Tanzania - Health and Nutrition Project

Document of THE WORLD BANK For Official Use Only Report No.: 19964 IMPLEMENTATION COMPLETION REPORT TANZANIA HEALTH AND NUTRITION PROJECT [CREDIT NO. 20980] DECEMBER 28, 1999 Africa Human Development Group 1 Eastern and Southern Africa Africa Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit = Tanzania Shilling Appraisal: US$1 = Tsh 193; Closing: US$1 = Tsh 680 Weights and Measures Metric System Fiscal Year of Borrower Central Government: July 1 - June 30; Local Government: January 1 - December 31 ABBREVIATIONS AND ACRONYMS AIDS - Acquired Immune Deficiency Syndrome APL - Adaptable Program Lending CHF - Community Health Fund(s) CMS - Central Medical Stores CSP - Cost Sharing Program DANIDA - Danish International Development Agency DCP- Drug Capitalization Program DHB - District Health Board(s) DHMT - District Health Management Team(s) DMDT - Department of Manpower Development and Training DMO - District Medical Officer GOT - Govermment of Tanzania HIV - Human Immuno-deficiency Virus HMIS - Health Management Information System H&N - Health and Nutrition HRD - Human Resource Development HSR - Health Sector Reform ICR- Implementation Completion Report IDA - International Development Association MOH - Ministry of Health MRALG - Ministry of Regional Administration and Local Government MPH - Master's in Public Health MSD - Medical Stores Department NHI - National Health Insurance PHC - Primary Health Care PIU - Project Implementation Unit PHN - Population, Health and Nutrition PS - Permanent Secretary PSN - Project Support Network QAG - Quality Assurance Group SAR - Staff Appraisal Report SDC - Swiss Development Corporation SDR - Special Drawing Rights SWAp - Sector Wide Approach TFNC - Tanzania Food and Nutrition Council USAID - United States Agency for International Development Managers and Staff Responsible Vice President: Callisto Madavo (AFRVP) Country Director: James W. Adams (AFMTZ) Sector Manager: Dzingai Mutumbuka (AFTH1) Task Team Leader: Oscar F. Picazo (AFTH1)  FOR OFFICIAL USE ONLY TABLE OF CONTENTS Preface Evaluation Summary ................................. i Introduction. i Project Objectives. i Implementation Experience and Results .................................1i Future Operations and Key Lessons Learned ................................. iii Part I: Project Implementation Assessment .................................1 j Introduction .I Project Objectives ..................................1 Achievement of Project Objectives .................................2 Implementation Experience .................................7 Major Factors Affecting the Project .................................9 Sustainability of Project Activities ................................. 10 Borrower Performance ................................. 12 Bank Performance ................................. 12 Assessment of Project Outcome ................................. 13 Future Operations ................................. 14 Key Lessons Learned ................................. 14 Part II: Statistical Annexes ................................. 16 Table 1 Summary of Assessments ................................. 16 Table 2 Related Bank Credits ................................. 17 Table 3 Project Timetable ................................. 17 Table 4A Credit Disbursements, Cumulative and Actual in U.S. Dollars ................ 18 Table 4B Credit Disburements, Quarterly and Cumulative in SDRs ........... ............ 19 Table 5 Key Indicators of Project Implementation ................................................... 21 Table 6 Key Indicators of Project Operation .......................................................... 23 Table 7A Studies Conducted Under the Project ....................................................... 24 Table 7B Long Term Training Sponsored Under the Project ................................... 25 Table 7C Short Tenn Training Sponsored Under the Project ................................... 26 Table 7D Workshops, Seminars and Conferences Supported by the Project ........... 27 Table 7E Study Tours Conducted Under the Project ................................................ 27 Table 7F Rural PHC Civil Works Completed Under the Project .............. ............... 28 Table 7G Rehabilitation and Equipping of District Hospitals Under the Project ..... 28 Table 8 Project Financing ......................................................... 29 Table 9A Project Costs by Expenditure Categories, Estimated and Actual ......... .... 30 Table 9B Project Costs by Project Components, Estimated And Actual .......... ....... 31 Table 10 Status of Legal Covenants ............................ .............................. 32 Table 11 Compliance with Operational Manual Statements .................. .................. 39 Table 12A Actual Bank Staff Inputs .......................................................... 40 Table 12B Bank Missions .......................................................... 41 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  Appendices Appendix A: Completion Mission's Aide Memoire Appendix B: Proceedings of the MOH and MRALG Internal Assessment Workshop for the H&N Project Appendix C: Borrower's Comments on the Draft ICR Appendix D: Map of Tanzania  IMPLEMENTATION COMPLETION REPORT TANZANIA HEALTH AND NUTRITION PROJECT (TZ-PE-2774) PREFACE 1. This is the Implementation Completion Report (ICR) for the Tanzania Health and Nutrition (H&N) Project, for which Credit No. 20980 (Sector Investment Loan) in the amount of SDR 36.1 million (US$47.6 million equivalent at the prevailing exchange rate) was approved on March 6, 1990, signed on March 7, 1990, and made effective on April 6, 1990. The credit closed on June 30, 1999; the original closing date of June 30, 1996 was extended for two years in 1996 and for another year in 1998. The last disbursement was made on November 3, 1999. An undisbursed balance of SDR 2.4 million (US$ 3.3 million at the prevailing exchange rate) will be cancelled once funds due from the Government are received and the accounts are closed. Any undisbursed balance will be cancelled. 2. The ICR was prepared by Mr. Oscar F. Picazo, Economist of the Human Development Group 1, Eastern and Southern Africa (AFTHI), with assistance from Mr. David Robalino, Young Professional, AFTH1. It was reviewed by Mr. Dzingai Mutumbuka, Sector Manager, AFTHI, and Mr. James W. Adams, Country Director for Tanzania. Additional comments were received from Dr. Philip Gowers, Principal Health Specialist, Ms. Chiyo Kanda, Task Team Leader of the project, and Mr. Andrew Follmer, Operations Analyst, all of AFTH1. The draft tables in the Annex were prepared by Dr. Faustin Njau, the Project Coordinator for Component I; Mr. Barney Laseko, Financial Management Specialist, AFMTZ; Ms. Emily Mwai, Information Technology Assistant, AFTH1; Mr. Robalino, and Mr. Picazo. The borrower reviewed the report before it was finalized. 3. The report was initiated during an ICR mission from September 17 to 21, 1999, comprising of Dr. Emmanuel Malangalila, Social Sector Advisor, AFMTZ; Mr. Picazo; and Mr. Laseko. The ICR is based on information collected during that mission, discussions with borrower representatives, and data from project files. The borrower assisted in the preparation of the ICR by contributing views as reflected in the Completion Mission's Aide Memoire (Appendix A) and by holding an internal assessment workshop jointly sponsored by the implementing agencies, the Ministry of Health (MOH) and the Ministry of Regional Administration and Local Government (MRALG) (Appendix B). This ICR reflects many of the inputs, comments, and observations made during this workshop. The Government also provided comments as it evolved and formally responded to the draft ICR (Appendix C).   IMPLEMENTATION COMPLETION REPORT TANZANIA HEALTH AND NUTRITION PROJECT (CR. TZ-PE-2774) EVALUATION SUMMARY Introduction 1. The Government of Tanzania (GOT) undertook the Health and Nutrition (H&N) Project to raise the quality, coverage and effectiveness of family planning, nutrition and basic health services in the country. The Project is the first IDA credit in the sector, amounting to SDR 36.1 million (or US$ 47.6 million). It was made effective on April 6, 1990, and closed on June 30, 1999, after two extensions of the closing date. The Project was designed to have three components: Component I involved the strengthening of national population, health and nutrition (PHN) systems; Component II provided health interventions in ten rural districts; and Component III was planned to cover similar interventions in urban Dar es Salaam. Eventually, Component III was spun off as parallel-financed set of activities funded by the Swiss Agency for Development and Cooperation (SDC). Project Objectives 2. The objectives of the H&N Project were to raise the quality, coverage and effectiveness of family planning, nutrition and basic health services through provision of support to critical and strategic elements of the PHN sector with particular emphasis on strengthening the institutional capacity for health planning and policy formulation; manpower development and training in the health sector; sustainable provision and financing of pharmnaceuticals and medical supplies; control of micronutrient deficiency; implementation of the national population policy; rural primary health care (PHC) through a trial implementation of the revised PHC strategy; and urban PHC through reform and rehabilitation of the urban health system in Dar es Salaam. Implementation Experience and Results 3. The project suffered from poor quality at entry: it was prepared with very ambitious objectives, complex implementation and coordination arrangements, poorly- phased covenants, and ill-considered cross-conditionalities between different implementing agencies. Weak GOT ownership at the start and poor project management (in part due to lack of previous IDA experience) resulted in poor project performance in the first three years. A midterm review led to project restructuring in 1996, resulting in clearer project direction, more feasible workplans, and more streamlined implementation. MOH staff training early on also contributed to the improved implementation. With the restructuring, planned activities were completed and project objectives were largely achieved. Project achievements in policy reform (health financing and pharmaceuticals) are highly satisfactory and these have begun to yield the expected resource-mobilization  Tanzania Health & Nutrition Project ICR Page ii of iv and efficiency impacts. The full range of benefits from these policy changes (in terms of sustainability and continuously improving services) will be realized in the long term. Other areas of project involvement were rated satisfactory including the procurement of drugs, training and manpower development, civil works/rehabilitation, rural PHC services, and health management information system. The implementation of the national population policy, however, was less satisfactory. 4. The closing date was extended twice in recognition of the project's development potential and the much-improved pace of implementation. The project was first extended in June 1996 to complete the activity backlog during the first three years; the second extension in April 1998 was made primarily to complete the pretest of alternative financing schemes and to support the GOT's Health Sector Reform initiatives. Both extensions proved to be warranted. By the original closing date, only 42% of the credit (reckoned in SDRs) had been disbursed; the three-year extension resulted in a cumulative disbursement of 55% so that by project's closing, a total of 97% of project funds had been disbursed. Despite initial difficulties with project management and covenant compliance, all civil works under this project were completed, all the planned training were undertaken, all the drugs and supplies were procured, and all the health financing pretests were carried out. 5. In addition to the poor quality at entry, the project suffered from factors outside the Government's control, including the transfer of the Tanzania portfolio from one World Bank division to another, frequent changes in IDA task manager, and the overall shortage of technical and managerial skills in Tanzania. However, there are factors within GOT's control that could have improved project performance including adequate allocation of counterpart funds, political acceptance of the need to restructure the pharmaceutical subsector, and to institute user fees to generate resources and improve efficiency, swift action to improve project management including contracting of management functions, more adequate planning especially in civil works, and a stronger procurement capacity. 6. Borrower performance showed a slow but consistent improvement in project ownership, management, and capacity building. According to the internal GOT project assessment workshop, project preparation had much to be desired, but project implementation was on the whole satisfactory largely due to the dogged commitment of the implementing agencies, the Ministry of Health (MOH) and the Ministry of Regional Administration and Government (MRALG), to the objectives of the project. Government leadership in the sector increased dramatically as the project matured, donor coordination improved, and MOH's commitment to pursue certain reforms, despite initial Parliamentary difficulties, eventually paid off with key legislative actions in health financing and MSD autonomy. Donor coordination and Government leadership in the sector also showed dramatic improvement during project implementation, paving the way for a multidonor-supported Health Sector Reform Program which the follow-on IDA credit will support. 7. Despite extensive background studies during project identification, Bank performance during project preparation and appraisal were deficient: the Bank's task team could be faulted for designing an overly complex project including cross-ministerial  Tanzania Health & Nutrition Project ICR Page iii of iv project responsibility that made coordination difficult; being oblivious to critical project risks (e.g., weak government ownership at the start, cross-conditionalities and covenants that are outside the purview of implementing agencies) and inadequate supervision during the first three years of implementation. Project supervision dramatically improved after the midterm review followed by formal restructuring in 1996, and the entry of a new task team. Thus, overall project supervision was satisfactory, according to the internal GOT assessment workshop as well as the Bank's Quality Assurance Group assessment. 8. Without doubt, the sustainability of the Tanzanian health sector and project- initiated activities more specifically, have been considerably enhanced relative to the situation in the late l 980s when the project was designed. Budget allocations to the health sector have doubled, and there is a significant and increasing source of extra-budgetary revenues from user fees, community health funds, and hopefully a health insurance program for civil servants. Drug and medical supplies are better financed and managed now than in the past. The substantial number of project-trained staff continue to serve in civil service, and there is stronger Government commitment to support human resource development even as it seeks greater cost-sharing for staff undergoing training. Communities have been shown to be a critical element in sustaining health investrnents and for providing additional resources for the health system. Yet, even with these positive developments, Tanzania remains a poor country, and economic growth is expected to continue at a modest pace. Thus, the Government budget for health and household contributions for health services will remain inadequate for the foreseeable future, and donor support will continue to be needed. Future Operations and Key Lessons Learned 9. Future IDA operation should preserve and expand the gains achieved under the H&N Project and should continue to assist the GOT in the further development of the health sector, with households, communities, local governments and the private sector having a more active role in the financing, delivery, and/or management of health services. On the basis of this principle, the successor IDA credit, the Health Sector Development Program, is being designed as an Adaptable Program Loan (APL) to support, in coordination with other donors, the Government's HSR Program. The first phase of the 12-year APL focuses on: further strengthening of health service delivery through phased decentralization of health services; further human resource development through the rationalization of training institutions; strengthening of central support systems; and expansion of alternative health financing initiatives. 10. The H&N Project provides the following useful lessons: * Project preparation should take account of project risks as comprehensively as possible and should reduce the complexity and extent of the project on the basis of the magnitude and probability of these risks. * The relationship between IDA and the host government during project negotiation should be weighed carefully. Cash-strapped poor countries often find themselves  Tanzania Health & Nutrition Project ICR Page iv of iv with little countervailing power and IDA should avoid the temptation to exploit this vulnerability by imposing too many conditionalities detrimental to implementation. * The involvement of multiple agencies/ministries make project implementation more difficult. Thus, the benefits of an inter-ministry project must be weighed carefully against the cost of coordination. * Critical policy reforms (e.g., health financing and pharmaceutical sector reforms) should be pursued as project outcomes rather than upfront conditionalities. If certain conditionalities are required, their timing for compliance should be carefully planned depending on the nature and objective of the conditionality, as well as it being reflective of government commitment. Compliance to these covenants and conditionalities should be within the purview of the implementing agencies. * Project components should be designed with an internal logic behind them and not merely used as baskets to put activities in need of funding. The synergy and mutually reinforcing nature of the project components should be taken into account. - Economic liberalization and related macroeconomic reforms provide strong underpinnings for sector reforms and significantly enhance their achievement. Analytical and policy work between the two should be done hand in hand and in mutually supportive manner. 3 Baseline data and performance targets should be defined at project appraisal. These enhance monitoring of project progress; their absence inhibits supervision and makes impact evaluation virtually impossible. Data gathering pertaining to these indicators should be made an intrinsic part of the project. * For skill-scarce countries or sectors, training should be conducted as soon as possible so that the project can benefit from staff's acquired management and technical expertise. Procurement should be made a central concern of project management. * Counterpart funds should be calculated globally for the whole project, rather than on a contract-by-contract basis, and should take into account extra-budgetary resources generated from project-supported initiatives such as user fees, prepayments and membership contributions or premia from risk-pooling or other community health financing schemes, and community inputs into village health initiatives.  PART I: PROJECT IMPLEMENTATION ASSESSMENT Introduction 1. The Government of Tanzania (GOT) implemented the Health and Nutrition (H&N) Project over a nine-year period from March 6, 1990 to June 30, 1999. It was the first IDA project in the health sector and was prepared during a period of fiscal difficulty. The project saw the transition in Tanzania's economy and the social sectors from its socialist orientation to one relying on a more pluralistic and competitive environment with private medical practice being allowed and with households and communities being accorded a critical role in the demand for health services. The project was extended twice (in June 1996 for two years and again in April 1998 for another year); in the last year, the project was closely linked with the preparation of a follow-on IDA credit, the Health Sector Development Program, within the context of the GOT's Health Sector Reform (HSR) Program to be funded by a consortium of donors and the GOT itself. Project Objectives 2. The objectives of the H&N Project were to raise the quality, coverage and effectiveness of family planning, nutrition and basic health services through provision of support to critical and strategic elements of the population, health and nutrition (PHN) sector with particular emphasis on strengthening the institutional capacity for health planning and policy formulation; manpower development and training in the health sector; sustainable provision and financing of pharmaceuticals and medical supplies; programs to control micronutrient deficiency; implementation of the national population policy; rural primary health care (PHC); and urban PHC through reform and rehabilitation of the urban health system in Dar es Salaam. 3. To achieve these objectives, three project components were implemented which, in hindsight, proved to be mere pegs to hang unrelated activities for funding, making the project unwieldy. Component I, Strengthening National PHN Systems (US$38.5 million, according to the SAR), was based in the Ministry of Health (MOH) and coordinated by the Project Implementation Unit (PIU). It covered health planning, policy formulation and information system; manpower development and training; pharmaceutical and medical supplies financing and procurement; micronutrient deficiency control programs; and national population policy. 4. Component II, Strengthening Rural PHC (US$11.6 million in the SAR), was based in the Ministry of Regional Administration and Local Government (MRALG), formerly in the Prime Minister's Office, and coordinated by the Project Support Network (PSN). This component was for the most part amorphous and poorly delineated. It involved the implementation of the then-recently revised PHC Strategy, including community mobilization for PHC, support to village PHC initiatives, rehabilitation and strengthening of health facilities, improved water supplies, strengthening of district health management, improved supply and maintenance systems, rehabilitation of district hospitals, and monitoring and evaluation for feedback to PHC policy formulation. The  Tanzania Health & Nutrition Project ICR Page 2 component focused on 10 poor and hard-to-reach districts where there was no other donor present. It was also designed to benefit from policy and prograin innovations in Component I, but the different pace of implementation in these two components as well as the different loci of responsibility made such synergy difficult to achieve. 5. Component III, Strengthening Urban PHC (US$11.5 million, in SAR), was planned as parallel-financed set of activities based on the bilateral agreement between the GOT and the SDC. It was implemented by the City Commission as the Dar es Salaam Urban Health Project (DUHP Phase I), supported by expatriate technical advisers. The project covered three urban districts and involved the improvement of management and supply systems, rehabilitation of health facilities, establishment of a maintenance system, and operations research. Although Component III was described in the SAR as part of the H&N Project, it operated as a separate project and was completely spun off by DUHP Phase II. 6. Evaluation of Project Objectives: The original project objectives were too ambitious. Some of these objectives were translated too strictly as inflexible project conditionalities and covenants which project managers saw as benchmarks to be met rather than directions to be pursued. A basic design flaw was made in making these policy changes (on user fees and on pharmaceutical reforms) conditions of project disbursement, resulting in stalled activities during the first three years. The absence of outcome measures and baseline data also made it difficult for project managers, the implementing Ministries, and other stakeholders to assess progress. As the 1994 midterm review indicates, the project "was perennially in noncompliance with these covenants, always facing the danger of cancellation under the Bank's effort to restructure its portfolio," and by implication, always facing the risk of not meeting its development objectives. Finally, the project design severely underestimated the political constraints of introducing reforms, the local management and technical capacity to put them in place, and the time it took for policy changes to be implemented once the political constraints have been eased. The midterm review in 1994 and the project restructuring that followed clarified the direction for the project, developed key performance indicators and feasible targets, distinguished project activities that are priority and less priority, provided directives to strengthen coordination between Components I and 11, and emphasized health financing initiatives. The restructuring proved to be the project's turning point, paving the way for much better implementation performance. Achievement of Project Objectives 7. Although little was accomplished during the initial years, the project restructuring expedited many activities from 1996 onwards. As a result, project development objectives were largely achieved by project's end. Project achievements in policy reform are highly satisfactory, impact has been considerable, and the full benefits of these will be realized in the longer term. Other areas were rated satisfactory: supply of pharmaceuticals and reforms in this subsector, training and manpower development, civil works, rural PHC services, and HMIS. However, project support for the implementation of the national population policy was less satisfactory.  Tanzania Health & Nutrition Project ICR Page 3 Component I: Strengthening National PHN Systems 8. Health Planning: Project training inputs were substantial and are bearing fruits in terms of stronger planning at the central MOH and better management by district medical officers (DMOs). In all, 30 persons received Masters degrees in public health, health planning and policy, pharmaceuticals management, construction management, information technology, human resources, and nutrition. No fewer than 66 persons also received short-term training on project management, food fortification, health financing, equipment maintenance, and medical stores management. Most of the project-trained government personnel are still in civil service. Most observers note the much improved planning capacity and more satisfactory policy and technical discussions at MOH and the focus districts compared to the situation at project inception. The project also supported the development of district health planning guidelines which have become standard requirements for government financing of district health services. 9. Policy Formulation and Health Financing: Through conditionalities, the project influenced greater budget allocation to the health sector, with the proportion of the budget devoted to health dramatically increasing from 6-7% in the late 1 980s to 12% in 1998/99. More importantly, the project enabled the shift in the Tanzanian social-sector paradigm from a free-care govemment monopoly enshrined in the 1967 Arusha Declaration to a system in which households contribute to the financing of care and the private sector has an important role in service provision. Such change was underpinned by project- supported policy and legislative reforms including Cabinet Paper No. 25/6 (June 25, 1993) establishing the Cost Sharing Program (CSP) in government health facilities, the Parliamentary Act establishing the Medical Stores Department as a semi-autonomous body in 1993 and concomitant reforms in drug supplies and financing, and the draft Cabinet paper on community health funds (CHFs) to make it a national policy, drawing on the pretesting experience in 10 project-supported districts. Another policy achievement was the enactment of the National Health Insurance in April 1999, which will initially cover 50,000 central government employees and their dependents. 10. Health Management Information System (HMIS): The project aimed to support the development of an HMIS to register data from maternal and child health and immunization programs and to eventually cover all programs in all districts. Up to February 1997, progress had been less than expected in this area due to frequent changes in the HMIS workplan. The project itself provided minor financial support; most of the HMIS activities were funded by other donors. During 1997-99, HMIS activities picked up, and modules for PHC were developed and introduced nationwide. All 114 District and 20 Regional Health Management Teams (D/RHMTs) have been trained on the HMIS, 20 regions have computerized their HMIS, and 2 districts are being pilot-tested for computerization. These were made possible by DANIDA (the major donor), the H&N Project, GTZ, and UNICEF. DANIDA plans to conduct an evaluation in December 1999 of the impact of these HMIS initiatives on management capacity; it is also supporting the development of the hospital HMIS module. 11. Manpower Development: The project initiated the reform of the Department for Manpower Development and Training through the creation of an Inspectorate Unit responsible for monitoring staff performance as well as the establishment of technical,  Tanzania Health & Nutrition Project ICR Page 4 administrative and logistics units within each MOH division. The project provided consultants to design the logistics and administrative systems; sponsored staff training programs at central, zonal, and district levels; supported the Continuing Education Division for the upgrading and equipping of 5 of the 6 zonal training centers including provision of textbooks and other learning materials; and funded the completion of the Human Resource Development Plan. The impact of these interventions on service delivery remains to be quantified. 14. Financing and Procurement of Pharmaceuticals and Medical Supplies: The project provided 31% of the US$ 32.1 million required to address immediate shortfalls in the supply of medical supplies and drugs. Difficulties were experienced under this subcomponent in the early 1990s due to the stringent project conditionalities that required reforming the institutional structure of CMS and that delayed the disbursement of IDA funds. In September 1993, CMS was made a semi-autonomous Medical Stores Department (MSD) by an Act of Parliament; the new entity was jointly capitalized by IDA, DANIDA and GOT to make it financially viable. GOT reformed drug pricing to reflect replacement costs and directed all government health facilities to channel their drug funds through MSD. The project rehabilitated the central MSD office as well as 5 of the 7 MSD zonal stores, and provided training in procurement and financial control. MSD now has a relatively well-trained management staff that can procure drugs and supplies locally and internationally, selling them at replacement costs. Procurement and distribution are supported by an MIS between center and zonal stores and a transport and logistics system integrated with procurement. While the MSD has established a well- capitalized revolving fund, it suffered from the demand constraints by public health facilities due to the inadequate and erratic funding from the Treasury. To address this issue, the project pilot-tested the Drug Capitalization Program (drug revolving fund) for hospitals. Drug availability has improved dramatically especially in districts with CHFs and facilities with drug revolving funds. 13. A GOT-sponsored external review of MSD in March 1998 suggests that improvements still need to be made in the following areas: (a) Procurement - the introduction of a supplier database, computerized tender scheduling package, and review of quality assurance system are needed to eliminate leakages and wastage; (b) Inventory control - some tracer items were still out of stock 13% of the time, necessitating introduction of new inventory control system; (c) Financing - distribution and personnel costs can be reduced further; (d) Information system - there is a need to build in-house capacity and reduce dependence on consultants. Management problems related to the duplication of functions between MSD's Board of Trustees and the Medical Tender Board have been solved by June 1999. 14. Micronutrient Deficiency Control Program: The project supported activities to train rural medical aides in the recognition of clinical requirements for vitamin A and iron; educate the public and promote consumption of foods rich in vitamin A and iron; and conduct research on the effects of vitamin A and iron deficiency on health, as well as studies on the socioeconomic determinants of the consumption of foods rich in vitamin A and iron. Key achievements include the establishment of a micronutrient surveillance system (25 sentinel centers) which generates semestral indicators on nutritional status; the development and distribution of policy guidelines and information, education, and  Tanzania Health & Nutrition Project ICR Page 5 communication (IEC) materials on the control of micronutrient deficiency to public and private health facilities; training of extension staff on micronutrients in all health units in 24 districts; managerial capacity building; holding of national nutrition workshops and radio campaigns; completion of nutrition research studies; and the diffusion of a more efficient technology in palm-oil extraction. A 1994/95 impact evaluation of this subcomponent in 14 districts in 8 regions provided mixed results: a slight increase in the use of vitamin A capsules but low utilization of ferrous sulfate relative to earlier trends. However, it appears that malnutrition in under-5 children has been considerably reduced, although other non-project interventions could also have helped. The micronutrient program has been mainstreamed in the health system and the Tanzania Food and Nutrition Council (TFNC) enjoys govermnent support. 15. National Population Policy (NPP): In 1990, the Government began implementing the NPP through the Tanzanian Council for Population and Development which was created and received project support through the provision of consultants; supplies, equipment and the renovation of office space; staff training in demography; and a research program on population and development. The National Population Information Center was created in the Planning Commission as a research and resource- information base for population issues. Project support came in the form of books, periodicals, supplies, computer equipment, vehicles, staff training, and consultant services. Annual plans were developed and disseminated, and workshops with parliamentarians and religious groups were held. However, many of the planned activities were never executed, and implementation delays occurred due to frequent changes in the management of the Population Planning Unit (PPU). After the initial spurt of activities, little follow-on work has been done. Thus, this subcomponent was repeatedly rated unsatisfactory and consensus is that outcome has been negligible especially during the second phase of the implementation (1996-98). Overall, project involvement in this area has been minor. One positive outcome of project support is that the PPU is now part of the Planning Commission with a dedicated budget. PPU also managed to implement some of the activities of the population policy under the GOT's Family Planning Program using resources outside the H&N Project. Component II: Strengthening Rural PHC 16. Support of Village PHC Initiatives: Activities in this subcomponent included training of district-level staff in community mobilization, IEC campaigns in selected villages in the 10 focus districts to inform them of the types of support they could get from their district for PHC activities, and district health planning. All of the 10 project districts produced district health plans that have received funding for implementation. Overall, supervision reports indicate positive impact of community involvement in rehabilitating and maintaining primary health facilities. Field reports suggest that communities that actually participated in planning and rehabilitation have a high commitment to and sense of ownership of their primary-care facility. Moreover, health infrastructure that relied partly on village funds appear to be better maintained. 17. Village-level PHC civil works include 34 shallow wells, 16 grain storage facilities, and 3 day-care centers. The October 1998 internal evaluation of Component II shows that in general, smaller-scale village level investments appear to have experienced  Tanzania Health & Nutrition Project ICR Page 6 more difficulties relative to larger-scale investments. For instance, most of the shallow wells have dried up. Also, most of the community-involvement projects (chosen by villagers themselves) has focused on rehabilitation of health facilities at the expense of nutrition-related activities, e.g., fish farning, seed oil extraction, vegetable and fruit gardening, or day-care centers. Finally, training was only implemented up to the ward level and did not go deeper to the grassroots. 18. Strengthening the District PHC System: This component involved the rehabilitation and maintenance of 87 dispensaries and 22 rural health centers using community based approaches in 10 districts. Earmarked equipment were mostly in place. The subcomponent also involved strengthening the district health management team (DHMT) through training in community mobilization, management, planning, and financial control (5 DMOs and 1 RMO were trained, all have been promoted); provision of one vehicle for each of the 10 districts; provision of bicycles and other forms of transport to key staff; provision of office furniture and supplies for the DHMT; and strengthening of the facility maintenance workshops through the provision of tools, equipment and supplies. While many of the planned outcomes were realized - all of the project districts now have health plans - supervision reports for Singida, Iramba, Igunga, and Nzega indicate that a number of problems remain including bat infestation of dispensary roofs, poor water provision, underutilized kitchen facilities, and intermittent undersupply of drugs especially for chloroquine-resistant malaria and schistosomiasis. Many health centers have reported increased utilization, but some have figures lower than expected. 19. Strengthening the District Health Referral System: This component involved the rehabilitation and equipping of 10 regional/district hospitals. The districts were selected based on poverty indicators and the absence of any other donor supporting the district. All physical rehabilitation was completed and earmarked equipment and furniture are in place. However, some of the hospital instruments and equipment procured appear to be of substandard quality. Also, problems remain regarding shortfalls of staff in some rehabilitated facilities. Component III: Dar Es Salaam Urban Health Project (DUHP) 20. DUHP was implemented in three phases between 1990 and 2000. Phase 1 (1991 to mid-1 993) focused on the rehabilitation of facilities in poorest state, equipping and provision of medical supplies to these facilities, and staff training. In Phase II (mid-1993 to 1996), better-off buildings were repaired, and IEC was introduced to enhance access. Cost sharing was also introduced. Although staff salaries remained, staff morale significantly improved due in part to the better workplace and working conditions. In Phase III (1996-2000), the remaining batch of facilities were repaired. Three district- focused roving maintenance teams were organized, each team consisting of one full-time technician and four artisans (electrical, civil works, plumbing, carpentry), which proved to be successful in keeping facilities in good repair. In the mid-1990s, long before the Local Government Reform (LGR) Program was geared for implementation, DUHP proposed alternative organizational forms to devolve health services. The organogram involving facility boards and district health boards was approved by MOH and MRALG and is in keeping with both the HSR and LGR. Board members are chosen from actual  Tanzania Health & Nutrition Project ICR Page 7 users of the health facilities (not necessarily influential persons) and serve a term of one year on a voluntary basis. 21. DUHP' s outcomes are quite impressive. In all, 60 urban health facilities (dispensaries, health centers and district hospitals) were rehabilitated, staffed, and supplied. The rehabilitation program combined quality and cost effectiveness, i.e., the average cost for the rehabilitation of a dispensary stands at about Tsh 16 million, which seems reasonable given the size of the structure and the amount of work done. The roving teams are keeping facilities well maintained; evaluation reports indicate staff satisfaction with their state of repair. A pilot drug revolving fund at dispensaries appears working well. DUHP has also provided working models of Facility Boards and District Health Boards which are very active. Each District Health Board now holds its own meetings; receives reports from all facilities in the district; formulates the district's health plan; develops the budget; and allocates health expenditures within the urban district. The project has thus successfully established an annual process resulting in the production of comprehensive but standardized district health plans. Implementation Experience 22. Overall, the H&N Project is rated satisfactory and all of the planned activities were eventually carried out. However, implementation in the early years was unsatisfactory due to poor quality at entry (see below). The project enjoyed a satisfactory rating in the mid-1990s with the revamped effort by the Government, a new task team of the Bank, and project restructuring. However, the project reverted back to unsatisfactory status in 1996/97 due to problems with counterpart funds, delays in civil works, the stalled rehabilitation of the MSD, and inability to disburse the second and third portion for pharmaceuticals due to noncompliance to a legal covenant. Most of these problems have been resolved by 1997, paving the way for satisfactory implementation for the rest of the project's life. 23. Quality at Entry: For the first three years, implementation was unsatisfactory due to weak project management and problems with procurement, audit compliance, and compliance with legal covenants which were made as cross-conditionalities, making implementation of one component dependent upon the other. The project also suffered from inadequate supervision from IDA due to frequent changes in the task manager. Poor handling by IDA management of the transfer of Tanzania from the Southern to the Eastern Africa country department also contributed to poor project supervision during this period. In hindsight, "quality at entry was unsatisfactory," according to a World Bank Quality Assessment Group report, and "there seems to be general agreement among all parties, including the government and earlier project managers, that the project design was too complex" for a sector that had no previous experience managing an IDA credit, and faced serious scarcity in project management skills. Moreover, ownership of the project was very limited, if at all, at project inception. Staff who have institutional memory of the project stated that no stakeholder consultations were held, and that the GOT was in a highly vulnerable position due to a serious need for foreign exchange to purchase drugs. In the Quality Assurance Group's view, the project was presented to the IDA Board prematurely, "possibly in response to perceived pressure to lend".  Tanzania Health & Nutrition Project ICR Page 8 24. Extension of Closing Date: The Bank's and the GOT's recognition of the project's development potential, as well as the much-improved pace of implementation, provided the rationale for the two extensions of the closing date. The task team recognized that the large overhang of activities from 1990-94 meant that the project would not disburse fully by the original closing date. Thus, the project was extended in 1996 for two years until June 1998 in order to complete planned activities. At the same time, the project was restructured to accommodate the new health financing innovations developed during the project as well as to support Government initiatives for health sector reforms. The second extension, made in April 1998, was premised primarily on the need to complete the pretest of alternative financing schemes and support operationalization of GOT's health sector reform along the lines of a sector wide approach. Although it was clear that there would be an unspent balance at the end of the project, the Bank made a deliberate decision not to make any cancellation at the time of extension to allow for flexibility to support the pilots that are innovative but financially unpredictable in nature. Both extensions proved to be warranted: the backlog of activities were all completed, the health financing pretests were successfully carried out, and the project served to facilitate the development of sector-wide program. 25. Project Management: Project management initially suffered from serious weaknesses including poor understanding of the project concept, weak project coordination between PIU/Component I which was based in MOH, PSN/Component II which was based in PMO/MRALG, and other implementing agencies (TFNC, PPU and CMS); inadequate staffing; changes in MOH leadership and project coordinators; lack of basic office procedures, annual work planning, and budgeting processes; and inadequate understanding of Bank procedures. Midway through the project, these problems had been largely addressed through intensive staff training, hiring of more experienced project coordinator (the third and last), computerization of accounting functions, and GOT's persistent effort in operationalizing the project concept. 26. Procurement: Procurement, especially international competitive bidding (ICB), was difficult in the early years owing to project staff's inexperience and lack of familiarity with IDA rules and procedures. Although MOH capacity to procure drugs internationally has improved, similar capacity to procure medical equipment and instruments need to be further strengthened, based on one experience (Tender No. 7 of 1997/98) that resulted in the entry of less-than-qualified suppliers and the importation of some substandard goods. In this regard, there is a need to develop better and more stringent specifications and to strengthen Tanzania's currently limited pre-shipment and post-shipment inspection capacity. Selective procurement based on prequalified bidders need to be explored, and in fact this was the GOT Tender Board's preferred route for Tender No. 7 but was prevailed upon by IDA to go through ICB. Procurement of technical assistance still leaves much to be desired, especially for quality- and cost-based selection, and this has to be addressed in the follow-on IDA credit through more intensive capacity building, or in the interim, through a procurement agent. Project management also informed IDA that in certain cases, operating through IDA-mandated letters of credit was more expensive and laborious (due to the excessively high fees charged by local banks) than through direct payment against shipping documents.  Tanzania Health & Nutrition Project ICR Page 9 27. Disbursement and Financial Management: Due to problems with project management, disbursement-related covenant compliance, and generally poor work planning and budgeting, the rate of disbursements was very low for the first four years; only about a quarter of the credit (reckoned in SDRs) was disbursed during that period. The last five years, however, have been satisfactory as the turn-around time for workplan development and IDA approval was expedited, major covenants were met, and project management was dramatically improved. By closing date, 97% of the credit had been disbursed. In terms of financial management, all audit reports were submitted to the Controller and Auditor-General. The manual accounting system delayed submission of financial reports in the early years, but the computerization of the accounting system (using a simple spreadsheet) made financial reports more readily available. Financial management was generally adequate; a statement of expenses is being finalized. 28. Counterpart Funds: GOT had difficulty meeting its counterpart-fund obligations under the project in 1994-96 due to fiscal crises. Given that most African countries are prone to this problem, project managers have proposed that in the future, other in-country resources should be counted as counterpart funds, including user fees from cost recovery programs; fees from drug revolving funds; prepayment contributions in community health funds; membership premia under health insurance programs; and in-kind community contributions in the rehabilitation of facilities. The policy on what constitutes counterpart funds should be reviewed. 29. Civil Works: Although the project was not supposed to finance new construction, the absence of a thorough technical evaluation of the project's rehabilitation program resulted in the setting up of new structures which were deemed a technical necessity. Project preparation also failed to come up with unit-cost estimates for rehabilitation; these had to be generated during project implementation itself, but once they were developed (Tsh 10 million for a dispensary; Tsh 38 million for a health center; and Tsh 200 million for a district hospital), they proved to be useful negotiating tools with contractors. In some cases, the problems resulting from inadequate planning or poor performance of contractors led to several amendments of contracts and delayed completion of works. In the late 1990s, delay was also due to El Nino rains resulting in the poor accessibility of project sites. International contractors sometimes did work of inferior quality, according to project management. 30. Project Monitoring: The lack of baseline data severely constrained project managers' and Bank staffs understanding of what the project is about and what it is supposed to do. There were no simple indicators or targets with which to measure progress. Lack of household and facility-level baseline data made the assessment of impact virtually impossible. The project was "retrofitted" in 1997 through the development of a logical framework, and this assisted in the preparation of the ICR. An end-of-project evaluation by an external team of consultants is planned to take place in early 2000 which should provide information on the impact of project interventions. Major Factors Affecting the Project 31. Factors Outside Government Control: (a) Poor quality at entry, resulting in poor implementation at the outset, can be traced in part to Bank pressure to lend as much as  Tanzania Health & Nutrition Project ICR Page 10 the Government's lack of readiness for the project, a finding highlighted under the QAG review. (b) Project supervision also suffered from the transfer, in the early 1990s, of the Tanzania portfolio from one World Bank division to another. (c) Scarcity of technical and managerial skills of the kind required for the project was, and continues to be, a problem in Tanzania, and not one that the Government could have easily addressed. 32. Factors Within Government Control: (a) The intermittent budgetary constraints accounted for the inadequate allocation of counterpart funds to the project especially in the mid-1990s resulting in project implementation delays. (b) The reform process could have moved much faster if not for the initial political resistance to eliminate the government pharmaceutical monopoly and to implement cost sharing. As a result, major delays were experienced in the approval of the long term health financing strategy, the action plan on the domestic pharmaceutical industry, and reforms in drug pricing, financing, budgeting, and procurement. (c) Project management and coordination could have been improved much earlier, but PIU remained weak for most of the early years of the project. Recommendations were made to contract out some of the management tasks to outsiders including expatriates, but these were largely ignored. (d) More adequate planning, especially in the civil works component, could have prevented delays and improved contractor performance. (e) Stronger procurement capacity could have prevented the importation of inferior medical equipment and instruments. Sustainability of Project Activities 33. Largely as a result of the change in health policy, GOT representatives expressed greater confidence over the sustainability of activities supported by the project especially in alternative health financing schemes and pharmaceutical reforms. However, the continued implementation of HRD, HMIS, and rural PHC activities hinges on the robustness of the GOT health budget and therefore on the Tanzanian economy. 34. In terms of health financing, the CSP in government hospitals is well established; CHFs have gained support from the political leadership as well as households; the hospital drug revolving funds promise a more sustainable drug financing and supply pending resolution of policy constraints on cost recovery; and the NHI program for civil servants offers opportunity for greater cost recovery in lieu of the current arrangement of direct government subsidy. Modest cost sharing is also being implemented in training programs, though a more intensified cost recovery effort as well as adoption of other revenue mechanisms may be called for. Likewise, budget allocation to the health sector has dramatically increased. There is certainly a better prognosis for financial sustainability of the Tanzanian health system than it was at the onset of the H&N Project. 35. Major activities in the HRD Plan are of uncertain sustainability. Firstly, the Plan's implementation will continue to be highly reliant on external donors especially the training of DHMTs which has been made more urgent with the devolution of health services, and the MPH training of district medical officers. Secondly, the zonal training centers continue to suffer from the lack qualified staff, inadequate budget for maintenance and selective upgrading of zonal training centers, and lack of funds for the purchase of books and other leaming materials.  Tanzania Health & Nutrition Project ICR Page 11 36. HMIS activities are also likely to continue relying on donor support. In the medium term, the newly-signed DANIDA Health Sector Program Support, Phase 2 will provide Tsh 270 million annually until 2003 for logistics and equipment support. However, the longer-term sustainability of these activities is not ensured since the recurrent budget for supplies and other key inputs is inadequate. Decentralization also poses a challenge to HMIS in that local authorities need to be persuaded to include HMIS supplies in the district budget. Regional-level HMIS implementation may also be adversely affected by decentralization. 37. The institutional and financial sustainability of drug supplies in Tanzania is much better now than it was a decade ago. With respect to the sustainability of the Hospital Drug Capitalization Program, strengthening the capacity in hospitals (e.g., management and accounting of drug revolving fund revenues, better prescription and dispensing practice) would be critical, which should be dealt with in the next IDA credit. In addition, the existing policy of 50% cost recovery under the CSP, as well as the current absence of a mechanism for the Treasury to reimburse health facilities' expenses for drugs dispensed to waived and exempted patients, may inhibit the development of a fully refinancing mechanism. This issue certainly needs to be addressed in the planned program evaluation in January 2000. Short of a full cost-recovery policy for drugs, additional budgetary allocation is needed for the shortfall. 38. Rural PHC activities have mixed sustainability prospects. Institutional sustainability is being enhanced with the adoption of district rolling health plans in the 10 districts; the increasing GOT commitment to support DMOs' Master's training in public health and management; and the planned revision of the existing PHC strategy that will explicitly recognize the level of resources as a fundamental planning parameter, in addition to health patterns, technology, and management structure. Ranged against these positive factors are the continuing inadequacy of maintenance budget for rehabilitated PHC facilities and the lack of funding for PHC community mobilization. Some communities have already mobilized their maintenance funds, and the district block grants planned under the HSR Program as well as the CHFs can become important financing sources for such maintenance. Overall, these and other factors need to be taken into account in the development of a replication strategy for rural/district PHC, which still needs to be done. 39. Project-rehabilitated and -equipped health facilities also needs to be sustained. At present, the preventive maintenance budget for these facilities is very limited; the number of maintenance technicians is deemed inadequate; and the technically competent ones are often ill-motivated due to low salaries. Inadequate security in some facilities has led to losses of portable medical apparata. Moreover, recurrent funding for lab reagents, x-ray films, and other supplies are, in general, inadequate, thus reducing the usefulness of installed equipment. Districts have been empowered to use revenues from user fees for maintenance upkeep, but these are not enough. GOT needs to set aside a larger proportion of the budget for maintenance as a matter of policy, or if this is infeasible, use alternative mechanisms such as the block grants and the CHFs. In addition, a "maintenance culture" need to be inculcated at all levels. The project did set up preventive maintenance programs in the rehabilitated hospitals, including maintenance workshops for hospitals, but staffing continues to be a problem. Community involvement  Tanzania Health & Nutrition Project ICR Page 12 in maintenance is critical, and MOH needs to exploit this further. GOT also needs to examine the possibility of securing maintenance staff for health facilities at the district level in line with the overall decentralization policy, possibly under the local council. MOH and district authorities also need to learn the successful and innovative maintenance program implemented under Component III by the Dar es Salaarn City Commission. 40. Overall, the sustainability of the health system in Tanzania remains fragile. The combined effects of resource-mobilizing efforts (budgetary as well as extra-budgetary means), as well as the efficiency-enhancing interventions (district planning, stronger facility management), are still not likely to fully sustain the considerable investments made under the H&N Project as well as other donor projects in the sector. It is estimated that the GOT can only provide 39% of the recurrent cost requirements of the existing health system. Serious stock-taking within GOT needs to occur on the optimal size of the health system that the country and its donor-partners can afford. This issue needs to be comprehensively and consistently dealt with in the Health Sector Reform Program, the Civil Service Reform Program, and the Local Government Reform Program that the GOT is undertaking. Borrower Performance 41. According to the GOT internal assessment workshop,projectpreparation was deficient due to GOT's failure to adequately involve all stakeholders in the process, and inexperience with IDA preparation procedures. The Government also negotiated the project under very difficult circumstances; at that time, it faced a serious fiscal and foreign-exchange crisis making it unable to procure drugs and medical supplies in the international market. Thus, GOT had to accept difficult conditionalities for the use of project funds for drug procurement. 42. Project implementation was, on the whole, satisfactory. However, implementation in the first 3-4 years was unsatisfactory due to the project's inherent complexity; difficult coordination among several implementing agencies including MOH, MRALG, PPU, and TFNC; weak project management; and frequent changes in MOH leadership and project coordinators. Compliance with project covenants were considerably delayed because the critical ones (user fees, MSD autonomy) required Parliamentary action. Once the project management issues and policy covenants have been addressed, however, project implementation became satisfactory, especially in the last three years. GOT also successfully built on the achievements and experiences gained from the project to support its more ambitious Health Sector Reform Program. Bank Performance 43. The amount of background studies was considerable and sector analyses were on the mark, making project identification satisfactory. However, the GOT internal assessment workshop rated project preparation and appraisal as deficient, echoing similar findings from the QAG review. The project was overly complex with too many components and subcomponents (some not well defined); involved multiple  Tanzania Health & Nutrition Project ICR Page 13 implementing agencies that made coordination difficult; and entailed covenants that were complex, difficult to achieve, and some outside the purview of the implementing agencies. Quality at entry was poor. Although the project preparation team identified major risks such as weak administrative and technical capacity, poor staff motivation, and uncertain macroeconomic environment, their implications were not sufficiently taken into account. Less obvious risks were not considered, e.g., the sector's socialist background and the political will to transform it, weak GOT commitment to and ownership of the project at the start, and lack of prior IDA involvement in the sector. To make the best of the situation, the first three years were virtually used for ground work and capacity building (formulation of sector and subsector plans, staff training on IDA procedures and specific technical areas, building a constituency within MOH for critical reforms sponsored by the project). As a result of poor preparation, project supervision was unsatisfactory in the first three years, according to GOT's internal assessment workshop and the QAG review. Poor preparation was compounded by frequent changes in task managers occasioned by the transfer of Tanzania from the Southern to the Eastern Africa department in the Bank. A new task team took over the project in 1993-94, undertook a midterm review, and on this basis restructured the project. From then on, supervision has turned satisfactory. Thus, overall, supervision over the nine years of the project is rated satisfactory. Assessment of Project Outcome 44. Project impact on sector policies are substantial and will be the project's longer lasting legacy. The GOT has laid the foundation for a more sustainable financing and pluralistic delivery of health services, a difficult task given its socialist background. During the project, the health sector was opened to private medical practice, fees were adopted, and a semi-autonomous pharmaceutical agency was established - reforms aided by the economy-wide structural adjustments, including privatization and fiscal rationalization. It is difficult to measure the impact of these changes in terms of better access, quality of care, and improved health services, especially under continuing economic difficulties and the confounding factor of the AIDS epidemic. There is evidence, however, that access is better in project areas and that quality was enhanced with a more reliable drug supplies and rehabilitated and equipped facilities. The physical objectives of the project were substantially met, with all the planned civil works and equipping activities completed. However, given the extent of rehabilitation that needs to be done in Tanzania's 114 districts, the work done in the 10 districts barely scratched the surface. 45. Project impact onfinancial objectives and institutional development, though dramatic, were only partially achieved. GOT annual allocation to the health sector almost doubled in the nine years of the project, partly due to project covenants. User fees, pioneered in the project, have had demonstration effects in the education and water sectors. CSP now yields significant financial contributions to the sector (13% of nonsalary recurrent costs); without the project's pushing for this reform, underfunding in the sector would have been more persistent with adverse impact on the access and quality of health services. However, significant amounts of revenue continue to be uncollected (42% to 75%, according to the Revenue Targeting Study). The impact of other alternative  Tanzania Health & Nutrition Project ICR Page 14 financing schemes designed or pretested in the project remains to be shown; although they have a clear potential for mobilizing additional resources, there are concerns about their high startup and maintenance costs. Some ways of reducing these costs include greater use of local rather than expatriate consultants, better rate negotiation with consultants/staff, a tighter community mobilization and training strategy relying on roving "zonal" teams, and finalizing "best practice" manuals and guidelines so that they can be used more widely in roll-out areas. Project impact on institutional development with respect to MSD is large and yielding benefits, i.e., more reliable drug supply. The impact of enhanced MOH planning capacity is expected to yield improved policy formulation to ensure quality services. However, similar institutional-development impact with respect to district health boards remains to be realized. These boards have been set up in CHF districts but those in the remaining districts are yet to be established. Moreover, hospital management still leaves much to be desired even in focus districts. Future Operations 46. Future IDA operations should preserve and expand the gains achieved under the H&N Project and should continue to assist the GOT in the further development of the health sector. The successor IDA credit, the Health Sector Development Program (HSDP), is being designed as an Adaptable Program Loan to support, in coordination with other donors, GOT's Health Sector Reform initiatives. The HSDP aims to address the sector-wide capacity and management issues and tackle the negative effects of a fragmented project approach. The first phase of the 12-year project focuses on further strengthening of health service delivery in the context of decentralization and greater focus on cost-effective package of health services; further capacity building and human resource development especially the rationalization of training institutions; strengthening of central support systems in planning and budgeting, regulation, pharmaceutical procurement, and joint donor implementation; and expansion of health financing initiatives including CSP, CHF, drug revolving funds, and health insurance. Key Lessons Learned 47. The following are the key lessons learned from the project: * Project preparation should take account of project risks as comprehensively as possible and should reduce the complexity and extent of the project on the basis of the magnitude and probability of these risks. Key risks that should be taken into account are government project ownership, sector leadership, management and technical capacity, and the extent and pace of the reform program. * The relationship between IDA and the host government during project negotiation should be weighed carefully. Cash-strapped poor countries often find themselves with little countervailing power and IDA should avoid the temptation to exploit this vulnerability by imposing too many conditionalities detrimental to implementation. * The involvement of multiple agencies/ministries make project implementation more difficult. The separation of implementation responsibilities inhibits accountability,  Tanzania Health & Nutrition Project ICR Page 15 imposes a heavy burden on coordination, and engenders problems with respect to scheduling of related activities, especially if one agency's activity is contingent upon the completion of another agency's activity. Thus, the benefits of an inter-ministry project must be weighed carefully against the coordination costs. * Critical policy reforms (e.g., health financing and pharmaceutical sector reforms) should be pursued as project outcomes rather than upfront conditionalities. If certain conditionalities are required, their timing for compliance should be carefully planned depending on the nature and objective of the conditionality, as well as it being reflective of government commitment. Some conditions should be used upfront for negotiation, some are better suited for project effectiveness, and others can be applied for funds disbursement. * Compliance to these covenants and conditionalities should be within the purview of the implementing agencies. It is counter-productive to make these implementing agencies bear the adverse consequences of noncompliance or prolonged delay in compliance for something that is outside their control. * Project components should be designed with an internal logic behind them and not merely used as baskets to put activities in need of funding. The synergy and mutually reinforcing nature of the project components should be taken into account. * Economic liberalization and related macroeconomic reforms provide strong underpinnings for sector reforms and significantly enhance their achievement. Analytical and policy work between the two should be done hand in hand and in mutually supportive manner. 3 Baseline data and performance targets should be defined at project appraisal. These enhance monitoring of project progress; their absence inhibits supervision and makes impact evaluation virtually impossible. Data gathering pertaining to these indicators should be made an intrinsic part of the project. * For skill-scarce countries or sectors, training should be conducted as soon as possible so that the project can benefit from staff's acquired management and technical expertise. Procurement should be made a central concern of project management. * Counterpart funds should be calculated globally for the whole project, rather than on a contract-by-contract basis, and should take into account extra-budgetary resources generated from project-supported initiatives such as user fees from cost recovery programs, prepayments from community health funds, membership contributions or premia from risk-pooling or prepayment arrangements, and community inputs into village health initiatives.  Tanzania Health & Nutrition Project ICR Page 16 PART II: STATISTICAL ANNEXES Table 1 Summary of Assessments A. Achievement of Objectives Substantial Partial Negligible Not Applicable Macro policies x Health sector policies X Financial objectives X Institutional development X Physical objectives X Poverty reduction X Gender issues _ x Other social objectives X Environmental objectives x Public sector management X Private sector development X B. Prolect Sustainability Likely Uncertain Unlikely |B. Project Sustainability X ll Highly Satisfactory Deficient Highly Satisfactory Unsatisfactory C. Bank Performance Identification_x Preparation assistance x Appraisal _ x _ Supervision _ D. Borrower Performance Preparation _ X Implementation Covenant compliance X Operation X E. Assessment of Outcome X  Tanzania Health & Nutrition Project ICR Page 17 Table 2 Related Bank Credits Credit Title and Amount Purpose Year of Status Approval Past Operations - None Following Operations - APL program purpose: To March 2000 Appraisal mission made Health Sector improve access, utilization, (planned) in August 1999; under Development Program quality, and financing of health preparation (TZ-PE-58627) services through increased efficiency and effectiveness in allocation and use of resources to maximize their impact on health outcomes especially among the poor, women, and children Project development objective (Phase I): To improve resource management and quality of health services through sector reforms and institutional capacity building I Table 3 Project Timetable Steps in Project Cycle Planned Date Actual Date Identification March 1988 March 1988 Preparation March 1988 May 1988 Pre-appraisal July 1988 July 1989 Appraisal September 1988 October 1989 Negotiations December 1989 January 1990 Board Presentation March 1990 March 1990 Signing March 6, 1990 March 7, 1990 Effectiveness April 6, 1990 April 6, 1990 Midterm Review September 19, 1994 September 19 - October 6, 1994 Project Completion December 31, 1996 December 31, 1999 Credit Closing June 30, 1996 June 30, 1999  Tanzania Health & Nutrition Project ICR Page 18 Table 4A Credit Disbursements, Cumulative and Actual in U.S. Dollars Fiscal SAR Estimates Actual Actual Year Quarterly Cumulative Quarterly Cumulative Cumulative as (US$ Mn) (US$ Mn) (US$ Mn) (US$ Mn) % of Credit FY 90 Q4 0.20 0.20 0.12 0.12 0.25 FY 91 Q1 0.30 0.50 0.02 0.14 0.29 Q2 0.50 1.00 2.18 2.32 4.87 Q3 10.50 11.50 0.03 2.35 4.94 Q4 1.00 12.50 0.02 2.37 4.98 FY 92 Ql 1.00 13.50 1.67 4.04 8.49 Q2 1.00 14.50 0.00 4.04 8.49 Q3 10.60 25.10 0.02 4.06 8.53 Q4 0.69 25.79 0.13 4.20 8.82 FY 93 Ql 0.60 26.39 0.20 4.39 9.22 Q2 0.60 26.99 0.04 4.43 9.32 Q3 10.28 37.27 0.01 4.44 9.32 Q4 1.20 38.47 0.06 4.50 9.45 FY 94 Ql 1.20 39.67 1.33 5.83 12.25 Q2 1.20 40.87 0.12 5.95 12.50 Q3 0.90 42.07 0.51 6.47 13.60 Q4 0.90 42.97 5.91 12.37 26.00 FY 95 Q1 0.90 43.87 0.38 12.75 26.79 Q2 0.90 44.77 1.38 14.13 29.68 Q3 0.40 45.67 0.71 14.84 31.17 Q4 0.40 46.07 1.84 16.68 35.04 FY 96 QI 0.40 46.47 2.80 19.49 40.94 Q2 0.40 46.87 0.46 19.94 41.89 Q3 0.09 47.27 1.66 21.60 45.38 Q4 0.09 47.36 0.48 22.08 46.39 FY 97 Ql 0.09 47.45 0.11 22.19 46.62 Q2 0.15 47.60 0.62 22.81 47.92 Q3 ______ 3.82 26.63 55.95 Q4 3.10 29.73 62.46 FY 98 Ql 1.00 30.72 64.54 Q2 1.55 32.28 67.80 Q3 1.94 34.21 71.87 Q4 1.17 35.38 74.32 FY 99 Ql 1.86 37.24 78.24 Q2 3.72 40.96 86.05 Q3 _ 1.53 42.49 89.26 Q4 1.92 44.41 93.30 FY 00 Ql 2.88 47.29 99.35 Q2 __0.08 47.37 99.52 Note: The percentage of disbursement was calculated using the original total credit amount of US dollar equivalent, i.e., US$ 47.6 million. Due to exchange rate fluctuations which resulted in the increase in the US dollar value of the credit, the figures in this table do not necessarily match those of the SDR disbursement table.  Tanzania Health & Nutrition Project ICR Page 19 Table 4B Credit Disbursements, Quarterly and Cumulative in SDRs Fiscal Year Quarterly Cumulative (US$ Million) (US$ Million) FY 90 Q4 0.10 0.10 FY91 Ql 0.01 0.11 Q2 1.51 1.62 Q3 0.02 1.64 Q4 0.01 1.66 FY 92 Ql 1.26 2.92 Q2 0.00 2.92 Q3 0.01 2.93 Q4 1.00 3.03 FY 93 Ql 1.32 3.16 Q2 0.03 3.19 Q3 0.01 3.20 Q4 0.04 3.24 FY 94 Q1 1.00 4.18 Q2 0.08 4.27 Q3 0.37 4.64 Q4 4.17 8.81 FY 95 Ql 0.26 9.08 Q2 0.95 10.02 Q3 0.48 10.50 Q4 1.17 11.67 FY 96 Q1 1.86 13.53 Q2 0.31 13.83 Q3 1.13 14.97 Q4 0.33 15.30 FY 97 Ql 0.76 15.38 Q2 0.43 15.81 Q3 2.73 18.54 Q4 2.24 20.78 FY 98 Q1 0.73 21.51 Q2 1.13 22.64 Q3 1.44 24.08 Q4 0.87 24.95 FY 99 Q1 1.39 26.34 Q2 2.65 28.99 Q3 1.10 30.09 Q4 1.42 31.51 FY 00 Q1 2.12 33.63 Q2 0.06 33.69  Disbursement Profile for Tanzania Health and Nutrition Credit 40.00 35.00 - 1Actual Amt. Cumulative - - 30.00 + -Profile Amt. Cumulative 25.00 . & Original Amt. Cumulative a 2 0 . 0 0 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - --Re v i s e d 15.00 . .// .. Dt 10.00, ;W Ds. f / ~~~~~~~~~~~~~~~~~Original.En 5.001 Closing Date; 1 Fq C; 0.00 q>