monitor, and In the early 1990s, the government of Benin replicate these complex transactions. gave the health center of Menontin in conces- IFC provides a team of transaction and tech- sion to a private religious association for 10 nical specialists and, working with government, years. An evaluation is not yet available. prepares the tender documents, including con- tracts for the provision of services along with Possible Bank and IFC roles. IFC's Advisory details of staffing, service, equipment, quality Services department focuses on assisting govern- standards, performance bonds, insurance re- ments to make efficient use of current spending quirements, penalties for nonperformance, and for publicly funded health services. This is ac- other specifications. Potential service providers complished primarily through public-private are pre-qualified in preparation for the tender partnership projects designed to tender for the to ensure their strength, both clinical and finan- private provision of these publicly funded ser- cial. The relevant contracting agency (generally vices. IFC has successfully designed and a ministry of health or health insurer) works implemented health PPP projects using various alongside and is trained throughout the process 30 TRENDSAND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA by IFC. At the conclusion of the tender process, Possible Bank and IFC roles. Those experiences government is provided with model documents need to be evaluated to see if there are any les- for replication of the pilot projects. sons to be learned, and the World Bank will soon finance such an evaluation. The question to an- iv. Divestitures swer is whether such arrangements have improved access to quality essential drugs by the Divestiture is the selling of government assets poor. to the private sector, forever. This has not hap- pened in Africa, but the closest, described below, v. Franchising has been to transform a public entity into a pri- vate one; however, the assets remained Concept of franchising government property, so what is described here is a mix of a concession and a divestiture. Franchising is a type of business model in which The Central Medical Stores of several coun- a firm (the franchiser) licenses independent busi- tries, namely in Cote d'Ivoire, Guinea, Mali and nesses (franchisees) to operate under its brand Niger, became autonomous as `Etablissements name. In general, a firm chooses to franchise, Publics.' In Benin, Burkina Faso, Madagascar, rather than manage its outlets in different loca- and Cameroon, they were transformed into non- tions, when it wants to shift day-to-day profit organizations with the state being present management responsibilities to franchisees and in their `management board' and have been op- expand its business network more rapidly. Fran- erating as such since the early 1990s. Success chising is not a new concept in Africa. For has been mixed. example, 20 franchise systems are estimated to Box 6 Divestiture in the pharmaceutical sector, the case of GuineaXLVII In Guinea, in 1994, the state-owned "Pharmacie Centrale de Guinée" became an autono- mous body under the "Tutelle" of the MOH, with government representatives on its board, but with no other input from government. It operates as a business, and its staff is paid from its profits. Government uses it as any other private drug wholesaler, it bids along with other private sector entities. It has no privileges, paying the same taxes as any other private entity. It wins quite a few government bids because it buys in bulk with other similar agencies from the West African sub-region. It imports and distributes drugs to hospitals through a network of five depots distributed throughout the country. Both private or public health structures can buy from this wholesaler. Its turnover increased from US$1.5 million in 2001, to three times this amount in 2002, and increased again in 2003. However, it faces the following problems, which endanger its sustainability: · although its profit on drugs was 17% in 2002, it decreased to 5% in 2003 due to the devalu- ation of the Guinean franc, coupled with long delays in payment by its main client, the government. · the contract it signed with government is not adequate and does not protect the "Pharmacie Centrale de Guinée": there's no legal recourse mentioned, payments are made 100% after the drugs have been delivered, there's no limit for delays in payments by the client. Experiences and Trends in Public-private Partnerships in Africa 31 be operating in Cote d'Ivoire, mainly in fast food, addition, the clinic chains--such as those rap- automotive rentals and hotels, while approxi- idly developing in South Africa--were not mately 478 franchises are operating in South included in the case study, as they are often Africa, 82% of which are of domestic origin. owned centrally and do not fall under the defi- The franchising approach has been used in nition of franchise given above3. the African health sector in the past decade. A review revealed that more than a dozen coun- · Successful franchising requires standardiza- tries have launched a type of franchise in health tion of services that enables relatively (See Annex 2 for a description of five franchises). straightforward training and monitoring. While there are some variations, under a typical This is especially important when the capac- franchising structure in the health sector, the ity of potential franchisees is limited. Most franchiser sets performance criteria, trains fran- franchises started out by providing services chisees, monitors their performance, and that are goods-intensive and relatively easy markets the brand name to target population. to standardize, such as essential drug provi- Franchisees are accountable to the franchiser to sion (e.g. CFW in Kenya) and family planning provide services at specified level, and may pay (e.g., K-MET, TOP Reseau). member dues (Diagram 1). However, franchising is being applied in other areas of services, especially to address Franchising in Africa. In preparing this paper, the HIV/AIDS crisis in the region (e.g., the team conducted a desk-review of existing New Start). Some existing networks origi- franchising programs in Africa (see Annex 2) nally established for more "traditionally and made the following observations. These franchisable" services have ventured out to observations are meant to provide a snapshot include HIV/AIDS in their services as well. of the current situation. They are formed based (e.g., KMET, CFW). on existing articles and franchisers' Websites and · The franchisers tap into existing resources to not on the result of a rigorous evaluation. In expand service coverage, whether they are general practitioners, nurses, or community- based health workers. · The franchisers are not-for profit organiza- Diagram 1 tions. PSI, which is running a majority of A typical structure of franchising in health franchises in Africa, is building on its experi- ence in social marketing to expand its franchising network. Franchiser · The cost of running franchises can be high for the franchiser, and some franchisers, such · Training as K-MET, rely substantially on the work of · Certification · Performance reports · Performance monitoring volunteers. Franchisees may become self-sus- · Member dues · Bulk purchase/credit tainable--for example, 80% of the CFW · Brand marketing Franchisees 3 For more information on clinic chains in South Africa, see Palmer, Natasha et al, "A new face for · Payment · Services private providers in developing countries: what implications for public health?", Bulletin of the World Health Organization, 81(4) pp292­297, Target Population 2003. 32 TRENDSANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA franchisees operate in a self-sustainable man- the poor. Such partnerships have yet to be ner. However, franchiser's sustainability formed, however. The World Bank may play remains to be tested. Among the cases re- a role in helping governments establish a viewed, only one franchiser (CFW) has framework for partnership. This may include: estimated the scale of operation that allows it to become self-sustainable. - Establishing a policy environment that ac- · Successful franchises seem to offer franchi- knowledges the role of private practitioners, sees proprietary gain from the franchise. including informal service providers Anecdotal evidence shows that joining TOP - Establishing standards to ensure the qual- Reseau resulted in higher use of clinics for ity of services, and to clarify qualification some doctors, and the founder of the CFW required to offer certain kind of services. network concludes that it was key to the They may include licensing and accredita- network's success to date. tion · Successful franchises combine supply-side and - Clarifying procedures to select private demand-side interventions. On supply-side partners and establishing transparent interventions, all programs provide quality mechanisms to transfer funds control to service providers. On demand-side, - Aligning private incentives to policy goals. franchises conduct an aggressive mass media campaign, and at least one (TOP Reseau) mo- · Many franchises focus on providing primary bilizes demand through the work of peer health care. While government may provide educators who lead education sessions in the more cost-effective services by establishing community. partnerships with such franchises, potential · Unlike other types of PPPs, such as service cost savings and benefits of such partnerships contracts and management contracts, most will be minimal without a strong referral link- of the franchises have limited partnership age that allows for comprehensive coverage with the public sector. Most partnerships, if of health care. The Bank may play a role in any, come in the form of donation of free assisting a government establish effective re- commodities from the government (e.g., K- ferral networks. MET) or contribution from international · IFC can play an important catalytic role, donors. as it did with the SHEF project franchise in Kenya (described in Annex 2). IFC worked Possible Bank and IFC roles with the founder on the business concept, provided access to funds, business training · As mentioned previously, cost effectiveness and advice, and other start-up resources as and sustainability of franchises has not yet the franchise concept was refined, grew, been vigorously studied, especially as com- and spread to the current network of 64 pro- pared to other modes of service delivery. The viders. World Bank may finance operational research into the cost effectiveness of a franchise pro- gram. d. Facilitation · Franchising has potential to deliver services to the poor, as it mobilizes existing health In this type of intervention, government make practitioners who are located close to the sure the private sector is represented in policy poor. As such, government may have interest forums and that it can access information on in forming partnerships with franchises, and how to contribute to public health goals as well tap into their network to deliver subsidies to as obtain the necessary financing. This report Experiences and Trends in Public-private Partnerships in Africa 33 gives several examples of the private sector be- The problem will be to ensure that these agree- ing represented in policy bodies. This trend needs ments and eventual bonuses are really enforced to be strengthened, as it is not common. In ad- and applied. dition, unless the forums concerned are efficient, there's an opportunity cost to attend such meet- ii. Decentralization ings, which the private-for-profit sector usually considers. The private sector will also have to Most African countries have initiated a decen- fight to gain a place at the policy table, as not tralization process, but few have gone beyond all governments will offer it. deconcentration. However, this decentralization trend is an opportunity to increase public-pri- vate partnerships. Most communes, or other e. Shifts in paradigm, trends and decentralized entities, know better than anybody opportunities to grasp else who are the best service providers in their area. Decentralized entities could monitor their i. Public-public performance-based performance more easily than the central gov- contracting ernment would, and they are more accountable. However, local officials have not yet fully ac- Lately, following the decentralization trend in cepted the PPP concept. In Tanzania, for most of Africa, some governments are starting example, the 2005 Technical Review of PPPs to contract public entities, in the same way they'd noted that strategic health planning at the dis- contract private providers. Formal contracts-- trict level does not yet properly accommodate used when dealing with the private sector--are PPPs. There's thus a need to train decentralized replaced by "Memoranda of Understanding" public servants as well as elected officials, such when dealing with two public entities, but the as mayors and their teams, in PPP concepts. same principles as in a public-private contract apply. This trend is an attempt to increase effi- iii. Private sector human resources ciency, transparency, motivation, and successful outcomes in the public sector. Some argue that--considering the lack of hu- The trend is to sign "performance-based con- man resources in the public sector--it would be tracts," as was tried by Burkina Faso with some counterproductive to encourage the private sec- degree of success (see Box 7). This represents a tor, as this might constitute another channel of shift in paradigm in the sense that those agree- brain drain from the public sector. A counter- ments give incentives for the service provider to argument is that the private sector already exists, reach, or even go beyond, the stated objectives. and we just want to make it more efficient and Such developments might help in making pub- improve its quality. An additional argument is lic-private partnerships contracts more that the nature of the provider, whether public acceptable. or private, is not as important as the outcomes, The reason why such formal agreements seem as measured by MDGs and other agreed targets to be well received might be because they help for public health and human development. ensure that rewards are based on objective cri- The concern should be how to encourage teria. This is not the perception most civil employees to serve healthcare consumers bet- servants have in a non-performance-based con- ter; how to stop the brain-drain outside of the tracting environment, where rewards are often continent by providing better opportunities perceived as due to solidarity networks, ethnic throughout the entire health sector in Africa; and ties, and other factors. how to ensure that the existing private sector human resources contribute to public goals. Sev- 34 TRENDSAND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Box 7 The Burkina Faso experience with public-public contractingXLVIII There appears to have been growth in some output indicators during the period of the pro- gram (e.g., immunization coverage), while others remained static (e.g., assisted deliveries and use of curative care services). However, methodological limitations make it difficult to link the program to either of these phenomena. Some evidence suggests that deficiencies in the health system's delivery structure may also have influenced performance during this period. Greater attention needs to be paid to incorporating a systematic research evaluation com- ponent into programs of this kind. Research is necessary to examine the effect of institutional capacity and delivery structure variables on a variety of health system performance outputs that are linked to individual/household behavior change and ultimately to improved health status outcomes. Scientific evidence that performance-based management programs in the public sector can contribute to improved health system performance is urgently needed. The foremost challenge for the immediate future is whether this model--implemented with Bank financing--can be replicated by governments with the support of a bilateral donor. It is essential that the legal framework within which the program was implemented--in particu- lar, a waiver to the existing law of finance--be re-evaluated. Procedures and requirements will also need to be simplified. Priorities, as articulated in the performance indicators, will need to be re-examined to ensure compatibility with the new health sector development plan. A better balance must be achieved between financial and technical performance auditing. Increased engagement of technical managers of vertical disease-control programs at central level also must be obtained. To achieve better efficiencies, greater decentralization of pro- gram oversight and management responsibilities should be considered. Support to health facility-community management committees came late in the program and should therefore become a primary focus of attention. Many questions remain. Can greater coherency be achieved in internal and external assis- tance to sub-national levels, so that managers can exercise greater autonomy and control over the resources required to achieve their priority outputs? Institutional capacity is neces- sary but not sufficient in itself to achieve improved health system performance. There are a considerable number of delivery structure problems; can they be adequately addressed? An important lesson learned from this first experience is that decentralized management programs are likely to achieve only partial success without concurrent policy reforms and capacity building at central level in a wide range of systems, such as planning, budgeting, and health information. Those lessons can also be applied to public-private partnerships. eral solutions are outlined below, each based on management training, franchisee training, and two principles: motivation of personnel and in- encouraging franchisers to set up business in creased accountability. drug retail or selected health service delivery Ways to increase motivation: · Facilitate access to credit for private provid- ers who want to work in certain underserved · Help public employees (those who already areas work in the private sector for themselves any- · Provide training opportunities to private sec- way) set up their business by giving them tor providers Experiences and Trends in Public-private Partnerships in Africa 35 · Work on the basis of performance-based con- private sector. An assessment identified govern- tracts, be it with the private or the public ment incentives to encourage private-for-profit sector, and stimulate competition by making providers to participate in health care provision. results known. Monetary incentives range from tax credits to full tax exemptions for some services, such as Ways to increase accountability: health financing by Medical Aid Societies. Non- monetary incentives, such as training, were more · Encourage medical graduates to set up their inclined to ensure quality of care. Contracting private practice in rural areas as country doc- out of some public services was the most tors. Mali has done this and their 80 country unexploited incentive, and was therefore recom- doctors are providing quality services to their mended as a way forward to creating efficiency communities at public facilities with minimal cost to the · Hire people who come from, or live in one governmentXLIX. area, to work in that area · Support demand-based programs where lo- iv. Private Internet communication cal communities want to work with health technology helps service delivery providers and will sign contracts with them (community health mutuals, community Private telecommunication firms could help set health centers, for example). up or partly subsidize communication technol- ogy, which would improve health care delivery Zimbabwe allows public-sector doctors to run (see Box 8). In addition, phone technology could their own private clinics after hours, and this be used as vouchers, which would eliminate the constitutes a large proportion of the private pro- high administrative costs of vouchers and limit vision. More than 50% of doctors work in the leakage. Box 8 Examples of the use of private sector Internet technology to improve health service delivery in Kenya and South Africa In Kenya, "Afriafya" was set up in 2000 as a consortium of seven partner agencies, all NGOs, and financed by the Rockefeller Foundation. It helped to set up computer communication equipment and train health personnel in seven dispensaries and it targeted AIDS. After 18 months of implementation in one of the districts, the information-coverage of reporting rose from 49% to 92%L. Improved reporting provides adequate and timely feedback to the ser- vice providers and is vital to better planning. A South African NGO, `Cell Life," financed by the Vodacom Foundation and in partner- ship with the University of Cape Town and Cape Technikon, another training institution, provides help to 500 HIV-positive patients. The strategy is as follows: "Severe limits to re- sources mean that the flow of information between doctors, hospitals, and patients has always been a challenge in rural South Africa. The extent of this challenge is illustrated by the fact that there are only 18 telephone lines per 1,000 people in some parts of the country. How- ever, cellular networks cover more than 90% of South Africa, and with a third of all South Africans currently using cellular phones, this information gap can be closed. (continued on next page) 36 TRENDS ANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Box 8 (continued) As an example, when an HIV-positive person is at a level suitable for treatment, the indi- vidual is interviewed for acceptance into an ARV treatment program. Once in the program, the person is assigned to a therapeutic counselor. Typically, counselors handle from 15 to 20 HIV- positive people. They are trained to gather vital information about these people, using a cellular phone equipped with sophisticated application software developed by Cell-Life engineers. The cellular phone provides a menu-based, real-time system to capture treatment relevant patient data, such as symptoms, drug adherence, and socio-economic factors. This information is logged immediately and directly to a central database. Efficient and user-friendly, this technology elimi- nates the need for cumbersome paperwork and provides a solution to logging accurate data about AIDS patients on a large scale, with minimum cost, maximum efficiency, and signifi- cantly less human error." (www.cellife.org). v. Globalization of public-private Another example is South Africa, where a pri- partnerships for health service delivery vate service provider, Netcare, sells services to England for ambulatory eye care clinics. These days health care delivery cannot be lim- Netcare personnel go to England a few months ited to just national frontiers. Africa is at a time and return enriched by the experience. developing centers of excellence that can Thus, human resources and technical plateau serve populations of the entire continent, and problems should start to be analyzed on a re- beyond. A couple of examples were given pre- gional and global basis, no longer solely on a viously for Mali and Kenya (see Para. II.d.). national one. CHAPTER 4 Private sector and Governments' challenges and possible responses a. Private sector challenges reach the whole network of NGOs. Although many professional associations exist, they are P rivate providers in Africa face a range still too fragmented and numerous. For ex- of challenging circumstances--lack of ample, there are associations of private clinics, infrastructure, limited access to sup- of midwives, of medical doctors, of nurses, plies and equipment, loans at very high etc., which all need to come together under interest rates, and other elements that can drive some sort of umbrella representation. This up the cost of services. Still, many recognize that will allow them to (i) have a stronger voice they can cost-effectively serve a larger market, in the dialogue with government and donors; including the poor, if they can develop a reliable (ii) be organized to receive funds; and (iii) paying market at a variety of service levels. In find ways to avoid delayed payments to pri- order to reach this viability, private providers vate providers for reimbursements by require reliable purchasers at rates that cover government. costs. This response from the private sector must · Gain a place at the policy table so that the be met with an equally committed effort from private sector is represented, informed, and government. can take advantage of opportunities that A number of important challenges to the pri- present themselves. This can be done through vate sector were identified by the mostly business coalitions or professional chambers. private-sector participants in the Kenya PPP Some governments have already taken actions workshop organized by the authors in June to improve partnership with the private sec- 2004. The challenges included: tor. For example, in Burkina Faso a permanent committee was created in 2001 by the Minis- · The need for the private sector to get orga- try of Health with the private sector. In 2002, nized so that a government does not have a a sub-directorate for Private Health Care was multitude of partners to deal with. An ex- created, and now there are private sector rep- ample is the way NGOs got organized and resentatives on the monitoring committee of are now regrouped in many countries under the National Health Program. one to three umbrella organizations. Such · Professional associations could also claim the organization allows government to quickly role of peer reviewers to ensure quality of care 37 38 TRENDSANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA in the private sector. Such peer reviews are labor force adequately is a sine qua non. usually quite effective, and a periodic exter- Government can also encourage private pro- nal evaluation could be set up by Government viders to work for poor communities as has to ensure sustained results. Those "semi-ac- been done in Mali, with community control. creditation" bodies could be contracted by Government should also enforce regulations government to maintain the quality of care so that the private sector provides a minimum based on government standards. standard of health services, like the South · Once the private sector gets organized, gains African accreditation system. Public-private its place at the policy table, and receives partnerships are a way to improve standards funds, it will have to be careful not to be- in the private sector, to cross-subsidize in or- come too dependent on government subsidies. der to give more money to the poor, and to In most African countries such subsidies can- liberate the public sector from serving the rich not always be expected to continue. so it is available to the poor. · Pooling of procurement for drugs and sup- plies is done by some private entities such as One of the most notable supply tools tried in religious associations. This example should recent years in Africa and elsewhere is Output- be followed by the for-profit sector so as to Based Aid (OBA). OBA is a way of allocating be able to obtain less expensive drugs. public resources, such that government contracts an entity to deliver services and ties payment to the services actually delivered. This has the po- b. African Governments' challenges tential to address issues related to ineffective targeting and lack of accountability. The government has two main types of tools OBA forces government to specify outputs that it must use together to ensure that poor clearly, which is not always the case with input- people have access to adequate health care: based approaches. Second, by tying disbursement of public funds to specific outputs, · Empowering tools give the consumer the OBA provides a clear framework for account- power to decide where to seek care and to ability and strong incentives for efficiency. Third, require quality care. Today, poor people tend the link between disbursement and outputs also to avoid the public sector because of the way means that it provides incentives for innovation. they are treated and because they have to pay Finally, under OBA, government screens service a high price for unsure outcomes. Health in- providers who are qualified to offer specific ser- surance, health mutuals, performance-based vices. By tying qualification of providers to contracts, and vouchers are all demand-side eligibility to receive public funding, OBA pro- tools that can help the poor increase its vides an incentive for service providers to raise choices. In addition, giving a place at the ne- their standards of services. The German KfW gotiation table to private providers as well has started to implement OBA programs in as to consumer groups is another way to Kenya and Uganda. empower civil society. This can only be done Other supply tools include: by working with private-sector or consumer associations, federations, or coalitions. · Making it easier for private providers to ob- · Supply tools must be put in place by govern- tain their licenses. For example, in Ethiopia ment to ensure it does not act as a predator the private providers obtain their licenses only on the poor, such as it is happening now when from regional health bureaus. This is per- corruption prevents the poor from benefit- ceived by them to be overly centralized, and ing from public services. Paying its own public most believe it should be delegated to the Private Sector and Governments' Challenges and Possible Responses 39 Graph 7 Uganda's government financing to NGOs Allocations of Government Funds to PNFP health sector 20 Hospitals 19.31 18 Lower Level Units 16 Health Training Schools 16.6 14 Drugs Sh. 12 Total 13.06 Ug. 10 10.4 10.08 8 Billions 6.07 6 7.04 5.3 3.89 4 3.03 3 4.04 1.6 2 1 1.07 2.02 2.03 0.04 0.3 0.6 0.07 1 1.01 0.76 0 97/98 98/99 99/00 00/01 01/02 02/03 03/04 Source: presentation by Dr. George Bagambisa, Coordinator of the PPP Unit, MOH, Uganda, on "PPP in Health: Uganda's Progress," PPP Workshop organized by the World Bank, Nairobi, Kenya, June 2004. zonal level. License renewal is also bureau- 2. Recognize there is an array of public-private cratic and often takes a long timeLI. partnerships programs the government can · Helping training institutions provide courses use, depending on what responsibilities and on PPP to public and private professionals. risks it would like to delegate to the private · Finally, financing the private sector through sector; different PPP programs. Already, quite a few 3. Accept that the partnership is not a one-time countries have been financing the private sec- affair, but is a relationship that needs to be tor with subsidies, but very often without nurtured from the inception to the termina- asking for results (Graph 7). It is time for this tion of the partnership. This requires to change. sustained effort; 4. Do not underestimate difficulties, and make Lessons from existing experiencesLIIhighlight sure that the state uses all possibilities to help elements that policy makers need to think about it manage these partnerships. For example, when considering public-private partnerships: using delegated contract management agen- cies to manage the multitude of contracts; 1. Political commitment, which is needed for 5. The need for public-private partnerships poli- any new way of doing things. It will help cies to provide the proper environment; make the new mechanisms more acceptable 6. The role of the state in investing in training, to practitioners, particularly by explaining regulation, capacity building, monitoring, that such partnerships do not necessarily and mediation; mean privatization or disengagement by the 7. The need for quantitative evaluations to en- state; sure that more lessons are learned as soon as 40 TRENDS ANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA possible. It is important that third-party evalu- role as steward, this report outlines an array of ation be organized to assure objectivity. policies governments can use to work with the private service providers, depending on their These elements reinforce the government's goals. These policies are summarized in Table stewardship role, which is essential to success- 11, which also spells out some elements for im- ful PPPs. To help governments carry out their provement based on what exists now in Africa. Table 11 How each policy and intervention can be used to reach public health goals, and examples on how to im- prove on what exists What government Type of policy What governments usually Reason for PPP wants to do most used need to improve a. To increase Motivate private providers Formal PPP · use OBA for formal PPP coverage by using to serve the poor through Financing · give faster feedback on issues existing providers targeted financing raised by private sector · provide more information and communicate better with private sector (open more communica- tion channels; institutionalize dialogue between govt. and private providers) b. To increase Motivate private providers Financing · pay contracts faster coverage where to move to those places Regulation · improve health referral system govt. is not Facilitation · ensure contracts are well present Formal PPP designed c. To increase Motivate private providers Financing · facilitate registration coverage by pro- to provide new services Regulation · let private managers manage viding services Facilitation that govt. is not Formal PPP providing d. To improve ser- Motivate private providers Regulation · improve supervision or vice quality of to apply the standards Facilitation inspection by MOH private providers · allow more private representa- tion Source: Author's own design. CHAPTER 5 The World Bank's and IFC's challenges and possible responses a. Portfolio review of what the World ducers/distributors of bed nets, contracep- Bank is doing on PPPs in Africa tives, food, pharmaceuticals, and oral rehydration solution were involved in 31% T he study team reviewed 40 health of projects projects in the Africa region approved · Health services from the private sector in- between Fiscal Year 1995 and Fiscal cluded primary care (34%), STD/HIV Year 2003 to determine the nature and prevention and treatment (26%), hospital extent of private sector involvement. The results services (20%), reproductive health care showed that: (14%), nutrition services (14%), and mater- nal and child health care (11%) · 23% of projects from 1994 to 1998 had high · In one-fifth of projects, the private sector private sector involvement, compared with played a role in financial protection by 57% of projects approved between 1999 and establishing rural or community health funds 2003 · Partnerships were implemented mainly · In 13 out of 40 projects (33%), no mention through service contracts (66%), manage- was made of the private sector ment contracts (33%); there was no leasing, · In 19% of projects, the private sector--NGOs divestiture, or franchising or community-based organizations (CBOs)-- · 43% of projects involved some form of train- was expected to help finance project activi- ing for private sector actors ties. The amount of co-financing expected · Regulation was used to engage the private from NGOs and communities ranged from sector in 29% of projects. 5­25%. No project specified co-financing from private actors other than NGOs or Notable Successes CBOs · The for-profit private sector played a role in · High involvement of NGOs and communi- 40% of projects reviewed, with physicians, ties in successful nutrition projects pharmacies, and clinics each involved in 22% · Community-run projects--additional ex- of projects. "Industry" actors, including pro- amples of success in reproductive health. 41 42 TRENDS AND OPPORTUNITIESIN PUBLIC-PRIVATE PARTNERSHIPS TO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Challenges identified sector participation specifically in child healthLIII, and its results were similar to the review de- · Government resistance scribed above. It revealed that about half of the · Contract delays plagued many of the projects reviewed projects engaged the private sector to · Lack of private sector expertise among Bank's improve child health outcomes (see Graph 9), task team leaders however, sub-Saharan Africa was one of the re- · Lack of engagement of informal private gions with comparatively high private sector health sector. involvement (Graph 8). What's needed in the Bank to more effectively engage the private sector b. What IFC is doing in PPPs · Further information on the size and scope of IFC has a long and successful history of advis- the private health sector in countries ing and implementing successful PPPs in Africa · Additional support from private-sector ex- for infrastructure and related sectors. Based on perts, and better-trained task team leaders in this experience and the needs in the social sec- terms of PPP tors, as expressed by government, IFC is seeking · Institutional commitment by the Bank's man- opportunities for PPPs in health. To date, IFC agement to work with the private sector has provided technical and financial support for one PPP project in health: the SHEF project Another review of Bank projects from 1993 (summarized in Annex 2) for a franchise of drug to 2002 throughout the world looked at private- shops across rural Kenya. To increase the vol- Graph 8 Proportion of the total number of projects financed by the World Bank with child health components that involved the private sector, by Region 80 70 60 50 40 Percent 30 20 10 0 Latin American Sub-Saharan Middle East Europe and South East Asia and Caribbean Africa and North Central Asia and Pacific Africa Asia Source: Henrik Axelsson, Flavia Bustreo, April Harding, "Private sector participation in child health," HNP discussion paper, World Bank, May 2003. The World Bank's and IFC's Challenges and Possible Responses 43 ume and scope of projects, IFC has in early 2005 in World Bank offices could help implement dedicated increased staff and resources to dis- this suggestion. covering and developing PPP opportunities in · Start considering health-service providers, Africa. Such opportunities will include both based on the public health functions that need stand-alone IFC and joint Bank-IFC projects. to be performed rather than on ownership. This means developing budget allocation guidelines, which explicitly include the pri- c. The World Bank's challenges vate sector. to contribute to PPPs · Conduct systematic assessments of the pri- vate sector's potential to contribute to What to do Internally program's objectives and monitor the private sector's involvement. In light of the Africa Region Portfolio Review · Raise the importance of PPP in higher policy and the trends noted in this report, if the Bank forums such as those on Poverty Reduction is to be successful in helping its country clients Programs' and Credits' matrix discussions. to work well with the private sector, there's a · Set up a PPP fund to complement the Bank's need for the Bank to: operating budget and thus ensure that this issue will not be bypassed. · Simplify its procedures to facilitate access to and use of IDA and IFC resources by small What to do Together with health service providers the Bank's Clients · Help document and evaluate existing experi- ences and share them across continents Based on the trends identified in this report and · Build on successes. The Bank has been suc- the comparative advantages of other institutions, cessful in working with communities for it is recommended that the Bank: prevention services, such as nutrition. It has not, however, been able to expand much on · Provide training to its clients through the those successes until now. World Bank Institute (www.worldbank.org/ · Train its staff on PPP, or ensure that there's wbi). This has started, but is only a begin- personnel who can assist country teams to ning. There's also the need to develop the work on this issue by providing focal persons capacity in African institutes to implement with enough time to work in this area. Such such training. persons should establish a PPP network to · Monitor, evaluate, and disseminate the results ensure that the best consultants are made of lessons learned. available to the country teams when needed. · Help clients develop partnership mechanisms For this, the Bank should work with other such as contracting, concessions, and fran- institutions such as WHO (for policy devel- chising. opment), with ILO (for mutuals), and with · Study and reinforce health mutuals, which private entities (for franchising and conces- are a promising venue. sions). · Finance training and capacity-building in · Establish direct communication channels with PPPs within countries. the private sector. Having PPP focal persons CHAPTER 6 Conclusion L everaging the private sector can be an It is important to note that the authors do important component in an overall not see public-private partnerships as a pana- poverty-reduction strategy. Data pro- cea, but just as one of the ways to improve health vided here did not offer any surprises systems. or news in the fact that the wealthier use the One of the major contributions of this report private sector. What is news is that the poor also is that it emphasizes the need to consider the use the private sector extensively. health system in its entirety. Planning, financing, Some of the myths about the private sector regulating, and formal PPPs all need to be set up were destroyed, many challenges identified, and based on the packages of public health functions some practical recommendations were provided that a government wants to deliver to different to the private and public sectors, as well as to population groups, which must involve all ser- the World Bank and the IFC. vice providers in the country. 45 ANNEXES Annex 1: Where Do People Go for Services? Care-Seeking For Young Children namely Mali 2001 (with 21% "other"), Rwanda with Recent Illness, Diarrhea, and/or 2000 (12% "other"), and Senegal 1997 (11% Respiratory Infection in the Two Weeks "other") before the Interview Analysis of Data from Demographic By the Sara Project, Academy for Educational and Health Surveys Development, Washington, D.C., March 2004 Countries with more than 10% "other source of care" were eliminated from the analysis, Benin 1996 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 64.19 55.81 61.68 Shop 8.90 5.65 7.93 Traditional healer 4.77 4.56 4.70 Private pharmacy 0.00 1.86 0.56 Private doctor 0.56 1.82 0.93 Private health facility 3.26 10.01 5.28 Public health facility 17.56 18.15 17.74 Other 0.77 2.14 1.18 Total (n) 932 399 1,331 47 48 TRENDS AND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 62.69 63.07 63.36 65.00 48.85 61.68 Shop 9.01 9.30 11.38 4.28 2.55 7.9 Traditional healer 6.62 6.41 3.38 4.68 0.00 4.70 Private pharmacy 0.38 0.00 0.00 1.01 2.30 0.56 Private doctor 0.30 0.00 0.26 2.41 2.92 0.93 Private health facility 3.66 2.38 2.71 4.97 19.18 5.28 Public health facility 17.35 17.10 17.96 14.89 23.73 17.74 Other 0.00 1.73 0.97 2.75 0.46 1.18 Total (n) 322 310 288 249 161 1,331 Benin 2001 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 58.16 53.65 56.77 Shop 8.53 7.16 8.11 Traditional healer 1.80 1.38 1.67 Private pharmacy 0.48 3.66 1.46 Private doctor 1.74 2.23 1.89 Private health facility 5.10 11.35 7.03 Public health facility 21.82 17.60 20.52 Other 2.37 2.97 2.56 Total (n) 1,052 469 1,521 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 59.66 56.76 54.61 55.76 56.59 56.77 Shop 8.48 10.66 11.96 7.24 0.44 8.11 Traditional healer 3.00 2.18 0.97 1.62 0.00 1.67 Private pharmacy 0.34 0.34 0.94 0.65 6.26 1.46 Private doctor 1.34 1.56 2.09 1.32 3.57 1.89 Private health facility 3.58 6.06 5.80 6.93 14.99 7.03 Public health facility 22.29 20.08 19.84 23.94 15.07 20.52 Other 1.31 2.35 3.78 2.54 3.08 2.56 Total (n) 352 314 308 304 242 1,521 Annexes 49 Burkina Faso 1992­1993 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 79.12 66.36 77.35 Traditional healer 5.03 2.62 4.69 Private pharmacy 0.07 0.15 0.08 Private doctor 0.00 1.03 0.14 Private health facility 0.56 2.23 0.79 Public health facility 14.94 27.32 16.66 Other 0.28 0.29 0.28 Total (n) 2,006 322 2,328 Quintile data was not included in the survey. Burkina Faso 1998­1999 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 73.46 46.34 70.64 Traditional healer 7.11 3.58 6.74 Private pharmacy 0.34 0.87 0.39 Private doctor 0.12 1.96 0.31 Private health facility 0.34 1.90 0.50 Public health facility 16.73 43.25 19.49 Other 1.91 2.10 1.93 Total (n) 1,976 230 2,206 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 79.92 69.05 76.35 72.92 51.13 70.64 Traditional healer 6.42 10.72 6.98 5.69 2.61 6.74 Private pharmacy 0.00 0.23 0.64 0.18 1.20 0.39 Private doctor 0.23 0.23 0.00 0.00 1.25 0.31 Private health facility 0.46 0.68 0.36 0.18 0.84 0.50 Public health facility 11.82 16.11 14.17 19.30 40.77 19.49 Other 1.15 2.98 1.50 1.73 2.21 1.93 Total (n) 509 515 368 454 361 2,206 50 TRENDS AND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Cameroon 1991 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 66.10 54.95 61.61 Traditional healer 3.53 1.85 2.86 Private pharmacy 0.88 5.65 2.80 Private doctor 0.71 0.11 0.47 Private health facility 9.67 9.52 9.61 Public health facility 18.75 26.52 21.88 Other 0.35 1.40 0.78 Total (n) 742 500 1,242 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 75.40 69.72 64.17 53.54 49.83 61.61 Traditional healer 5.22 3.80 3.78 1.97 0.42 2.86 Private pharmacy 0.00 0.46 1.63 5.70 5.14 2.80 Private doctor 0.69 0.00 0.62 0.94 0.19 0.47 Private health facility 7.56 8.48 6.59 12.54 11.55 9.61 Public health facility 11.13 17.53 23.22 24.59 30.08 21.88 Other 0.00 0.00 0.00 0.72 2.77 0.78 Total (n) 191 285 212 279 275 1242 Cameroon 1998 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 56.78 48.80 54.56 Shop 6.22 4.82 5.83 Traditional healer 5.42 3.35 4.85 Private pharmacy 2.23 6.15 3.32 Private doctor 0.96 1.12 1.00 Private health facility 7.34 12.22 8.69 Public health facility 19.30 22.33 20.14 Other 1.75 1.22 1.61 Total (n) 800 308 1,108 Annexes 51 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 62.43 61.39 55.01 44.07 47.73 54.56 Shop 3.97 7.97 6.35 6.64 4.12 5.83 Traditional healer 5.96 7.22 4.04 3.73 2.67 4.85 Private pharmacy 0.00 1.14 2.35 7.46 6.66 3.32 Private doctor 1.99 0.00 0.40 0.55 2.17 1.00 Private health facility 4.47 7.97 9.25 12.19 10.46 8.69 Public health facility 19.70 12.60 20.25 24.06 25.05 20.14 Other 1.49 1.71 2.35 1.30 1.13 1.61 Total (n) 257 224 217 233 176 1,108 Central African Republic 1994­1995 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 71.82 55.02 64.65 Shop 0.51 0.19 0.37 Traditional healer 0.64 0.27 0.48 Private pharmacy 0.30 0.68 0.46 Private doctor 0.19 0.29 0.23 Private health facility 4.26 5.10 4.62 Public health facility 22.17 38.47 29.13 Other 0.10 0.00 0.06 Total (n) 988 735 1,723 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 72.28 73.89 62.08 63.26 52.03 64.65 Shop 0.74 0.49 0.00 0.57 0.00 0.37 Traditional healer 1.51 0.89 0.00 0.00 0.00 0.48 Private pharmacy 0.28 0.59 0.26 0.00 1.19 0.46 Private doctor 0.27 0.29 0.00 0.00 0.60 0.23 Private health facility 2.41 4.61 2.73 6.69 6.21 4.62 Public health facility 22.23 19.25 34.94 29.48 39.97 26.13 Other 0.28 0.00 0.00 0.00 0.00 0.06 Total (n) 351 332 303 385 352 1,723 52 TRENDS AND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Chad 1996­1997 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 77.15 61.68 73.69 Shop 2.97 1.62 2.67 Traditional healer 1.76 0.21 1.41 Private pharmacy 0.00 2.99 0.67 Private health facility 9.13 5.49 8.31 Public health facility 5.54 25.62 10.04 Other 3.45 2.39 3.21 Total (n) 1,910 551 2,460 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 89.05 78.28 73.93 68.90 59.60 73.69 Shop 1.20 1.78 6.12 3.26 1.71 2.67 Traditional healer 0.30 1.68 0.99 2.55 1.15 1.41 Private pharmacy 0.00 0.00 0.00 0.00 3.32 0.67 Private health facility 4.35 9.87 10.24 10.29 6.18 8.31 Public health facility 2.09 6.49 4.09 10.67 25.33 10.04 Other 3.01 1.89 4.63 4.32 2.72 3.21 Total (n) 430 615 390 530 496 2,460 Comoros 1996 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 37.79 31.82 36.28 Shop 4.88 2.27 4.22 Traditional healer 12.08 12.88 12.28 Private pharmacy 4.37 15.91 7.29 Private doctor 1.80 5.30 2.69 Private health facility 1.03 4.55 1.92 Public health facility 32.65 21.97 29.94 Other 5.40 5.30 5.37 Total (n) 389 132 521 Annexes 53 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 43.85 28.44 38.46 35.42 32.93 36.28 Shop 5.38 8.26 1.92 3.13 1.22 4.22 Traditional healer 15.38 16.51 14.42 7.29 4.88 12.28 Private pharmacy 3.08 4.59 5.77 10.42 15.85 7.29 Private doctor 0.77 0.92 3.85 1.04 8.54 2.69 Private health facility 0.77 1.83 2.88 2.08 2.44 1.92 Public health facility 26.15 33.03 24.04 35.42 32.93 29.94 Other 4.62 6.42 8.65 5.21 1.22 5.37 Total (n) 130 109 104 96 82 521 Côte d'Ivoire 1994 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 47.56 35.53 43.14 Traditional healer 13.10 11.13 12.37 Private pharmacy 2.44 5.40 3.53 Private health facility 2.48 2.62 2.53 Public health facility 23.42 39.96 29.49 Other 11.00 5.37 8.94 Total (n) 1,047 607 1,654 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 53.66 48.39 43.47 36.63 30.27 43.14 Traditional healer 15.26 12.04 13.13 14.08 5.96 12.37 Private pharmacy 1.18 0.34 4.29 4.98 7.96 3.53 Private health facility 2.56 2.20 2.79 2.05 3.19 2.53 Public health facility 13.69 26.73 27.04 36.49 48.14 29.49 Other 13.65 10.30 9.28 5.76 4.49 8.94 Total (n) 365 347 327 341 275 1,654 54 TRENDS AND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Côte d'Ivoire 1998-1999 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 36.42 37.93 36.88 Traditional healer 24.17 8.24 19.37 Private pharmacy 10.60 10.04 10.43 Private doctor 0.33 0.96 0.52 Private health facility 0.99 2.47 1.44 Public health facility 22.19 37.56 26.82 Other 5.30 2.80 4.54 Total (n) 548 237 785 Quintile data was not included in the survey. Ethiopia 2000 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 78.78 50.03 76.46 Shop 1.81 0.89 1.74 Traditional healer 0.75 0.08 0.70 Private pharmacy 4.07 9.24 4.49 Private doctor 1.83 10.49 2.53 Private health facility 0.61 2.51 0.76 Public health facility 9.32 26.25 10.68 Other 2.83 0.50 2.64 Total (n) 4,362 383 4,745 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total (%) No care sought outside the home 81.97 80.09 76.47 79.40 58.71 76.46 Shop 1.98 2.09 1.44 2.10 0.81 1.74 Traditional healer 0.62 0.65 1.04 0.88 0.05 0.70 Private pharmacy 2.91 3.02 5.21 3.71 8.98 4.49 Private doctor 1.32 0.91 2.31 2.33 7.42 2.53 Private health facility 0.55 0.65 0.99 0.25 1.60 0.76 Public health facility 7.91 8.93 10.22 8.97 20.59 10.68 Other 2.73 3.67 2.31 2.36 1.84 2.64 Total (n) 982 1,038 1,109 943 672 4,745 Annexes 55 Ghana 1993 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 36.73 36.18 36.59 Shop 1.48 1.01 1.36 Traditional healer 7.98 3.52 6.78 Private pharmacy 17.81 11.56 16.12 Private doctor 3.34 1.51 2.85 Private health facility 3.15 5.53 3.79 Public health facility 23.01 36.68 26.69 Other 6.49 4.02 5.83 Total (n) 539 199 738 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total(%) No care sought outside the home 42.34 38.07 30.43 36.30 36.44 36.59 Shop 2.19 1.70 0.62 1.37 0.85 1.36 Traditional healer 10.95 6.82 8.70 4.11 2.54 6.78 Private pharmacy 19.71 17.61 17.39 14.38 10.17 16.12 Private doctor 1.46 2.27 5.59 2.05 2.54 2.85 Private health facility 2.92 3.41 3.11 4.79 5.08 3.79 Public health facility 14.60 23.30 27.33 30.82 39.83 26.69 Other 5.84 6.82 6.83 6.16 2.54 5.83 Total (n) 137 176 161 146 118 738 Ghana 1998 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 48.00 46.77 47.69 Shop 0.89 0.46 0.78 Traditional healer 3.24 1.28 2.75 Private pharmacy 21.28 21.76 21.40 Private doctor 0.22 1.05 0.43 Private health facility 3.17 6.53 4.00 Public health facility 19.11 20.02 19.34 Other 4.10 2.14 3.61 Total (n) 811 269 1,080 56 TRENDS AND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 51.97 51.45 44.40 40.24 46.23 47.69 Shop 1.59 0.00 0.51 0.66 0.85 0.78 Traditional healer 3.45 2.57 2.61 3.49 0.86 2.75 Private pharmacy 15.23 22.57 27.73 24.36 19.63 21.40 Private doctor 0.00 0.78 0.55 0.74 0.23 0.43 Private health facility 1.95 2.71 4.29 5.69 8.10 4.00 Public health facility 20.30 15.00 18.31 21.49 23.29 19.34 Other 5.51 4.93 1.59 3.32 0.81 3.61 Total (n) 312 232 222 170 144 1,080 Guinea 1999 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 51.92 31.81 46.92 Traditional healer 12.75 4.30 10.65 Private pharmacy 1.69 6.16 2.80 Private doctor 1.24 1.28 1.25 Private health facility 1.10 1.31 1.15 Public health facility 28.53 53.67 34.78 Other 2.77 1.49 2.45 Total (n) 1,493 495 1,988 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total(%) No care sought outside the home 55.61 56.19 48.91 39.00 29.99 46.92 Traditional healer 13.07 17.18 10.49 6.69 4.18 10.65 Private pharmacy 1.32 2.86 1.18 4.99 4.23 2.80 Private doctor 1.50 0.78 1.85 0.48 1.63 1.25 Private health facility 1.01 0.84 1.66 0.24 2.22 1.15 Public health facility 23.90 21.13 32.10 46.32 56.65 34.78 Other 3.60 1.01 3.81 2.28 1.09 2.45 Total (n) 507 388 369 389 335 1,988 Annexes 57 Kenya 1993 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 36.38 19.24 34.47 Shop 20.47 16.23 20.00 Traditional healer 1.99 1.09 1.89 Private pharmacy 1.61 5.47 2.04 Private doctor 4.13 6.85 4.43 Private health facility 9.00 18.73 10.08 Public health facility 26.33 32.40 27.01 Other 0.09 0.00 0.08 Total (n) 2,380 298 2,678 Quintile data was not included in the survey. Kenya 1998 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 37.17 25.44 35.19 Shop 9.44 4.35 8.58 Traditional healer 2.85 0.60 2.47 Private pharmacy 5.73 9.28 6.33 Private doctor 4.63 8.77 5.33 Private health facility 10.33 18.11 11.64 Public health facility 28.14 32.09 28.81 Other 1.71 1.37 1.65 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total(%) No care sought outside the home 36.08 40.50 35.19 32.13 29.07 35.19 Shop 9.76 8.56 11.16 7.04 4.99 8.58 Traditional healer 4.24 2.39 2.44 0.49 1.69 2.47 Private pharmacy 3.71 4.59 7.45 11.83 5.96 6.33 Private doctor 4.50 3.99 4.98 4.34 10.45 5.33 Private health facility 7.92 13.15 9.32 10.15 20.83 11.64 Public health facility 30.23 26.39 28.80 32.76 25.30 28.81 Other 3.57 0.42 0.66 1.27 1.71 1.65 58 TRENDS AND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Madagascar 1992 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 54.92 49.73 54.18 Shop 1.43 0.53 1.31 Traditional healer 4.92 2.85 4.62 Private pharmacy 0.96 0.71 0.92 Private doctor 3.14 14.44 4.76 Private health facility 2.19 7.49 2.94 Public health facility 28.28 22.82 27.50 Other 4.17 1.43 3.77 Total (n) 1,843 308 2,151 Quintile data was not included in the survey. Madagascar 1997 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 59.17 48.02 56.81 Shop 5.07 3.62 4.77 Traditional healer 0.29 0.00 0.23 Private pharmacy 0.94 3.03 1.38 Private doctor 6.64 14.90 8.39 Private health facility 2.22 3.83 2.56 Public health facility 24.49 24.98 24.60 Other 1.16 1.61 1.26 Total (n) 1,488 399 1,887 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total(%) No care sought outside the home 55.43 59.18 61.50 57.05 49.19 56.81 Shop 8.00 6.54 4.13 1.90 0.23 4.77 Traditional healer 0.37 0.34 0.30 0.00 0.00 0.23 Private pharmacy 1.69 0.85 2.54 1.12 0.20 1.38 Private doctor 7.04 7.25 5.57 5.43 21.34 8.39 Private health facility 2.37 2.07 1.31 1.71 6.75 2.56 Public health facility 24.78 21.60 23.62 30.92 20.86 24.60 Other 0.31 2.17 1.02 1.86 1.43 1.26 Total (n) 541 369 363 362 251 1,887 Annexes 59 Malawi 1992 Source of Care Rural (%) Urban (%) Total (% No care sought outside the home 36.40 35.73 36.33 Shop 12.88 9.19 12.48 Traditional healer 3.86 2.11 3.67 Private pharmacy 0.00 0.25 0.03 Private doctor 0.72 2.18 0.88 Private health facility 13.26 10.08 12.91 Public health facility 32.56 40.46 33.42 Other 0.32 0.00 0.29 Total (n) 1,733 210 1,944 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 41.16 36.89 40.60 32.56 28.34 36.33 Shop 10.69 12.05 14.84 15.63 8.07 12.48 Traditional healer 2.87 6.10 2.25 3.66 4.01 3.67 Private pharmacy 0.00 0.00 0.00 0.00 0.16 0.03 Private doctor 0.37 0.13 0.32 0.40 3.67 0.88 Private health facility 13.36 11.95 9.92 14.61 15.58 12.91 Public health facility 31.43 32.43 32.08 33.14 39.10 33.42 Other 0.12 0.43 0.00 0.00 1.06 0.29 Total (n) 399 360 462 393 329 1,943 Malawi 2000 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 42.88 30.78 41.56 Shop 30.10 28.27 29.90 Traditional healer 4.57 2.06 4.30 Private pharmacy 0.13 3.09 0.45 Private doctor 0.32 1.70 0.47 Private health facility 6.41 9.45 6.74 Public health facility 14.88 23.87 15.86 Other 0.72 0.78 0.73 Total (n) 5,026 616 5,641 60 TRENDS AND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 47.30 45.86 41.70 40.31 30.81 41.54 Shop 28.49 29.58 29.05 32.65 29.69 29.92 Traditional healer 6.00 3.82 3.93 4.35 2.94 4.28 Private pharmacy 0.00 0.03 0.35 0.12 1.99 0.45 Private doctor 0.23 0.19 0.75 0.07 1.26 0.47 Private health facility 4.26 5.10 6.93 5.44 13.07 6.74 Public health facility 12.85 14.98 16.61 16.32 19.33 15.87 Other 0.87 0.45 0.68 0.74 0.91 0.73 Total (n) 1,252 1,133 1,041 1,206 1,006 5,638 Mali 1995­1996 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 62.43 39.47 57.12 Shop 6.59 7.64 6.83 Traditional healer 8.93 6.38 8.34 Private pharmacy 6.27 20.01 9.45 Private doctor 0.00 0.86 0.20 Private health facility 2.55 1.34 2.27 Public health facility 10.64 23.95 13.72 Other 2.60 0.34 2.08 Total (n) 1,748 526 2,274 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 69.48 60.13 60.21 54.09 35.02 57.12 Shop 6.41 7.41 6.17 8.35 5.29 6.83 Traditional healer 9.72 8.64 9.39 7.42 5.78 8.34 Private pharmacy 1.92 6.53 8.58 10.60 24.07 9.45 Private doctor 0.00 0.00 0.00 0.00 1.35 0.20 Private health facility 3.14 3.09 2.40 1.34 0.96 2.27 Public health facility 7.26 10.97 11.38 15.90 27.17 13.72 Other 2.07 3.23 1.87 2.30 0.35 2.08 Total (n) 475 500 478 486 334 2,274 Annexes 61 Mozambique 1997 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 59.98 41.59 54.61 Shop 0.78 0.39 0.67 Traditional healer 9.11 1.75 6.96 Private pharmacy 0.40 1.91 0.84 Private health facility 0.91 0.38 0.76 Public health facility 26.66 48.59 33.06 Other 2.15 5.40 3.10 Total (n) 1,178 486 1,664 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 63.45 52.77 59.01 53.22 42.28 54.61 Shop 0.40 1.22 0.34 0.98 0.68 0.67 Traditional healer 10.61 16.70 7.44 2.91 1.07 6.96 Private pharmacy 0.31 0.29 0.23 0.48 2.68 0.84 Private health facility 0.34 0.00 2.87 0.18 0.33 0.76 Public health facility 23.14 28.35 28.81 39.36 45.07 33.06 Other 1.75 0.66 1.30 2.86 7.89 3.10 Total (n) 434 187 325 360 359 1,665 Namibia 1992 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 29.69 26.09 28.66 Shop 1.59 2.31 1.80 Traditional healer 3.23 0.48 2.44 Private pharmacy 0.34 1.56 0.69 Private doctor 0.10 4.66 1.41 Private health facility 0.16 0.49 0.25 Public health facility 62.34 61.67 62.15 Other 2.55 2.75 2.61 Total (n) 1,215 490 1,704 62 TRENDS AND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 32.85 28.79 30.12 25.54 24.72 28.66 Shop 1.36 1.01 0.86 3.21 3.01 1.80 Traditional healer 2.70 3.52 3.11 2.38 0.00 2.44 Private pharmacy 0.22 0.52 0.18 0.00 2.76 0.69 Private doctor 0.00 0.00 0.00 0.40 7.65 1.41 Private health facility 0.22 0.00 0.33 0.00 0.80 0.25 Public health facility 60.73 63.87 63.67 64.39 57.70 62.15 Other 1.91 2.30 1.75 4.07 3.36 2.61 Total (n) 380 406 319 301 298 1,704 Niger 1992 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 80.19 53.26 76.65 Shop 5.49 2.44 5.09 Traditional healer 4.42 3.32 4.28 Private pharmacy 0.21 0.00 0.19 Private doctor 0.00 0.11 0.01 Private health facility 0.43 0.45 0.43 Public health facility 8.70 39.73 12.77 Other 0.57 0.68 0.59 Total (n) 1,982 299 2,281 Quintile data was not included in the survey. Niger 1998 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 60.23 41.87 57.34 Shop 15.87 5.93 14.30 Traditional healer 5.75 2.43 5.23 Private pharmacy 0.91 3.32 1.29 Private health facility 0.86 2.39 1.10 Public health facility 13.19 43.13 17.91 Other 3.19 0.95 2.84 Total (n) 1,811 339 2,150 Annexes 63 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 64.57 66.32 58.43 49.26 43.28 57.34 Shop 13.13 11.71 18.70 20.64 7.45 14.30 Traditional healer 6.11 6.27 5.41 5.35 2.29 5.23 Private pharmacy 0.44 0.91 0.00 1.36 4.18 1.29 Private health facility 1.34 0.85 0.29 1.22 2.01 1.10 Public health facility 12.73 11.13 12.03 18.79 39.89 17.91 Other 1.68 2.81 5.13 3.38 0.88 2.84 Total (n) 393 570 413 411 362 2,150 Nigeria 1990 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 44.38 34.67 42.65 Shop 13.87 11.63 13.47 Traditional healer 7.06 3.02 6.34 Private pharmacy 3.95 8.45 4.75 Private doctor 1.41 6.79 2.37 Public health facility 27.66 35.28 29.01 Other 1.67 0.15 1.40 Total (n) 1,774 384 2,158 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 49.08 50.35 40.24 38.51 29.71 42.65 Shop 10.11 12.83 12.06 18.55 14.34 13.47 Traditional healer 10.47 6.79 5.91 5.37 1.55 6.34 Private pharmacy 2.54 1.37 10.97 4.78 6.11 4.75 Private doctor 0.39 1.91 1.92 1.09 7.98 2.37 Public health facility 23.93 25.55 27.29 31.60 40.14 29.01 Other 3.49 1.20 1.62 0.10 0.17 1.40 Total (n) 482 523 380 431 342 2,158 64 TRENDS AND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Nigeria 1999 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 45.97 29.99 41.91 Shop 2.06 0.61 1.69 Traditional healer 4.61 0.39 3.54 Private pharmacy 16.18 17.76 16.58 Private doctor 0.69 2.83 1.24 Private health facility 6.96 9.68 7.65 Public health facility 23.10 37.72 26.82 Other 0.42 1.00 0.56 Total (n) 740 252 992 Quintile data was not included in the survey. Rwanda 1992 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 59.11 47.90 58.62 Traditional healer 3.77 2.69 3.72 Private pharmacy 3.68 6.29 3.79 Private doctor 0.66 1.80 0.71 Public health facility 26.88 35.63 27.26 Other 5.91 5.69 5.90 Total (n) 2,602 118 2,719 Quintile data was not included in the survey. Senegal 1992­1993 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 65.22 53.22 61.68 Shop 2.70 0.51 2.05 Traditional healer 3.62 2.88 3.40 Private pharmacy 0.64 1.86 1.00 Private doctor 0.28 0.51 0.35 Private health facility 3.26 5.76 4.00 Public health facility 22.50 32.71 25.51 Other 1.77 2.54 2.00 Total (n) 1,409 590 1,999 Quintile was not included in the survey. Annexes 65 South Africa 1998 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 36.22 26.24 31.36 Shop 0.40 0.30 0.35 Traditional healer 0.05 0.11 0.08 Private pharmacy 2.05 2.42 2.23 Private doctor 11.66 27.41 19.33 Private health facility 1.02 3.48 2.22 Public health facility 48.09 39.39 43.86 Other 0.50 0.65 0.58 Total (n) 914 867 1,781 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 35.51 31.61 32.90 29.74 24.92 31.36 Shop 0.62 0.00 0.22 0.50 0.48 0.35 Traditional healer 0.12 0.23 0.00 0.00 0.00 0.08 Private pharmacy 1.33 3.00 1.55 0.36 5.27 2.23 Private doctor 5.67 14.15 15.69 22.97 46.84 19.33 Private health facility 1.32 0.91 1.23 2.40 6.45 2.22 Public health facility 54.36 49.97 48.03 43.30 15.35 43.86 Other 1.06 0.12 0.37 0.72 0.68 0.58 Total (n) 401 419 366 310 285 1,781 Tanzania 1992 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 38.76 27.60 36.17 Traditional healer 3.98 0.53 3.18 Private pharmacy 2.80 7.90 3.99 Private health facility 7.47 10.88 8.26 Public health facility 46.26 51.77 47.54 Other 0.73 1.31 0.87 Total (n) 1,899 575 2,475 Quintile data was not included in the survey. 66 TRENDS AND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Tanzania 1996 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 33.96 21.75 31.60 Traditional healer 1.65 0.10 1.35 Private pharmacy 10.63 9.44 10.40 Private health facility 5.20 11.99 6.51 Public health facility 47.48 55.99 49.12 Other 1.08 0.73 1.01 Total (n) 1,805 432 2,237 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 41.93 33.69 30.39 26.44 25.33 31.60 Traditional healer 2.38 1.95 1.04 1.34 0.00 1.35 Private pharmacy 10.19 11.96 9.86 10.50 9.75 10.40 Private health facility 4.32 6.68 5.60 6.08 10.37 6.51 Public health facility 39.70 45.73 52.50 53.93 53.57 49.12 Other 1.47 0.00 0.61 1.72 0.97 1.01 Total (n) 485 372 457 500 424 2,237 Togo 1998 Source of Care Rural (%) Urban (%) Total (%); No care sought outside the home 60.87 47.85 57.93 Shop 7.78 6.79 7.56 Traditional healer 3.16 3.97 3.34 Private pharmacy 0.08 1.86 0.48 Private doctor 0.18 1.19 0.41 Private health facility 1.34 7.87 2.82 Public health facility 17.36 21.71 18.34 Other 9.23 8.76 9.13 Total (n) 1,523 445 1,968 Annexes 67 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 64.13 62.83 57.14 55.76 42.79 57.93 Shop 8.40 8.82 8.47 4.85 6.09 7.56 Traditional healer 3.40 3.09 3.65 3.81 2.56 3.34 Private pharmacy 0.00 0.00 0.30 0.43 2.48 0.48 Private doctor 0.18 0.00 0.00 0.43 2.06 0.41 Private health facility 0.52 1.44 1.55 3.49 10.21 2.82 Public health facility 16.23 15.33 18.26 20.13 24.89 18.34 Other 7.14 8.49 10.64 11.09 8.90 9.13 Total (n) 487 446 413 350 272 1,968 Uganda 1995 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 26.38 16.10 25.33 Shop 17.13 8.51 16.25 Traditional healer 2.52 1.31 2.39 Private pharmacy 1.80 7.05 2.34 Private doctor 3.51 6.50 3.82 Private health facility 27.21 43.48 28.88 Public health facility 20.07 16.00 19.65 Other 1.38 1.05 1.35 Total (n) 2,736 313 3,049 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 26.41 27.08 29.29 24.24 18.01 25.33 Shop 18.56 17.82 17.56 13.14 12.98 16.25 Traditional healer 5.40 2.25 0.87 2.04 0.52 2.39 Private pharmacy 1.18 1.91 2.01 2.47 4.80 2.34 Private doctor 4.17 2.60 2.87 5.02 4.62 3.82 Private health facility 21.13 29.22 27.20 32.13 37.66 28.88 Public health facility 21.05 17.85 19.48 19.44 20.49 19.65 Other 2.09 1.28 0.72 1.52 0.92 1.35 Total (n) 710 656 594 593 495 3,049 68 TRENDS AND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Uganda 2000­2001 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 19.31 8.59 18.17 Shop 7.98 2.01 7.35 Traditional healer 1.03 0.73 1.00 Private pharmacy 10.71 16.54 11.32 Private doctor 0.81 0.73 0.80 Private health facility 34.92 57.21 37.28 Public health facility 21.51 12.31 20.54 Other 3.73 1.87 3.54 Total (n) 3,192 378 3,570 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 22.26 20.28 20.09 15.33 9.48 18.17 Shop 10.98 8.08 6.89 5.66 3.03 7.35 Traditional healer 1.07 0.11 2.03 1.19 0.75 1.00 Private pharmacy 9.42 8.78 12.11 12.86 15.44 11.32 Private doctor 1.12 0.23 0.91 1.17 0.60 0.80 Private health facility 30.54 33.52 34.89 40.27 53.23 37.28 Public health facility 21.11 25.82 18.52 19.49 15.26 20.54 Other 3.51 3.18 4.55 4.04 2.23 3.54 Total (n) 867 830 670 666 537 3,570 Zambia 1992 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 28.44 17.91 23.69 Shop 3.90 7.86 5.69 Traditional healer 5.90 1.94 4.12 Private pharmacy 1.01 6.32 3.41 Private doctor 0.50 2.84 1.56 Private health facility 8.39 14.67 11.22 Public health facility 43.22 45.38 44.20 Other 8.62 3.08 6.12 Total (n) 1,626 1,336 2,962 Quintile data was not included in the survey. Annexes 69 Zambia 1996 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 33.86 24.25 29.74 Shop 5.10 8.89 6.72 Traditional healer 6.88 1.01 4.36 Private pharmacy 0.26 4.98 2.28 Private doctor 0.54 1.75 1.06 Private health facility 4.16 11.25 7.20 Public health facility 44.22 45.98 44.98 Other 4.98 1.89 3.65 Total (n) 1,881 1,412 3,293 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 36.02 35.32 32.49 24.27 21.32 29.74 Shop 3.84 4.98 6.44 10.12 8.04 6.72 Traditional healer 8.00 7.35 4.22 1.76 0.53 4.36 Private pharmacy 0.28 0.00 0.72 5.60 4.09 2.28 Private doctor 0.00 0.43 0.00 2.36 2.27 1.06 Private health facility 2.96 4.79 5.09 1.78 22.85 7.20 Public health facility 43.76 43.44 45.97 51.79 38.83 44.98 Other 5.14 3.69 5.08 2.32 2.07 3.65 Total (n) 825 512 552 762 642 3,293 Zambia 2001­2002 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 30.00 32.22 30.72 Shop 3.72 4.25 3.89 Traditional healer 4.48 1.54 3.54 Private pharmacy 0.99 5.73 2.51 Private doctor 0.49 0.77 0.58 Private health facility 11.88 6.98 10.31 Public health facility 46.10 47.10 46.42 Other 2.33 1.41 2.04 Total (n) 1,848 873 2,721 70 TRENDS AND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Zimbabwe 1994 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 34.06 32.68 33.75 Shop 15.87 7.48 13.98 Traditional healer 3.70 2.38 3.40 Private pharmacy 1.76 5.67 2.64 Private doctor 0.74 10.74 2.99 Private health facility 6.70 2.12 5.67 Public health facility 33.97 37.47 34.76 Other 3.21 1.46 2.81 Total (n) 983 286 1,269 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 39.63 33.79 31.67 28.86 32.22 33.75 Shop 12.88 16.09 19.13 13.28 8.12 13.98 Traditional healer 4.79 4.61 1.84 3.61 0.91 3.40 Private pharmacy 0.87 1.60 1.87 4.62 5.34 2.64 Private doctor 0.00 0.45 0.91 3.85 13.02 2.99 Private health facility 4.91 8.40 7.77 3.65 3.96 5.67 Public health facility 33.40 30.70 32.93 40.55 36.41 34.76 Other 3.52 4.37 3.89 1.57 0.00 2.81 Total (n) 349 237 228 266 189 1,269 Zimbabwe 1999 Source of Care Rural (%) Urban (%) Total (%) No care sought outside the home 40.56 42.70 41.21 Shop 10.53 9.46 10.21 Traditional healer 1.35 0.00 0.94 Private pharmacy 1.42 6.44 2.94 Private doctor 0.22 13.31 4.20 Private health facility 5.04 5.63 5.22 Public health facility 36.32 22.47 32.11 Other 4.56 0.00 3.17 Total (n) 999 436 1,436 Annexes 71 % of people using different sources of care by quintile of wealth index Source of Care Lowest Second Middle Fourth Highest Total No care sought outside the home 44.06 41.89 38.21 40.35 41.38 41.21 Shop 8.58 9.62 12.13 10.19 10.68 10.21 Traditional healer 1.83 0.55 1.82 0.41 0.00 0.94 Private pharmacy 0.79 1.25 1.61 5.66 5.48 2.94 Private doctor 0.26 0.37 0.00 6.65 14.49 4.20 Private health facility 5.43 2.20 4.95 9.83 2.97 5.22 Public health facility 35.65 40.05 35.41 24.53 25.00 32.11 Other 3.40 4.08 5.88 2.37 0.00 3.17 Total (n) 307 277 280 311 261 1,436 Use of public and private services among people from the highest income quintile who sought care when the child was sick Among those who sought care outside the home, % who went to: Country and % seeking care year of DHS Private sector Public sector Other Total outside home Burkina 99 12% 83% 5% 100% 48.9 CAR 95 17% 83% 0% 100% 48.0 Guinea 99 18% 81% 2% 100% 70.0 Mozambique 97 8% 78% 14% 100% 57.7 Namibia 92 19% 77% 4% 100% 75.3 Tanzania 96 27% 72% 1% 100% 74.7 Niger 98 28% 70% 2% 100% 56.7 Cote d'Ivoire 94 25% 69% 6% 100% 69.7 Chad 97 31% 63% 7% 100% 40.4 Ghana 93 33% 63% 4% 100% 63.6 Cameroon 91 34% 60% 6% 100% 50.2 Nigeria 90 43% 57% 0% 100% 70.3 Malawi 92 44% 55% 1% 100% 71.7 Ethiopia 00 46% 50% 4% 100% 41.3 Zambia 96 48% 49% 3% 100% 78.7 Comoros 96 49% 49% 2% 100% 67.1 Cameroon 98 50% 48% 2% 100% 52.3 (continued on next page) 72 TRENDSANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Use of public and private services among people from the highest income quintile who sought care when the child was sick (continued) Among those who sought care outside the home, % who went to: Country and % seeking care year of DHS Private sector Public sector Other Total outside home Benin 96 53% 46% 1% 100% 51.2 Togo 98 41% 44% 16% 100% 57.2 Ghana 98 55% 43% 2% 100% 53.8 Mali 96 58% 42% 1% 100% 65.0 Madagascar 97 56% 41% 3% 100% 50.8 Gabon 00 56% 38% 7% 100% 79.0 Kenya 98 62% 36% 2% 100% 70.9 Benin 01 58% 35% 7% 100% 43.4 Malawi 00 71% 28% 1% 100% 69.2 Uganda 95 74% 25% 1% 100% 82.0 S. Africa 98 79% 20% 1% 100% 75.1 Uganda 01 81% 17% 2% 100% 90.5 Gross average: 44% 52% 4% 100% 62.9 Annex 2: Case Studies of Franchises in Africa TOP Reseau, Madagascar are paid US$150 per month (from donor funds) and must lead at least 40 education sessions per I n 2000, Population Services Interna- month. At typical educational session, peer edu- tional, a NGO, launched "TOP cators would distribute literature, discuss Reseau," a franchise of reproductive reproductive health issues, and lead condom health care, with a US$1 million grant demonstrations. Peer educators play an impor- from the Bill and Melinda Gate Foundation. PSI tant role in improving franchisee clinics' use rate. acts as a franchiser. There are 17 clinics in the "My business has increased 25% since I joined network, all in the city of Tamatave and sur- the franchise, and 60% to 70% of my clients rounding rural areas. are now youth," said one doctor. The clinics were invited to join the franchise Diagnosis and treatment for sexually trans- after interviews and evaluations. A detailed mitted infections (STIs) draw many of the youth. contract specifies the franchise arrangement. While the HIV prevalence rate is less than 1% Extensive and continual training is provided nationwide, the STI rate is as high as 30% in to improve both technical skills and manage- some areas, a concern for many youth. Other ment capacity, with an emphasis on services include family planning counseling, youth-friendly services. PSI has created and pregnancy testing and counseling, as well as gen- refined a training manual on youth-friendly eral services such as immunizations, physical services, as well as other operational manuals, exams, breast exams, and pap smears. The pro- curricula, client kits, and flip-chart visual aids viders currently pay a modest membership fee for providers and youth clients (primarily 15 to TOP Reseau, but indicated a willingness to to 24 years of age). pay more for franchise services as their business The clinics in the network generally provide continues to increase. a full range of services through general practi- In the first two years, 17 clinics served total tioner physicians. The clinics, all of which of 2,500 clients, with about half of the visits display a matching TOP Reseau logo, benefit related to reproductive health issue. The project from marketing and media promotion on tele- faces high costs, however, and large payments vision and radio, and from referrals from other to peer educators may not be replicated else- components of the project coordinated by PSI. where. In order to sustain and expand the TOP Reseau builds its client base through the project, the franchiser plans to (1) develop rela- work of peer educators in the schools, commu- tionships with the private sector and other nity events, and other activities. Peer educators community organizations; (2) document the 73 74 TRENDSANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA impact of franchising techniques; and (3) explore This was extended for a second 5-year period; ways to diversify funding and generate revenue. current support from FPIA is approximately US$120,000 per year. In addition, KMET re- Source: LaVake, Steven, "Applying Social Fran- ceives US$25,000 per year for specific activities chising Techniques to Youth Reproductive to support two member providers and related Health /HIV Services," Youth Issues Paper 2, community-based distribution of family plan- Family Health International, 2003. ning products on the remote Mfangano Island Contact: www.psi.org in Lake Victoria. KMET also receives free in- kind family planning commodities from the government, which is passed on at no cost to KMET, Kenya participating members and placed in condom- distribution boxes in bars and other locations In 1996, KMET, a NGO, started operation of a targeting at-risk populations. reproductive health franchise in Kisumu, West- ern Province, and is established in hard-to-reach Expenditures areas of five of Kenya's eight provinces. KMET acts as a franchiser and started the operation by The bulk of KMET's expenditures is for staff training private doctors and consultants in safe cost. KMET employs four full-time staff, an abortion practices and post-abortion care. The executive director, and support staff. KMET program quickly grew to include midwives, clini- rents one building near the town of Kisumu, and cal officers, and nurse practitioners, and as of two provincial satellite offices. KMET owns two 2001 all new providers were mid-level provider 4-wheel drive pickup trucks, which are used to cadres. transport staff and to deliver family planning Providers are required to meet set facility stan- supplies to members. dards. In exchange, they receive free training, a free initial MVA kit, and regular delivery of Activities contraceptive commodities to their clinic. Fran- chisees pay some token annual membership fee, In addition to having 125 franchisees, KMET which makes them eligible to participate in a operates one health clinic itself, which employs revolving loan program. KMET has extended two nurses and one paramedic, and is further medical training to include family planning and supported by volunteers from the nearby com- other reproductive health services, including STI munity. There are approximately 200 registered management and home-based care for HIV/ members of the KMET network working in five AIDS. s a result of their involvement in HIV/ provinces.4In addition to the provider network, AIDS, they started training community-based KMET has produced models in advocacy and workers (CBWs) in early 2002 in a number of policy development; in participatory learning pilot locations. CBWs receive instruction from practices and collaborative community net- KMET trainers in home-based care, with a fo- works; and in curriculum development, peer cus on support for AIDs patients. The CBW education, and provider-training programs. The network is currently active only in the Kisumu government, USAID, and other health organi- area, but expansion is planned. Financial support 4 Although this is the number of registered mem- bers, initial survey results suggest that some of KMET received a 5-year grant from Family Plan- these members are not currently active, and so ning International Assistance (FPIA) in 1996. these must be verified by the ongoing survey. Annexes 75 zations have replicated these models. The AIDS counseling, home based care, and de- projects are funded minimally, and operate pri- worming of children. The Clinics are operated marily through volunteers organized by the by nurses, who receive a four week training, KMET staff. primarily in business management. The CFW Clinics sell 81 essential drugs, and are also per- Cost-effectiveness mitted to charge for diagnosis and treatment of more general illnesses. As with the shops, only KMET has not calculated or kept records spe- drugs bought from CFW headquarters are per- cifically for cost-effectiveness. Nonetheless, using mitted in the franchise outlets. Prices are clearly their MIS data on provider activities, together advertised, and regular controls are imple- with their volume of free-condom distribution, mented. As of December 2003, SHEF/CFW it is possible to calculate a Cost per Couple Year operated 40 franchised drug shops and 20 fran- of Protection (Cost/CYP) for their activities of chised clinics. approximately US$4.11. While this figure is based on a number of assumptions, it does rep- Financial support resent a cost/CYP value at the lower end of all supported family planning programs in Sub-Sa- SHEF/CFW currently operates with an annual haran Africa.5 budget of approximately US$250,000. Over 80% of the shops are operating in a self-sus- Source: Montagu, Dominic, "Output based taining manner. All CFW shops pay franchise services for health and their potential applica- fees to the SHEF/CFW (15% markup on drugs, tion in Kenya," June 2003 and 5% of gross income). SHEF/CFW estimates Contact: Solomon Orero, KMET Solorero@ that it could be financially self-sufficient as a hotmail.com franchiser with 500 franchised shops. With the recent introduction of nurse-run CFW Clinics, which have higher incomes and therefore pay Sustainable Healthcare Enterprise higher franchise fees, this calculation may Foundation (SHEF), Kenya change. In April 2000, the Sustainable Healthcare En- Result to date terprise Foundation's Child and Family Welfare Project launched a franchise of drug shops, all SHEF/CFW franchisees had treated approxi- operating under the Child and Family Welfare mately 250,000 patients as of March 2003. The (CFW) brand. The shops are owned and run founder evaluates that keys to their success by community health workers who have un- include strict implementation of rules and stan- dergone training by the Anglican Church, and dards (e.g., disqualification of franchisees if have received an additional four weeks train- there is violation), and making sure that ing from CFW upon initiation. CFW Shops sell there is proprietary value in the business for 18 essential drugs, all of which must be pur- franchisees. chased from the CFW headquarters. No non-CFW drugs are allowed to be sold from a franchise. 5 Assumptions are made as to use of free condoms, In April 2003, the network expanded to in- calculated as 1/3 of the rate of purchased condoms. clude medical clinics staffed by nurses, also In addition, many of the services included, such operating under the Child and Family Welfare as surgical sterilization, would likely occur to some (CFW) brand. New services added include HIV/ extent without the involvement of KMET 76 TRENDSANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Source: Montagu, Dominic, "Output based ser- seen 57,000 clients through December 2001, and vices for health and their potential application the average number of monthly clients has in- in Kenya," June 2003. Kimbo, Liza, presenta- creased from 230 in 1999 to more than 4,000 in tion at Engaging the private Sector in Franchising 2002. PSI estimates that 120,000 cases of pri- for the Public Good: Technical Seminar Series, mary and secondary HIV cases were averted in May 2003. CFW Shops Website, http:// 2002. www.cfwshops.org. New Start Centers are established in high-traf- Contact:Liza Kimbo liza@cfwshops.org fic areas or at institutions that already provide health-related services, in order to help reduce the uneasiness many VCT clients feel due to the New Start, Zimbabwe stigma of HIV. Ten of New Start centers inte- grate VCT services into existing health service In order to respond to the HIV/AIDS crisis, the delivery institutions, such as public clinics and HIV/AIDS and TB Unit of the Zimbabwean hospitals, non-governmental organizations, and Ministry of Health and Child Welfare began a private health facilities. The other two centers national franchise of Voluntary Counseling are "free-standing" sites, which are operated and and Testing (VCT), New Start, in 1999. The managed directly by PSI. U.S. Agency for International Development Quality is maintained through agreements (USAID) provides funding, while Population between PSI/Zimbabwe and the health centers Services International (PSI) provides technical that stipulate the standards to which the centers and managerial assistance. New Start targets must adhere, and by regularly conducting "mys- high-risk population such as young couples, tery client surveys," in which researchers pose adolescents, commercial sex workers, transport as clients. New Start counselors undergo exten- industry workers, and other "mobile" popula- sive initial and follow-up training. tions. PSI says one factor that contributed to the At New Start test centers, counselors advise success of the project is strategic and culturally clients before and after each HIV test. The coun- sensitive multi-media communication campaigns selors are trained in protocols and procedures through TV, radio, and print. PSI operates simi- to provide on-site, state-of-the-art, rapid HIV lar franchising on family health and VCT in testing, combined with pre- and post-test coun- Angola, Botswana, Burkina Faso, Madagascar, seling sessions that include referrals--when Mali, Mozambique, Namibia, Rwanda, appropriate--to community support groups. Swaziland and Togo. For the first two years, New Start used stan- dard testing technology, which required clients Source: PROFILE : New Hope with New Start, to wait a week before getting the results. How- PSI, May 2002. Presentation made at Social ever, because of the time lag, a large portion of Franchising of TB and TB/HIV in Low Resource New Start clients did not return for their re- Settings Meeting at the Rockefeller Foundation, sults. To improve the situation, New Start April 2003. network introduced new rapid HIV tests in 2001. The new tests permit clients to receive their results within an hour, which increased the ProFam, Zimbabwe portion of clients who receive their results from 77% to 97%. PSI/Zimbabwe began the Professional Family New Start was launched with only one cen- Planning Services Project (ProFam) in October ter, and has since grown into a network of 12 1997 to increase demand for and supply of re- centers throughout the country. The network has productive health services in the private sector. Annexes 77 The project trains private-sector, medical-service As the declining value of local currency forced providers such as doctors, independent nurses, the price of ProFam products beyond the reach and pharmacists, who in turn provide quality of its target consumers, the initiative was rede- reproductive health services and contraceptive signed to provide products at subsidized prices products at affordable prices. to ensure affordability. 78 TRENDSANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Annex 3: Case Studies of Vouchers in Africa Vouchers for Insecticide-Treated Bed through a network of public and private pro- Nets in Tanzania viders, including shopkeepers, health workers, and village leaders. These agents were given a T he KINET project in Tanzania employed credit of 550 shillings (including a 50-shilling a competitive voucher program to sub- handling charge) for each voucher by the whole- sidize insecticide-treated bed nets, which saler or by the project when they purchased the have been shown to give substantial pro- nets. Wholesalers purchased nets directly from tection from malaria, in two districts of southern the project and were given a credit of 600 shil- Tanzania. The vouchers were targeted to preg- lings for each voucher returned to the project. nant women and women with children under Overall, 23% of the more than 65,000 so- five years of age. The project distributed dis- cially marketed mosquito nets sold during the count vouchers through mission and project period were purchased with vouchers. government maternal and child health clinics, Voucher return rates were extremely high: a to- with the aim of: tal of 8,000 vouchers were issued, and of these, 7,720 were returned. There was some evidence · reducing the price of treated nets for preg- of higher voucher use among least poor women, nant women and women with young children as opposed to the poorest women (many of · drawing attention to the group most at risk whom could not afford a net, even with a of severe disease voucher). However, two years after the program · promoting increased equity among pregnant was instituted, awareness of the program among women and young children. target groups was only 43%, and only 12% of women had used a voucher to purchase a net Vouchers were distributed to women when (although some claimed to have already pur- they came to the clinics for treatment. Clinic staff chased one). There was some indirect evidence wrote the name, date, and contact details on the of misuse; however, the extent was difficult to part of the voucher given to the women and the measure. Nonetheless, it seemed likely that some part returned to the project team. A mark was vouchers were used by ineligible household made on the health card of the mother or child members and households from other areas. to indicate that a voucher had been given. Each Evaluation of the project revealed that a sub- voucher could be used as part payment (500 stantial amount of time--several years--was Tanzanian shillings) for a treated net (total cost needed for people to understand and participate of 3,000 shillings). The nets were available in the program. A promotional campaign that 79 80 TRENDS ANDOPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA involves multiple communication channels may agents. All groups were trained to provide in- be warranted. The voucher program did, how- formation on emergency contraception and ever, serve as a strong link between the public distribute vouchers for ECPs. Staff of the public health system and private sales agents, optimiz- clinics and private pharmacies was also trained ing the skills of each to increase the coverage of to provide the pills, but regulatory restrictions treated nets. Finally, the program also served as prevented peer counselors and sales agents from a promotional tool to heighten awareness of the directly providing ECPs. health benefit of treated nets--and to identify When potential users were given information the group most at risk of severe malaria--among about emergency contraception by any partici- family members and members of the commu- pating health worker, they were given a voucher, nity in general. which could then be redeemed for a pack of ECPs, either by the same health worker or by Source: Mushi, Adiel K., Joanna RM Armstrong any one of the project's participating providers. Schellenberg, Haji Mponda, and Christian On each voucher, the information provider in- Lengeler, "Targeted subsidy for malaria control dicated the age of the client and whether the with treated nets using a discount voucher client was in or out of school, and could also scheme in Tanzania," Health Policy and Plan- indicate the type of provider he/she was. Once ning, 18(2): 163­71, 2003. the card was redeemed for a pack of pills, the supplier noted on the back of the card the cat- egory to which he/she belonged. Clients who A voucher program for delivering came to a supplier without a card were asked emergency contraception to young where they had received information about people in Zambia emergency contraception, and the card was filled out appropriately. If the supplier was also the This voucher program was undertaken as part initial source of information, the supplier com- of a study in Lusaka and its environs to deter- pleted both sides of the card. The number of mine the most effective channels for delivering vouchers given to each category of health work- "youth-friendly" emergency contraceptive ser- ers was recorded to keep track of the number of vices. Two earlier studies had shown that (i) contacts made. access to family planning services is severely lim- Providers were supplied with stocks of ECPs. ited for many potential users, including youth, At the public clinics vouchers could be redeemed in part due to cultural norms and provider bi- for ECPs free of charge. The private pharmacies ases; and (ii) that young people were not seeking were permitted to charge, and the maximum out emergency contraception in part because of price was set at 500 Kwacha (US$0.13). The lack of anonymity or privacy at local clinics or normal market price was 15,000 Kwacha counseling centers. The study thus aimed to (US$3.75). Throughout the study, redeemed gauge, both quantitatively and qualitatively, vouchers were collected on a monthly basis. The preferences for sources of information and ser- data were then processed, thereby allowing the vices among youth. research to track over time the frequency and The study identified four different types of patterns by which the vouchers were issued and health workers to provide information on emer- redeemed. gency contraception and/or actual emergency Between February 26 and September 30, contraception pills (ECPs). The four groups were 2000, 3,517 vouchers had been distributed to peer counselors, based at public sector health the four types of workers. Of those, 1,798 facilities; clinic-based outpatient health care pro- vouchers had been given to clients, and 421 were viders; private pharmacists; and community sales redeemed by women between the ages of 12 Annexes 81 and 45. Pharmacists clearly emerged as the most · create a reproductive health information and frequently used provider of ECPs, although they service environment responsive to the needs were also one of the most difficult to coordi- of young people. nate and organize for training purposes. Their fairly anonymous approach, though perhaps The project was piloted for three years and less than ideal from an informational perspec- had two components: information and referral, tive, made them more attractive sources of and service provision. Adult educators were used information and ECPs. Young people tended in the information component. Young parents to distrust more traditional providers of family were recommended from the community to serve planning services. The second choice of provider as "Friends of Youth" (FOY), who were paid varied with age, with peer counselors being an honorarium for their services. They were more popular among younger clients and clinic/ trained and assigned an area of operation. They outpatient nurses being more popular among reached young people through existing groups older clients. and institutions, or they formed young people into their own groups. By the end of the pilot Source: Skibiak, John P., Mangala Chambeshi- period (1998-2000), about one-third of young Moyo, and Yusuf Ahmed, "Testing alternative people in the project area had had direct con- channels for providing emergency contraception tact with a FOY. to young women," 2001. In the service component, a network of pri- vate and public service providers were used to reach the largest number of young people. The A voucher program in the Nyeri Youth providers' skills were updated, and they were Health Project in Kenya trained on "youth friendliness." In all, there were 12 service outlets participating in the project-- The Nyeri Youth Health Project (NYHP) is a including seven private clinics, two public pilot community-based project for young people facilities, a lab, one chemist, and one counselor. living in Nyeri, the capital of Kenya's Central FOYs issued vouchers and a list of service pro- Province, which is the homeland for the Kikuyu, viders to young people. In addition, service Kenya's largest ethnic group. The project is a providers used the voucher to refer young people collaborative effort between the Population to other providers, such as a lab for tests, if Council and the Family Planning Association of needed. The voucher entitled the young person Kenya. Earlier studies in Nyeri revealed that to services at a subsidized cost. The program young people and parents preferred that adults, was designed so that all parties contributed to rather than peers, give sexual and reproductive the cost of providing the service. The young per- health information to young people (this is con- son paid KSH 50 (about US$0.60), the providers sistent with local Kikuyu culture). waived their consultation fee, and FPAK paid The goals of the NYHP, targeted at unmar- for any additional costs over and above that, ried people between the ages of 10 and 24 (as including drugs for treating STIs and lab tests. well as the adults who influence their environ- If the young person could not pay the fee, it was ment), were to: waived. The voucher itself had three sections: one sec- · delay the onset of sexual activity among youth tion was retained by the FOY as a record of not yet sexually active who had been referred; the other two sections · prevent sexually experienced youth from suf- were given to the young person who, in turn, fering negative consequences of sexual gave them to the provider. One of these two sec- activity tions was submitted to the Family Planning 82 TRENDSAND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA Association of Kenya (FPAK) with the diagno- received by girls (15%) and male circumcision sis and the treatment given so that FPAK could for boys (15%). Originally, male circumcision reimburse the cost of services. The coupons had was not one of the services offered, but young serial numbers. If a coupon was issued by a FOY, men in the project area requested that it be pro- but not returned to FPAK--implying that the vided, as it was part of a socially recognized rite young person did not go for the services--the of passage into adulthood. The cost to FPAK of issuing FOY was given the serial number of the subsidizing the services was approximately $7 coupon and asked to follow up with the young per contact. person. During the pilot phase of the project, An evaluation of the program revealed that nearly 100% of the vouchers were used, because in 2001, compared with 1997, fewer Nyeri boys young people saved money by using the coupon and girls had initiated sex, more had abstained and because there was a mechanism to follow or used condoms, fewer had multiple sexual up if services were not sought within a reason- partners, and more communicated with a par- able time. ent or other adult. In contrast, behavior in a During the 3-year project, 2,800 young people neighboring control area worsened over the pe- were referred and received services using this riod. system. The majority of young people received treatment for STIs (about 55%). After that, fam- Source: Erulkar, Annabel, Personal communi- ily planning was the most popular service cation relayed through Anna Gorter, 2003. Annex 4: Case Studies of Contracting for Health Service Provision in Africa Contracting for Hospital Services in The contract did not manage to contain costs, Zimbabwe a result of overuse and the fee-for-service na- ture of the contract. Approximately 70% of I n Zimbabwe, the Ministry of Health has provincial non-salary recurrent expenditure was maintained a long-standing contract allocated to the hospital, yet the hospital was with Wankie Colliery, a 400-bed hospi- accessible to only a minority of the population. tal, to provide the services of a district The absence of skilled personnel to appropri- hospital. Patients who are characterized as "gov- ately monitor the contract and screen patients ernment responsibility" (patients exempt from was an important constraint affecting the out- user fees at public hospitals) may be treated at come of the arrangement. The contract was the hospital on a fee-for-service basis. A gov- ultimately terminated due to disagreements over ernment official stationed in the hospital certifies revisions (Mills 1997). the eligibility of the patient, the patient is treated, and the hospital sends the bill for treatment to Source: McPake, Barbara and Charles Hongoro, the Ministry of Health. "Contracting Out of Clinical Services in Zim- The original contract was signed in the 1950s, babwe," Social Science and Medicine. 41(1): without a competitive tendering process and 13­24, 1995. prompted by the lack of government hospitals Mills, Anne (1997). "Contractual relation- in the area. The contract does not specify charge ships between Government and the commercial rates, although a fee-for-service system has op- private sector in developing countries", in Pri- erated since the inception of the contract. There vate health providers in developing countries: is no formal monitoring of quality of services or serving the public interest, ed. Sara Bennett, operation of the contract. Barbara McPake, and Anne Mills, Zed Books: An analysis of this contract by McPake and London & New Jersey, p. 189­213. Hongoro in 1995 revealed several important lessons. Because there was no competition and the government did not retain its capacity Contracting with General to offer services, the monopoly position of Practitioners in South Africa Colliery in the district was entrenched, rather than challenged by the contract. This and In both the Western and Eastern Cape of South the failure of the government to appropriately Africa, provincial governments have long-stand- screen patients for their ability to pay and need ing contracts (dating to the apartheid era) with for hospital services led to excessive provision. general practitioners, referred to as part-time 83 84 TRENDSAND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA district surgeons, to provide clinical and other specifying and monitoring complex services, medical services in a defined geographical area, particularly in rural areas. The weakness of the predominantly rural. South Africa has a well- formal contract meant that practitioners de- developed private health sector, and few general cided for themselves how to deliver services, practitioners are willing to work for the public with professional and ethical standards the key sector, particularly in rural areas. The depart- motivation. ments of health of provincial governments therefore contracted with private physicians to Sources: Palmer, Natasha and Anne Mills, "Clas- work part-time provide predominantly curative sical versus Relational Approaches to primary care services in rural areas. Contracts Understanding Controls on a Contract with In- entail a basic fee-for-service agreement, specify- dependent GPs in South Africa," Health ing a general list of medical services to be Economics 12: 1005­20, 2003. provided. Palmer, Natasha, "Does Duty Call: Contracts A careful analysis of this contracting program and GPs in South Africa," Insights Health, (Palmer 2003; Palmer and Mills 2003) found March 2003. that the contracts were incomplete and open to interpretation by the practitioners. No standards for quality were outlined in the contract. Sanc- Contracting for Basic Health Services tions (termination of contract) for negligence, in the Democratic Republic of Congo breach of material conditions, or misconduct were included in the contract; however, examples This US$49 million, Bank-financed health of what such misbehavior entailed were not project is a sub-component of an emergency given. Many of the practitioners interviewed multi-sector rehabilitation and reconstruction claimed that there was very little monitoring or program (EMRRP) for the Democratic Repub- supervision. Rather, interactions between pur- lic of Congo (DRC). Approximately 90% of the chaser and provider were governed more by funding came in the form of an IDA grant. Of interpersonal relationships, and a sense of mu- the total, US$42 million covered 67 health zones tual dependence increased the tendency of the and 8 million people. The remaining US$7 mil- two to cooperate. lion was used to provide institutional support Provincial governments depended on the prac- and strengthening at the central and provincial titioners for service, as there were not sufficient level. The project was launched in September providers to promote competition, and provid- 2003. ers were dependent on the contract for their Healthcare in the DRC is allotted according income. Further, the providers' impetus to pro- to a health zones system, of which 50% were vide quality care was driven by a necessity to co-managed by NGOs or other agencies. The maintain their reputation, particularly among role of contracting was to use these health zones, closely knit communities, and, for many, per- scale-up existing projects to deliver a "minimum sonal values and an obligation to the package of care," and help the government build communities they served. The quality of care the institutions necessary to deliver healthcare often varied on the basis of the extent of this in the long term. commitment. In fact, many practitioners were The project is being implemented in two operating without an effective contract. phases. Phase I entails reaching the target health In this case, a kind of relational contract, zones, and Phase 2 involves developing a long- rather than classical contract developed, as the term strategy and supporting central and provincial governments' capacity and resources provincial capacity building. to monitor were limited by the difficulty of Annexes 85 For Phase 1, nine NGOs were selected, based supervision with the high cost; and ensure that on their presence and success in providing health financial management compliance remains a care in the target health zones, mainly along priority. National Road One (HZ). An interest in expand- ing assistance to the surrounding HZ and the Source: Eva Jarawan, Lead Health Specialist, ability to provide a broad spectrum of develop- AFTH3, World Bank. ment assistance were also important. Other criteria included a maximum 12% administra- tive fee; good working relationship with local Contracting for Nutrition Services: communities and authorities; a successful track The Secaline Project in Madagascar record in resource management; and a willing- and the Community Nutrition Project ness to co-manage projects with regional/district in Senegal medical offices. The selected NGOs were given an initial The Secaline project in Madagascar and the US$80,000 contract to: Community Nutrition Project (CNP) in Senegal are two examples of contracting out · visit each of the health zones programs in which a delegated contract man- · discuss the project with local and regional agement approach was used to carry out authorities large-scale nutrition projects targeted at chil- · visit the direct medical offices to discuss dren and pregnant and lactating women. project implementation Secaline was funded primarily by UNICEF · complete a rapid inventory for each health (91%), with user contributions, the Government zone of Madagascar, the World Bank, the World Food · assess the need for rehabilitation Program, and the Government of Japan con- · develop the overall project proposal and bud- tributing the rest. The CNP was funded by the get World Bank, the World Food Program, KfW, · write a one-year action plan and budget. the government, and community contributions. The two projects targeted women and children Two agencies had oversight over the NGOs. living in poor and semi-urban areas, which were The MOH held a supervisory role and moni- not adequately served by the government or tored technical aspects of the project. The MOH, other providers. in turn, signed a memorandum of understand- The Secaline project covered 534 villages in ing with the Bureau Central de Coordination two regions, serving more than 300,000 (BCECO), who served as the contracting agency people, and the CNP covered 14 cities, serving with the NGOs and handled compliance with more than 200,000 people. Both projects pro- procurement and fiduciary issues. vided the following services at the community The project is in its early stages; however, level: several issues have already emerged. Procure- ment became problematic due to differences of · monthly growth monitoring of children in the interpretation in procurement guidelines. This program was resolved with the help of a Bank procure- · weekly nutrition and health education ses- ment specialist. The concept of an Annual sions for women Procurement Plan was also introduced in 2004. · referral to health services for unvaccinated The other challenges include the need to moti- children and pregnant women, for severely vate poorly paid civil servants to effectively malnourished children, or for sick beneficia- monitor contracts; balance the need for close ries 86 TRENDSAND OPPORTUNITIESINPUBLIC-PRIVATE PARTNERSHIPSTO MPROVE I HEALTH SERVICE DELIVERYINAFRICA · home visits to follow up on beneficiaries who Contracts given to NGOs and GIEs specified were referred or who did not use the avail- the work to be done as well as the performance able services expected. A minimum number of services to be · food supplementation to malnourished chil- delivered was specified, as was attendance in dren weekly educational sessions. Management infor- · improved access to water stand pipes mation systems in both projects served to (Senegal) or referral to a social fund for in- regulate performance. Close supervision by come generating activities (Madagascar). Agetip and project units lessened the risk of fal- sified data. In Senegal, competition was In Madagascar, services were provided by a encouraged among local youth groups as well Community Nutrition Worker (CNW), gener- as NGOs to win the bid, and then to perform ally a woman from the village, and in Senegal, adequately to have their contract renewed. In by a Groupement d'Interet Economique (GIE), Madagascar, because the project was located in a legal entity consisting of a group of four young rural areas with few competitors, quality was people, typically unemployed, living in the tar- ensured by allowing a center to remain open only get neighborhood. These workers are trained by if it met strict standards. project staff (Madagascar) or by local consult- The Secaline project and CNP were able to ants or training institutions (Senegal); they were achieve strong results by using the private sec- supervised by NGOs (Madagascar) or NGOs tor and community members--who were not or other GIEs (Senegal). All relationships were healthcare professionals--to provide nutrition contractual. services. Both programs lowered malnutrition In both projects, the functions contracted out rates substantially (from 20%­30% to 5%­ included overall management, service delivery, 10% in Madagascar; and from 60%­70% to supervision, and research and training (con- 20%­30% in Senegal). In Senegal, severe mal- tracted to local consultants or institutions). In nutrition disappeared in children aged 6­11 Madagascar, a project management unit linked months, and moderate malnutrition for to the office of the prime minister and two re- ages 6­35 months dropped from 28% to 24%. gional management units were established to The heavy donor contribution to these pro- manage contracts for the government. NGOs grams raises the issue of sustainability, although were contracted (via an open tender process) to both programs have addressed this through ca- manage the CNWs (chosen on the basis of strict pacity-building at the community and criteria; verbal contract established). institutional level. Technical coordination and project manage- ment accounted for 13% of the project's cost. Source: Marek, Tonia, Issakha Diallo, Biram In the CNP, Agetip, an NGO, signed a conven- Ndiaye, and Jean Rakotosalama, "Successful tion with the government to execute the project. contracting of prevention services: fighting mal- It contracted local NGOs and GIEs (via open nutrition in Senegal and Madagascar," Health tender), who in turn oversaw (other) GIEs (also Policy and Planning, 14(4): 382­89, 1999. contracted through open tender) delivering ser- Contact: Tonia Marek, Lead Public Health vices. Agetip charged the government 17% of Specialist, AFTH2, The World Bank, the project's cost for operation, monitoring, and tmarek@worldbank.org. evaluation. References I Guinea Country Status Report on Health Ghana: fiscal sustainability and strategic and Poverty, The World Bank/Ministry of purchasing of priority health services," Health of Guinea, June 2003. World Bank, Washington, D.C., 2004. II Islamic Republic of Mauritania, Country IX Adama Ouédraogo, "Etude sur l'apport des Status Report, The World Bank/Ministry structures sanitaires privées à la vaccina- of Health of Mauritania, 2004 tion de routine," MOH/European Union, III Marek Tonia, Eichler Rena, Schnabl Islamic Republic of Mauritania, 2003. Philip, "Resource Allocation and Purchas- X Natasha Palmer, Anne Mills, Haroon ing in Africa: what is effective in improving Wadee, Lucy Gilson, Helen Schneider, "A the health of the poor?", The World Bank, new face for private providers in develop- Africa Region Human Development ing countries: what implications for public Working Paper Series--No. 74, August health?", Bulletin of the WHO, 81 (4), 2004. 2003. IV Nezhir Ould Hahmed, Mohamed and XI Adapted from Sandra Mac Donagh, Su- Soucat, Agnès, "Santé et Pauvreté en san Murray, Tim Ensor, "Examining the Mauritanie: Analyse et cadre stratégique role of private maternity services in Nepal, de lutte contre la pauvreté," Région India and Tanzania," Draft literature re- Afrique, Banque Mondiale, June 2004. view, King's College, London, 2003. V World Bank, "Guinea Country Status Re- XII MOH/IMCI Unit, Republic of Uganda, port," Draft, Washington, D.C., 2003. "Utilizing the potential of formal and in- VI World Bank, "Santé et Pauvreté au formal private practitioners in Child Burkina Faso," Washington, D.C., June survival in Uganda," August 2001, citing 2003. Omaswa, Francis, "Quality of care in pri- VII Presentation made by the Kenyan associa- vate medical practice," 1999. tion of private hospitals, nursing homes XIII United Republic of Tanzania, "Technical and clinic owners, KAPH, at the Nairobi Review 2005: Public private partnership Workshop on PPP, June 28­29, 2004. for equitable provision of quality of health VIII De Roodenbeke, Eric, "Purchasing hospi- services," Final Report, Independent Tech- tal services in mutual health insurance in nical Review on behalf of the Ministry of 87 Health, the President's Office, Regional XXV Ministry of Health, "Analyse des Administration and Local Government, indicateurs de santé et population dans le and the Government of Tanzania, March cadre de la stratégie nationale de lutte 2005. hera@herabelgium.com contre la pauvreté au Sénégal," Draft, XIV World Bank, "Improving Health, Nutri- Sénégal, 2002. tion, and Population Outcomes in XXVI Communication by Dr. Mame Cor Ndour, Sub-Saharan Africa: The Role of the World from the Ministry of Health of Senegal, at Bank," Washington, D.C., December 2004. the PPP Workshop in Nairobi, Kenya, June XV Palmer et al, WHO, 2003. 28-29, 2004. XVI Abdelkrim Smine, Khady Diouf, Nancy XXVII Kenya's Director of Medical Services, Blum, "Antimalarial drug quality in Opening Remarks to the PPP Workshop Senegal," USP drug Quality and Informa- in Nairobi, Kenya, World Bank, June tion Program, U.S. Pharmacopeia, 2002. 28­29, 2004. uspdqi@usp.org , XXVIIITHETA, "Innovation or re-awakening? XVII Charles Maponga, Clive Ondari, "The Roles of traditional healers in the manage- quality of antimalarials: a study in selected ment and prevention of HIV/AIDS in African countries," WHO.EDM/PAR/ Uganda," Participatory Evaluation Re- 2003.4, WHO, Geneva, May 2003. port, October 1998. XVIIIOmaswa, Francis, "Quality of care for un- XXIX April Harding , Alexander Preker, eds., der-five children in private medical practice "Private participation in health services in urban and peri-urban areas in Uganda," handbook," HNP Series, World Bank, Ministry of Health, Kampala, July 1999. Washington, D.C., 2002, citing Hanson & XIX Vicki Marsh, Wilfred Mutemi, "Private Berman, 1998. sector drug retailers and malaria control XXX MOH and GTZ, "Rapport final de l'étude in Kenya," Insights Health, March 2003. sur les pratiques parallèles dans les www.id21.org hopitaux nationaux en Guinée," Conakry, XX Palmer Natasha et al, Bulletin of the Guinea, 2003. WHO, 2003. XXXI Stephen Jan, "Public sector doctors with XXI "Compte rendu de l'atelier de restitution second jobs," Insights Health, March des résultats de l'étude sur la sur- 2003. www.id21.org tarification et les pratiques parallèles dans XXXII Stephen Jan, Ying Bian, Manuel Jumpa, les hôpitaux nationaux de Conakry," Min- Qingyue Meng, Norman Nyazema, Phusit istry of Health, Guinea, June 2003. Prakongsai, Anne Mills, "Regulating dual XXII Kuawab Business consultants, "Study on job holding amongst public sector health malaria financing and expenditures in workers in developing countries," Draft, Ethiopia," Addis Ababa, Ethiopia, August London School of Hygiene & Tropical 2002. Medicine, 2004. XXIIIAlfred Obuobi, M. Pappoe, S. Ofusu- XXXIIIAlex Preker, "Voluntary health insurance Amaah, Prince D.Y. Boni, "Private health in development: review of role in the Af- care provision in the greater Accra region rica Region and other selected developing of Ghana," PHR/Abt, Small Applied Re- country experiences," Concept Note, search Paper No. 8, July 1999. World Bank, October 2004. XXIV MOH Senegal/ADIRA Etudes et Conseils, XXXIV Gorter A, P. Sandiford, Z. Segura, C. "Mise en place d'une politique pérenne de Villabella, "Improved health care for sex motivation des personnels de la santé," workers: a voucher programme for female July 2003. sex workers in Nicaragua", in Research XLV Benjamin Loevinsohn, April Harding, for Sex Work, Nbr. 2, August 1999. "Contracting for the delivery of commu- XXXV Gorter A, P. Sandiford, Z. Segura, C. nity health services: a review of global Villabella, "Improved health care for sex experience," World Bank, September workers: a voucher programme for female 2004. sex workers in Nicaragua", in Research XLVI Marek, T., Yamamoto, C., Ruster, J., "Pri- for Sex Work, Nbr. 2, August 1999. vate Health," Viewpoint Note No. 264, XXXVI "A Guide to Competitive Vouchers in The World Bank, June 2003. Health," World Bank, 2005. XLVII Communication by Source: Dr. Housseini XXXVII Ngalande-Banda and Walt, 1995. Bah, Pharmacie Centrale de Guinée. XXXVIII Bennet, Ngalande-Banda, 1994. Pharma_centrale@yahoo.fr XXXIX Lilani Kumaranayake, Sally Lake, Phare XLVIII Joseph Naimoli, "Partnerships within the Mujinja, Charles Hongoro, Rose public sector to achieve health objectives: Mpembeni, "How do countries regulate Performance-based management in an the health sector? Evidence from Tanza- evolving decentralized public health sys- nia and Zimbabwe," Health Policy and tem in West Africa: The case of Burkina Planning, vol. 15(4), pp. 357­67, 2000. Faso," World Bank, Washington, D.C., XL Simon Delay, Lucy Gilson, David March 2003. Hemson, Keith M. Lewin, Mphela XLIX Oliver Mudyarabikwa, Denford Motimele, Rebecca Scott, Haroon Wadee, Madhina, "An assessment of incentive "South Africa: study of non-state provid- setting for participation of private for- ers of basic services," Un. of Birmingham, profit health care providers in London School of Hygiene & Tropical Zimbabwe," PHR, November 2000. Medicine, Loughborough Un.; Un. of Sus- L Kigondu John, Naisho Joyce, "Afriafya sex, Un. of Witwatersrand, Draft, 2004. programme evaluation report," Kenya, XLI Hilary Standing, Gerald Bloom, "Where May 2003. there is no regulator," Insights Health, LI Kuawab Business Consultants, "Study on March 2003. WWW.id21.org malaria financing and expenditures in XLII Adapted from J. Perrot and E. Ethiopia," Addis Ababa, Ethiopia, Au- de Roodenbeke, editors, "La contractua- gust 2002. lisation dans les systèmes de santé: pour LII Adapted from Jean Perrot and une utilisation efficace," to be published. Eric de Roodenbeke, editors, "La XLIII Mills, A. and J. Broomberg. 1998. Expe- contractualisation dans les systèmes de riences of contracting: An overview of the santé: pour une utilisation efficace," to literature. Macroeconomics, Health and be published. Development Series Nbr. 33. World LIII Henrik Axelsson, Flavia Bustreo, April Health Organization. Harding, "Private sector participation in XLIV Broomberg, Masobe, and Mills, 1997. child health," World Bank, May 2003.