Document of The World Bank Report No: ICR00001106 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IDA-36840) ON A CREDIT IN THE AMOUNT OF SDR 29.1 MILLION (US$ 36 MILLION EQUIVALENT) TO THE REPUBLIC OF BURUNDI FOR A MULTISECTORAL HIV/AIDS CONTROL AND ORPHANS PROJECT June 29, 2009 Human Development 3 Country Department CE1 Africa Region CURRENCY EQUIVALENTS (Exchange Rate Effective as of April 2009) Currency Unit = FBu 1069.60= US$ [1.00] US$ 1.00 = [SDR x] FISCAL YEAR January 1 ­ December 31 ABBREVIATIONS AND ACRONYMS Vice President : Obiageli Katryn Ezekwesili Country Director : John Murray McIntire Sector Manager : Lynne Sherburne-Benz Project Team Leader : Pamphile Kantabaze ICR Team Leader : Pamphile Kantabaze AFR Africa Region of the World Bank AM Aide Memoire ARV Antiretroviral Treatment BIF Burundian francs BSS Behavioral Study Survey CSO Civil Society Organization DCA Development Credit Agreement DO Development Objectives EU European Union FMR Financial Management Report GF Global Fund for AIDS, Tuberculosis and Malaria ICR Implementation Completion and Results Report IDA International Development Association IEC Information, Education and Communication IMF International Monetary Fund IP Implementation Progress ISR Implementation Status and Results Report KPIs Key Performance Indicators M&E Monitoring and Evaluation MAOP Multisectoral AIDS Control and Orphans Project MAP Multicountry AIDS Program MTR Mid-term Review MOH Ministry of Health MSHAP Multisectoral HIV/AIDS Program MWMP Medical Waste Management Plan NAC National AIDS Council NASP National AIDS Strategic Plan NGO Non-governmental Organization OI Opportunistic Infections OVC Orphans and Vulnerable Children PAD Project Appraisal Document PBC Performance Based Contract PDO Project Development Objectives PHRD Population and Human Resource Development (Grant) PES Permanent Executive Secretariat of the National Council PLWHA Persons Living with HIV/AIDS PMTCT Prevention of Mother-to-Child Transmission PSO Public Sector Organizations PSR Project Supervision Report QER Quality Enhancement Review SIDA HIV/AIDS SDR Special Drawing Rights STD Sexually Transmitted Disease TTL Task Team Leader QSA Quality of Supervision Assessment VCT Voluntary Counseling and Testing BURUNDI MULTISECTORAL HIV/AIDS CONTROL AND ORPHANS PROJECT CONTENTS Data Sheet A. Basic Information B. Key Dates C. Ratings Summary D. Sector and Theme Codes E. Bank Staff F. Results Framework Analysis G. Ratings of Project Performance in ISRs H. Restructuring I. Disbursement Graph 1. Project Context, Development Objectives and Design........................................................... 1 2. Key Factors Affecting Implementation and Outcomes .......................................................... 7 3. Assessment of Outcomes...................................................................................................... 14 4. Assessment of Risk to Development Outcome..................................................................... 22 5. Assessment of Bank and Borrower Performance ................................................................. 22 6. Lessons Learned.................................................................................................................... 25 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners....................... 27 Annexes Annex 1: Project Costs and Financing......................................................................................28 Annex 2: Outputs by Component ...............................................................................................32 Annex 3: Economic and Financial Analysis..............................................................................43 Annex 4: Bank Lending and Implementation Support/Supervision Processes..........................45 Annex 5: Beneficiary Survey Results .........................................................................................47 Annex 6: Stakeholder Workshop Report and Results ................................................................47 Annex 7: Summary of Borrower's ICR and/or Comments on Draft ICR...................................48 Annex 8: Comments of Co financiers and Other Partners/Stakeholders ..................................56 Annex 9: List of Supporting Documents ....................................................................................57 Annex 10: Summary of Key Findings-Three-Province Pilot Project .......................................58 Annex 11: Achievement of Project Development Objectives (details ) 61 MAP...........................................................................................................................................70 Tables Table 1: Burundi MAOP: DCA Agreed Core KPIs........................................................................ 3 Table 2: Changes in Distribution of MAOP Credit Proceeds, by Disbursement Category6 Table1.1Project Cost By Component ..........................................................................................27 Table 1.2: Source of Funds in Million USD................................................................................28 Table 1.3: Project Final Expenditure in USD...........................................................30 Table 2.1: MAOP Expansion of Core Services............................................................................31 Table 2.2: MAOP Number of Providers Trained.......................................................32 Table 2.3: MAOP Beneficiaries of Priority HIV/AIDS Interventions ..............................32 Table 2.4: Institutional Outputs of the Public Sector...................................................................34 Table 2.5: Public Sector Responses by Institution .....................................................35 Table 2.6: CSO Grant Subprojects financed and approved..........................................36 Table 2.7: Civil Society Subprojects by type of Beneficiaries........................................37 Table 2.8: Beneficiaries of Civil Society Subprojects by type38 Table 2.9: Priority Services Provided to OVC .......................................................39 Table 2.10: Number of OVC Subprojects financed...................................................40 Table 2.11: Proportion of subprojects targeting OVC ................................................40 Table 2.12: Key Project Outputs benefitting Youth ...................................................42 Table 2.13: Expenditures on Procurement...............................................................42 Table 11.1: HIV and Syphilis Prevalence/Pregnant Women .........................................61 Table 11.2: Persons tested for HIV/AIDS ..............................................................62 Table 11.3: HIV/AIDS Prevalence Rate among Blood Donors.......................................63 Table 11.4: Condom distribution..........................................................................63 Table 11.5: Pregnant Women Receiving PMTC........................................................63 Table 11.6: Newborns under ARV treatment ...........................................................64 Table 11.7: Expansion of ARV treatment sits and beneficiaries .....................................64 Table 11.8: Number of Providers trained in OI Treatment.............................................65 Table 11.9: Orphans' Well-being..........................................................................66 Figure Figure 1.1: MAOP Total Disbursements per Year (in 000 Burundian Francs)............................ 29 A. Basic Information Multisectoral Country: Burundi Project Name: HIV/AIDS Control and Orphans Project Project ID: P071371 L/C/TF Number(s): IDA-36840 ICR Date: 06/30/2009 ICR Type: Core ICR GOVERNMENT OF Lending Instrument: APL Borrower: BURUNDI Original Total XDR 29.1M Disbursed Amount: XDR 24.7M Commitment: Environmental Category: B Implementing Agencies: Conseil National de Lutte contre le VIH/SIDA Cofinanciers and Other External Partners: B. Key Dates Process Date Process Original Date Revised / Actual Date(s) Concept Review: 05/03/2001 Effectiveness: 10/07/2002 Appraisal: 11/19/2001 Restructuring(s): Approval: 06/27/2002 Mid-term Review: 03/14/2005 03/14/2005 Closing: 12/31/2006 11/30/2008 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Moderately Satisfactory Risk to Development Outcome: Substantial Bank Performance: Moderately Satisfactory Borrower Performance: Moderately Satisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Satisfactory Government: Moderately Satisfactory Quality of Supervision: Moderately SatisfactoryImplementing Agency/Agencies: Moderately Satisfactory Overall Bank Overall Borrower Performance: Moderately SatisfactoryPerformance: Moderately Satisfactory i C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments Performance Indicators (if any) Rating Potential Problem Project Yes Quality at Entry None at any time (Yes/No): (QEA): Problem Project at any Quality of No Moderately Satisfactory time (Yes/No): Supervision (QSA): DO rating before Satisfactory Closing/Inactive status: D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) General public administration sector 14 14 Health 30 30 Other social services 36 36 Primary education 10 10 Vocational training 10 10 Theme Code (as % of total Bank financing) Conflict prevention and post-conflict reconstruction 11 11 HIV/AIDS 22 23 Participation and civic engagement 22 22 Social safety nets 22 22 Vulnerability assessment and monitoring 23 22 E. Bank Staff Positions At ICR At Approval Vice President: Obiageli Katryn Ezekwesili Callisto E. Madavo Country Director: John McIntire Emmanuel Mbi Sector Manager: Lynne D. Sherburne-Benz Arvil Van Adams Project Team Leader: Pamphile Kantabaze Malonga Miatudila ICR Team Leader: Pamphile Kantabaze ICR Primary Author: Karen Lashman ii F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) Support the implementation of Burundi's Multi Sectoral HIV/AIDS Program, through : (i) Slowing down the spread of HIV/AIDS in the general population; and (ii) mitigating the damage of HIV/AIDS on individuals and families. Revised Project Development Objectives (as approved by original approving authority) (a) PDO Indicator(s) Original Target Formally Actual Value Indicator Baseline Value Values (from Revised Achieved at approval Target Completion or documents) Values Target Years Percentage of (i) sex workers, (ii) Young men (age 15-25), (iii) young women Indicator 1 : (iv) displaced persons, (v) military who report using a condom during their last contact with a non-regular partner (Source Global average 15-49 old, Value all groups: 19.81 (15-49 old, quantitative or # 43% (PAD); >60% all groups NA all groups, BSS Qualitative) # 12.7 % (NASP, 2001 BSS) 2007) Date achieved 10/07/2002 11/30/2008 12/31/2007 12/31/2007 Comments (incl. % achievement) Indicator 2 : % of girls 15 years old or younger who report being sexually active (including forced encounters due to war conditions) Value 18% (PAD); quantitative or 32.5% (NASP, 2001 10% 18% 21.46% (2007 BSS) Qualitative) BSS) Date achieved 10/07/2002 01/30/2008 12/31/2007 12/31/2007 Comments (incl. % achievement) Indicator 3 : Percentage of orphans placed with families by the pilot component who advance from one grade of primary school to the next. Value quantitative or 0 80% 65.1% Qualitative) Date achieved 10/07/2002 11/30/2008 11/30/2008 Comments Original targeting was based on 3 provinces pilot project. Closing achievement (incl. % was calculated based on the extension of the Orphans component#s activities to iii achievement) all 17 provinces (b) Intermediate Outcome Indicator(s) Original Target Formally Actual Value Indicator Baseline Value Values (from Achieved at approval Revised Completion or documents) Target Values Target Years Percentage of ministries which have incorporated HIV/AIDS prevention into Indicator 1 : their work plan and budget (total number of ministries reduced from 27 to 20 plus 3 Institutions following elections) Value (quantitative 20% 100% 100% or Qualitative) Date achieved 10/07/2002 11/30/2008 11/30/2008 Comments (incl. % achievement) Indicator 2 : Number of orphans placed in foster families Value (quantitative 0 518 163,229 163,229/year or Qualitative) Date achieved 06/30/2002 11/30/2008 12/31/2006 12/31/2006 Comments (incl. % No. of Orphans placed in families and receiving scholar fees per year (PES/NAC achievement) 2006 Report) Indicator 3 : Number of civil society organizations working in HIV/AIDS receiving implementation support the MPLSAO Value (quantitative 0 400 530 (PES/NAC or Qualitative) 2008 Report) Date achieved 06/30/2002 11/30/2008 11/30/2008 Comments (incl. % achievement) Indicator 4 : Annual Voluntary Counseling Value 236,988/year (quantitative 31,000 50,000/year (PES/NAC 2008 or Qualitative) Report) Date achieved 10/07/2002 11/30/2008 12/31/2008 Comments (incl. % achievement) Indicator 5 : Number of PLHA under ARV treatment Value (quantitative 1200 12,500 14,343 (PES/NAC or Qualitative) 2008 Report) Date achieved 07/22/2003 11/30/2008 12/31/2008 iv Comments (incl. % Cumulative number of persons under ARV Therapy (Target = 10% of estimated achievement) G. Ratings of Project Performance in ISRs Actual No. Date ISR Archived DO IP Disbursements (USD millions) 1 11/20/2002 Satisfactory Satisfactory 1.50 2 12/16/2002 Satisfactory Satisfactory 1.50 3 06/17/2003 Satisfactory Satisfactory 2.49 4 07/22/2003 Satisfactory Satisfactory 3.34 5 03/26/2004 Satisfactory Satisfactory 7.78 6 10/26/2004 Satisfactory Satisfactory 11.70 7 05/24/2005 Satisfactory Satisfactory 18.65 8 12/20/2005 Satisfactory Satisfactory 25.19 9 03/30/2006 Satisfactory Satisfactory 26.57 10 08/25/2006 Satisfactory Satisfactory 31.13 11 06/20/2007 Satisfactory Satisfactory 37.82 12 12/18/2007 Satisfactory Satisfactory 40.74 13 06/20/2008 Satisfactory Satisfactory 42.58 H. Restructuring (if any) Not Applicable I. Disbursement Profile v 1. Project Context, Development Objectives and Design 1.1. Context at Appraisal 1. Burundi's long civil strife exacted an extremely high social as well as economic toll. Between 1993 and the late 1990s alone, more than 300,000 persons had died, and some 800,000 had been forced to emigrate. Pervasive poverty, with 70 percent of Burundians living on less than US$1 a day, and a GDP of about US$100 per capita/year, underpinned and exacerbated development challenges, among the most serious of which were: (i) A significant deterioration in health status as reflected in such key indicators as the infant mortality rate which rose from 110 to 125 per 1000 live births between 1993 and 1998; prevalence of chronic malnutrition among children under five years of age which increased from less than 6 percent in 1993 to over 20 percent in 2001; and life expectancy at birth which declined from 55 years in 1993 to 53 years in 1999 (Burundi : Poverty Reduction Strategy Paper, 2007). (ii) Limited institutional capacity to deliver basic social services. Major gaps in social sector financing and in skilled professionals, and the poor state of most public sector facilities and equipment, coupled with limited capacity within NGOs and provincial and communal authorities, severely limited access to health, education and other social services essential for prevention, control and treatment of HIV/AIDS and mitigation of its impact on those infected and affected. Low levels of education and literacy, particularly among women, and social disruption perpetuated the vicious cycle of poverty and high fertility, exacerbating health risks and slowing development. (iii) A seemingly high prevalence of HIV/AIDS. During project preparation, the prevalence of adult HIV in Burundi was believed to be 12 percent based on UNAIDS estimates, representing a sharp increase from less than 1 percent in 1990 (UNAIDS estimate derived largely from epidemiologic bulletins from selected sentinel sites on prevalence among pregnant women 15-24 years of age and national survey data from 1989-90--the only prevalence data then available). Data from a nationally representative sero-prevalence survey conducted in 2001 whose results became available after the project had commenced, revealed a significantly lower prevalence rate of 3.9 percent among those 18 and older, putting the original estimates into question. Significant urban and rural differentials in prevalence rates also existed. Nationally, an estimated 19,000 children under 15 and 360,000 adults were living with HIV/AIDS, with women disproportionately affected (55.8 percent of all adult cases). HIV/AIDS led to an estimated 11,000 to 39,000 deaths annually, and had become the major cause of adult deaths, threatening to erode all development gains of the 1970s. (iv) Hundreds of thousands of orphans, to which HIV/AIDS was a major contributor. Nationally, UNICEF and ISTEEBU (Institut des Statistiques et des Etudes Economiques du Burundi) survey data indicated some 558,000 children under 14 years of age--one in five in this age group -- had lost at least one parent, and 77,000 both parents. According to UNICEF and WHO estimates, some 160,000 to 230,000 of those orphaned were due to HIV/AIDS. 2. Despite efforts by the government to change behavior around HIV/AIDS (including the first National Program launched in 1986, three years after diagnosis of the first cases, and the formulation and implementation of a 1999-2003 Strategic Plan), successive behavioral surveys revealed limited change in general knowledge of HIV/AIDS, prevalent high risk behavior, and low use of condoms. Uncoordinated public and private sector activities hindered the Program's ability to reduce HIV/AIDS prevalence or to effectively address HIV/AIDS-associated health and social problems. 3. In July 2001 an autonomous National AIDS Council was created under the Office of the President, and in November 2001 a designated Ministry of HIV/AIDS was created. These 1 entities brought attention to the HIV/AIDS epidemic and served as a major impetus to the development of the 2002-2006 National AIDS Strategic Plan (NASP), developed by the government. The Plan espoused a multisectoral approach and a highly decentralized public- private partnership for delivering priority services. Solidly endorsed by donors, the NASP focused on: (i) prevention; (ii) expanded treatment of PLWHA and activities to mitigate adverse socioeconomic effects on PLWHA; (iii) intensive capacity building endeavors across public and private sectors to accelerate the decentralization of the program to provinces, communes, and local areas; (iv) studies to enrich knowledge of HIV/AIDS; and (v) the institution of an integrated national AIDS monitoring and evaluation system. 4. Burundi thus met the four key eligibility criteria for access to IDA resources within the framework of the Multi-Country HIV/AIDS Program (MAP) for the Africa Region [(a) satisfactory evidence of a strategic approach to HIV/AIDS; (b) existence of a high level coordinating body; (c) government agreement to use exceptional implementation arrangements, including channeling funds directly to implementing agencies, contracting out activities where appropriate, and adopting simplified administrative procedures; and (d) use and funding of multiple implementation agencies]. Financing of an HIV/AIDS project also promised important synergies with the ongoing IDA-supported Second Health and Population Project, Second Social Action Project, and the Emergency Recovery Credit. Finally, HIV/AIDS was one of the pillars of IDA's 2002 Transitional Support Strategy. 1.2. Original Project Development Objectives (PDO) and Key Indicators (as approved) 5. The project supported implementation of Burundi's Multisectoral HIV/AIDS Program (MSHAP) and its underlying 2002-2006 NASP. Overarching objectives of both the program and the project were to: (i) slow the spread of HIV/AIDS in the general population; and (ii) mitigate the impact on those infected with or affected by HIV/AIDS. The project indicators reflected a mix of outcome, output and process measures. 2 Table no. 1: Original Performance Indicators Agreed Indicators Baseline Closing Targets Outcome PAD DCA NASP (*) 1. Percentage of adults who are using a condom during their last 43 43 12.7 60 contact with a non regular partner (2001 BSS) 2. Percentage of Girls 15 years old or younger who report being 18 18 32.5 10 sexually active (2001 BSS) 3. Percentage of Orphans placed with families by the pilot 0 0 80 component who advance from one grade of primary school to the next N.A. 4. Percentage of women who test positive for syphilis during 3 3 2.4 1 antenatal consultations (MoH Epi Stat) Output 5. Percentage of ministries which have incorporated HIV/AIDS 20 20 30 100 prevention in their work plan and budget (8/26) source: MoH 6. Percentage of health facilities which have improved blood safety N.A. 0 0 100 and safe injection practices 7. PAD: Percentage of registered orphaned children who receive 5 5 70 social support at community level; 8. Number of civil society organizations working in HIV/AIDS N.A. 0 400 receiving support to implement their sub-projects 9. Percentage of funds flowing to communities N.A. 0 - 60 (PAD); 92 (DCA) 10. DCA: Percentage of supervision and support missions by staff of 0 - 75 coordinating units to implementing agencies and communities N.A. carried out; PAD: Percentage of sub-projects supervised by the staff of CPLS (Provincial AIDS Committees) 75 (*) As spelled out in the Supplemental Letter dated July 25, 2002. The appraisal baseline values were incorrect for the two first PDO indicators as opposed to those in the NASP which are based on the BSS results. No source was found for the appraised data and the ICR mission did not find any explanation to this anomaly. The PAD target value for indicator no. 9 is different from that in the DCA. Indicator no. 10 was phrased somewhat differently between the PAD and the DCA albeit the concept of what needed to be measured is the same in both cases. 3 1.3. Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification 6. Over the project life, no formal revisions were made to the PDOs or to the 10 core KPIs. However, a spirit of "learning by doing" pervaded the MAOP, as in many first generation MAP projects. TTLs exercised considerable discretion in selection and use of indicators to report on PDO progress as better HIV/AIDS information became available, the M&E system matured, and project-supported activities evolved. The mix and number of KPIs reported in PSRs/ISRs over time varied, using most of agreed KPIs and adding new indicators, but at that time formal approval for such revisions was not required by Bank guidelines. All original KPIs continued to be monitored ­ among others ­ by the national M&E system throughout the project. 7. Modifications to ISR reporting on KPIs represented important enhancements to performance monitoring, allowing measurement of project progress in priority investment areas not captured in the original KPIs. As of March 2004, ISRs reflected (i) 2 new indicators related to the numbers of persons receiving respectively VCT and ARV treatment (at the start of the project ARVs were considered too costly to provide to PLWHA; with decreased prices and improved treatment sites and trained personnel, ARV treatment was begun and monitored); and (ii) the number of orphans placed in families, all of which had been being monitored within the NASP M&E framework. The ICR assessment is based on the 10 original KPIs as well as on these 3 additional indicators. 8. The paucity of timely, reliable, nationally representative HIV/AIDS data at the time of project preparation and appraisal led to erroneous baseline values provided by the Borrower and consequently unrealistic targets for two of the core outcome indicators, as mentioned above. This hampered the assessment of project outcomes based on those values, as discussed later in the ICR. Furthermore, in some cases where innovative CSO grants were developed, there were no precedents for determining realistic target indicators. As new reliable data became available, the team modified these outcome indicator targets informally. In today's Bank this would have constituted a reason for first order restructuring under the "opportunistic restructuring" label. However, at the time it was not common practice to restructure a project in order to adjust the indicators, but rather reflect those changes in the Aide memoirs and back to office reports. Therefore the Mid-term Review (MTR) changed the end-of-project target value for the percentage of girls 15 years or younger who reported being sexually active--one of the core outcome indicators--was increased from 10 percent to 18 percent (as compared to an estimated baseline of 18%) from the 12/20/ 2005 ISR forward solely with a cryptic notation in that ISR's KPI matrix that the original target was "no longer realistic because of the increase due to war conditions." It was reported for this ICR that the base line value of 18 per cent was taken from studies done prior to the conflict, and that the results from the 2001 Behavioral Study Survey (shared in 2002 after project effectiveness) suggested that the real number was 32.5 percent. 9. Further, the reliability of KPI values reported in ISRs was uncertain, given the lack of adequate data quality control measures in the M&E system. Considerable effort was expended on reconciling various data inconsistencies in the process of preparing this ICR. This underscored continuing challenges the M&E system faces to become a fully effective Program management information tool as discussed in Section 2.3 below. 1.4. Main Beneficiaries 10. The project was intended to prevent the spread of infection among all groups in Burundi with special attention to high risk groups, including orphans and other extremely vulnerable children, women of reproductive age, commercial sex workers, youth in and out of school, long distance truck drivers, people 4 infected with STIs, armed forces and displaced persons and refugees. Substantial benefits were also expected to accrue to those infected with or affected by HIV/AIDS through project-supported interventions to mitigate social and economic impacts. This primary target group remained unchanged throughout the project. 11. In addition, through its support for implementation of multisectoral HIV/AIDS efforts through the public and private sectors, the project was expected to strengthen national capacity to deal effectively and efficiently with both the current HIV/AIDS epidemic as well as other potential epidemics. Original Components Component 1: Strengthening the public sector response to the HIV/AIDS crisis (US$9.3 million). To expand HIV/AIDS activities implemented by public sector entities. Component 2- Increasing the civil society response (US$13.4 million): To increase HIV/AIDS control activities initiated and implemented by grass-roots organizations and communities (as well as some initiatives of private sector enterprises) in the areas of prevention, mitigation and capacity building proposed in the National Strategic Plan. Component 3 - Orphans Integration & Education Pilot (US$ 7 million): To (i) support the placement of orphans with their extended family; or with an unrelated family; (ii) provide orphans with scholarships for primary education and vocational training; (iii) strengthen the economic and nutritional well being of families that care for orphans through investments to supplement nutrition and income- generating activities; and (iv) provide training for the informal job market to older adolescent orphans. Component 4 - Strengthening the Program Strategic Formulation, Implementation, Monitoring ,and Evaluation capacity (US$4 million): to reinforce the capacity of public and private agencies in the design, implementation, and monitoring and evaluation of HIV/AIDS control activities. Component 5 - Project coordination (US$3 million): to ensure efficient coordination of the project's HIV/AIDS control activities by all concerned parties. Revised Components 12. No components were substantially revised. The definition of orphans targeted in Component 3, however, was changed from children up to 16 (with one or no parents) to orphans 16 to 18 years of age, a subgroup highly vulnerable to HIV/AIDS, particularly if they are heads of household and/or unemployed. This change was made based on the project launch workshop's recommendations, without any amendment of the DCA. Other significant changes 13. There were three DCA amendments during the life of the project. They all answer to changing realities in the country situation. The First Amendment dated March 4, 2005, increased the percentage of IDA financing 100 percent of all categories of Credit expenditures. The Amendment was approved by the country director in view of the serious counterpart financing problems due to a major deterioration in the country's macroeconomic situation. The Amendment, not applicable retroactively, required the Government to meet all current outstanding counterpart financing commitments and debt (subsequently done). It also modified procurement arrangements, consistent with simplified Bank procurement procedures introduced after Credit signature, and in recognition that goods such as ARVs, drugs and other related products could only be purchased from a limited number of suppliers; increased thresholds for national competitive bidding and shopping for selected works and goods; and 5 permitted, under certain thresholds, procurement of goods including drugs and medical supplies from UN agencies. 14. The second and third amendments reflected two reallocations of Credit proceeds (March 30, 2006 and June 1, 2007, respectively). Both reallocations were approved by the regional vice-president. 15. This reorientation of the project proved crucial to advance the PDOs. Extremely weak capacity within the public sector, increasingly evident over the early implementation period, precluded it from assuming the central, direct role in implementation envisaged by the project design. In sharp contrast, the civil society actively and positively responded to availability of grant financing, with the volume of quality grant requests received rapidly exceeding original project targets. After the two reallocations, financing for grants for CSO subprojects increased to 76.6 percent, or 46.7 percent of total project financing (Table 2). This substantial increase enabled the PES/NAC ultimately to fund 564 (98.6 percent) of the 572 approved CSO subprojects by project's end (see Annex 2). 16. The reallocations also better aligned Credit financing with evolving program funding needs, particularly in light of: (i) availability of alternative sources of financing to procure priority goods originally to be financed by MAOP such as ARVs, other drugs and condoms, including increased support from UNDP, UNICEF and the EU and new financing from the Global Fund, a new program partner from 2003; (ii) increased demand for financing for non-drug related costs of treatment and follow up care of PLWHA by CSOs, for which IDA was the primary source; (iii) reduced need for financing for works subsequent to the decision to rehabilitate existing small facilities nationwide to serve as youth centers rather than construct two new large ones initially planned; and (iv) need for financing of additional studies to both evaluate program performance and inform preparation of the National AIDS 2007-2011 Strategic Plan. 17. The reallocations contributed to: (i) a major scaling up of the breadth and depth of CSO participation in the National AIDS Program, thus successfully deploying much of their unexploited potential to expand access to priority services; (ii) mobilization of supplementary Program resources, with CSOs sharing costs of subprojects via resources they mobilized from other sources, with Credit-supported grant payments circumscribed predominantly to costs associated with direct service provision; and (iii) intensification of community-designed and community-led initiatives, thus increasing likelihood of responsiveness to local needs and, importantly, sustainability of actions. Table No. 2: Changes in Distribution of MAOP Credit Proceeds, by Disbursement Category Expenditure Original Category as March 2006 Category as June 2007 Category as Percentage Category Allocation (000 Percentage of Reallocation Percentage of Reallocation Percentage of Change between SDR) Total Original (000 SDR) Revised March (000 SDR) Total Final Original and Allocation 2006 Allocation Approved Final Allocation Allocation Works 1,540 5.3 560 1.9 500 1.7 -67.5 Vehicles, 4,300 14.8 1,300 4.5 1,150 4.0 -73.3 equipment and office supplies Drugs, tests, 8,500 29.2 6,360 21.9 6,300 21.6 -25.9 medical supplies Consultants' 3,800 13.0 6,460 22.2 5,700 19.6 +50.0 services, audits and training Grants for CSO 7,700 26.5 12,650 43.5 13,600 46.7 +76.6 subprojects Operating Costs 1,800 6.2 1,770 6.1 1,850 6.4 +2.8 Unallocated 1,460 5.0 0 0 TOTAL 29,100 100.0 29,100 100.0 29,100 100.0 6 Sources: July 25, 2002 Development Credit Agreement for Credit 3684 BU; and March 30, 2006 and June 1, 2007 Approved Reallocations Note: Percentages may not add to 100.0 due to rounding. 18. Other significant changes included: (i) A major expansion of the scale and scope of project support for ARV treatment and for orphans. These changes were requested by the government, endorsed in principle by a Bank Quality Enhancement Review Panel convened March 8, 2005, prior to the Mid-term review, and made effective during that mission. Demand for ARV treatment had risen significantly after the government introduced its policy of universal access to treatment (the required reallocation resulted in a 70 percent increase in financing for consultants' services, audits and training (Category 3) to support increased capacity building and performance-based contract activities; a 63 percent increase for grants for CSO subprojects, consistent with Mid-Term review recommendations to scale up CSO activities; and a 1.7 percent increase in operating costs, to be accommodated by decreases in all other expenditure categories.) Extension nationwide of coverage of the OVC component responded in part to mounting political pressure amid large unmet needs of hundreds of thousands of OVC, but also highly positive, multiple improvements in the well-being of orphans and their host families found in an independent, project-financed impact evaluation of the three pilot provinces, summarized in Annex 10. (ii) Extension through the end of secondary school of MAOP financial support for school fees and related education expenses for OVC, with special attention to identifying and providing secondary school support to young girls. 2. Key Factors Affecting Implementation and Outcomes 2.1. Project Preparation, Design and Quality at Entry 19. Project preparation proceeded satisfactorily. A fast track approach was adopted, supported by a $682,400 PHRD (Japanese) grant approved in March 2001, and with the MOH as the main counterpart. The approach recognized the magnitude of Burundi's HIV/AIDS crisis and significant risks of an increase in prevalence with greater population mobility in the post-conflict period. The March 2001 Washington visit of a high level government delegation to solicit Bank financing for the National Program, in which representatives of NGOs and civil society notably were included, was viewed by Bank management as an "exceptional" reaffirmation of national HIV/AIDS commitment, and provided additional push. 20. The design process was highly participatory with: (i) regular consultations with key public and private sector personnel, including CSO representatives, elected officials, and international stakeholders; and (ii) the organization of focus groups at the central, regional, and local levels, all working to develop appropriate strategies at each level. 21. The design responded well to the need to scale up Burundi's HIV/AIDS response rapidly by providing strong incentives, including both financial and technical assistance, to fully engage the public and private sectors and all administrative levels of the nation in the Program. The project design and the national plan it supported were informed by a series of surveys and studies commissioned by the government to better understand local attitudes and barriers to behavioral change, particularly among the most vulnerable groups--children, women, and youth. One example was a survey of the population 15 years of age and younger in which 9 provinces participated and in which 38 focus group discussions were held with over 2000 persons participating. The project design also drew extensively on lessons learned from first generation Bank MAP-supported projects, 7 from successful HIV/AIDS programs in Uganda and Senegal and, even from Burundi's own decentralized, community-based operations (Social Action Projects and two community nutrition projects). 22. Although there were several qualitative studies, quantitative data was scarce because the country's ongoing conflict impeded the collection of good quality base-line data. Corrections were not included in the restructuring or formally corrected during project implementation. This resulted in a discrepancy in PDO indicators between the original project outcome and the final outcome. 23. A special design challenge arose in mid-2001, with the joint Bank and Government decision to fully integrate into the HIV/AIDS project a separate project, then under preparation, to support the social integration and education of orphans. This merger, which brought an additional US$7 million to the orphan's component, was expected to permit realization of important synergies between the two projects as well as to reduce administrative costs. Consistent with the original thrust of the orphans' project, the new component would target only the most vulnerable orphans, regardless of their HIV/AIDS status, and provide them the already agreed mix of benefits. Needs of the larger group of orphans were to be addressed by other HIV/AIDS project components. 24. The relatively late integration of this new component on the orphans in the design phase of the project presented substantial risks. The project executing agency, the National Council (PES/NAC), had no specialized expertise or experience in the area of orphans. The risk was largely mitigated by the Council's recruitment of an expert team involved in preparation of the original orphans' project to guide the execution of the component (for which Norwegian Trust Funds were used until the project became effective in order to ensure continuity of preparation and leadership). The Orphans Steering Committee, established as part of the initial orphans' project preparation process, was also tapped to review all CSO subprojects serving orphans. Its agreement to the appropriateness and consistency of targeting criteria was a condition of the PES/NAC's final approval of the financing of these subprojects. 25. Project Quality at Entry, was compromised, by the preparation team's significant underestimation of the high risks associated with serious capacity constraints at all levels, particularly shortages of qualified professionals, as discussed in Section 2.2. 2.2. Implementation 26. The Credit became effective on October 7, 2002, approximately two months later than expected, followed immediately by a highly participatory project launch with over 140 representatives from all sectors as well as beneficiary groups affirming broad "buy-in" for the project. Initially project implementation advanced steadily but at a far slower pace than projected, amid institutional capacity constraints not adequately identified during preparation. 27. Incomplete project documentation hampers assessment of the extent to which the implementation strategy pursued by the government and the Bank was appropriate, including whether sufficient priority was assigned to those interventions most likely to reduce HIV transmission, especially among high risk groups. The vagueness in project documents of what "actually happened on the ground" was underscored by the September 2006 QSA7 Panel. 28. Public sector response: The KPI target for this component was attained. However, the quality and number of action plans and activities varied significantly across central ministries. This was due in large part to the exodus of many highly qualified public sector staff to CSOs, external assistance agencies, and neighboring countries during the conflict. Nonetheless, by project end, all ministerial departments had some definable Action Plan established. 8 29. Contrary to the PAD's assumption that the MOH "should have little problem in scaling up its operations," (Para 4.1, p. 25) its HIV/AIDS sectoral unit had limited capacity, which led the MOH at mid-term to recruit four additional specialists -- a physician/health planner, an epidemiologist, an infectious disease physician, and a pharmacist. With this improved capacity, the unit produced solid guidelines for treatment, certification of facilities and staff training, and improved projections of pharmaceutical needs. In 2006 an in-depth cross- ministerial strategic planning exercise resulted in improved plans. 30. Despite implementation problems at the central level, the eight Ministries key to the fight against HIV/AIDS, had integrated a solid HIV/AIDS action plan, with budget, into their sectoral strategy by the end of the project, leaving a solid base for sustainability beyond the MAOP Project. 31. At the decentralized level, the project preparation team's assumption that existing capacity in provincial, communal and local committees was "good enough to ensure speedy implementation from project launch" (PAD p. 26) proved to be overly optimistic. Trained personnel constraints slowed the decentralization process and perpetuated inequities in access to quality HIV/AIDS care and services (March 2005 Mid-Term Review). 32. Notwithstanding, over time, decentralized public entities, particularly provincial authorities, assumed a major role in project execution, not as implementing agencies, as had been initially envisaged, but rather in mobilizing, coordinating and guiding the response of local CSOs, in strengthening their capacity, and in monitoring and evaluation, including data collection and analysis. Local authorities were represented in all provincial and communal committees reviewing subproject proposals and were closely involved in monitoring subproject performance. They were, thus, instrumental in forging the dynamic public-private partnership that propelled the project forward. 33. MAOP also provided essential technical and financial support for accelerating the decentralization of the (PES/NAC). By the project's second year, Provincial Committees had been established and were operative in all 17 provinces. Communal Committees were also operating in all of 17 provinces. By 2004, AIDS Communal Committees were in operation in all 129 communes. These decentralized Council Committees provided an institutional base for full engagement of local authorities in project-supported activities. 34. The civil society response was exceptionally strong, with subproject grant requests greatly exceeding original project expectations and targets. Fueled by the availability of grant financing, CSOs rapidly scaled up the breadth and depth of delivery of HIV/AIDS education and services. Engagement of the CSO sector in the HIV/AIDS Program fostered substantial innovation and experimentation. It also promoted a highly participatory, inclusive approach with PLWHA, women and youth, among others, actively engaged in subproject activities. 35. The Orphan Component scaled up quickly. By mid-term, responding to its good results and the demand from communities and local authorities, it was expanded nation-wide. Over the life of the project, this component was largely mainstreamed into the public sector and civil society components of the project. CSOs were particularly responsive to addressing OVC needs, with 181 OVC subprojects financed in total, representing one-third of the total 564 CSO grants provided by MAOP (see Annex 2). 36. The Mid-Term Review (October-November 2005) marked a critical juncture in the project. On March 8, 2005, a Quality Enhancement Review (QER) was held, attended by both the Bank project team and their government counterparts. Suggestions were made regarding: (i) essential preconditions to ensure the success of the proposed expansion of coverage of the OVC component and of ARV treatment (approved at Mid-Term); (ii) the desirability of addressing human resource constraints cross-sectorally within the Burundi portfolio; and (iii) the high priority that must be assigned to establishing a truly national, integrated M&E HIV/AIDS system. The QER panel also underscored the need for the Bank team to significantly improve ISR documentation, with a wide divergence found between the quality of the Aide Memoires and the ISRs. 9 37. The MTR team assigned a generally "satisfactory" rating to overall implementation performance and noted that several components had advanced well. It also identified potential challenges to full achievement of DOs (though maintaining a "satisfactory" rating for both DO and IO in the ensuing ISR). Among priority next steps were: (i) improvements in institutional arrangements, the clarification of discrete roles and responsibilities of all executing agencies, and the enhancement of activity coordination; (ii) full compliance with mandated financial management requirements, including timely submission of FMRs; and (iii) solid, more systematic monitoring and evaluation of all project activities, with special attention to CSO-executed subprojects which had scaled up rapidly. With improved monitoring of these steps following the MTR, project implementation progress accelerated substantially. 38. Despite the generally positive outcome of the MTR, it became evident in mid 2006 that the closing date of the project would have to be extended in order to fully expend the credit proceeds. 39. Disbursements lags over the project period were primarily due to : (i) Favorable exchange rate fluctuations over the life of the project that increased total Credit financing to almost US$43 million equivalent, an increment of 19 percent (about one year of implementation) over the US$36 million original amount based on appreciation of the SDR to dollar exchange rates from the time of Board approval. (ii) The slow process of putting in place essential conditions for effective implementation including intensive capacity building across the public sector and CSOs and the finalization of procedural manuals, financing code and forms for submission and evaluation of CSO grants and training of decentralized Council Committees in their use. (iii) The initially weak financial planning of the PES/NAC leading to a substantial overdraw of the sub-project category and the consequent stop of disbursements for this category until a reallocation was formally proposed by government and approved. (iv) Local banks' periodic freezing of funds within the project's Special Account to cover outstanding Letters of Credit, thereby obstructing payment to foreign suppliers upon delivery of procured goods. This practice, specific to Burundi, resulted in periodic, unanticipated shortfalls in project funding. Eventually, the World Bank agreed to lower thresholds for IDA direct payment of foreign suppliers, while PES/NAC improved its financial management, allowing more frequent replenishment of the Special Account. 40. In November 2006, the government requested a one year extension of the credit's closing date. The Bank team, however, advised and received a 23 month extension, which proved to be helpful in bridging unforeseen financing gaps. Indeed, the volatility of Global Fund (GF) resources had a substitution effect on government use of project resources. Specifically, when the second tranche of Round 5 of the GF was delayed, and Burundi's proposal for Round 7 was not approved, the government was compelled to allocate project proceeds away from fast disbursing activities (drugs) that could be covered by the GF remnant resources toward slow disbursing operational costs that were supposed to be covered by the GF. While IDA's flexibility ensured the PES/NAC and the national program continuous operation, as recognized by GF reports, it had a deleterious effect on the pace of disbursement. 41. During implementation two other problems arose, the risks of which were not foreseen. First, periodic shortages in the initial project years of antiretroviral medicines and reagents threatened discontinuity of care for PLWHA and reduced testing capacity. The government attributed these disruptions to long and often arduous procurement procedures. A March 8, 2005 DCA amendment permitting the government to use Bank simplified procurement methods alleviated this problem. Second, significant geographic and gender disparities in access to 10 key HIV/AIDS interventions emerged. Uneven geographic distribution of HIV/AIDS service delivery sites, due to both capacity and security constraints, left some provinces significantly under- if not un-served. M&E system data on beneficiaries of youth centers and school-based Stop AIDS Clubs-- key National AIDS Program modalities to reach this key target group--revealed that girls 15 years of age and younger were substantially underserved--a group at substantial risk of HIV/AIDS because of their relatively high rates of sexual activity, as highlighted in the KPIs in Table 1. In future, activities will be considerably more targeted toward at-risk populations. 2.3. Monitoring and Evaluation (M&E) Design, Implementation and Utilization 42. The project invested substantially in building the capacity of the new National AIDS Monitoring and Evaluation System. 43. M&E Design: The M&E system for the multisectoral HIV/AIDS Control Program was designed, and established with project funds and IDA's technical assistance. Financial support included financing of: (i) salaries of M&E staff; (ii) training of the staff; (iii) technical assistance; and (iv) decentralization of the M&E system, with a multisectoral M&E team established in 2006 in each of the 17 provinces to enhance local data collection and analysis. 44. The indicators identified to monitor progress toward PDO were adequate, given the knowledge available at the time of appraisal. The seventh indicator ­ regarding orphans -- in the DCA referred to access of services for orphans at the community level, while the seventh indicator in the PAD referred to the placement of orphans into families. The NASP' s arrangements for results-monitoring included the latter, plus another indicator on access to psychosocial support. The base-line data given in the DCA was collected during a period of conflict and proved to be both outdated and not fully representative of the overall country situation. This was first made clear in 2002 when the government released data from the 2001 National HIV/AIDS Survey, and subsequently as data from new surveys of greater extent became available. 45. M&E Implementation: The challenges of establishing a national AIDS M&E system which would respond in a reliable and timely manner to the demanding HIV/AIDS Program were not fully recognized in the project design. The quantity and quality of data improved over the life of the project: initially, as the M&E system was being designed and set-up, data collection was spotty; following the mid-term review, more attention was given to the system and improvements were made. Nonetheless, to date, although project data were gathered using adequate collection methods, no data audit has been carried out to verify quality. There were data inconsistencies while gathering data for this report. Among key challenges faced were: (i) The lack of previous M&E system experience in the country to build upon, and hence weak institutional capacity in M&E to support the new system (ii) Increasing demands on the M&E team, over time, by the Bank and other Program partners to expand the number of indicators monitored and the sophistication of analysis (iii) The need for systematic compilation and coordination of data across diverse administrative levels and key entities, e.g. EPISTAT, the Sector Unit of the MOH, and the PES/NAC's own teams managing the diverse MAOP components. Though in theory one national integrated M&E system existed, in reality, parallel collection points and even duplication of data frequently occurred (iv) Use of a plethora of data collection instruments, complicating data entry and management (v) Incomplete reporting of requisite data, including reports from many MOH health centers and hospitals on STDs and OI treatment 11 (vi) Inadequate base-line data which impedes assessing the impact of specific program interventions (vii) Insufficient transfer of skills by the external technical expert to fully prepare the M&E team to assume M&E system responsibilities when that technical assistance ended. 46. Serious M&E problems were flagged in Bank supervision reports throughout much of the project life. The October 2004 supervision mission, noting significant shortfalls at that point in the M&E system -- a full two years into implementation--reached agreement with the Council to streamline data collection, simplify data entry and, notably, "recruit full-time M&E staff to coordinate M&E activities." Nevertheless, more than a year later, the December 2005 supervision mission underscored continuing significant lags in development of the M&E system that "threatened timely, reliable measurement of the final project outcome." 47. The quality and timeliness of data reporting was enhanced over time by training partner organizations in data management and analysis, and introduction of more standardized data collection tools. The resultant far more systematic reporting enabled the PES/NAC to prepare and widely disseminate quarterly M&E reports. Better mobilization and use of GAMET (Global AIDS Monitoring and Evaluation Team) support during supervision missions, and additional technical assistance from ASAP (AIDS Strategy & Action Plan, a joint UNAIDS /World Bank program) in the latter years of the project, also helped strengthen overall M&E capacity. 48. M&E Utilization: Data from the M&E system have been used for strategy definition, policy orientation and resource mobilization. Specifically, results from the M&E system shaped the 2007-2011 strategy as follows: (i) preventive strategies focused on at risks groups; (ii) prioritizing prevention of mother to child transmission efforts; (iii) expanding orphan efforts to the whole territory; (iv) improving ARV needs forecasting; (vi) preparing the different rounds of the Global Fund documents and the IDA second MAP. Other uses of data generated by the M&E system were: (i) preparation of regular, six month Program status reports shared with the Bank and other Program partners and used extensively by Bank supervision missions to monitor project progress; and (iii) preparation of proposals and definition of strategic directions and annual planning for other external support to the Program, including for several rounds of the Global Fund. 49. However, the system was not effectively utilized as a day-to-day management tool for the National AIDS Program. For instance it is only in the last year of the project that data were used to actively redirect investments to underserved areas and hot spots. 50. In conclusion, although designing and implementing the M&E system represents a significant, solid project achievement, there are still weaknesses fragile in select areas. Major challenges ahead are: (i) to refine the system, as needed to improve the timeliness and quality and reliability of routine inputs; and (ii) further enhance team capacity to manage the large volume of data compiled and translate data into relevant reports, policy analyses, etc. to inform and guide Program management decision making. The issues facing the M&E system and the consequent measures needed to improve it will form an integral part of the institutional strengthening component of the Second HIV/AIDS Project. 2.4. Safeguard and Fiduciary Compliance 51. Environment: The project was originally assigned a "B" category. Accordingly Provision 5.01 of the DCA stipulated that the Borrower prepare, within three months of project effectiveness, a medical waste management plan (MWMP). Thereafter, however government encountered difficulties in implementing the MWMP and ensuring the collection, disposal and management of medical waste, which posed some health and environmental risks. Lack of progress on this front was not actually flagged to management until some three years into project implementation. Under increased monitoring, by 2007 selected health staff had 12 been trained on appropriate disposal of medical waste, and guidelines and norms had been developed by MOH for handling waste by health facilities. 52. Subsequently, the MWMP was revised and updated. It was publicly announced in February 2008 and approved by the government in March 2008. Its implementation is to go into effect within the Second HIV/AIDS Project. Recently, the MOH and PES/NEC jointly identified priority medical waste management activities and committed to the quick implementation of the 2009 Plan of Action to fulfill this outstanding Covenant. Overall, the difficulties in executing the MWMP stemmed more from post-conflict capacity limitations than from lack of commitment by the Borrower. During the MOAP, the country had no national environment strategy. Environmental capacity is often weak in post-conflict settings, and future projects should take into account the unique limitations such countries face. Procurement, Disbursement and Financial Management (FM) 53. Procurement: Project-supported procurement proceeded smoothly overall, with a total of 116 tenders over the project life, all successfully concluded within budget. Disaggregating total procurement by category, almost 90 percent of the project's $13.5 million was spent on equipment and supplies. 54. Procurement was facilitated by: (i) recruitment and retention, from preparation through the end of the project, of a two-person procurement team well versed in Bank procedures from prior work on a Bank- supported health project; and (ii) continuous technical support from the procurement specialist at the Bank country office. Intensive capacity building across the diverse line ministries, CSOs and community organizations involved in pharmaceutical and medical supplies procurement by the PES/NAC's Procurement Specialist helped to improve procurement processes nationally, one of the most important project legacies. In 2007 delayed communications between the Technical and Finance departments of the NAC led to a temporary overestimation of available funds for project implementation. Examples included (i) delayed equipment supply and training of personnel for the implementation of the MWMP; and (ii) delayed equipment and furniture supply for the National Center for Blood Transfusion. These issues were quickly remedied with Bank support during the February 2007 supervision mission and thereafter. 55. Financial Management. Intensive Bank training of the PES/NAC's financial management team (intended to reduce the high financial management risk generally associated with Bank-financed MAP projects) had an important multiplier effect: the PES/NAC team conducted extensive training of executing agencies of the diverse project components. 56. A major financial management problem was the PES/NAC financial team's failure to submit the four DCA mandated quarterly financial management reports (FMRs) for 2004 or for the first quarter of 2005. This reporting failure, coupled with difficulties of the government to secure counterpart financing over this period and inadequate supervision (by the PES/NAC) of the accounting function, as reported in the ISRs, led to Bank downgrading of financial management in the May 2005 ISR to "moderately unsatisfactory." Notably, this was the first ISR to assign less than a "satisfactory" rating to financial management and the first to alert Bank management about the several missing FMRs (all of which were finally received in June 2005). An Action Plan for improved financial management was agreed with the government as part of the Mid-Term review and from that point, close monitoring of financial management is evident in all ensuing ISRs. Though all outstanding FMRs had been issued by the end of June 2005, from March 2006 until June 2008 the financial management rating was upgraded to, but sustained at "moderately satisfactory." Over this period, stronger linkages were established between financial disbursements and actual activities, and communications were improved between the PES/NAC's financial management and procurement teams to ensure adequate funds were available to support planned procurement. With these improvements, in June 2008, albeit less than six months prior to project closing, 13 the rating for financial management was deemed "satisfactory." It is noteworthy that satisfactory, unqualified audit reports were received each year. 57. Examining disbursement flows, Bank and project executing agency data provide quite distinct patterns. As noted in the ICR Basic Data Sheet, actual Bank disbursements over the project period closely followed projections. While the graphs in the Bank system are based on cumulative disbursement flows, those prepared by the PES/NAC are based on actual disbursement flows per year as shown in Annex 1(e). These more closely mirror the early implementation progress (with a relatively slow rate of disbursements over the initial few years), a peak disbursement rate in 2005 (particularly after the Bank assumed 100 percent of financing), and a gradual decline in 2006 with lower but steady disbursements thereafter. 2.5. Post-completion Operation/Next Phase 58. The need to sustain the significant HIV/AIDS Program momentum created by the first project and address newly emerging needs is well understood. The country's continued heavy reliance on external support to meet HIV/AIDS Program needs is also fully recognized. Continued Bank involvement is viewed as crucial to consolidate and scale up effective interventions under the first project and to fill continuing serious financing gaps confronting the National HIV/AIDS 2007-2011. IDA is the only source willing to fund select, priority prevention and treatment activities delineated in that Plan and to reinforce multisectoral activities. With the Bank, the government wants to benefit from the comparative advantages of the public sector and civil society gains to progressively mainstream HIV/AIDS activities in key ministries. It wants to further decentralize implementation in order to maximize the Program's longer term sustainability. To ensure the requisite financial bridge for these priority actions through 2009, IDA financing was made available through a Second HIV/AIDS Project (preparation of which began in December 2007). This three-year project in the amount of SDR 9.4 million (US$15 million equivalent) was financed by IDA-14 funds was approved by the Board on May 13, 2008, had its DCA signed on July 16, 2008, and was declared effective October 9, 2008, before the first Project closed. 59. Building on lessons learned in the MAOP and other MAP projects, the Second Project adopts a sectoral approach, financing a slice of the overall costs of the 2007-2011 National HIV/AIDS Strategy. Through its four components, the project will support: (i) reduction of transmission among high risk groups through carefully tailored IEC specifically programs promoting behavioral change and increasing access to preventive services including male circumcision; (ii) implementation of the MWMP rolled over from MAOP; (iii) expansion nationwide of PLWHA access to ARV and OI treatment, with services to be provided via performance-based contracts signed between the MOH and public and private health facilities; (iv) provision of small grants for decentralized subprojects to provide monetary and other support to high risk groups and HIV-affected families, including support to protect rights of HIV/AIDS orphans and widowed women; (v) strengthening the capacity of key ministries and local authorities to implement and monitor HIV/AIDS control activities; and (vi) operational costs for managing the National Program including improving the M&E system. 3 Assessment of Outcomes 60. A preliminary assessment of project performance was shared with the government in an April 2009 Aide Memoire, based on initial findings from a March 23-29, 2009, mission to Burundi. The paucity of data available during that mission precluded drawing firm conclusions on project performance. The ensuing sections take into account government comments on that preliminary assessment as well as substantial project data subsequently provided by PES/NAC. 3.1 Relevance of Objectives, Design and Implementation 14 61. The relevance of the project's development objectives and implementation is rated "satisfactory." The PDOs are as relevant now as during MAOP preparation. A MAOP-supported national sero- prevalence survey in 2007 found select subgroups of the population were at particularly high risk of HIV/AIDS, with a 4.46 percent rate among displaced persons and 38 percent prevalence found among female sex workers. Greater mobility of the population as political stability deepens poses a real threat of an increase in prevalence rates unless intensified prevention and control activities are undertaken. This could lead to an increase in the number of HIV/AIDS orphans, currently estimated to be about 25 percent of the some 800,000 orphans and vulnerable children nationally, or about 200,000 children. 62. Despite strong government commitment to attain growth with equity, Burundi's still fragile national reconstruction process and slow economic growth seriously constrain social development. It remains one of the poorest countries in the region as revealed by such key measures as its US$110 per capita annual income and a maternal mortality rate of 1100 per 100,000 live births, among the highest in the Region. 63. Combating the HIV/AIDS epidemic was designated as one of four strategic pillars of the 2006 Poverty Reduction Strategy Paper (PRSP) presented to the IDA and IMF Boards in March 2007. Furthermore, the attainment of the other three pillars depend largely on slowing the epidemic and mitigating its effects. The PDOs remain consistent with the overall Bank Country Assistance Strategy approved in 2008, and in specific with its objectives to improve access to social services and consolidate social stability. The PDO also is consonant with two key Bank sector strategies -- the 2007-2011 HIV/AIDS Agenda for Action and the 2007 HNP strategy. 64. HIV/AIDS prevalence rates continue to be high in rural areas, where 90 percent of Burundi's population resides. Its impact will continue to exert a powerful downward drag on development and exact high associated economic, social, and political costs, significantly threatening progress toward the Millennium Development Goals (MDGs). 3.2 Achievement of Project Development Objectives 65. Achievement of the project's two primary PDOs (slowing the spread of HIV/AIDS and mitigating its effects) is rated "Moderately Satisfactory." Projected causal linkages between key activities financed by the project, output indicators and anticipated outcomes were reasonable. There is some evidence that the project helped reduce or stabilize HIV transmission, and its positive impact on the education of orphans (i.e. mitigation of risks) was substantial, given that they had higher enrollment rates than their non-orphaned peers. There appears to have been large improvement, from a low base, in the number of people receiving VCT and ARV treatment. Given questions about the data, a marginally satisfactory rating has been given. 66. Attribution: Given that IDA funds accounted for between 80 and 90 percent of the national program resources for HIV/AIDS activities until 2006 and for about 60 percent in 2007 (see graph 1.1 in Annex 1) much of the progress made during the project period can be attributed to project activities. HIV prevalence appears to have stabilized or to have declined. 67. Survey data suggests that the prevalence of HIV/AIDS may be at least stabilizing if not declining among the general population in Burundi, with a slight decrease in prevalence observed among youth 15- 24 years old (from 3.5% in 2001 to 3.1% in 2007)--a key target group for the National AIDS Program and the project. HIV prevalence among pregnant women receiving prenatal care at sentinel sites also 15 seemed to decline between 2002 and 2007 (see Table 11.1 in Annex 11). It should be noted, however, that when ARV treatment began being provided, international health authorities agreed that HIV prevalence in the general population was no longer a viable measure of HIV transmission (although if prevalence goes down in the face of increasing access to effective ARV treatment which would extend the lives of PLWHA, then incidence must be declining). Slowing the spread of HIV/AIDS 68. Public sector commitment to increasing HIV/AIDS awareness among staff and constituents. Progress was made in developing a multisectoral approach towards educating the entire population about HIV/AIDS. By project end all 30 central level public sector institutions had instituted annual HIV/AIDS Action Plans (up from 8 out of 20 in 2002). Through these plans, heavily focused on IEC, about 308,065 civil servants, their families, and sector constituents, were reached, not only raising their awareness of the epidemic but also of behavioral changes needed to protect against infection (see Annexes 2 and 11). 69. Improved safety and management of the nation's blood supply. Over the project period, the prevalence of HIV among those donating blood fell from 0.50 percent to 0.36percent although there are possible issues with the quality of the data (see Table 11.3 in Annex 11). 70. Behavioral change has occurred among some key groups. The percentage of adults using a condom during contact with a non-regular sexual partner rose from 12.7 percent in 2001 to 19.8 percent in 2007 (National Behavioral Survey data). Over the same period, the percentage of girls 15 and under sexually active decreased from 32.5 percent to 21.5 percent. Nonetheless, survey data also shows that by the project's end, the risk of HIV infection of these same populations was still quite high (see Core KPI table, p. 2) compared to other countries in the region. 71. Increases in access to voluntary counseling and testing and preventive services. Over the course of the project, the number of functioning VCT sites increased four-fold. Almost 3000 persons were trained in counseling and testing, increasing the number of informed persons seven-fold, and disaggregated data by age and gender are available (see Annex 2 and Annex 11). By project end, 15-17 percent of Burundi's population may have been tested (depending on how often individuals had themselves re-tested). 72. Youth-oriented HIV/AIDS education scaled up significantly. The 4410 young people trained under the project to serve as peer educators (over 30 for each of the 129 communes) has created a strong cadre with substantial promise of reaching successive cohorts of youth with crucial information to impact their behavior and reduce the risk of infection. 131 youth centers were created (covering all 129 communes) and 972 Stop SIDA Clubs were started nationwide. 73. Condom distribution. Consistent with the project's approach of reaching the general population with key preventive activities, over 52 million condoms were distributed over the life of the project (Table 6, Annex 11). Mitigating the adverse effects of HIV/AIDS. 74. Significant expansion of PMCT sites to serve both mothers and newborns. The number of PMCT sites increased from 1 in 2002 to 63 in 2008, contributing directly to a ten-fold increase in the number of pregnant sero-positive women treated over this same period. (Annex 2). Despite this major scale up of effort, unmet needs for PMCT are significant, with less than 8 percent of HIV infected pregnant women covered by the close of the project (see table 11.5, Annex 11). 16 75. Similarly, 1,299 newborns of sero-positive mothers were treated in 2008 compared to only 274 in 2002 (Annex 2), but coverage remained far below needs (see Table 11.6, Annex 11). 76. Rapid expansion of ARV and OI treatment capacity: The number of PLWHA receiving ARV treatment increased from 600 in 2002 to 14,343 in 2008. 68 centers for ARV treatment were established, ARV treatment providers were trained, home-based care systems were developed and drugs and medical supplies were provided. The 43.7 percent ARV coverage rate attained in 2007 substantially exceeded the average 30 percent for the Sub-Saharan Africa region, although such comparisons are complicated. (see Table 11.7 and 11-8, Annex 11). 77. Substantial improvements in well-being of orphans: The project contributed to a major expansion of access by orphans and vulnerable children to education through 2006 (when school support was assumed by UNICEF), medical and psychosocial support (see Table 11.9, Annex 11). 78. Project inputs generated numerous positive outcomes for both orphans and their families. Primary school enrollment increased, while rates of repetition and dropout decreased. Between school year 2002-3 and 2004-2005 the number of orphans receiving school assistance went from 5245 to 19,565, i.e. virtually the totality of the estimated 19,800 orphans of primary school age in the three provinces at the time. Enrollment, rates of promotion, and non-drop out rates among assisted orphans were clearly better than those of their non-orphan peers in the three pilot provinces. Moreover, although the overall numbers of orphans enrolled at the secondary school level was significantly lower than those attending primary school, major advances also were achieved in this key age group: the number of orphans assisted increased from 176 in the 2002-3 academic year to 1655 in the 2004-5 academic year, and their rate of promotion increased from 60.2 percent in the 2002-2003 school year to 83.4 percent in the 2004-5 school year. 79. These advances contrast sharply with the findings of a 2003 social analysis in the pilot area indicating that only 43.2 percent of orphans of school age were enrolled in school; that dropouts were common due to lack of funds to pay school fees and other educational needs; low motivation to continue schooling; halting of schooling, particularly by girls, to carry out domestic chores including caretaking of brothers and sisters; interruption of schooling during the national conflict coupled with low reintegration in schools by those older orphans; and non-enrollment of orphans because the support families, with limited resources, favored enrollment of their own children. 80. Other important outcomes that the 2006 orphans component evaluation revealed included: attainment of new vocational/technical competencies among older orphans attending project-supported workshops; improvements in the living conditions of orphans as a result of project supported economic and food supports to the families in which the orphans resided; and increased identification and resolution of cases of orphan exploitation and abuse, as highlighted in Annex 10. 81. In summary, the above indicators reflect substantial progress made towards achievement of the PDOs over a relatively short time period and under difficult country circumstances. These indicators may be better gauges of project performance than the original KPIs because: (i) two of the original baseline values presented in the DCA were incorrect (see para. 6), thus underestimating the magnitude of improvement expected in these indicators by the end of the project (the baseline for percentage of adults using condoms with non-regular partners actually was 12.7 percent, not 43 percent; and the percentage of girls 15 and younger who were sexually active was 32.5 percent, not 18 percent); (ii) two KPIs related to new OVC initiatives (percentage of orphans advancing from one grade of primary school to the next and percentage receiving psychosocial support within their communities) for which there were no baseline data and hence no solid basis for setting targets; and (iii) the low starting point from which the MAOP embarked (post conflict and inexistent HIV/AIDS infrastructure). In aggregate, these factors strongly 17 suggest that many KPI targets were unrealistic. Based solely on DCA KPIs, one of the four outcome targets, two of the three output targets and one of the three process targets were fully met; one process target was partially met (see Table 1). 3.3 Efficiency 82. Project efficiency is rated as moderately satisfactory using criteria applied in other MAP projects. 83. Slow start-up of implementation in first few years. Because the public sector was weaker than expected at project outset, considerable effort had to be expended over the first few years to organize and train key executing agencies at all levels. Secondly, the fast track preparation process did not allow sufficient time to prepare the ground for project execution prior to project launch. Preparation activities -- such as design refinement, evaluative instrument testing and fine-tuning, and development of financing and monitoring and evaluating systems -- had to be finalized during the implementation period. This slowed the pace of overall project implementation and contributed to disbursement lags in the first years. 84. Phased capacity-building of public sector institutions. The project pursued a three-phased strategy to assist PSOs in developing HIV/AIDS Action Plans: (i) organization of training workshops on elaboration of plans with representatives from the HIV/AIDS sector units within each Ministry; (ii) mobilization of resource persons to assist these sectoral units, when needed, to finalize their plans; and (iii) validation of the sectoral plans through consultations with sectoral staff and their beneficiaries. This systematic approach resulted in final Action Plans for all 30 PSOs by project end compared to 8 at the project outset. 85. Adoption of a generalized rather than targeted approach. Based on project results, it was found that the substantial project resources spent for mass media-delivered IEC, mass condom distribution and promotion of the wide utilization of VCT was not the most efficient use of funds. A 2007 socio-behavioral study showed that the HIV/AIDS radio transmissions reaching out to an estimated 4 million Burundians, actually had limited impact. Taking these lessons into account, the Second HIV/AIDS Project will adopt a far more targeted approach, delivering interventions that have been proven to respond to risks among the most vulnerable groups. 86. Within the public sector component, the effort expended by the PES/NAC to engage all line Ministries in the HIV/AIDS Program was inefficient. Many of the ministries were not committed to the program despite NAC/PES staff time and effort. While seven Ministries incorporated HIV/AIDS plans and budgetary resources into their sectoral programs, many Ministries committed only part-time staff to their HIV/AIDS sectoral unit and appear to have limited their HIV/AIDS actions to IEC. The Second HIV/AIDS Project has adopted a more strategic, targeted approach to engagement of line Ministries, with identification of lead Ministries to promote and facilitate actions by others within their cluster. 87. Efficient Management of CSO subprojects. Management of the CSO subprojects was efficient, and since almost one-half of total project disbursements went into this component, it translates into project efficiency. The evaluation of financed subprojects consistently revealed high payoffs from invested funds in the first few years of project implementation. A 2007 financial audit of 66 subprojects revealed overall satisfactory performance (and led to the exclusion of some grant recipients who failed to meet performance standards). A 2008 evaluation of subprojects also indicated a strong performance. A recent independent audit of the 564 subproject grants funded in the course of the project revealed only 12 poorly performing subprojects. Those 12 projects represented just 2.3 percent of the total number of subprojects, and 1.6 percent of the total financing disbursed under this category. 18 88. Underlying these impressive results were: (i) careful development of the financing code guiding grant applications that required extensive analysis of subproject plans, expected outcomes and indicators and strict application of subproject selection criteria by the PES/NAC and its provincial and communal Committees; (ii) circumscription of CSO overheads to 10 percent of subproject costs; (iii) substantial accountability, with systematic internal and external audits to assess performance and compliance with contracts and elimination of CSOs not performing at acceptable standards from eligibility for future grants; and (iv) transparency of the process, with detailed information on all grants disseminated publicly. 89. Performance-based contracts. Another efficient change took place in late 2006, when the project introduced performance-based contracts (PBCs) for grants provided to CSOs delivering regular, continuous medical services, while maintaining standard grant agreements for CSOs carrying out local initiatives of a defined duration. Use of PBCs led to major efficiencies in grant management due to: (i) clarity regarding beneficiaries and expected results from the outset which facilitated performance monitoring and the release of grant tranches; (ii) the reduced need for PES/NAC support to CSOs; (iii) direct flow of funds to provincial heads of CSOs and more rapid disbursement of subproject financing; (iv) reduced cost per beneficiary, as all costs not directly linked to the service to be delivered were eliminated from financing; (v) cost efficiencies through the sharing of experiences across CSOs; and (vi) an increased sense among CSOs of partnership with the PES/NAC through greater collaboration and interaction in the course of implementation. 90. Approval and oversight of both PBCs and standard CSO grants were highly decentralized. The respective NAC Provincial Committees played a major role ensuring satisfactory technical execution, appropriate accounting, and timely and effective monitoring and reporting. 91. Finally, efficient use of project funds lay in the wide use of relatively low-cost, community-based, often innovative models to deliver priority HIV/AIDS education and services. Peer education among youth, community educators or animateurs, and community home care providers to ensure continuity of ARV and OI patients are among these services. The latter also reduced the need for hospitalization and additional health system costs. 92. One notable, but controversial inefficiency in the MAOP was the decision by the PES/NAC to limit CSO contracts to just one year from the beginning until the of the project closing. While this facilitated high quality control over contracts and reduced breaks in core medical services from year to year if the CSOs planned carefully and ensured that successive annual agreements were in place before the current one ended, ICR mission discussions with CSO representatives in-country indicate this modus operandi made it difficult for them to reliably project their annual funding streams and hence to plan efficiently. Since many CSOs participating in the project were quite small with limited alternative sources of financing, unpredictability of MAOP funding from year-to-year threatened possible discontinuity of the services they were delivering, including, for example, ARV treatment. It also raised overall transaction costs in terms of staff time of CSOs to prepare annual proposals to mobilize continued funding and of the PES/NAC and its decentralized Committees to review and approve them. 3.4 Justification of Overall Outcome Rating Rating: Moderately Satisfactory 93. Overall, the PDOs remain highly relevant and their attainment essential to accelerating Burundi's economic and social development. They are consonant with the current Bank country assistance strategy. Although end of project coverage levels of priority services appear low, the project directly contributed to the government's establishment of preconditions to carry out an expanded, multisectoral, coordinated and sustained 19 attack on the HIV/AIDS epidemic. The project worked with the government to develop a strong, decentralized, institutional framework, a significantly expanded infrastructure, a body of skilled professionals, and additional external and national finances to advance the fight against AIDS, as discussed throughout this ICR and its Annexes. The project's emphasis on community-led and community-based models of delivery of HIV/AIDS education and preventive and curative-oriented services, represents an efficient use of national resources. Independent evaluations of the CSO subprojects, the largest single expenditure category, consistently revealed high quality projects. All things considered, the Project outcome was deemed moderately satisfactory. 3.5 Overarching Themes, Other Outcomes and Impacts Poverty Impacts, Gender Aspects, and Social Development 94. The project revealed that: (i) PLWHA should have a strong role in HIV/AIDS Program design, execution, monitoring and evaluation and more broadly in HIV/AIDS policy making. In this project PLWHA were represented in all key aspects of the National Program, from key PES/NAC entities at the central and decentralized levels to direct participation in HIV/AIDS education and promotion of prevention and control activities. (ii) Women (especially widows, elderly women and young female orphans) have a strong crucial contribution to make in the provision of support to orphaned children and youth. (iii) Youth bring substantial value added to all aspects of AIDS programs, particularly as as advisors and peer educators. Two important innovations of the MAOP was the national AIDS festival in which 820 young artists presented works on AIDS themes and a concert that drew 800 young people. (iv) Child labor is more common among children who have lost both parents and must work to maintain themselves and their siblings. (v) Females -- particularly young girls -- are likely to be under served if close monitoring of the gender balance among beneficiaries of HIV/AIDS interventions is not enforced. They should be well represented in stakeholder consultations and advisory capacities. (vi) Community-based Child Protection Committees are invaluable in monitoring and reporting on child abuse, including violations of widows' and children's property and inheritance rights. These committees were known to exert collective political pressure, where needed, to review existing laws and regulations and to formulate new national policies to protect these vulnerable groups. Institutional Change/Strengthening 95. From the perspective of long-term investment impact and institutional development, three major achievements of MAOP should be noted. First, the project directly contributed to the transformation of a nascent National AIDS Council into a mature, strong multi-sectoral, autonomous, highly decentralized entity to guide and coordinate National Program activities. The PES/NAC has now become a well-reputed institution, staffed by highly qualified professionals with a strong commitment to results. Initially PES/NAC played a major catalytic role in resource mobilization for the national program, attracting substantial funding for priority HIV/AIDS activities from other national and external sources. Presently, the GF is in the process of negotiating with the government a new proposal for which Burundi has requested US$140 million, with allocations to be distributed between the public sector and civil society. The successful leveraging of MAOP resources under the first HIV/AIDS Project, with an extensive group of national and international partners mobilized to support the National Program, as highlighted in Annex 1 (c) augurs well for the National Program's ability to respond to the still largely unmet HIV/AID education and service needs nationwide. 96. Second, MAOP instilled a new culture of strategic planning, monitoring and evaluation across the public and private sectors, including CSOs. Further, MAOP significantly enhanced transparency across the National 20 AIDS Program. This transparency has helped create substantial, heretofore somewhat limited, trust between the public sector and CSOs. In this sense, it contributed directly to good governance objectives of the country's poverty reduction strategy. 97. Third, MAOP catalyzed a major decentralization of the National AIDS Program. This process extended far beyond establishing key institutional structures such as the provincial and communal committees with their broad representation. Actual program implementation was mainstreamed to these decentralized levels. Via the PES/NAC, project funds were directly transferred to these decentralized entities. They in turn mobilized CSOs and communities to participate in the Program, reviewed and approved local proposals for grant financing, channeled project funding to CSOs for executing approved grants, and monitored performance of those subprojects. The project thus led to a significant devolution of power to local authorities, setting a positive precedent which will, hopefully, be continued. 98. Strengthened CSO capacity in planning, management, monitoring and evaluation of HIV/AIDS projects had enormous spillover effects. The institutionalization of this process is evident in several major CSOs that participated in MAOP such as Réseau Burundais des Personnes Vivant Avec Le VIH (RPB+) which now has its own multi-year strategic plan, annual work plans and monitoring indicators. 99. Access to subproject grant financing as well as considerable technical assistance for capacity building, proposal preparation and project implementation and evaluation itself promoted a substantial expansion of the CSO presence nationally. Whereas only some 35 CSOs were working in HIV/AIDS in the late 1990s, 530 CSOs were active nationally by project end. CSO activities complemented public sector actions, filling many key gaps in public sector capacity to deliver core services, even in geographic areas still in conflict and inaccessible to government entry. Other Unintended Outcomes and Impacts (positive or negative) 100. The highly decentralized PES/NAC structure, with high levels of community participation, strongly promoted experimentation and innovation. CSOs proved particularly creative and well poised to test new models such as use of community promoters or animateurs for IEC, with a network of over 6000 animateurs in place across the nation by project end. CSOs also introduced new community-based HIV/AIDS service delivery approaches and filled crucial service gaps, such as conduct of psychosocial counseling, a priority need to which the public sector could not adequately respond because of its own critical shortages of psychologists and social workers. 101. Concomitantly, CSOs' sense of real ownership and stake in the outcome of the National AIDS Program was significantly expanded. For the first time, they gained an important voice in the Program. Their influence extended beyond their role in service delivery to their direct representation in key positions including the National Council and its Coordinating Committee, Permanent Executive Committee and decentralized provincial and communal committees. 102. Finally, MAOP helped identify, through its M&E system and a special follow-on study, the challenges the Program faces in reaching young girls--a key target group if the epidemic is to be slowed-- through youth centers and through their recruitment for participation in youth-oriented activities. Information from the study will help develop new girl-friendly prevention and control activities under the Second HIV/AIDS Project. 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops 103. Not applicable 21 4. Assessment of Risk to Development Outcome Rating: Significant 104. Substantial political, security, and economic uncertainties in a nation only recently emerging from conflict and among the poorest in the world, makes the risks to National AIDS Program sustainability significant, despite major program advances under MAOP and strong commitment from the government. Combating HIV/AIDS is one of four strategic pillars of the country's Poverty Reduction Strategy Paper, and a well articulated 2007-2011 NASP is in place to guide policy and investment decision-making. MAOP's inclusive and participatory approach has integrated key stakeholder groups, particularly PLWHA, women, and youth, in all aspects of the program. Nonetheless, the country's heavy dependence on external aid to advance its development agenda makes it highly vulnerable amid the current global economic downturn. 105. Institutionally, the now well-experienced PES/NAC, along with a dynamic public-private partnership, are major assets upon which to build. However, the roles and responsibilities of three major "lead" entities--the PES/NAC, the Ministry of Health, and the Ministry of HIV/AIDS -- have not yet been clearly determined. (The Ministry at the President's Office in charge of HIV&AIDS was established from 2001 to 2007; in November 2007, a Vice-Ministry in charge of AIDS was created within the Ministry of Health and HIV/AIDS; in January 2009 the Ministry of HIV/AIDS and the Ministry of Health were recreated). Other unclear areas are: (i) the pace at which the PES/NAC will now turn their attention from implementation responsibilities to a coordinating one; to do this, they will have to work with other public and private entities, building up their capacity to take over implementation; (ii) whether the MOH has fully acquired the expertise and absorptive capacity to fulfill the broad technical leadership role the HIV/AIDS Program requires while, at the same time, meet its broader health sector responsibilities; (iii) whether the delivery of priority HIV/AIDS prevention and promotion activities, can be effectively and efficiently assumed by other entities; and (iv) whether a stronger MOH presence in the HIV/AIDS Program will undermine the multisectoral achievements of the first project. 106. The continuing large HIV/AIDS Program financing gap -- presently on the order of US$ 104 million -- obviously puts sustainability at risk. The 2002-2006 NASP secured only 30 percent of the financing required to support planned National Program activities, even though successfully leveraging Bank resources to attract many other financiers, including the Global Fund (see Annex 2). The PDOs must be attained, particularly in light of the 2007 national sero-prevalence survey showing that prevalence rates remain high, especially among select vulnerable groups. Bank financing for Burundi's Second HIV/AIDS project is less than one-half that provided for MAOP, and the project itself extends only until June 30, 2011. This makes it all the more imperative that the government rapidly attract additional external resources, with only a little more than two years left to accomplish the 2007-2011 goals. The significant financing shortfall has had one important, positive effect, however. It has forced prioritization of the Plan's multiple programs, active pursuit of evidence-based interventions, and better targeting of activities to highest risk groups--all of which promise more cost-effective use of scarce financial and technical resources. 5. Assessment of Bank and Borrower Performance 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Satisfactory 107. Bank performance in project design through MAOP approval was satisfactory overall. The Bank, mobilized a substantial PHRD grant to enable fast tracking of project preparation. The Bank team worked fervently and extremely closely with the government to prepare the project, moving from Concept Review to appraisal in a remarkably short time, particularly given the difficult country situation over this period. The overall project design was sound, drew well on Bank and Burundi prior project experience, and rightly identified 22 the wisdom of expanding civil society participation in the HIV/AIDS program, in close partnership with the government. 108. Early on, the Bank recognized the possibility of realizing synergies and significant reductions in administrative costs by merging an Orphans Project currently under separate preparation, with the HIV/AIDS Project. While this decision carried some calculated risks, the merger proved to have many benefits. It permitted a quick start and rapid scale-up of OVC activities by enabling the OVC program to draw upon the strong PES/NAC institutional capacity, and to utilize its M&E systems rather than having to create a new institutional framework. It facilitated mobilization of extensive CSO participation in identifying needs and delivering OVC services via the project's CSO subproject grant mechanism. Most importantly, the integration of the two projects identified OVC with HIV/AIDS -- among the most vulnerable OVC -- thus allowing the project to target specific interventions to improve their lives. (b) Quality of Supervision Rating: Moderately Satisfactory 109. Bank supervision performance was uneven over the project life. A September 8, 2006 Quality of Supervision Assessment rated overall supervision as "moderately satisfactory." 110. Four changes in TTLs between preparation and project close, created some lags in project oversight. Turnover in other Bank Headquarters staff, particularly from 2005-2006, appears to have slowed implementation in terms of delayed approvals of government requests for non objections and reallocation of Credit proceeds. 111. From project initiation through mid-term, ISRs did not always meet Bank standards for supervision reporting. It was noted, upon occasion,: (i) the failure to flag selected issues needing decisive action (e.g. missing FMRs and failure to implement the MWMP); (ii) some cursory entries from one ISR to the next of updated information within the "Summary of Issues and Actions" section, suggesting no progress had been made in addressing key implementation issues (PSR of 10/26/04), whereas Aide Memoires from corresponding ISRs revealed significant project advances; (iii) ISRs that appeared to have replicated information with no explanation of missed deadlines; and (iv) some inconsistencies and errors found in the reporting (ISR of 3/26/04). It was also noted that project performance indicators, reported within the "status of agreed outcomes" matrix of the ISR, were added, revised or deleted without adequate explanation. Over time, four core KPI ceased to be reported on in the ISR. Revisions in wording of some indicators created uncertainty in what statistics actually were being reported. 112. The quality of Aide Memoires (AMs) on which ISRs were based varied considerably over the course of the project. Most provided solid, thoughtful analyses of project status and recommendations to redress implementation problems, but some were more narrative than strategic, exceeded 30 pages in length, and included several pages of non-prioritized actions for the government to take subsequent to the mission. 113. Inadequate ISR reporting was mentioned in two separate Bank Panel Reviews. The need for "future ISRs to capture more effectively relevant implementation status points and issues highlighted in the Aide Memoires written in French and in back to office reports" was raised by the March 8, 2005 QER Panel prior to the Mid-term review. The September 2006 QSA commended the core health team for "performing very well in identifying issues and addressing them," but, the report concluded that the team was "not always successful in documenting and reporting on the measures adopted and the results achieved." Though documentation improved over the remaining project period, it did not fully attain satisfactory standards. The documentation shortfalls limited Bank management access to timely, comprehensive updates on project performance and hence effective decision making. 23 114. Following the March 2006 supervision mission, the Bank initiated a strategic planning exercise, with expert input, to assist each sector to improve its HIV/AIDS action plan. The Bank also conducted an institutional assessment to help better define capacities and comparative advantages/optimum roles of distinct organizations involved in the National Program. This information was also used in the preparation of the 2007- 2011 National Strategy. (c) Justification of Rating for Overall Bank Performance Rating: Moderately Satisfactory 115. Consistent with ICR guidelines, one rating of satisfactory and one rating of moderately satisfactory dictates that the overall rating assigned be moderately satisfactory. 5.2 Borrower Performance (a) Government Performance Rating: Moderately satisfactory 116. Throughout the project, the National AIDS Program benefitted from exceptionally strong commitment from the Office of the President, but not always accompanied by a corresponding commitment from some line Ministries. As the project neared its end, the Borrower's institutional framework for advancing the National AIDS Program became unclear when significant changes were made in Ministerial responsibilities and mandates, as mentioned earlier. These organizational changes were not expected by the HIV/AIDS program stakeholders, since previous ministerial decrees had just been issued in the summer 2007 clarifying the respective roles of the Ministry of Health, the Ministry of HIV/AIDS, and the National Council with regards to the 2007-2011 National AIDS Strategy. As reported in the 6/20/2008 ISR, the resultant uncertainties about respective roles of the various ministers and vice ministers hampered management of HIV/AIDS activities in several sectors, including health. Weighing the strong commitment at the Presidential level against the organizational uncertainties of the MOH and the reduced commitment of other line Ministries to fully commit staff and resources to the National AIDS Program results in a "moderately satisfactory" rating. (b) Implementing Agency or Agencies Performance Rating: Moderately Satisfactory 117. Overall performance of the executing agency was commendable. The PES/NAC was pro-active in identifying implementation problems and coordinating closely with the Bank team to recalibrate the project plan of action when needed. It quickly identified capacity constraints in the public and private sector and decided to scale up activities more gradually than originally planned in order to strengthen that capacity. Similarly the PES/NAC submitted, through the government, a timely request for amending the DCA when the country's macroeconomic deterioration precluded meeting counterpart funding commitments. It also requested a reallocation of Credit proceeds to respond to evolving project financing needs, including the much higher than anticipated demand for grant financing by CSOs that held much promise of accelerating progress toward the National Strategy goals. When needed, it also increased demands for funds to support consultants' services, audits and training. 118. On the down side, the PES/NAC experienced two changes of its Permanent Executive Secretary during the six years of the project. Additionally, a 10-month leadership hiatus occurred between the first and second secretaries. The acting secretary was eventually confirmed in her position. Though each permanent executive secretary selected was a highly qualified professional, the lack of continuity of leadership, compounded with the changing TTLs from the Bank, had an unsettling effect on project implementation. The relatively low level of turnover among PES/NAC staff overseeing such key areas as financial management and implementation of selected components helped to ease the transitions in PES/NAC leadership but did not fully offset their effects. 24 The fluidity in PES/NAC management may also account in part for some slippages that occurred in such crucial areas as financial management reporting and finalization and implementation of the MWMP. 119. Notwithstanding, the executing agency is to be recognized for its balanced project performance assessment, reflected in its ICR, that identifies substantial project strengths and achievements but also significant challenges and obstacles confronted in implementation, and a substantial unfinished agenda to still pursue to achieve goals of the National HIV/AIDS 2007-2011 Strategic Plan. (c) Justification of Rating for Overall Borrower Performance Rating: Moderately Satisfactory 120. Consistent with ICR Guidelines, with both government and Executing Agency performance ratings moderately satisfactory, the overall rating must be moderately satisfactory. 6. Lessons Learned 121. Many key lessons that emerged in the execution of MAOP are consistent with those found in other MAP projects by the external MAP Project evaluation and the IEG Review of the Multi-Country AIDS Program in Africa, Select Countries, 2006-2008: Lessons Learned. Those lessons were taken into account in the design of the second MAP Project. Among the most important lessons to be drawn from MAOP follow. 122. Regarding project preparation and planning: (i) Adoption of a fast track preparatory process has hidden costs that cannot be discounted. The extensive effort needed over the first few years of the project period essentially to complete preparation of core inputs required for actual project execution significantly slowed implementation and disbursements and ultimately contributed to the need to extend the MAOP closing date by almost two years. (ii) Investments made in the careful preparation of procedural manuals, a financing code, detailed application and evaluation forms for selecting CSO grants, use of performance-based contracts, training of decentralized committees, and good monitoring and evaluation systems had high returns. The PES/NAC's conduct of financial and technical audits of CSO projects imbued a strong sense of accountability for results (see section on efficiency). Circumscribing CSO grants to just one year, presents many problems, not only in terms of creating substantial uncertainties among CSOs for programmatic and financial planning purposes (and the risk of breaks in service delivery), but also in terms of generating relatively high transaction costs for both the CSO and the government and local authorities in the preparation and review of annual proposals. With the introduction in 2006 of performance based contracts for medical services delivered by CSOs, adequate protections are in place to enter into multi-year contracts while maintaining good quality and financial controls. (iii) The risks to implementation from capacity weaknesses, particularly pervasive shortages of skilled professionals, can be and indeed were in Burundi significantly underestimated during project preparation. These shortages presented major obstacles to effective, efficient project implementation. Tackling major, widespread human resource constraints may be best done cross- sectorally rather than on a sector-by-sector basis since the underlying dynamics of low morale and high turnover among civil servants tend to be common across public sector entities. 123. Regarding the design: 25 (i) The integration of the orphans' project was desirable but added complexity to project design. This component introduced additional project objectives that were not part of the National AIDS Strategic Plan or the HIV/AIDS components of the project. Meeting the OVC objectives demanded specialized expertise unique to this component to support interventions in such areas as education and vocational/technical training, income generating activities and home gardens and protection of property and inheritance rights of orphans and widows. This demanded close interaction with ministries outside the general AIDS community, including the ministries in charge of Gender and Human Rights and National Solidarity, Repatriation and Reintegration of Refugees and Internally Displaced Persons. (ii) Initial (and erroneous) base-line data suggested that the epidemic was generalized, so the project was designed accordingly (general population-oriented mass IEC, VCT, condom distribution and treatment of STDs). It was later found that HIV/AIDS in Burundi was heavily concentrated in select population groups and geographic areas. In such a case, evidence-based interventions, targeted and tailored to those at highest risk, appear to be the more cost-effective way to reduce transmission. Projects need to be sufficiently flexible to accommodate changes in information regarding the epidemic profile. 124. Regarding Monitoring and Evaluation: (i) In selecting key performance indicators to monitor project progress, extreme care must be taken to ensure that baseline data exist and are verified, and that the suggested methodology for obtaining data over time to track progress is feasible and cost-effective. Similarly, the baseline data must be adjusted quickly if new, more reliable data sources become available and the targets relating to adjusted indicators must be amended then as well to ensure that they reflect reasonable magnitudes of change, given the specific mix of project interventions and the timeframe for project execution. In the case of this project, the failure to restructure the project in order to formally update baseline values (and subsequently their targets for two core outcome indicators) adversely impacted assessment of project performance, as discussed earlier. (ii) The preparation of both the 2007-2011 National Strategy and the second MAP highlighted the importance of including ongoing surveillance, epidemiological and analytical work in any project. , Impact assessments and M&E studies are needed to effectively guide and adapt policies. Systematic sharing of M&E results with key stakeholder groups is instrumental in gaining their support for financial and strategic programs. 125. Regarding Institutional Arrangements for Implementation: (i) In pursuing a truly multisectoral strategy, balance must be maintained in distribution of responsibilities, human and financial resources, and hence power, between high level coordinating bodies like Burundi's National AIDS Council and its permanent secretariat (formerly PES/NAC; then PES/NAC), and other public agencies and local authorities whose buy-in and own concrete actions are essential to attain HIV/AIDS Program goals. Substantial risks exist that such coordinating bodies become widely perceived as "supra-Ministries," particularly given their proclivity over time, especially in the face of serious capacity constraints in the public and private sectors, to assume core Program implementation responsibilities, thus extending their role well beyond their fundamental coordination mandate. This inevitably creates friction with and resistance to collaboration from other entities participating in the Program. (ii) In the long run, every effort must be made to ensure that the Ministry of Health in particular has both the commitment and capacity to play the central role it must for successful HIV/AIDS Program 26 execution. The MOH's exercise of technical leadership is crucial to ensure quality of care and treatment, including establishment of HIV/AIDS treatments protocols and standards for ARVs, STDs and OIs. With the primary responsibility of National AIDS Program activity largely centered at the PES/NAC, some partner organizations have suggested that the MOH assumed, detrimentally, that it was largely "off the hook" in terms of its own accountability for the HIV/AIDS Program. Indeed wide experience reveals that every effort must be made over time to fully integrate HIV/AIDS care and treatment within existing health services in order to ensure institutional sustainability and rationalize use of scarce resources. 126. Regarding Project Management - High turnover of Bank TTLs presents a major obstacle for smooth project implementation. The five TTLs over the MAOP life led to serious shortfalls in project performance, monitoring, and reporting. Continuity of Bank project leadership, particularly in challenging operational environments is essential. Similarly, having two Executive Secretaries at the NAC was disruptive. In such circumstances, increased attention must also be given to identifying alternative approaches to attract and retain project managers in executing agencies. 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners (a) Borrower/implementing agencies 127. The Government has prepared a comprehensive final evaluation report which is summarized in Annex no. 7 below. (b) Cofinanciers (c) Other partners and stakeholders Not Applicable 27 Annexe 1 Project Costs and Financing (a) Project Cost by Component (in USD Million equivalent) Table 1. 1: Project Cost by Component (in USD Million equivalent) Appraisal Estimate Actual/Latest Components Percentage of (USD millions) Estimate (USD millions) Appraisal EXPANDING PUBLIC SECTOR RESPONSE TO THE CRISIS 9.00 16.55 183.89 INCREASING PARTICIPATION OF GRASS-ROOT ORGANIZATIONS AND 13.00 8.19 63.00 COMMUNITIES ORPHANS INTEGRATION & EDUCATION PILOT 7.00 6.16 88.00 STRENGTHENING THE PROGRAM STRATEGIC FORMULATION AND 4.00 5.93 148.25 MONITORING CAPACITY PROJECT COORDINATION 3.00 6.16 205.33 Total Baseline Cost 36.00 42.99 119.42 Physical Contingencies 0.00 0.00 0.00 Price Contingencies 0.00 0.00 0.00 Total Project Costs 0.00 42.99 Front-end fee PPF 0.00 0.00 .00 Front-end fee IBRD 0.00 0.00 .00 Total Financing Required 0.00 NOTE: This table significantly overstates the relative share of project financing actually supporting the public sector response and significantly understates the importance of the civil society organizations in project implementation because financing for the component "participation of grass root organizations and communities" actually was disbursed against all five project components, as highlighted in the table on MAOP final expenditures by component and category presented in Annex 1 (d). 28 (b) Financing Table 1. 2: Source of Funds in Million USD Appraisal Actual/Latest Source of Funds Type of Co- Percentage of financing Estimate Estimate (USD millions)(USD millions) Appraisal Borrower 0.70 0.00 .00 International Development Association (IDA) 36.00 42.99 119.42 Note: significant increases in total project financing reflect major improvement in the SDR to dollar exchange rate over the project life. (c) National AIDS Program Financing by Source of Funds 1. Full implementation of the 16 programs within the five-year National AIDS Strategic Plan 2002- 2006 was estimated to cost US$233 million. The PES/NAC played a major catalytic role in resource mobilization for the National AIDS Program. The recognized strong managerial and technical capacity of its team was a strong magnet to attract financing from both national and external sources. Nationally, these resources included cost-sharing of CSOs that were recipients of project-financed grants. A multiplicity of other financiers ultimately participated in that Program, providing in aggregate 3.73 billion BIF over the project life. Of special note are contributions from the GF which provided US$8.657 million in complementary support for the period April 2003 to March 2006, and US$32 million for a five-year period commencing April 2006. 2. Notwithstanding this strong resource mobilization effort by the PES/NAC, as highlighted in the Graph below, IDA support channeled via the MAOP represented the dominant source of National AIDS Program financing from 2002 through 2007. 3. The successful leveraging of MAOP resources allowed the National Program to scale up activities significantly over the project period. Notwithstanding the large number of partners mobilized, and the substantial resources they infused, the financing available for the National Program fell significantly short of needs defined in the National Strategic Plan 2002-2006, leaving a large unfinished agenda for action for the 2007-2011 Plan, now being supported by IDA via the Second HIV/AIDS Project. 29 Figure 1. 1: Financing for the National HIV/AIDS Program, 2002-2008 20,000,000 18,000,000 16,000,000 14,000,000 12,000,000 IDA 10,000,000 Others Global Fund GOB 8,000,000 6,000,000 4,000,000 2,000,000 0 2002 2003 2004 2005 2006 2007 2008 Source: National AIDS Council, Financial Management System NOTE: The above table includes only those financial resources that were managed directly by the National Council but they represent the majority of National AIDS Program funding. Excluded from the table are relatively modest supplementary funds channeled directly to specific HIV/AIDS activities by a few external organizations. 30 (d) MAOP Final Expenditures by Component and Category (BIF and US$) Table 1. 3: Project Final Expenditure in USD Multisectoral HIV/AIDS Control and Orphans Project Final Expenditures by Component and Category in BIF and US$ Public 17,702,772,090 BIF US$ 16,550,789.24 Response Works 893,564,394 BIF 835,416.96 Vehicles, Equipment and Office Supplies 1,020,828,678 BIF 954,399.70 Drugs, Tests & Medical Supplies 8,956,946,179 BIF 8,374,085.58 Consultants' services, audits and training 3,070,817,421 BIF 2,870,988.32 Grants for CSO Subprojects 3,742,144,598 BIF 3,498,629.83 Operating Costs 18,470,820 BIF 17,268.86 Response of Civil Society Organizations 8,760,595,149 BIF 8,190,511.82 Vehicles, Equipment and Office Supplies 34,250,519 BIF 32,021.71 Drugs, Tests & Medical Supplies 75,729,690 BIF 70,801.69 Consultants' services, audits and training 78,100,065 BIF 73,017.81 Grants for CSO Subprojects 8,571,632,875 BIF 8,013,845.99 Operating Costs 882,000 BIF 824.61 Orphans and other Vulnerable Children 6,588,512,238 BIF 6,159,774.13 Vehicles, Equipment and Office Supplies 62,963,050 BIF 58,865.82 Consultants' services, audits and training 236,151,761 BIF 220,784.52 Grants for CSO Subprojects 6,259,087,755 BIF 5,851,786.48 Operating Costs 30,309,672 BIF 28,337.31 Strengthening of Capacity 6,341,775,730 BIF 5,929,093.65 Vehicles, Equipment and Office Supplies 617,230,930 BIF 577,065.50 Drugs, Tests & Medical Supplies 820,000 BIF 766.64 Consultants' services, audits and training 2,357,812,357 BIF 2,204,381.05 Grants for CSO Subprojects 3,020,930,122 BIF 2,824,347.37 Operating Costs 344,982,321 BIF 322,533.08 Management and Coordination 6,591,310,186 BIF 6,162,390.00 Works 366,930 BIF 343.05 Vehicles, Equipment and Office Supplies 269,858,695 BIF 252,298.02 Drugs, Tests & Medical Supplies 880,000 BIF 822.74 Consultants' services, audits and training 3,291,112,219 BIF 3,076,947.75 Grants for CSO Subprojects 31,979,445 BIF 29,898.43 Operating Costs 2,997,112,897 BIF 2,802,080.02 - TOTAL 45,984,965,393 BIF 42,992,558.83 1,069.60 =BIF/USD FX rate Source: National AIDS Council, Financial Management Information System 31 Annex 2: Outputs by Component Overview 1. A major objective of the first HIV/AIDS Project in Burundi was to establish an institutional framework and operational base to develop a timely, effective and sustainable national response to the epidemic. The MAOP financed numerous studies that helped inform the 2007-2011 NASP. MAOP financing also helped support the expansion of service sites and centers for prevention, control, and treatment of HIV/AIDS. It helped upgrade the National Center for Blood Testing; paid for programs allowing for the systematic testing of donated blood and improvements to safe handling of injections; provided funds for the intensive training of providers of crucial services. 2. The project's contributions to the establishment of a basic infrastructure to deliver core HIV/AIDS interventions and to the development of a skilled cadre to deliver priority services, is described in the following two tables. Table 2. 1: Multisectoral HIV/AIDS Control and Orphans Project (MAOP) Expansion of Core Services Provision for HIV/AIDS Control and Treatment 2002 ­ 2008 Functioning Sites by Service Offered (Cumulative # per 2002 2003 2004 2005 2006 2007 2008 year) VCT 64 80 105 128 151 180 266 PMCT 1 9 11 12 27 38 63 ARV Treatment 0 0 23 26 36 46 68 Communal Youth Centers (Total # 82 98 130 131 131 Communes=129) Communal Committees to Defend Rights of 129 129 129 131 PLWHA Local Committees to Defend Rights of 21 pilot OVC (Total # Zones zone 2353 2353 2353 2353 = 2353) Source: National AIDS Council, Monitoring and Evaluation System 32 Source: National AIDS Council, Monitoring and Evaluation System * This contrasts with the total of 133 pregnant women who received prophylaxis in 2001 for prevention of mother- to-child transmission. As highlighted in this Table, the project allowed a significant increase in coverage over its 6 years of implementation. 3. Further, the large body of MAOP-financed studies, surveys and evaluations, included in Annex 9, contributed to a major expansion of knowledge of the evolving HIV/AIDS situation in the country as well as that of OVC, though crucial knowledge gaps remain. Among the most significant MAOP contributions was the 2007 national sero-prevalence survey urgently needed to inform the National Program. Its findings have contributed substantively to the design of the Second HIV/AIDS Project, and strongly influenced that project's adoption of a more targeted strategy to combat the epidemic, focusing intensively on the highest risk groups identified in that survey. Public Sector Response (Component 1) 4. This component focused on the following main areas: (i) establishment of HIV/AIDS focal sector units within each line Ministry; (ii) preparation of HIV/AIDS Action Plans by all 30 public sector institutions, including not only line the then 26 line Ministries but also the Office of the President, the Senate, the National Assembly, and the University of Burundi; (iii) decentralization of the National PES/NAC to the provincial, communal and local levels; (iv) conduct by public institutions of IEC targeted to their employees, families of employees and sectoral constituents; and (v) guarantee of the safety of the nation's blood supply and transfusions, including construction of the National Blood Transfusion Center (completed in 2005), strengthening of 4 regional transfusion centers, training of both professional and traditional health providers (including birth attendants), surveillance of blood donors via 5 centers, and provision of materials and equipment. 5. Several new strategies were effected, including: (i) definition of an HIV/AIDS communication strategy and a strategy for condom distribution by the Ministry of the Presidency; (ii) a guide for training of peer educators and teachers on HIV/AIDS by the Ministry of Education; and (iii) a national strategy for accelerating access to ARV, together with simplification, standardization and validation of ARV treatment, consistent with WHO norms. The MAOP also catalyzed formulation and passage of several new national policies, including: (i) a 2004 PMCT policy, with protocols distributed in all health facilities offering PMCT; (ii) national condom policy in 2004; (iii) a national policy to protect PLWHA; and (iv) a national policy to assist OVC and to protect the rights of orphans and widows. 6. Key achievements over the project period by the central level are captured in the following tables. 33 Table 2. 2: Key 2002-2008 Institutional Outputs of the Public Sector Component by Number and Percentage Key 2002-2008 Institutional Outputs of the Public Sector Component by Number and Percentage Indicators 2002 2003 2004 2005 2006 2007 Number of sector units for HIV/AIDS functioning in Line Ministries 8 (30%) 26 (100%) 26 (100%) 26 (100%) 26 (100%) 24 (100%) Number of Line Ministries with an HIV/AIDS Action Plan 8 (30%) 26 (100%) 26 (100%) 26 (100%) 26 (100%) 24 (100%) Number of Provincial Committees of the National AIDS Council created an 0 (0%) 17 (100%) 17 (100%) 17 (100%) 17 (100%) 17 (100%) Number of Communal Committees of the National Council( COCOLS ) cre 0 (0%) 17 (14%) 129 (100%) 129 (100%) 129 (100%) 129 (100%) Number of Local Committees of the National Council created and function 0 0 0 0 0 0 Source: National AIDS Council, Public Sector Component Team Notes: *At project launch in 2002, 8 line Ministries already had an HIV/AIDS Action Plan **A Government restructuring in 2007 eliminated two line Ministries, with their responsibilities integrated into other public entities ***The Senate, National Assembly, Office of the President and University of Burundi also elaborated HIV/AIDS Action Plans ****Local Committees of the National AIDS Council already have been created but none are yet functioning pending mobilization of two animateurs per locality of which at least one must be a woman Operationalization of the diverse Action Plans has been supported as well by other HIV/AIDS partner organizations, including notably BIT, UNICEF, UNESCO, WHO, ActionAid, and GTZ 34 Table 2. 3: Public Sector Responses Component Beneficiaries by Institutions Multisectoral HIV/AIDS and Orphans Project ­ Public Sector Response Component Beneficiaries of IEC interventions by public sector institutions 2002 - 2008 Public Institution 2002 2003 2004 2005 2006 2007 2008 Total 1 Ministry of Transport, Mail & Telecommunications 600 600 300 1,500 2 MInistry of Solidarity, Reintegration and Reinstallation of Patriots and Refugees 300 300 120 720 3 Ministry of Public Works and Equipment 240 240 60 3 600 4,140 4 Ministry of Public Operations 180 240 60 480 5 Ministry of Labor and Social Security 60 240 600 600 1,500 6 Ministry of National Defense 15,000 18,000 19,000 20,000 72,000 7 Ministry of Public Health (Promotion of Health by Nurses and Technicians) 3,000 3,000 3,000 9,000 8 Ministry of the Interior and Public Security 120 240 240 600 9 Ministry of Communication 80 240 100 420 10 Ministry of Development Planning and Reconstruction 120 120 240 11 Ministry of National Education 600 6,500 200,000 207,100 12 Ministry of Youth and Sports 200 960 240 2,000 3,400 13 Ministry of Good Governance 60 60 14 Ministry of Finance 120 120 240 15 Ministry of Energy and Mines 180 240 420 16 Ministry of Territorial Planning, Environment and Tourism 120 180 300 17 Ministry of Commerce and Industry 180 180 18 Ministry of Agriculture and Livestock 120 360 480 19 Ministry of External Relations and Cooperation 60 120 150 330 20 Ministry of Justice 120 350 470 21 Ministry of Communal Development 120 120 Ministry of the Presidency for the fight against AIDS (including the personnel of the 22 National AIDS Council and Permanent Executive Secretariat) 150 150 150 450 23 Ministry of Institutional Reforms, Human Rights and Parliamentary Relations 50 50 24 Ministry of Mobilization for Peace and National Reconciliation 55 55 25 Ministry of Vocational Education and the Craft Industry 420 420 26 University of Burundi 3,000 3,000 27 Presidency of the Republic 160 130 290 28 Senate 50 50 29 National Assembly 50 50 TOTAL 15,060 24,010 33,715 229,680 2,000 0 3,600 308,065 7. To help actualize the multisectoral approach promulgated in the National AIDS 2002-2006 Strategy, all public sector institutions benefitted from technical and financial support from MAOP or other Program partners to develop and institute HIV/AIDS Action Plans. The public sector response was rapid, with 100 percent of public institutions having annual Action Plans in place by project mid-term. The intensive IEC efforts these institutions mounted reached in aggregate 308,065 civil servants, their 35 family and sector clients, as highlighted above. The intensity of effort varied strikingly between Ministries, with the Ministries of Education and of National Defense accounting for 91 percent of all beneficiaries reached, as also noted in the above table, albeit their employees and constituents were key target groups to stem the HIV/AIDS epidemic,. Additionally, treatment solidarity funds were established in public institutions and parastatal enterprises to support ARV treatment for PLWHA who were employed in these entities. Civil Society Response (Component 2) Table 2. 4: MAOP : CSO Grant Subprojects financed and approved, by year Subprojects 2003 2004 2005 2006 2007 2008 Financed : New 41 84 309 27 99 4 Financed : Cumulative 41 125 434 461 560 564 Approved : New 41 84 419 3 13 12 Approved Cumulative 41 125 544 547 560 572 Source: National AIDS Council, Civil Society Component Team 8. Performance under this component was exceptionally strong in terms of mobilization of CSO participation as a major actor within the National AIDS Program, significantly exceeding original project targets. The number of CSOs working in HIV/AIDS related activities increased from about 35 at project appraisal, per the PAD, to over 500 at project closing. The project also promoted entry of new actors, encouraged by access not only to financing but to also to technical assistance from proposal writing to implementation and M&E. This TA notably was carried out by some relatively large, established CSOs operative in HIV/AIDS as well as the Council and its decentralized structures. Access to grants and capacity building also empowered communities to identify and respond to their own needs, thus creating a more sustainable model over the long term to address HIV/AIDS. 9. With the Credit reallocations, grants for civil society subprojects ultimately accounted for almost half (47 percent) of total project disbursements--the largest expenditure category. Investments in this component in aggregate were 13.8 billion BIF. The total 564 subprojects financed were well dispersed geographically. This underscores the strong comparative advantage of CSOs to operate in communities nationally, including many areas in which social unrest precluded delivery of public services. In aggregate, 765,794 beneficiaries were served under this component over the project life. Almost 90 percent of the total number of subprojects financed over the project life served four key target groups: PLWHA (34.2 percent); OVC (32.1 percent); youth (13.5 percent); and the general population (7.8 percent), as highlighted in the tables below. 10. Existing documentation on this component focuses almost exclusively on the outputs achieved in terms of number and distribution geographically and by type of beneficiary of projects and their general performance and overall monitoring results, including those of financial audits. As the September 2006 QSA Panel noted, it was "unable to get a clear picture from project documentation, including ISRs, of the results of targeted interventions or what worked and what did not in the efforts of NGOs and other organizations supported under Component 2." Among the CSOs that received MAOP grant financing under performance-based contracts signed with the NAC to deliver medical services, their operations were carried out in close collaboration with public health centers and hospitals. Thus it is not possible to fully disaggregate what portion of HIV/AIDS related medical services they directly provided. One proxy for assessing the contribution of CSO activities is provided in a technical evaluation of subprojects executed by CSOs. That evaluation, released in November 2008, and based on a sample of subprojects, 36 indicated the following most frequent motivations cited by beneficiaries for participating in subprojects: (i) to receive psychosocial counseling (80.5 percent); (ii) to receive food support (74.5 percent); (iii) to receive financial support (68.5 percent); (iv) to obtain medical care (68.0 percent); and (v) to obtain information on HIV/AIDS (59.2 percent). Table 2. 5: Number of Civil Society Subprojects by type of Beneficiaries over the Project Life Type of beneficiaries 2003 2004 2005 2006 2007 2008 Total # Pygmies 1 2 1 4 1 2 Female heads of household 5 8 Military Wives 1 1 Sex workers 1 3 1 1 6 Single mothers (rape victims) 3 3 Traditional warriors 1 2 3 Youth 14 10 39 5 8 76 Female leaders 1 1 2 Other leaders (religious, AC, PE) 2 4 3 9 Infants 1 1 Orphans & vulnerable children 6 34 96 6 39 181 Health personnel 1 1 General population 9 11 22 1 1 44 PLWHA 7 24 118 7 34 3 193 Associations & Private 1 enterprises(Capacity Strengthening) 2 3 Workers of enterprises 1 1 Disaster victims 1 1 2 1 1 6 Widows 1 12 4 17 Road workers 1 1 Prisoners 1 1 Handicapped 1 1 HIV infected infants 1 1 2 Total 41 84 309 27 99 4 564 Source: National AIDS Council, Permanent Executive Secretariat, Civil Society Team 37 Table 2. 6: Number of Beneficiaries of civil society subprojects by type and year Total Type of beneficiaries 2003 2004 2005 2006 2007 2008 Beneficiaries Pygmies 15 30 20 65 Female household heads 523 50 45 618 Military wives 335 335 Sex workers 50 102 50 52 254 Single mothers (rape victims) 884 884 Traditional warriors 360 100 460 Youth 12115 1942 9332 2189 385 25963 Female leaders 60 500 560 Other leaders (religious, AC, PE 281 385 338 1004 Infants 16 16 Orphans & vulnerable children 892 11561 34120 7376 3494 57443 Health personnel 40 40 General population 534 1400 26489 4062 188 32673 PLWHA 6038 7880 28503 11569 1317 1300 56607 Associations & Private enterprises(Capacity Strengthening) * 2 1 3 Enterprise workers 3600 3600 Disaster victims 40 84 759 1100 100 2083 Widows 28 683 145 856 Road workers 7600 7600 Prisoners 220 220 Handicapped 50 50 HIV infected infants 59 60 119 *These activities focused on strengthening the coordination capacity of diverse entities. Hence, these outputs cannot be added to the beneficiary numbers. Source: National AIDS Council, Permanent Executive Secretariat, Civil Society Team Orphans and Vulnerable Children (Component 3) 11. The project attained impressive levels of support to orphans and their families, although wide variations in numbers of beneficiaries occurred per year, as highlighted below. 38 Table 2. 7: Multisectoral HIV/AIDS Control and Orphans Project (MAOP) Priority Services Provided to Orphans and Vulnerable Children (OVC) Number of Beneficiaries, 2002 ­ 2008 Type of Support 2002 2003 2004 2005 2006 2007 2008 Provided School Support* 2,500 12,500 64,820 153,292 163,229 0 0 Medical Care 0 0 25,000 25,649 3,000 14,704 23,704 Psychosocial Counseling 0 1,454 12,151 165,000 10,826 26,382 43,507 Source: National AIDS Council, Monitoring and Evaluation System * No MAOP financing for school support was provided in 2007 and 2008 because of competing project financing needs and the availability of resources from other donors, including UNICEF and DFID, to finance school fees in those two years. 12. The OVC subprojects were unevenly distributed across the country. This reflects in part the concentration of project effort in the initial years until the Mid-Term Review in three pilot provinces-- Karuzi, Kayanza and Muramvya--particularly via the OVC component itself, which was expanded nationally thereafter. 39 Table 2. 8: Number of OVC Subprojects Financed by MAOP, by Province 2003-2008 Province 2003 2004 2005 2006 2007 2008 Total Bubanza 1 1 3 5 Buja M 1 2 5 1 9 Buja R 2 8 9 19 Bururi 1 9 6 16 Cankuzo 2 1 3 Cibitoke 2 2 3 2 9 Gitega 1 5 1 4 11 Karuzi 6 6 1 13 Kayanza 8 18 3 29 Kirundo 1 4 1 6 Makamba 5 3 8 Muramvya 5 8 1 14 Muyinga 1 5 1 1 8 Mwaro 1 1 6 8 Ngozi 1 5 1 7 Rutana 1 4 3 8 Ruyigi 2 2 Multiple provinces1 1 2 1 2 6 National2 Total 6 34 96 6 39 0 181 Source: National AIDS Council, OVC Component Team Note: The OVC projects are a subset of the CSO subprojects tables provided above and hence should not be added to those totals. Table 2. 9: Proportion of subprojects targeting OVC financed vs. approved 2003-2008 Province 2003 2004 2005 2006 2007 2008 Cumulative number 6 40 136 142 181 181 financed Cumulative number 6 40 181 182 185 185 approved Percentage of those approved that were financed 100 100 75.1 78.2 97.8 97.8 Source: National AIDS Council, OVC Component Team Note: The OVC projects are a subset of the CSO subprojects tables provided above and hence should not be added to those totals. 13. In terms of impacts, MAOP's contribution was quite limited in terms of placement of orphans in family or community homes compared to actual and growing OVC needs. Notably, only 1672 orphans 1 Subprojects that cover more than one province but not all provinces. 2 Subprojects not limited to specific provinces, national in scope. 40 were placed over the project life; this contrasts with an estimated 77,000 children in-country who had lost both parents. Moreover, the placement process moved exceptionally slowly, with the number of orphans placed failing to increase over a few years, as revealed in M&E indicators. Youth (The Cross-cutting Component) 14. Activities targeting youth formed a centerpiece of the National AIDS Program. In the public sector, youth-oriented activities were concentrated in two Ministries. Within the Ministry of Education, the project supported: (i) Preparation of a personal life skills book for 3rd grade students, with a teacher guide, with 300,000 copies distributed to school districts nationally, coupled with training of 1447 school directors and inspectors on the guide (financed in partnership with UNICEF) (ii) Training of 6322 5th and 6th grade teachers on this guide who then taught peer educators, also funded in collaboration with UNICEF (iii) Establishment of 972 Stop AIDS Clubs in secondary schools nationwide (iv) In the Ministry of Youth, Sports and Culture actions concentrated on prevention, participation and development of youth out of school, including: (v) Preparation of a personal life skills book (vi) Training of a pool of 30 youth to serve in advisory capacity at the national level and 510 youth to serve in provincial areas (vii) Training of 4410 local peer educators (over 30 for each of the 129 communes) corresponding also to the number of youth centers operational nationally (viii) Establishment of a National Network of Youth Active Against AIDS that functions via a national committee, 17 provincial committees and 129 communes. These committees assure the representation of youth in all institutional aspects of the AIDS Program including in the PES/NAC and in its decentralized provincial and community committees (ix) Promotion of youth centers, facilitated by preparation of a guide for creation and management of youth centers that helped mobilize 129 youth centers nationally 17. That only 76 (13.5 percent) of the 564 subprojects (financed through civil society grants) went to youth-targeted activities explains the relatively small impact of the project on youth, as revealed in the following tables. On the other hand, many of the outputs did establish a strong institutional base for future activities targeted to this key risk group. Of particular note in this regard are the National Network of Youth Against AIDS which plays a critical advisory role to the National Program and the now extensive national network of youth centers and Stop SIDA Clubs, established under project auspices. Also of note is the existence of a cadre of 4410 peer educators across the nation trained with MAOP resources to help expand youth knowledge of HIV/AIDS and promote the adoption of requisite attitudes and behavioral practices to protect them against infection. 41 Table 2. 10: Key Project Outputs Benefitting Youth, by Year Output 2003 2004 2005 2006 2007 2008 Total Comments Subprojects carried out by civil 14 10 39 5 8 76 25 963 youth have been reached with society preventive activities (associated largely with income generating activities) via the subprojects Beneficiaries by year 12 115 1942 9 332 2 189 385 Number of new youth centers 0 0 111 0 0 18 129 Note : separate data from the M&E team functioning reported 131 Youth Centers operative Audio-visual equipment for 0 0 106 0 0 0 106 Other equipment has been furnished youth centers for personal life and completed via contributions of other skills training partner agencies such as UNICEF. Numbers of peer educators 0 540 3 870 0 0 0 4410 UNICEF has upgraded skills and trained trained on the personal life new peer educators for the ensuing skills approach years which has resulted in an overall total of 5300 peer éducators now in place Number of new Stop AIDS 0 346 540 86 0 0 972 Clubs functioning Source: National AIDS Council, Civil Society Component Procurement Table 2. 11: MAOP Expenditures on Procurement by Category and Procurement Method Expenditure Category Procurement Method Budget Forecasts Actual Expenditures (US$) (US$) (1) Works ICB 600,000 675,594 NCB 91,000 82,698 2 (a) Vehicles, Equipment ICB 5,957,231 4,719,516 & Office Supplies; NCB 2,104,902 1,834,761 2 (b) Drugs, Tests & Shopping (International) 4,402,400 3,836,947 Medical Supplies Shopping (National) 1,910,727 1,578,342 SSS 92,000 90,000 Services QBS 18,000 16,040 LCS 260,000 257,293 SSS 470,000 456,007 TOTAL 15,906,260 13,547,198 Source: National AIDS Council, Procurement Unit 42 Annex 3: Economic and Financial Analysis 1. In the absence of data for economic and financial analysis, this section adopts the IEG methodology for evaluation of the first generation of MAP projects which bases the project economic and financial results on the counterfactual of no assistance and particularly on its contribution to: (i) deepen political commitment to contain the epidemic; (ii) expand and strengthen national and sub-national AIDS institutions for the long-run response; (iii) mobilize NGOs to participate and reinforce their capacity to provide prevention and care among high-risk groups most likely to contract and spread the infection; and (iv) enhance the efficiency of the national AIDS programs. Political Commitment 2. The political commitment at the highest levels of the government to redress the epidemic is clear. Combating HIV/AIDS is a strategic pillar of the Burundi PRSP. The Strategic Plan 2007-2011 provides a widely recognized framework to sustain the national Program and delineates priority HIV/AIDS actions. Creation of the PES/NAC reflected the priority being assigned to stopping the epidemic within the national development agenda. Though the PES/NAC has further capacity strengthening needs, it is widely recognized as a strong, decentralized autonomous body with a multisectoral mandate that is well prepared to coordinate the National AIDS Program. 3. Among many competing development priorities within the post-conflict setting, the government has agreed to contribute US$3 million in its budget for the first year of implementation of the Second HIV/AIDS Project, although to date only US$100,000 has been committed. 4. Expanded and strengthened national and subnational AIDS institutions. The project's investments in the National Council and its highly decentralized entities (provincial, community and local Committees) have created an exceptionally strong and previously absent institutional base for effectively and efficiently coordinating and advancing national program activities across the public and private sectors. Project technical and financial support to public institutions via Component 1 have led to establishment of HIV/AIDS sector units in all ministries and introduced the mandatory preparation of HIV/AIDS Action Plans to mobilize each institution to play a role. These actions to substantially strengthen the capacity of national public institutions reflect the government's recognition of the adverse development consequences of the HIV/AIDS crisis and firm commitment to launch and sustain a solid, multisectoral attack on it. 5. Mobilization and strengthening the capacity of NGOs to provide priority services. As highlighted previously, the dynamic public-private partnership for HIV/AIDS that emerged from the project provides a substantially expanded network to provide quality HIV/AIDS services. Intensive capacity building across the CSO sector has significantly improved their ability to design, execute, monitor, and evaluate HIV/AIDS activities. It also has improved national monitoring of the HIV/AIDS situation by compilation of data from the now extensive network of CSOs operating in the country. CSOs consulted for this ICR revealed that MAOP filled critical revenue gaps to advance their work, such as financing ARVs and medicines to treat OIs. Limiting subproject financing to predominantly direct costs of HIV/AIDS programmatic interventions, maximized use of limited project resources. 6. Enhanced efficiency. Numerous challenges exist to determining the actual efficiency of the project with current data. One proxy measure of efficiency that may be considered is the extent to which project funds were expended on appropriate interventions, contextualized and adapted to the needs of the local communities to ensure resources were well spent. In this respect, the project empowered 43 communities to play stronger advocacy roles and to identify and respond to their own priority needs. A key mechanism for doing this was by providing the decentralized committees with financing and with the authority to approve and manage small subproject grants for community-based activities. It also encouraged significant experimentation and innovation, one example of which was the introduction of low cost models of service delivery, which filled important gaps in health promotion, and preventive and curative care, as has been noted previously in this ICR. 44 Annex 4: Bank Lending and Implementation Support/Supervision Processes (a) Task Team members Names Title Unit Responsibility/ Specialty Lending Miatudila Malonga Senior Health Specialist AFTH3 TTL Michael Azefor Seior Public Health Specialist AFTH3 TTL Betty Oyella Bigombe Consultant AFTH3 Anne Anglio Senior Language & Program Assistant AFTH3 Anne Marie Bodo E T Consultant AFTH2 TTL (Orphans Menahem Prywes Senior Poverty Economist AFTH3 Integration & Education Project) Bella Lelouma Diallo Sr Financial Management Specialist AFTFM Serigne Omar Fye Consultant AFTH1 Astania Kamau Language Program Assistant AFTH3 Pamphile Kantabaze Senior Operations Officer AFTH3 Susan Opper Senior Education Specialist AFTH3 Prosper Nindorera Senior Procurement Specialist AFTPC Nadine Manirambona Temporary Team Assistant AFMBI Isabela Michali Drossos Senior Counsel LEGAF Joseph Kizito Mubiru Sr Financial Management Specialist LCSFM Leoncie Niyonahabonye Program Assistant AFMBI Seraphine Nsabimana Program Assistant AFMBI Michael Fowler Senior Finance Officer LOAG2 Supervision/ICR Otieno Ayany Financial Management Specialist AFTFM Aurelien Serge Beko E T Consultant AFTP3 Betty Oyella Bigombe Consultant AFTH3 David M. Blankhart Consultant HDNGA Anne Marie Bodo E T Consultant AFTH2 Rene Bonnel Consultant HDNGA Bella Lelouma Diallo Sr Financial Management Specialist AFTFM Serigne Omar Fye Consultant AFTH1 Astania Kamau Language Program Assistant AFTH3 Pamphile Kantabaze Senior Operations Officer AFTH3 Robert John Ivo Leke Consultant AFTH3 Alessandra Macri Consultant AFTH3 Nadine Manirambona Temporary AFMBI Myrina D. McCullough Consultant AFTH3 Joseph Kizito Mubiru Sr Financial Management Specialist LCSFM 45 Adjaratou Diakhou Ndiaye Consultant HDNGA Prosper Nindorera Senior Procurement Specialist AFTPC Seraphine Nsabimana Program Assistant AFRVP Esra Pelitozu Consultant HDNGA Christine Lao Pena Senior Human Development Econo LCSHH Dominique Puthod E T Consultant AFMBI Rosalia Rodriguez-Garcia Sr Monitoring & Evaluation Spe HDNGA Clarette Rwagatore Team Assistant AFMBI (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle No. of staff weeks USD Thousands (including travel and consultant costs) Lending FY01 30 100.50 FY02 56 251.13 FY03 0.00 FY04 0.00 FY05 0.00 FY06 0.00 FY07 0.00 FY08 0.00 Total: 86 351.63 Supervision/ICR FY01 0.00 FY02 0.00 FY03 26 167.69 FY04 15 95.48 FY05 34 125.12 FY06 36 131.85 FY07 19 75.97 FY08 16 43.40 FY09 5 0.00 Total: 151 639.51 46 Annex 5: Beneficiary Survey Results (if any) Not applicable Annex 6: Stakeholder Workshop Report and Results (if any) Not applicable 47 Annex 7: Summary of Borrower's ICR and/or Comments on Draft ICR BURUNDI EXECUTIVE SUMMARY OF THE MAP1 COMPLETION REPORT (PMLSAO) 1. The Multisectoral HIV/AIDS Control and Orphans Project (PMLSAO) started in October 2002 after a launch workshop conducted from October 15­18, 2002. The total credit was SDR 29,100,000 (US$36,000,000). 2. The project development objective was to "slow the spread of HIV/AIDS in the general population and mitigate the impact of HIV/AIDS on families." 3. The project was divided into 16 programs and executed through five components, namely: (i) Strengthening Public Sector Response to the Crisis (US$9.3 million); (ii) Increasing Participation in Grassroots Organizations and Communities (US$13.4 million); (iii) Orphans Integration and Education (US$7 million); (iv) Strengthening the Program's Strategic Formulation, Implementation, Monitoring and Evaluation Capacity (US$4 million); and (v) Project Management (US$3 million). 4. Program 1 sought to reduce high-risk sexual behaviors through Information, Education, and Communication (IEC) activities. A national communication policy to change behavior was formulated. IEC activities geared toward reducing high-risk behaviors that lead to HIV/AIDS contamination were conducted in all AIDS prevention sectoral units. A number of them were conducted for staff members themselves, while others targeted the population groups served by the relevant sector, in the form of seminars, workshops, and radio or television broadcasts in French and Kirundi. 5. Supported by the project, IEC material and educational guides were produced and training sessions organized for various groups, such as government officials and employees and more specifically, inter alia, teachers, community facilitators, uniformed services, caregivers, HIV testing counselors, women, youth, and sex workers. 6. Special emphasis was placed on training the trainers and peer educators including youth, uniformed services, sex workers, prisoners, displaced persons, child heads of households, and female heads of households (Ministry for Cooperation). In schools, the Ministry of National Education was able to promote awareness building through the Stop AIDS Clubs (972), while the Ministry of Youth, Sports, and Culture organized a similar activity at youth centers in all communes across the country (129). Media practitioners also benefited from training in advocacy. A national communication policy to change behavior was formulated within the context of coordinating and standardizing IEC interventions. 7. IEC activities were also conducted through civil society organizations (CSOs). The project financed 163 subprojects for the various high-risk behavior groups, and religious groups assisted greatly with prevention activities. The radio stations with the largest listenership were also used to build awareness. In all, almost four million persons were reached by radio broadcasts. Two studies were also conducted on the impact of Radio Culture broadcasts and 48 another on knowledge, attitudes, and practices of members of religious groups toward HIV/AIDS and sexually transmitted infections (STIs). 8. The "Youth Response" component is cross-sectoral in relation to the various other components of MAP1. It helped achieve the first project development objective: "Slow the spread of HIV/AIDS in the general population." The component's objective was to slow the spread of HIV/AIDS in the youth population. 9. At the outset, the youth component was not included in the Project Appraisal Document (PAD); it was incorporated during the implementation phase into the organizational chart of the Permanent Executive Secretariat of the National AIDS Council [Secrétariat Exécutif Permanent du Conseil National de Lutte contre le SIDA -- PES/NAC]. Given its cross-sectoral role, civil society organizations and public sector entities supported the implementation of these strategies. The Ministry of Youth, Sports, and Culture was called upon to provide leadership for other sectors. Some of the activities conducted were surveys on girls' participation at youth centers; support to link prevention activities to revenue-generating activities (RGAs) and the distribution of condoms and IEC material; support for the operation of youth centers; and the training of youth peer educators, all with a view to involving youth as actors and not as mere beneficiaries. 10. Several additional interventions were conducted in collaboration with a number of development partners, such as UNICEF, UNFPA, UNAIDS, and UNESCO. Activities were conducted in and around youth centers, in schools, and through media campaigns and/or entertainment events. 11. The component focused on the package of prevention services including awareness building, condoms, RGAs, and training. These activities were accompanied by the identification and training of youth leaders, the standardization of prevention messages, peer education, and the retraining of peer educators of over 3,870 boys and girls. 12. A consolidated action plan for all stakeholders involved in directing the logical framework of the project document was drafted at a participatory workshop. As a result, both ministries with responsibility for youth (the Ministry of National Education and the Ministry of Youth, Sports, and Culture) coordinated the prevention activities for youth through 972 Stop AIDS Clubs and 129 youth centers, respectively. A survey of girls' participation was conducted. 13. Program 2 focused on the promotion of condom use. A national condom policy was formulated. Training in condom use was organized and male and female condoms distributed: 54.7 million male condoms in comparison to 102,730 female condoms from 2002­2008. 14. Program 3 dealt with HIV/AIDS testing. Achievements included regular voluntary counseling and testing (VCT) accreditation (266 VCTs in 2008), testing at VCT sites (791,050 cases tested from 2002­2008), the regular supply of reagents, and the organization of training in pre- and post-test counseling at the various care facilities. Seventy-one civil society subprojects were financed during the implementation of this program, of which 10 focused exclusively on this area, and the other 61 focused simultaneously on other areas. 49 15. Program 4 sought to cover all health-related training in the area of STI diagnosis and treatment. The project facilitated the training of 1,903 caregivers and the production and distribution of 6,000 algorithms at care facilities. Twenty-one accredited associations helped implement the program while 17 subprojects covering aspects of STI treatment were financed for 3,661 patients. 16. Program 5 focused on reducing the risks of transmission by blood. The program included the construction and equipping of the National Center for Blood Transfusion [Centre National de Transfusion Sanguine CNTS], the purchase of equipment for the Regional Centers for Blood Transfusion [Centres Régionaux de Transfusion Sanguine CRTS], and the supply of care facilities with protective gear and material for the handling of blood exposure incidents. With a view to further increasing blood transfusion safety, prevention guidelines were made available. The quality and reliability of tests were routinely monitored at blood transfusion centers while blood donors were recruited and retained, which helped to ensure that the blood donor seroprevalence rate did not surpass 0.3 percent. 17. Program 6 supported antenatal testing and the prevention of mother-to-child transmission of HIV/AIDS (PMCT). Achievements of this program include the obtaining of accreditation for PMCT sites (63), the production of IEC material, the training of service providers (441) within the context of the PMCT, and the organization of briefing and awareness- building sessions for government employees and associations. Five associations were supported by the financing of 20 subprojects. A total of 5,660 women and 4,663 newborns benefited from this program from 2002­2008. 18. Program 7 dealt with psychosocial care for People Living with HIV/AIDS (PLWHA). This activity was primarily organized and structured around civil society organizations with a view to supplementing full medical care. A counseling guide for STI/HIV/AIDS was produced and used for training, and 738 persons were trained in psychosocial care. 19. Program 8 addressed prophylaxis, the diagnosis and treatment of opportunistic infections (OIs), and the care continuum. The program provided medication for OIs and the treatment of patients, in addition to training for caregivers in the treatment of OIs. A legal and regulatory framework for home-based care was developed and 500 stakeholders were trained in this area. In addition, 131,489 OI cases were treated at public, private, and community facilities. More specifically, associations benefited from the financing of 55 subprojects. The purchase of medication for OIs has continued. 20. Program 9 dealt with the improvement of access to antiretrovirals (ARVs) for HIV/AIDS. The results from this program are the establishment of new ARV sites, which totaled 68 in 2008, care for 14,343 PLWHA in 2008 in comparison to 600 in 2002, 1,210 in 2003, 3,900 in 2004, 6,416 in 2005, 8,048 in 2006, and 10,928 in 2007. 21. Health mediators were also provided: they serve as an interface between the physician and the patient, and facilitate the continuity of care and observance of treatment. To support this aspect of health care, CD4 counters were made available to users. In addition, treatment support funds [caisses de solidarité thérapeutique] were established within government and parastatal 50 entities; however, the Military Support Fund [Caisse de Solidarité des Militaires CASSOM] is the best organized and most efficient of them all. 22. Program 10 focused on the promotion of health links - human rights, the protection of People Living with HIV/AIDS, and the other vulnerable groups. In this context, Decree-Law No. 1/018 of May 12 was enacted and translated into Kirundi, while the ILO Code of Practice on HIV/AIDS and the World of Work was disseminated to personnel at public, private and community facilities. The Leagues of Human Rights and PLWHA associations and networks, headed by the RBP+, have continued to advocate for the defense and protection of the rights of PLWHA, orphans, and other vulnerable children. 23. Program 11 was executed as a specific component for orphans and other vulnerable children (OVC). Initially the program covered three provinces (Karusi, Kayanza, and Muramvya). Its activities extended across the entire country because of the growing needs increasingly expressed by the provinces not covered. It focused on the care of orphans and other vulnerable children. One hundred and eighty-seven associations and groups received support. In all 1,672 orphans were placed in families, while socioeconomic support activities for 32,281 families have helped keep children in school. Many more activities were conducted with a view to better documenting the situation of orphans: a study on the situation of orphans in the pilot zone (2003); identification of OVC across the country (2005); establishment of an OVC module in the AIDS INFO database; formulation of the national policy for OVC (2005); and the formulation of a national plan for OVC. With the expansion of activities, the program supported the enrollment of 396,341 children, medical and nutritional care, and the protection of rights. 24. Program 12 financed revenue-generating activities for the poorest PLWHA. Therefore, within the context of reducing the impact and the other determinants of HIV, revenue-generating activities were financed for poor PLWHA; 446 members of RGA committees were trained and 258 PLWHA benefited from socioprofessional reintegration. 25. Program 13 focused its interventions on the strengthening of civil society organizations divided into national, regional, and local associations, NGOs/nonprofit organizations, and private sector organizations. 26. As a result of the sponsorship approach, large associations (or groups) received support to bolster community mobilization, and organized or sponsored groups of persons living with HIV/AIDS through three projects. These groups currently operate in every commune in the country and are estimated to be 129 in all, with approximately 8,000 members. The Burundi Alliance against AIDS [Alliance Burundaise contre le Sida ABS] was financed to strengthen the capacities of 85 member associations. 27. In terms of the private sector, five subprojects geared toward awareness-building activities and mobilization for voluntary testing and care in 55 companies were executed through two large associations, namely the employers' association of Burundi, and the Greater Involvement of People Living with HIV/AIDS (GIPA) Center. The activity allowed for the establishment and strengthening of 30 of the 55 AIDS committees planned. 51 28. Program 14 focused on strengthening capacities to formulate, coordinate, monitor, and evaluate sectoral plans and decentralized action plans. In 2003, only eight ministries had HIV/AIDS action plans. In 2005, each ministerial department had an action plan, along with the Senate, the Office of the President of Burundi, and the University of Burundi. Ministries received support for the formulation of action plans and eight of them were provided with vehicles, while 20 received information technology and office equipment (computers and accessories, and photocopiers). At the decentralized level, plans were formulated in provinces (17) and communes (129). 29. Program 15 focused on planning, monitoring and evaluation, and the information management system. In 2003, a monitoring and evaluation system was developed and approved at all levels. Users were trained to operate it. Standardized STI/HIV/AIDS data collection tools were developed. A project planning, monitoring, and evaluation training module was also developed, and trainers trained in this module. An AIDS INFO database, a documentation center, and a CNLS website were established, while a macroeconomic impact study on HIV/AIDS in Burundi and two seroprevalence studies were also conducted. 30. Program 16 covered project coordination and management. It provided inputs for the efficient implementation of the project. "Procurement" was the tool used for the acquisition of goods and services, and the execution of works for the PMLSAO. As a result of frequent disruptions in the supply of antiretrovirals and reagents, one of the causes of which was the length of the supply procedures, an amendment to the credit agreement was signed on March 8, 2005 to allow for the acquisition of ARVs and other medical products at an estimated cost of less than US$500,000 per contract, using the shopping procurement method. 31. PMLSAO project expenses totaled US$42,292,558.82 with an allocation of 38 percent for the public sector response, 19 percent for the civil society component, 14 percent for the OVC component, 13 percent for capacity building, and 14 percent for project management and coordination, respectively. 32. Cumbersome administrative and protracted national procurement procedures were the main causes of frequent disruptions in the supply of antiretrovirals and reagents. In order to rectify this situation, IDA agreed to modify the procurement procedures. As a result, an amendment to the credit agreement was signed on March 8, 2005 to allow for the acquisition of ARVs and other medical products using the shopping procurement method. Assessment of the Roles of Actors 33. The objectives formulated in the project were relevant because they were tailored to the problems they were intended to solve. At times, several of these objectives were not feasible, because of their formulation and the lack of criteria for assessing the level of attainment. Moreover, the lack of baseline data for specific indicators has not facilitated the assessment of progress made. 34. Implementation has at times been thwarted by the unavailability of funds on one hand, and the lack of commitment on the part of a number of the actors on the other. The programs were restructured at mid-term to ensure greater coherence between them. 52 35. Youth were involved as actors and not mere beneficiaries. On average, 29 percent of the CSOs' activities were initiated by the youth themselves. The activities of the AIDS Sector Units [Unités Sectorielles de Lutte contre le SIDA USLS] that targeted the youth accounted for approximately 16 percent, and those of the Provincial AIDS Committees [Comités Provinciaux de Lutte contre le SIDA CPLS] accounted for 35 percent. This demonstrates the importance of youth as a target group and as actors. 36. Working through youth centers and Stop AIDS Clubs proved to be a good practice although the decentralized administrative entities (schools and communes) need to assume more responsibility for monitoring. The promotion of initiatives mobilized the support of the other partners (UNICEF supported youth centers in eight provinces) and the management of the youth centers by the Ministry of Youth, Sports and Culture. 37. The contribution of the PMLSAO through the civil society component is significant based on the current level of commitment shown by associations and the private sector to AIDS prevention efforts. 38. The assessment of the financial management of subprojects implemented by civil society organizations and associations enabled the PES/NAC to determine to what extent the funds made available to CSOs were used for the purposes stipulated in the Grant Agreements and in accordance with these agreements. 39. Likewise, the technical evaluation of the work done by associations revealed the benefits provided to target groups by the activities of each subproject, especially the relevance in relation to the PMLSAO objectives, the effectiveness and efficiency in the use of financial resources by the PMLSAO, and the feasibility and impact of activities conducted for the target groups indicated in the Burundi Action Plan for the Struggle against HIV/AIDS. 40. With regard to the support for OVC, it must be acknowledged that the government has worked hard to pilot the national OVC policy, formulate the national action plan for 2007­2011, analyze the subprojects submitted by CSOs, and mobilize funds. 41. CSOs have played the role of implementing agencies for activities on the ground and in the analysis of OVC subprojects. They have been involved in advocacy for and the identification of OVC. Religious organizations have been heavily involved in the care of OVC, either by submitting and implementing subprojects, or by implementing subprojects submitted by CSOs. Performance of the Bank and the Borrower 42. The World Bank has played a leadership role in the financing of AIDS prevention activities in Burundi. The other donors have followed the trail blazed by the World Bank by adding their own contributions to those of the World Bank. As a result, a national entity was organized and established (CNLS and its decentralized agencies). 43. The supervisory missions conducted by the World Bank have enabled the strengthening of the national entity's operational capacities, especially in terms of planning, financial management, monitoring and evaluation, and management in general. 53 Lessons Learned 44. Leadership and management style are determining factors for the optimal operation of the entities and organizations involved in the HIV/AIDS response. 45. A fully operational Management Information System (MIS) with a database is essential for quality monitoring and evaluation that reflect the prevailing situation, and planning that is based on objective information describing the real problems to be solved. 46. Effective decentralization, aided and supported by the other development partners, has proven to be a sure and effective means of providing national response coverage. 47. Sound management of resources justified by the audit reports is an indication of the confidence placed in the national project management entity and the professionalism of the managers. Recommendations 48. The recommendations made seek to ensure the sustainability of the gains made. (i) Strengthen the capacities of the CNLS and its agencies (ii) Formulate a resource mobilization strategy (iii) Diversify the national program's financing sources and emphasize resource mobilization at the national level (iv) Observe the three principles in the national HIV/AIDS response (v) Continue advocacy among political, religious, and community leaders (vi) Strengthen the organizational and technical capacities of CSOs (vii) Improve the condom distribution networks (viii) Present a range of Information, Education, and Communication/Communication for Behavior Change (IEC/CCC) messages (ix) Strengthen medical care (x) Develop the PMCT program (xi) Simplify procurement procedures (xii) Strengthen decentralization especially at the commune level; and participate in the local government decentralization process currently underway (xiii) Continue awareness-building activities geared toward the most at-risk groups (xiv) Strengthen the information management system (collection, processing, dissemination, and feedback) (xv) Prioritize joint planning and reviews with the various participants in HIV/AIDS prevention efforts (xvi) Improve the quality of the data collected on OVC (xvii) Train CSOs in the use of the AIDS INFO database (xviii) Conduct more studies to better understand the evolving issue of children affected by AIDS (xix) Improve the technique for reporting violations of children's rights 54 (xx) Gradually organize commune-level forums on the protection of OVC (xxi) Improve the care of children affected by AIDS. 55 Annex 8: Comments of Co financiers and Other Partners/Stakeholders Not applicable 56 Annex 9: List of Supporting Documents Etude d'Impact Qualitatif et Quantitatif de l'Expérience Pilote de Prise en Charge des OEV Dans Les Provinces de Karuzi, Kayanza et Muramvya: Rapport Définitif, Ministère à la Présidence Chargé de la Lutte Contre Le SIDA, Conseil National de Lutte Contre Le SIDA, Novembre 2006 Etude d'Impact Qualitatif et Quantitatif de l'Expérience Pilote de Prise en Charge des OEV Dans Les Provinces de Karuzi, Kayanza et Muramvya: Annexes du Rapport, Ministère à la Présidence Chargé de la Lutte Contre Le SIDA, Conseil National de Lutte Contre Le SIDA, February 2007 Etude d'Evaluation Technique des Sous Projets Exécutés par les Organisations de la Société Civile du Burundi, Novembre 2008: Health Focus GmbH, Allemagne Audit Technique des Agences d'Exécution et Mise en oeuvre de la composante « Orphelins et Enfants Vulnérables » du PMLSAO Enquête Nationale Socio-Comportementale sur l'Infection VIH/Sida au Burundi, Ministère de la Santé Publique, Projet Santé & Population II, December 2001 ; Etude de la Prévalence des Hépatites B, C, de la Syphilis et les Co-Infections avec le VIH/Sida au Burundi, Etude de mise en place d'un Système de Gestion des Déchets Bio-Médicaux dans 29 hôpitaux et 1 centre de santé modèle 57 Annex 10: Summary of Key Findings of the 2005 Evaluation of the Three-Province Pilot Project for Orphans and Vulnerable Children 1. As the Bank required, prior to expanding the geographic coverage of this component beyond the pilot provinces of Karuzi, Kayanza and Muramvya, the government undertook a 2005 qualitative and quantitative impact analysis of the pilot activity based on a survey of 2602 beneficiaries, one-half of whom were children birth to 18 years of age; and the other half heads of households where they resided. 2. Multidimensional impacts of the pilot project were found that resulted in major overall direct and indirect improvements in the lives of the beneficiaries of support provided via MAOP. A high degree of satisfaction with OVC services provided was documented. Among the most significant impacts on the OVC were: (i) an increased rate of schooling due to enrollment of new students, reintegration of those who had interrupted education and pursuit of schooling by those who had left due to lack of resources; (ii) reduced rate of grade repetition and school abandonment among assisted orphans compared to those not assisted; (iii) acquisition of new technical/professional competencies via different workshops attended by OVC who never attended school or dropped out previously; (v) improvements in the living conditions of OVC due to economic support provided to host families, principally through increasing household resources for home and nutritional improvements; (vi) increased identification and resolution of cases of exploitation and abuse, principally through enhancing OVC knowledge of their rights, also leading to a significant reduction in the occurrence of new cases. These findings, while encouraging, must be interpreted with caution, as there was no baseline information regarding the vulnerability of the assisted children prior to the project, as is crucial to determine actual impact. 3. Importantly, the study found no health impact on beneficiaries. Widespread problems of access to health care were indicated with many public and private health providers reportedly refusing to accept the public health insurance mechanism (CAM) to pay for services rendered. The evaluation warned that lack of access to health care placed orphans at high risk given their prevalent morbidity and mortality. 4. Three groups were found to be the most vulnerable: children heads of households, abandoned children including street children, and children infected or affected by HIV/AIDS. 5. Overall the study recommended the continuation and expansion of coverage of the OVC component. Among the main conclusions emanating from evaluation of the pilot provinces was that the multidimensional demands orphans and their families face -- coupled with the generally high costs of response -- demand improvements in performance of discrete interventions. With specific reference to school support, it was suggested that the PES/NAC reorient funds previously allocated to payment of school fees to other facets of school support also needed to ensure regular attendance. It also was recommended that a better mechanism be put in place within the school system to monitor the academic performance of those OVC.3 Subsidies provided for primary education and vocational training were found to be especially important to encourage attendance, particularly as coverage under this component was expanded to those 16 to 18 years of age--a highly vulnerable subgroup--as noted earlier in the ICR. The subsidies went predominantly towards primary school fees, but funding was also needed for associated requirements (e.g. uniforms and textbooks) whose costs in aggregate represented the major category of expenditure within the OVC component. The evaluation revealed that these subsidies served 3Note: This was particularly important in terms of providing reliable data for the Project's Key Performance Indicator that sought measure increases in the proportion of assisted orphans who advance to the next grade of primary school 58 as strong incentives for school attendance, even though they went directly to the schools. Investments in small-scale income generating activities and gardening to enhance the economic and nutritional well- being of families and communities hosting orphans proved to be a powerful inducement for the social integration of orphans, particularly among the poorest subgroups. 6. In terms of the next phase of operations, it was urged that high priority be assigned to prevention, treatment, and care and support to children infected or affected by HIV/AIDS. This should include identification of those at highest risk in the community, provision of quality psychological and medical care, including treatment of opportunistic infections and potentially ARV treatment; and refinement of the HIV/AIDS program to respond to vulnerable children in general and OVC in specific. 7. Specific suggestions to improve management of the component were for the PES/NAC to: improve its collaboration with the OVC component-executing agencies, including helping them improve their capacity; instituting a more rigorous selection process with a view toward obtaining a true "professionalization" of executing agencies; identifying alternatives to the CAM to ensure OVC access to health care and services; promoting children's rights, including strengthening the capacity of local level committees to protect their rights; creating a national cadre who will respond to OVC needs; and establishing and maintaining an interface between OVC activities and those to protect children against vulnerabilities. 8. It was further recommended that streamlined procedures be introduced to facilitate CSO participation in OVC activities. The evaluators also suggested increased quality control of the component, including conducting an internal audit of performance of all CSOs carrying out OVC projects to identify the best and worst performers. This was viewed as crucial given the priority being attached to accountability for use of grants. The audit would serve a dual purpose: it would cull poor performers while also helping to inform the Council's Permanent Executive Secretariat about needed revisions to its criteria to guide future selection of CSOs to deliver OVC services and, thus, enhance overall component performance. 9. These OVC pilot project findings were complemented by a subsequent independent technical evaluation of all subprojects executed by CSOs. The November 2008 report of this later evaluation found significant improvements in the status of orphans who were beneficiaries of the project financed grants, including: a 58.8 percent rate of school attendance and reduction of school dropout; a 57.5 percent improvement in the health status of the orphan and family; a 55.5 percent improvement in behavior; a 49.7 percent increase in new knowledge; and a 24.8 percent increase in family revenues/income. 59 Annex 11: Achievement of Project Development Objectives (details) 1. Achievement of the Project's two primary PDOs--slowing the spread of HIV/AIDS and mitigating its effects--is rated "Moderately Satisfactory". The core KPIs outlined in Section 1.2 included four "outcome" measures. Of these, three sought to measure extent of achievement of selected behavioral changes essential to slow the rate of new infections, as anticipated from project investments in orphans' progression in primary school, as anticipated from Project-financed education subsidies and other social and economic assistance provided to orphans and their host families. Two of the output indicators sought to measure progress in creating selected essential pre-conditions for slowing the spread of HIV/AIDS, while the third assessed extent to which orphans received crucial supports. All three process indicators focused on execution of project-financed grants to civil society organizations to carry out HIV/AIDS activities. Projected causal linkages between key activities the project financed, output indicators and anticipated outcomes were reasonable. 2. HIV prevalence appears to have stabilized or be declining. When ARV treatment began being provided, it was generally agreed that HIV prevalence in the general population was no longer a viable measure of HIV transmission (because prevalence could increase due to longer survival by PLWHAs alone). Moreover, the 2002 national sero-prevalence survey used a different methodology from the 2007 survey, making comparisons difficult. Nevertheless, the survey data suggests that the prevalence of HIV/AIDS may be at least stabilizing if not declining, In 2001, the sero-prevalence rate was 3.2 percent in the general population 12 years of age and older while in 2007 the prevalence rate among people 18 years of age and older was 2.97 percent. More importantly as a measure of HIV transmission, HIV prevalence was 3.5 percent among those 15-24 years of age in 2001 and was found to be 3.1 percent among the same age group in 2007. 3. Data from seven sentinel surveillance sites monitoring patients continuously since 1999 and an eighth site which resumed operation in 2005 suggest that HIV prevalence among pregnant women may have declined between 2002 and 2007, as shown in Table 11.1. Clearly, these data must be interpreted with caution. As with most surveillance data of this type there is considerable variability from year to year and an apparent sharp rise in HIV prevalence rates in 2006 is troubling. , Nevertheless, the sentinel site data provide some indication of trend and some encouragement that HIV prevalence is at least stabilizing. The observed decrease in between 2002 and 2007 in the prevalence of syphilis among pregnant women--one of the core project KPIs-- is also encouraging. 60 Table 11.1: HIV and Syphilis Prevalence rates Among Pregnant Women Coming to Sentinel Sites A) HIV Prevalence: Year # of sites 15-49 years 15-49 years 15-24 years 15-24 years Sample size % HIV+ sample size % HIV+ 2002 7 1934 5.38% 881 5.11% 2003 7 2162 5.50% 930 4.62% 2004 7 3400 4.44% 1510 2.78% 2005 8 3041 5.23% 1369 4.82% 2006 8 3295 10.11% 1440 9.65% 2007 8 3367 3.98% 1494 2.28% B) Syphilis Prevalence Year # of sites 15-49 years 15-49 years 15-24 years 15-24 years Sample size % Syphilis + sample size % Syphilis + 2002 7 1881 2.39% 850 2.59% 2003 7 2162 1.80% 930 2.37% 2004 7 3215 2.12% 1513 1.65% 2005 8 3041 1.51% 1369 1.68% 2006 8 3295 1.12% 1404 1.35% 2007 8 3367 1.46% 1494 1.81% Prevalence of HIV among Pregnant Women 12 10 8 Age 15-49 % 6 Age 15-24 4 2 0 2002 2003 2004 2005 2006 2007 Year 61 Prevalence of Syphilis among Pregnant Women 3 2.5 2 Age 15-49 %1.5 1 Age 15-24 0.5 0 Year Slowing the spread of HIV/AIDS 4. Behavioral change has occurred among some key groups. Comparison of 2001 and 2007 National Behavioral Survey data for two DCA outcome indicators reveals progress made in reducing selected high risk behaviors by two target groups. The percentage of adults using a condom during contact with a non-regular sexual partner rose from 12.7 percent in 2001 to 19.8 percent in 2007. Over the same period, the percentage of girls 15 and under sexually active decreased from 32.5 percent to 21.5 percent. These data also reveal, however, that by the project's end, the proportion of the adult population and of young girls at high risk of HIV infection was still quite high compared to other countries in the Region.. 5. Increases in access to voluntary counseling and testing and preventive services (see Annex 2). The more than four-fold increase in the number of VCT sites functioning between project start and completion and almost 3000 persons trained in VCT have contributed to an almost seven-fold increase in the number of persons tested and informed of their status in 2008 versus 2002 (see Annex 2). The number of persons 15-49 years of age who were tested in 2007 and 2008, the only years for which disaggregated data by age and gender are available, represented 3.69 percent and 5.72 percent of the total Burundi population in that age group. By project end possibly 15-17 percent of Burundi's population had been tested (if they were only tested once). Table 11.2: Number of persons tested for HIV/AIDS in 2007 and 2008 Year 2007 Year 2008 Total M F Total M F Number of 135,972 59,987 75,985 221,205 93,509 127,696 persons tested Population from 3,680,597 1,860,035 1,770,892 3,865,870 1,860,035 2,005,835 1549 years of age % of 1549 year 3.69% 3.23% 4.29% 5.72% 5.03% 6.37% olds tested 62 NOTE: The number of persons reported as tested in the above tables includes only those 15-49 years of age and hence are less than those included in Annex 2 tables on beneficiaries of VCT which include all persons tested across all age groups. 6. Youth-oriented HIV/AIDS education scaled up significantly. The 4410 youth trained under the project to serve as peer educators (over 30 for each of the 129 communes) creates a strong cadre with substantial promise of reaching successive cohorts of youth with crucial information to impact their behaviors and reduce the risk of infection. The project supported establishment of an extensive network of sites to engage youth in HIV/AIDS education and dialogue, including 131 youth centers, covering all 129 communes, and 972 Stop SIDA Clubs dispersed nationwide. 7. A wide public sector commitment to increasing HIV/AIDS awareness among their staff and constituents. An important step toward slowing the spread of infection is the project's success in mobilizing a truly multisectoral approach, with all 30 central level public sector institutions having instituted annual HIV/AIDS Action Plans by project end, compared to only 8 with such Plans at the project's inception in 2002. These Plans, heavily focused on IEC, over the project period reached in aggregate 308,065 civil servants, their families and sector constituents, not only raising their awareness of the epidemic but also of behavioral changes needed to protect against infection. (Annex 2) 8. Improved safety and management of the nation's blood supply. A key goal of the NASP that MAOP supported was to reduce HIV transmission through systematic screening of the nation's blood supply. Over the project period, as noted below, the prevalence of HIV among those donating blood appeared to fall although the fact that the prevalence was reported as exactly 0.40 percent from 2003 to 2006 raises questions of data quality. 63 Table 11.3: HIV/AIDS Prevalence Rate among Voluntary Blood donors. 2002-2008 Year 2002 2003 2004 2005 2006 2007 2008 # of pouches of blood tested 14,541 17,887 18,940 21,163 18,172 20,256 20,253 Sero-prevalence rate of blood 0.50% 0.40% 0.40% 0.40% 0.40% 0.33% 0.36% donors Source: National AIDS Council, Monitoring and Evaluation System 9. Massive distribution of condoms: Consistent with the project's approach of reaching the general population with key preventive activities, over 52 million condoms were distributed over the life of the project. Table 11.4: Number of Condoms distributed, 2002-2008 Year 2002 2003 2004 2005 2006 2007 2008 Number of 5,146,551 7,042,784 8,857,129 1,0399,064 6,704,372 8,091,481 8,444,184 condoms distributed Mitigating the adverse effects of HIV/AIDS. 10. Significant expansion of PMCT sites to serve both mothers and newborns. The number of PMCT sites increased from 1 in 2002 to 63 in 2008, contributing directly to a ten-fold increase in the number of pregnant sero-positive women treated over this same period (Annex 2). Despite this major scale up of effort, unmet needs for PMCT are significant, with less than 8 percent of HIV infected pregnant women covered by the project's end, as highlighted below. Table 11.5: Number of Pregnant Women Received PMTC, 2007 and 2008 Year 2007 Year 2008 Number of pregnant sero-positive women who 1,102 1,488 received a complete ARV prophylaxis to reduce the risk of mother-to-child transmission Estimated number of pregnant sero-positive women 18,000 19,000 in need of PMCT. % of all pregnant women at risk who actually 6.12% 7.83% received PMCT 11. Similarly, the 1,299 newborns of sero-positive mothers provided ARV treatment in 2008 compared to only 274 in 2002 is an important step toward reducing new infections (Annex 2), but coverage at project end remained far below needs, as shown below. 64 Table 11.6: Number of Newborns under ARV Treatment, 2007, 2008 Year 2007 Year 2008 Number of newborns of seropositive mothers who 976 1,299 received prophylaxis treatment Estimated number of newborns of seropositive 18,000 19,000 mothers in need of ARV treatment % of newborns at risk who received ARV treatment 5.42% 6.84% 12. Rapid expansion of ARV and OI treatment capacity: Establishment over the project life of 68 centers for ARV treatment, coupled with the expansion of home-based follow-up and care by CSO grant- financed subprojects, training of providers in ARV treatment, and project financing of ARVs and other drugs and supplies, have contributed to a 23-fold increase in the number of PLWHA under treatment from 600 in 2002 to 14,343 in 2008. The 43.7 percent (of 10 percent target ) ARV coverage rate attained in 2007, substantially exceeded the average of 30 percent for the Sub-Saharan Africa region. Table 11.7: Expansion of ARV treatment sites and beneficiaries, by year Number of ARV treatment sites 80 70 68 60 sites 50 of 46 40 36 30 26 20 23 Number10 0 0 0 2002 2003 2004 2005 2006 2007 2008 Year 65 Number of people with ARV treatment 16,000 14,000 14,343 12,000 people 10,000 10,928 of 8,000 8,048 6,000 6,416 4,000 3,900 Number 2,000 1,210 0 600 2002 2003 2004 2005 2006 2007 2008 Year Source: National AIDS Council, M&E system Note: *the 43.7 percent coverage in 2007 is based on an estimated total of 250,000 HIV+ persons in Burundi in 2007 of which 10 percent or some 25,000 are clinically/biologically eligible for ARV treatment, consistent with annual coverage targets in the 2007-2011 NASP. Recent data from the national census and the lower than expected prevalence rates of the HIV/ADS survey suggest that in 2007 the number of HIV+ people was probably between 215,000-225,000. 13. The significant expansion of skills of health providers in appropriate ARV and OI treatment is highlighted below and in Annex 2. Table 11.8 Number of Providers Trained in Treatment of OIs 2003 2004 2005 2006 2007 2008 Number of health providers 32 160 719 438 488 264 trained in IO treatment by year Number of health providers 1198 1548 1898 2248 2598 2948 % of providers trained in treatment of IOs 2.67% 10.34% 37.88% 19.48% 18.78% 8.96% Number of health providers trained in ARV treatment 140 200 395 415 488 264 Number of providers 1198 1548 1898 2248 2598 2948 % of health providers trained in ARV treatment by year 11.69% 12.92% 20.81% 18.46% 18.78% 8.96% 66 Table 11.9: Multisectoral HIV/AIDS Control and Orphans Project (MAOP) Number of Providers Trained in Priority HIV/AIDS Interventions 2002 ­ 2008 Type of Training 2002 2003 2004 2005 2006 2007 2008 TOTAL Diagnosis & treatment of STDs 160 200 818 270 245 210 1903 VCT 60 60 667 192 259 210 1448 Treatment of OIs 32 160 719 438 488 284 2121 ARV treatment 140 200 395 415 488 264 1902 Source: National AIDS Council, Monitoring and Evaluation System, April 2009 15. The investments in infrastructure and human resources permitted an increase in the number of Burundians benefitting from priority services of the HIV/AIDS Program over the project life, as revealed below. Of special note is the almost 700 percent increase in the number of persons receiving VCT and the 2291 percent increase in the number of PLWHA under ARV treatment. The increased access to ARV treatment likely encouraged testing. Table 11.10: Multisectoral HIV/AIDS Control and Orphans Project (MAOP) Beneficiaries of Priority HIV/AIDS Interventions 2002 ­ 2008 Number of Beneficiaries 2002 2003 2004 2005 2006 2007 2008 VCT 30,412 50,303 71,729 103,951 150,092 147,575 236,988 PMCT (Pregnant sero+ women 145* 252 434 977 1,129 1,102 1,488 treated) Newborns of Sero+ women treated 274 215 400 774 725 976 1299 ARV Treatment (Cumulative # Adults and children 600 1210 3900 6,416 8,048 10,928 14,343 under treatment) PLWHA treated for OIs 25,851 28,557 38,146 38,935 PLWHA receiving Psychosocial 20,714 27,511 34,239 72,780 Counseling STDs Testing & treatment 19,483 17,751 21,878 21,036 31,790 33,781 14. Substantial improvements in well-being of orphans: The project contributed to a major expansion of access by orphans and vulnerable children to education and other priority services, as highlighted below. 67 Table 11.11: Orphans' Well-being, 2002-2008 2002 2003 2004 2005 2006 2007 2008 Number of OVC (girls and boys) under 18 years of age living with a support family who received school support, by year 2500 12500 64820 153,292 163,229 0 0 Estimated total number of OVC 701,665 721,136 740,606 760,603 781,139 802,230 803,617 % receiving school support 0.36% 1.73% 8.75% 20.15% 20.90% 0.00% 0.00% Number of OVC benefitting from a medical consultation, by year 25,000 25,649 3,000 14,704 23,704 Estimated total number of OVC 740,606 760,603 781,139 802,230 803,617 % receiving medical consultations 3.38% 3.37% 0.38% 1.83% 2.95% Number of OVC receiving psychosocial support by year 1,454 12,151 165,000 10,826 26,382 43,507 Estimated total number of OVC 721,136 740,606 760,603 781,139 802,230 803,617 % receiving psychosocial support 0.20% 1.64% 21.69% 1.39% 3.29% 5.41% Note: Commencing in 2007 financing of OVC school support was assumed by UNICEF, with MAOP funds shifted to other areas 15. As documented in a 2006 impact evaluation of key project-financed interventions in the three pilot provinces in which the orphans' component initially was concentrated, these inputs generated numerous positive outcomes for both orphans and their families. Among the most notable findings were increased rates of primary school enrollment and reduced rates of repetition and dropout, with supported orphans attaining much higher promotion and much lower school dropout rates than their peers who were not orphaned. Notably, over the three school years from 2002-3 through 2004-2005--the focus of the evaluation-- the number of orphans receiving school assistance almost quadrupled from 5245 in the 2002- 3 academic year to 19,565 in the 2004-5 academic year, the latter representing virtually the totality of the estimated 19,800 orphans of primary school age in the three provinces in 2004-5. Importantly, by the 2004-5 academic year: (i) primary school enrollment of assisted orphans had increased to 98.8 percent compared to an average enrollment rate among their non-orphan peers of 84.9 percent in the three pilot provinces; (ii) assisted orphans' rate of promotion from one primary school grade to the next was 80.7 percent compared to a promotion rate of between 65.1 percent and 77.7 percent in the general school age population in those provinces; and (iii) the rate of repetition was 16.97 percent and of dropout was 4.1 percent among orphans assisted under MAOP compared to average repetition rates ranging from 27.8 - 32.7 percent and of dropouts between 3.5 percent and 5.4 percent in the primary school age population in the three provinces. 4 Moreover, although the overall numbers of orphans enrolled at the secondary school level was significantly lower than those attending primary school, major advances also were achieved in this key age group: the number of orphans assisted increased from 176 in the 2002-3 academic year to 1655 in the 2004-5 academic year, and their rate of promotion increased from 60.2 percent in the 2002-2003 school year to 83.4 percent in the 2004-5 school year. 16. These advances contrast sharply with the findings of a 2003 social analysis in the pilot area indicating that only 43.2 percent of orphans of school age were enrolled in school; that dropouts were common due to lack of funds to pay school fees and other educational needs; low motivation to continue schooling; halting of schooling, particularly by girls, to carry out domestic chores including caretaking of brothers and sisters; interruption of schooling during the national conflict coupled with low reintegration 4Source: Table 25, p. 35 of "Etude d'Impact Qualitatif et Quantitatif de l'experience pilote de prise en charge des OEV dans les provinces de Karuzi, Kayanza et Muramvya." Rapport Definitif, November 2006. Ministere a la Presidence Charge de la Lutte Contre le SIDA et Conseil National de Lutte Contre de SIDA. Comparative data for the general school age population in the three provinces were provided to the evaluation group by the Bureau of Education Planning in Bujumbura. 68 in schools by those older orphans; and non-enrollment of orphans because the support families, with limited resources, had favored enrollment of their own children. 17. Other important outcomes that the 2006 orphans component evaluation revealed included: attainment of new vocational/technical competencies among older orphans attending project-supported workshops; improvements in the living conditions of orphans as a result of project supported economic and food supports to the families in which the orphans reside5; and increased identification and resolution of cases of orphan exploitation and abuse, as highlighted in Annex 10. 18. In summary, the above outcome indicators demonstrate progress made toward achievement of the PDOs over a relatively short time period and under difficult country circumstances. These may be better gauges of project performance than relying solely on changes in the 10 core KPIs for several reasons: (i) original baseline values presented in the PAD and ensuing DCA for two of the core outcome indicators were incorrect, thus significantly underestimating the magnitude of improvement expected in these indicators by the end of the project;6 (ii) two KPIs related to new OVC initiatives7 for which there were no baseline data and hence no solid basis for setting targets; and (iii) the low starting point from which the MAOP embarked, with years of conflict contributing to the large proportion of the population with limited, if any, access to HIV/AIDS information or to condoms or, importantly, to sites and trained providers to obtain priority care and services such as VCT, PMCT, ARV, among other HIV/AIDS associated health care needs. In aggregate, these factors strongly suggest that many KPI targets were unrealistic. Based solely on DCA KPIs, only one of the four outcome targets, two of the three output targets and one of the three process targets were fully met; and one process target was partially met.. 5The number of households in which orphans resided reporting having just one meal per day fell from 39.7% before the MAOP support to support families to 18.9% with MAOP support. Concomitantly the quality of the meal improved measurably in project-assisted households. This underscores the contribution of the income generating activities to the overall well-being of orphans and their host families. Source: Etude d'impact noted in Footnote 41. 6The baseline for percentage of adults using condoms with non-regular partners actually was 12.7 percent, not 43 percent; and the percentage of girls 15 and younger who were sexually active was 32.5 percent, not 18 percent. 7Percentage of orphans advancing from one grade of primary school to the next and percentage receiving psychosocial support within their communities. 69