58919 The World Bank Middle East and North Africa and the Global HIV/AIDS Program Health, Nutrition and Population CURBING THE HIV/AIDS EPIDEMIC IN DJIBOUTI October 2010 Photo: James Martone CURBING THE HIV/AIDS EPIDEMIC IN DJIBOUTI The World Bank Middle East and North Africa World Bank Global HIV/AIDS Program Discussion Paper October 2010 ii World Bank Global HIV/AIDS Program Discussion Papers This series is produced by the Global HIV/AIDS Program of the World Bank's Human Development Network, to publish interesting new work on HIV/AIDS quickly, make it widely available, and encourage discussion and debate. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations or to members of its Board of Executive Directors or the countries they represent. Citation and the use of material presented in this series should take into account that it may be provisional. Papers are posted at www.worldbank.org/AIDS (go to "publications"). For free print copies of papers in this series please contact the corresponding author whose name appears toward the bottom of the next page. Enquiries about the series and submissions should be made to Martin Lutalo. Submissions should have been previously reviewed and cleared by the sponsoring department, which will bear the cost of publication. The sponsoring department and author(s) bear full responsibility for the quality of the technical contents and presentation of material in the series. For questions, please contact: Martin Lutalo World Bank, 1818 H Street, NW, Washington DC, 20433. Tel. (202) 458-1406 Fax: (202) 522-1252 Email: mlutalo@worldbank.org Cover photograph: Laboratory technician doing HIV tests. James Martone © World Bank © 2010 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 All rights reserved. iii Curbing the HIV/AIDS Epidemic in Djibouti Sameh El-Saharty, 1 Wendy V. Ravano, 2 Omar Ali, 3 and Maissa Abdel-Rahmane 4 This study was undertaken by the Human Development Department, Middle East and North Africa Region. Abstract: This study outlines the initial challenge presented by the HIV/AIDS epidemic, describes Djibouti's response, reviews the results achieved and the enabling factors in curbing the spread of the epidemic, and identifies remaining challenges. Between 2002 and 2008, HIV prevalence among young pregnant women aged 15­24 was reduced from 2.7 percent to 1.9 percent, and among sentinel surveillance groups from 2.5 percent to 1.9 percent. HIV prevalence among tuberculosis patients was reduced from an estimated 22 percent to 12 percent. Condom use during last intercourse outside marriage increased from 27 percent to 55 percent and reached 95 percent among sex workers. Among the general population, awareness of HIV/AIDS increased to 95 percent and knowledge about transmission and prevention rose to 50 percent. Political commitment, engagement of community and religious leaders, rigorous communication, social marketing and the provision of an integrated package of medical and social services, and donor harmonization were among the key factors that contributed to the achievement of these results. Despite these impressive results in a relatively short period, Djibouti still has to address several challenges and consolidate program gains, but most importantly, funds are being mobilized from government resources to sustain the national AIDS control program. Keywords: HIV, AIDS, Tuberculosis, Malaria, Voluntary Counseling and Testing, Sexually Transmitted Diseases, Antiretroviral Therapy, Middle East and North Africa, Horn of Africa, Djibouti, World Bank. Disclaimer: The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors, and do not represent the views of the World Bank, its Executive Directors, or the countries they represent. Correspondence Details: Sameh El-Saharty, The World Bank, 1818 H Street, NW, Washington DC, 20433, US. Tel. +1 (202) 458-7014, fax: +1 (202) 522-2955, email: selsaharty@worldbank.org. Suggested Citation: El-Saharty, Sameh., Ravano, Wendy, Ali, Omar, and Abdel- Rahmane, Maissa. 2010. "Curbing the HIV/AIDS Epidemic in Djibouti." HNP/GHAP Discussion Paper. The World Bank, Washington, DC. 1 Senior Health Policy Specialist, South Asia Region, the World Bank. 2 Public Health Consultant, the Middle East and North Africa Region, the World Bank. 3 Former Executive Secretary, National AIDS Control Program, Djibouti. 4 Research Analyst, the Middle East and North Africa Region, the World Bank. iv ACRONYMS AND ABBREVIATIONS ADEPF L'Association Djiboutienne pour l'Equilibre et la Promotion Familiale ARV Antiretroviral CBA Community-based association CAB Coordination des Associations de Balbala CAMME Centrale d'Achat des Médicaments et Matériel Essentiels (Central Purchasing Agency for Essential Medicines and Supplies) CCAF Cabinet Conseils Appui et Formation CIS Cabinet Ingénierie Sociale CRS Comités Régionaux de Santé (Regional Health Committees) CREDES Centre de Recherche, d'Etude et de Documentation en Economie de la Santé (Center for Research and Studies in Health) CCMI Comité de Coordination Multisectorielle et Inter-partenariale (Multisectoral and Interpartner Coordination Committee) CISU Community Intervention Support Unit (Unité d'Appui aux Interventions Communautaires) CSW Commercial sex worker DASAP Dispositif d'Appui Social Acceleré aux Personnes vivant avec le VIH SIDA (Accelerated Mechanism of Support to Persons Living with HIV/AIDS) ES Executive Secretariat FGC Female genital cutting IC Interministerial Committee (to fight HIV/AIDS, Malaria and Tuberculosis) ITC Intersectoral Technical Committee (to fight HIV/AIDS, Malaria and Tuberculosis) M&E Monitoring and evaluation MDG Millennium Development Goal NACP National HIV/AIDS Control Program NASP National HIV/AIDS Strategic Plan OPS Office de Protection Sociale (Social Protection Office) NGO Nongovernment organization PLS Programme de Lutte contre le SIDA (AIDS Control Program ­ in line ministries) PLHIV People living with HIV PMTCT Prevention of mother-to-child transmission (of HIV) PVG Priority vulnerable group STI Sexually transmitted infection UDC Union pour le Développement et la Culture UNAIDS Joint United Nations Programme on HIV/AIDS UNDP United Nations Development Programme UNFD Union Nationale des Femmes Djiboutiennes UNFPA United Nations Population Fund UNICEF United Nations Children's Fund USAID United States Agency for International Development VCT Voluntary counseling and testing WHO World Health Organization v TABLE OF CONTENTS ACKNOWLEDGMENTS .............................................................................................. vii EXECUTIVE SUMMARY ........................................................................................... viii Initial Challenge ........................................................................................................... viii Response ...................................................................................................................... viii Results and Enabling Factors ....................................................................................... viii Remaining Challenges ................................................................................................... ix INTRODUCTION............................................................................................................. 1 HIV/AIDS SITUATION PRIOR TO 2002 ..................................................................... 1 HIV Risk Factors Specific To Djibouti .......................................................................... 1 DEVELOPMENT OF THE NATIONAL HIV/AIDS STRATEGIC PLAN ............... 2 Baseline Studies and Vulnerability ................................................................................. 2 Strategy Development ..................................................................................................... 4 NATIONAL RESPONSE TO THE EPIDEMIC ........................................................... 5 Institutional Capacity and Sector Management Response .............................................. 5 Health Response............................................................................................................ 16 Multisectoral Response ................................................................................................. 27 Community Response ................................................................................................... 31 RESULTS AND ENABLING FACTORS .................................................................... 36 REMAINING CHALLENGES...................................................................................... 38 ANNEX 1: COUNTRY AND SECTOR BACKGROUND....................................... 41 TABLES Table 1: Support from Development Partners to the National HIV/AIDS Control Program, 2002­2010 ..................................................................................................... 15 Table 2: Antiretroviral Therapy, 2004­2008 ................................................................ 21 Table 3: Priority Vulnerable Groups per Umbrella NGO, 2004­2008 ......................... 33 FIGURES Figure 1: National HIV/AIDS Control Program Institutional Structure, 2003............... 6 Figure 2: Cumulative Number of Condoms Distributed (millions) .............................. 11 Figure 3: Condom Sales Points ..................................................................................... 12 Figure 4: National HIV/AIDS Control Program Institutional Structure, 2009............. 14 Figure 5: HIV/AIDS Transmission Dynamics .............................................................. 17 vi Figure 6: Growth of Voluntary Counseling and Testing, 2004­2008 .......................... 19 Figure 7: Total Number of STI Cases Diagnosed and Treated, 2002­2008 ................. 23 Figure 8: HIV Prevalence among new TB patients, 2002­2008 .................................. 25 Figure 9: Utilization of PMTCT Services, 2003­2008................................................. 26 Figure 10: HIV Seroprevalence in Djibouti, 2002­2008 .............................................. 27 Figure 12: Distribution of Community-based Associations ......................................... 33 BOXES Box 1: Guides and Other Documentation ....................................................................... 8 Box 2 : Psychosocial Companions ................................................................................ 22 Box 3 : Soeur-à-Soeur Project ...................................................................................... 35 vii ACKNOWLEDGMENTS This case study was prepared by a team composed of Sameh El-Saharty, then Senior Health Policy Specialist, the Middle East and North Africa Region, the World Bank; Wendy V. Ravano, Senior Public Health Consultant; Omar Ali, then Executive Secretary of Djibouti's Executive Secretariat of the fight against HIV/AIDS, Tuberculosis, and Malaria; and Maissa Abdel-Rahmane, Research Analyst. The study was peer reviewed by Mariam Claeson, HIV/AIDS Advisor, South Asia Region and edited by Joy de Beyer, Senior Knowledge Management Officer, World Bank Global HIV/AIDS Program. The team would like to express its gratitude to the staff of the Executive Secretariat of the fight against HIV/AIDS, Tuberculosis, and Malaria in Djibouti for the valuable information that they provided for this study. viii EXECUTIVE SUMMARY INITIAL CHALLENGE The first AIDS cases were reported in Djibouti, a small country in the Horn of Africa, in 1986. The situation steadily worsened: at the end of 2000, there were 2,179 registered AIDS cases. Between 1994 and 2000, several surveys, including of pregnant women and blood donors, showed prevalence to be particularly high among certain groups. Information for the general population was inconsistent until closer analysis in 2002 revealed that prevalence was highest among 20­35-year-olds (5 percent); transmission was mostly heterosexual; and persons aged 15­29 represented 47.4 percent of registered AIDS cases, with women infected at younger ages than men. RESPONSE A rigorous situation analysis funded by the World Bank in 2002 enabled careful response planning. Djibouti's National HIV/AIDS Strategic Plan, National Malaria Strategic Plan, and National Tuberculosis Strategic Plan were implemented through a wide variety of public sector agencies, as well as private, nongovernmental, and community-based organizations. Effective December 2003 (and closed in September 2008), the Djibouti HIV/AIDS, Malaria and Tuberculosis Control Project funded by the World Bank sought to (i) prevent HIV infection by contributing to changes in behavior among the Djiboutian population, particularly among young people, using social communication, peer education, and multisectoral, civil society, and community initiatives; (ii) provide care, support, and treatment to people with HIV in Djibouti; and (iii) treat and control the spread of malaria and tuberculosis. A World Bank/International Development Association (IDA) grant contributed $12 million for the project. The project supported the national coordinating structure for the three diseases and strengthened public, private, and nongovernmental institutions. A Health Sector Development Project approved in 2002, also financed by IDA, helped jump-start implementation of the HIV/AIDS project: fiduciary staff, procedures, and mechanisms were already in place. In 2005, a grant from the Global Fund to Fight AIDS, Tuberculosis and Malaria (the Global Fund) used the structures set up under the IDA-funded project to rapidly increase access to antiretroviral (ARV) treatment. The Global Fund agreed to use the existing national authority as the grant recipient, to channel funds through the existing structures and processes, and to rely on the existing national HIV monitoring and evaluation system to track progress. From the start, all three organizations--IDA, the Global Fund, and the national program authority--were committed to working together in a fully cooperative, harmonized way. RESULTS AND ENABLING FACTORS Between 2002 and 2008, the National HIV/AIDS Control Program (NACP) achieved the following key results: ix · HIV prevalence among young pregnant women aged 15­24 was reduced from 2.7 percent to 1.9 percent, and among sentinel surveillance groups from 2.5 percent to 1.9 percent, thus suggesting the curbing of the epidemic. · HIV prevalence among tuberculosis patients was reduced from an estimated 22 percent to 12 percent. · Condom use during last intercourse outside marriage increased from 27 percent to 55 percent and reached 95 percent among sex workers; in total, more than 2.6 million condoms were distributed, up from only 30,000 condoms in 2004. · Among the general population, awareness of HIV/AIDS increased to 95 percent and knowledge about HIV transmission and prevention rose to 50 percent. · Voluntary counseling and testing increased more than fourfold, from 1,461 in 2004 to 7,158 tests in 2008. About 85 percent of women attending prenatal care services accepted counseling and 76 percent agreed to be tested for HIV. Various enabling factors contributed to these results. Political commitment was a key factor in mobilizing resources and implementing a multisectoral response. Another critical factor was engaging community and religious leaders in addressing cultural concerns and adapting messages to be consistent with social values. In fact, getting the religious leaders in a Moslem country to endorse the use of condoms was a remarkable breakthrough in changing the attitude of many religious segments of the population. Increasing awareness was successful mainly as a result of extensive and diverse use of communication channels focusing on radio, and on the community response targeting high risk and priority vulnerable groups such as sex workers, coupled with a vigorous condom distribution system at the national level to meet the increasing demand. Also, the provision of an integrated package of medical and social services was very effective, particularly the introduction of ARV therapy together with prevention interventions. In addition, the use of "umbrella nongovernmental organizations" was instrumental in building capacity and supporting the implementation of community projects. Finally, harmonized donor support for one national program executed through one national coordinating entity greatly facilitated successful implementation. REMAINING CHALLENGES Despite achieving many impressive results in a relatively short period, Djibouti still has a lot to do to address challenges, consolidate the gains, and sustain the NACP. Among these challenges is the need to develop alternative strategies to promote the use of voluntary counseling and testing services among people in uniform. Case management of sexually transmitted infections needs further strengthening. Regarding behavior change, communication and awareness campaigns need to be strengthened and social marketing needs to adopt a more rigorous and targeted strategy to increase condom use, particularly among high-risk groups. The multisectoral response was very complex and difficult to implement at the beginning of the program and should continue to focus on the few critical ministries and sectors. Finally, the NACP has been funded almost exclusively by donor funds, an approach that is not sustainable. Mobilizing funds from government resources is critical to sustain the minimum program components and ensure its viability. Curbing the AIDS Epidemic in Djibouti 1 Cur bing the HIV/AIDS Epidemic in Djibouti INTRODUCTION The Republic of Djibouti is a small country in the Horn of Africa with an estimated population of 830,000 (2008). The natural annual rate of population growth is around 2.4 percent. Two thirds of the population lives in the capital city Djibouti-Ville, most of the remaining population live in a handful of secondary towns. Illiteracy, unemployment and poverty are high -- 60% of Djiboutians live below the poverty line. Overall, life expectancy was 54 years in 2008. Djibouti's infant and child mortality rates were 67 and 94 per 1,000 live births in 2007, respectively. The maternal mortality ratio for 1998­2004 was 673 deaths per 100,000 live births. (Please refer to Annex I for more details on the country and the health sector background information). HIV/AIDS SITUATION PRIOR TO 2002 The first AIDS cases were reported in Djibouti in 1986. The situation worsened steadily, and at end-2000 there were 2,179 cases. Between 1994 and 2000, several surveys found particularly high HIV prevalence in commercial sex workers (CSWs), dock workers, the police, and the army (the last two together, "people in uniform"). Data on the general population were unreliable. HIV RISK FACTORS SPECIFIC TO DJIBOUTI Trade Djibouti is a highly urbanized state (more than 80 percent of the population live in urban areas). Economic activity is centered on the port, which serves the Horn of Africa and services French military bases. In addition, about 1,000 transport trucks enter and leave Djibouti's port each day to supply Ethiopia's needs (a country about 100 times larger in population). The estimated HIV prevalence rate in the adult population of Ethiopia was 11 percent at end-1999. Djibouti was highly susceptible to the spread of HIV through the transport sector, and there was the risk of the disease spreading in the sub-region along truck routes. In addition, Djibouti-Ville's traders and military bases attract many sex workers. Sexually transmitted infections (STIs), another factor contributing to the spread of the epidemic, were frequent, with the number of infections estimated at 25,000 per year. The population was (and remains) young, and the age of first sexual contact was low. Condoms were not readily available and were expensive. 2 Migration The situation was further complicated by the large influx of refugees and displaced persons as well as the movement of nomad populations across borders. There was very little cooperation with neighboring Ethiopia, Somalia, and Eritrea in assessing HIV/AIDS prevalence rates, providing voluntary counseling and testing (VCT), formulating consistent behavioral change communication messages, or treating TB, which is a common co-infection of HIV/AIDS. Gender Inequality Women are the most vulnerable segment of society. Economically, women have low labor force participation rates and low levels of education. (In the 20­29 age range, 60 percent of women, compared to 30 percent of men, have never attended school.) The maternal mortality ratio is one of the highest in the world. This is due to high fertility rates, anemia caused by malnutrition, and the widespread practice of female genital cutting (FGC). About 99 percent of women have been subjected to FGC (mostly infibulations, that is, the most extreme type), which often leads to other health problems. A law prohibiting the practice was adopted in 1995, but the private nature of the procedure makes it difficult to enforce. UNICEF, the Ministry of Health, CARITAS (a Catholic charity), and UNFPA jointly initiated a project to reduce FGC in 1999. A multi- institutional professional team conducted awareness meetings with decision makers at the central level. In addition, meetings to increase awareness were conducted with religious leaders to address the problem of FGC and to identify ways to limit or eradicate it. DEVELOPMENT OF THE NATIONAL HIV/AIDS STRATEGIC PLAN The initial HIV/AIDS response was carried out in a health system whose challenges and deficiencies were amplified by the epidemic. With assistance from the World Bank, the government of Djibouti commissioned the Centre de Recherche, d'Etude et de Documentation en Economie de la Santé (Center for Research and Studies in Health; CREDES) to do two major studies: (i) from May to December 2001, a situation and health system analysis and preparation of a national strategic health plan 2002­2011; and (ii) from March to September 2002, a situation analysis pertaining to HIV/AIDS and analysis of the national response, followed by the preparation of a national intersectoral HIV/AIDS strategic framework 2003­2007 and operational plan 2003­2005 (see CREDES 2003). BASELINE STUDIES AND VULNERABILITY Studies were carried out to provide baseline data for the general population and some specific groups, including STI patients, people in uniform, and CSWs. Baseline studies included (i) epidemiological studies in Djibouti on HIV/AIDS and other STIs (1986­ 2001), and (ii) an HIV prevalence study and HIV/AIDS and STI knowledge, attitude, 3 practice, and behavior study. The latter studies were carried out among the general population, school children, people in uniform, and dock workers from March to November 2002. People living with HIV (PLHIV) were involved in the design and implementation of a social assessment of the impact of AIDS on families and individuals. A national seroprevalence survey was conducted in 2002 to determine the HIV prevalence rate and guide the development of the control strategy. It revealed a national HIV prevalence rate of 3.1 percent for the whole population. This was lower than expected (expectations were partly founded on an earlier UNAIDS estimate among 15­ 49 year olds). Djibouti's HIV prevalence rate was below that of Sub-Saharan Africa (averaging 6 percent). HIV prevalence standardized by age and gender among 15­45 year olds was 3.4 percent (CI: 95 percent: 1.1­5.6) in Djibouti-Ville, and 1.1 percent (CI: 95 percent: 0.3­1.8) in other districts. However, further analysis by age group showed prevalence above 5 percent among persons aged 20­35, confirming that infection was affecting younger age groups, including the most economically productive age groups. This situation was cause for concern and called for a multisectoral response along with full recognition by the government of HIV/AIDS as a development issue. The national seroprevalence survey further revealed that (i) transmission was mostly heterosexual (95.6 percent of declared cases among women and 91.6 percent among men, and roughly even numbers of cases); (ii) persons aged 15­29 accounted for 47.4 percent of registered AIDS cases, which showed that people were infected at an early age; and (iii) women were infected at a younger age than men: women aged 15­29 represented 54.3 percent of declared cases, while men of the same age group represented 42.7 percent. Underlying the seriousness of the situation were data from an unpublished study conducted on 1,334 pregnant women in October­November 2002 in eight sentinel centers at prenatal care services in Djibouti-Ville. It revealed a seroprevalence rate of 2.5 percent (CI: 95 percent: 1.7­3.5). In the World Health Organization (WHO) categorization, seroprevalence among pregnant women over 1 percent is one of the criteria for a "generalized epidemic" (WHO 2000). Other available data, such as from the Office de Protection Sociale (Social Protection Office; OPS) in 2000 and 2002, also point in this direction. HIV vulnerability stems from a range of factors beyond the control of individuals, families, and communities who, without capacity building and tailored prevention measures, have less ability to avoid HIV infection. The national seroprevalence survey (as well as CREDES surveys and stakeholders consultation meetings during the project preparation phase) identified priority vulnerable groups (PVGs) for HIV/AIDS (as well as for tuberculosis) as follows: · Youth (aged 15­19), in and out of school · Truck drivers and dock workers (Ethiopia­Djibouti corridor personnel) · CSWs and high HIV-prevalence women who work in bars and their clients 4 · Persons in uniform · PLHIV (they are vulnerable to TB) · People living under the poverty line (they are also vulnerable to TB) · TB patients in the districts and Djibouti-Ville. (See also Table 3 below for a listing of groups at higher risk by agences d'encadrement, or "umbrella nongovernmental organizations".) 5 STRATEGY DEVELOPMENT Prior to the establishment of the National Program against HIV/AIDS, Malaria and Tuberculosis 6 in 2002--hereafter referred to as the National HIV/AIDS Control Program (NACP) 7--the lack of strategic planning had hindered the HIV/AIDS response. The following were the consequences: (i) insufficient focus on priority strategies and activities leading to unnecessary duplication, failure to analyze key sectors, and ineffective individual initiatives; (ii) poor understanding of roles, linkages, and responsibilities of actors at national level, thereby wasting resources; (iii) difficulty for donor agencies to contribute to the national HIV/AIDS response; and (iv) obstruction to mobilizing external financial resources caused by the lack of ownership of strategies and programs at higher levels of authority. In this context, the two major studies (mentioned above), for which CREDES provided technical assistance served as a foundation for strategy development. In June 2002, a national consensus workshop 8 validated the strategic and operational directions. A representative sample of local communities was also consulted. Several workshops were subsequently held in the five districts and in Djibouti-Ville to help finalize the National HIV/AIDS Strategic Plan (NASP) and to discuss the findings of knowledge, attitude, practice, and behavior studies. The NASP was approved during a second consensus seminar in December 2002 that included all partners. 9 The main strategic thrusts of the NASP were as follows: · Strengthen the coordination of all actors to achieve a broadened national, regional, and international response to the HIV epidemic 5 The translation of agence d'encadrement is "supporting agency" but the concept in the English literature is known as "umbrella NGOs," which the authors use in this document. 6 In French: Programme de Lutte contre le SIDA, le Paludisme et la Tuberculose (PLSPT). 7 In French: Programme National de Lutte contre le SIDA (PNLS). NACP was designed to fight HIV/AIDS, TB, and malaria, but given that the case study focuses on curbing the HIV/AIDS epidemic, the activities related to the malaria program are not presented. As an HIV coinfection, TB is covered. 8 Attended by representatives from line ministries including the Ministry of Religious Affairs, parastatal institutions such as the OPS, leaders of civil society, associations and NGOs, UNAIDS and other United Nations agencies, World Bank, and donor agencies. 9 In all, three national strategic plans to control HIV/AIDS, malaria, and tuberculosis (Plan stratégique national de prévention du VIH/SIDA, Plan stratégique national de lutte contre le Paludisme, et Plan stratégique national de lutte contre la Tuberculose) were approved for the period 2003­2008. 5 · Strengthen systems that allow the understanding as well as monitoring and evaluation (M&E) of interventions · Strengthen measures to prevent the transmission of HIV and other STIs · Improve the quality of case management of people with HIV · Reduce the socioeconomic impact of HIV/AIDS on individuals, families, and communities. NATIONAL RESPONSE TO THE EPIDEMIC The objectives of the NACP are to contribute to change in behaviors of the population in order to contain or reduce the spread of HIV; to mitigate its impact on infected and affected persons; and to contribute to the control of malaria and tuberculosis. The NACP has a multisectoral approach to prevent the spread of HIV by reducing transmission, in particular among high-risk groups; and to expand access to treatment of opportunistic illnesses, malaria, and tuberculosis, and provide care, support, and treatment to PLHIV; and to support multisectoral, civil society, and community initiatives for HIV prevention and care, as well as malaria and tuberculosis prevention. INSTITUTIONAL CAPACITY AND SECTOR MANAGEMENT RESPONSE Institutional and Management Structure of the National HIV/AIDS Control Program In 2002, the government created an Interministerial Committee to fight HIV/AIDS, Malaria, and Tuberculosis (Comité Interministériel de Lutte contre le Sida, le Paludisme et la Tuberculose; IC) as required in its NASP, which aimed to play a role in setting national policies to control the three diseases and to stimulate the multisectoral approach. The IC was chaired by the prime minister with the vice presidency held by the minister of health. In addition to the Ministry of Health, the IC included 12 ministries (Government of Djibouti 2003, Article 2). 10An Intersectoral Technical Committee for the fight against HIV/AIDS, Malaria and Tuberculosis (Comité Technique Intersectoriel de Lutte contre le Sida, le Paludisme et la Tuberculose; ITC) was established as the technical arm of the full IC to manage the response to HIV/AIDS. The ITC was presided over by the Secretary General of the Ministry of Health and the vice presidency was assumed by the representative of the prime minister (Figure 1). 10 Ministries of Economy, Finance and Planning; Youth, Sports and Tourism; International Cooperation; National Education and Higher Education; Promotion of Women, Social Affairs and Family Welfare; Defense; Employment and National Solidarity; Communication, Telecommunication and Culture; Justice and Penitentiary and Religious Affairs; Waqfs and Islamic Affairs; Interior and Decentralization; and Equipment and Transport. 6 Figur e 1: National HIV/AIDS Contr ol Pr ogr am Institutional Str uctur e, 2003 Source: HIV/AIDS, Malaria and Tuberculosis Control Project, Project Appraisal Document, Report No. 25264-DJI, World Bank, 2003. DEHP: Direction de l'Epidémiologie et de l'Hygiène publique; DEPCI: Direction des Etudes, de la Planification et de la Coopération international; DSSB: Direction des soins de Santé de Base; MSR: Programme de la Maternité et la sante de reproduction; ONUSIDA = UNAIDS; SEPS: Service d'Education Pour la Santé. In addition to representatives of 11 of the 12 other ministries in the IC (the exception was international cooperation), the ITC had representatives of the President of the National Assembly, the Executive Secretary of the Executive Secretariat (ES; see below), the Director of Hygiene and Epidemiology of the Ministry of Health, the three coordinators of the Ministry's programs for AIDS, TB and Malaria, the National Union of Djiboutian Women (Union Nationale des Femmes Djiboutiennes; UNFD), the Chamber of Commerce, and two from civil society (Government of Djibouti 2003, Article 6). With assistance from the Executive Secretariat (ES, described below), the role of the ITC encompassed reviewing annual work plans and budgets and monitoring their implementation in all sectors, 11 in accordance with the NASP. 11 Of the ministries represented in the IC and ITC, nine ministries and the Parliament were involved in implementing the multisectoral response in addition to the Comités Régionaux de Santé (Regional Health Committees; CRSs), which were added in 2005. Therefore, 11 ministries and sectors were responsible for the multisectoral response. 7 Executive Secretariat and its Coordinating Technical Units The ES, attached to the Prime Minister's Office, was set up as a high-level body to coordinate all prevention and mitigation programs implemented by multisectoral and civil society organizations. Its secretary was nominated by the Council of Ministers. The ES was designated to be responsible for project execution, coordination, follow-up, and monitoring of project interventions in line ministries and communities. This was deemed the most appropriate mechanism for implementing the project, given the multisectoral nature of the activities being proposed and the coordination efforts needed to ensure efficient use of funds and impact of project activities. The IC and the ITC were assisted by the ES. The ES was mandated to closely liaise with the Ministry of Health Planning Unit, the Directorate of Primary Health Care, and national technical experts, who could be called on to provide assistance either as technical assistance specialists or as specialist trainers. The ES took a year and half to establish itself as an operational institution. It developed six coordinating technical units (unités techniques de coordination), also built from scratch, through which it functioned: · Administrative and Financial Unit · Multisectoral Technical Coordination Unit (Unité Technique de Coordination de la Réponse multisectorielle; MTCU) · Community Intervention Support Unit (Unité d'Appui aux Interventions Communautaires; CISU) · Communication Unit · Social Marketing Unit · Monitoring and Evaluation Unit (MEU). Since inception, the ES has been reinforced by long-term technical assistance, including a management and procurement adviser, medical advisers for AIDS clinical treatment and for VCT, a medical adviser for the TB program, and a communication adviser. It was also supported by significant short-term technical assistance experts in different areas from various development partners. By early 2005, all the ES units were staffed by national professionals, paid by various donors. The start of the program was rough and slow. The situation was exacerbated by the establishment of the Country Coordination Mechanism in 2004 to oversee preparation of the Round 4 grant proposal submitted to the Global Fund to Fight AIDS, Tuberculosis and Malaria (the Global Fund). This Mechanism was further developed and reorganized in September 2004 as the Multisectoral and Interpartner Coordination Committee (Comité de Coordination Multisectorielle et Inter-partenariale; CCMI). The CCMI is chaired by the Minister of Health and consists of a vice president, a representative of civil society, a representative of education and operations research, and a secretariat. The CCMI has four technical committees: Resource Mobilization; Comprehensive Case 8 Management; Community Interventions and Social Mobilization; and Planning, Monitoring, and Evaluation. A technical working group prepared a document of standards and guidelines to be used by a pool of trainers to derive and adapt training materials. Because the NACP, including the ES, were launched from scratch, numerous training courses were held to build national capacities to execute and implement the new program. Training was both in the country and abroad, long term and short term. Training participants included line- ministry executives, technical managers, service providers, community leaders, and peer educators. Particular emphasis was put on training service providers in VCT, since knowing one's HIV status is an entry point for prevention. It took all of 2004 and into early 2005 to fully staff the different implementation units, develop all the guidelines and protocols, train people, improve national coordination, and harmonize the monitoring indicators (see "Monitoring and Evaluation," below). This delay in program start-up was, however, extremely useful in Box 1: Guides and Other Documentation building NACP capacity. During To ensure quality and provide a common technical this time, a set of critical products reference base, the following guides and manuals were was developed, such as guidelines developed by the National HIV/AIDS Control Program: and manuals for social 1. Procurement and Administrative Manual communication, peer education, 2. Social Communication Guide community projects, and clinical 3. Peer education guides for PVGs (in French and two treatment protocols for AIDS, TB, local languages) 4. Several flipcharts, flyers, and posters for PVGs and STIs, as shown in Box 1. 5. Information guide for religious leaders 6. Guides for various community audiences Despite the unavoidable initial 7. Social marketing manual 8. Manual for the management and quality control of delay in full-scale program community-based projects implementation, the ES went on to 9. Manual for case management of STIs accomplish most of its national 10. Protocol for voluntary counseling and testing 11. Protocol for case management of Opportunistic coordinating mandate in a timely Infections fashion and to execute activities 12. Manual for prescribers of ARV drugs funded by various donors. 13. Manual for epidemiological surveillance of HIV 14. Manual for case management of mother and children in PMTCT A review financed by the World 15. Prevention of HIV transmission in clinical settings Bank in May 2007 found that the 16. Monitoring and evaluation manual ES had implemented 75 percent of its action plan by end-2006 (Conseil Santé 2007). It noted, however, that the IC, chaired by the prime minister, did not meet regularly and was inadequately engaging in overseeing progress. The ITC and the CCMI, chaired by the minister of health and responsible for technical oversight, were more actively engaged and adequately carrying out their respective mandates. Over time, the CCMI played an increasing role in steering the NACP. The ES continues to operate as a "central command" of sorts, preventing duplication of interventions and improving resource management. 9 Administrative and Financial Unit This unit has responsibility for managing projects funded by development partners. Project funds are initially assigned to the Prime Minister's Office and then delegated to the ES. This unit comprises two sections, one for procurement and one for financial management. Multisectoral Technical Coordination Unit The MTCU coordinates the activities in the action plans implemented by the 11 different ministries and sectors. Each ministry established an AIDS Control Program unit (Programme de lutte contre le SIDA; PLS). Each PLS prepared a yearly action plan that was subsequently approved by the MTCU (which also supervised implementation of these plans). Funding for this component was provided by the World Bank and, from 2007, also by the Global Fund. Originally, the strategy included a phased-in plan starting with five ministries in 2004, increasing to eight in 2005, and to 11 in 2006. However, in 2004, the IC decided to start with all ministries from year one to create national multisectoral support and awareness, and then phase them out gradually. Unfortunately, that major decision failed to take into consideration both the nascent MTCU capacity and the lack of knowledge and experience in the line ministries. At the beginning, many focal points were insufficiently committed as they perceived their role as having additional duties without compensation. Their frequent replacement caused delays in implementation, submission of action plans, and in preparation of technical and financial reports, which led to overall delays in implementing this component. In 2005, the situation improved and activities related to training, increased awareness, and peer education were implemented, but with some delays. As of 2006, support to several ministries started to be phased out as planned. Two key ministries were supported through 2008, as they had good results, and were the most involved: the ministries of Education, and of Youth and Sports implemented all their planned activities and achieved more than 75% of their expected outputs. In addition to the increased awareness activities, the National Police, Defense, and the OPS provided clinical services, including VCT and treatment (but with a modest numbers of users, as most people preferred to use Ministry of Health services to ensure anonymity). Community Intervention Support Unit This was the most innovative component and the CISU succeeded as no organization had before in Djibouti in mobilizing the community against HIV/AIDS. A key NASP component was to support and promote civil society organizations and community initiatives. Community projects were to have been developed by beneficiary groups and selected on the basis of technical quality, cost-effectiveness, and likely impact. However, NGOs were very weak in 2002 and lacked expertise, professionalism, and know-how, which therefore required in-depth systematic institutional capacity 10 building. In consultation with the development partners, the ES decided to implement capacity-building activities before launching community interventions, via a two-tier system of umbrella NGOs and community-based associations (CBAs), 12 in order to ensure effective development, supervision, and implementation of community projects. CISU manages the umbrella NGOs that in turn supervise their CBAs. CISU took one year (2004) to strategize and develop a methodology and related normative framework, including an operations manual; a geographic map of the PVGs; a generic guide for peer educators, which was later adapted for each type of PVG and adopted by the PLS in each of the line ministries; a pilot project called Dispositif d'Appui Social Accéléré aux Personnes vivant avec le VIH SIDA (Accelerated Mechanism of Support to Persons Living with HIV/AIDS; DASAP); and a social support guide. CISU started recruiting and training umbrella NGOs in early 2005, and began a quality control program for CBA activities in August 2005. In late 2005, the selected umbrella NGOs in turn each recruited and trained about 200 peer educators in different CBAs, which started their activities in early 2006, monitored and supervised by the umbrella NGOs. In 2006, about 75 community projects were financed and implemented; disbursement mechanisms to CBAs were working, albeit with a few delays that led to the disruption of some activities between two financing tranches. In 2007, CISU conducted a second round of quality control on the umbrella NGOs; strengthened, consolidated, and rolled out the DASAP; and developed a partnership with UNICEF and the Ministry of Women Promotion, Social Affairs, and Family Welfare for components for orphans and other vulnerable children, prevention of mother-to-child transmission (of HIV) (PMTCT), and HIV/AIDS programs on the Djibouti­Ethiopia transport corridor. By end-2008, community mobilization was very strong: 170 CBAs across the country were involved in the response to HIV, marking the birth of the "CBA movement" in Djibouti. CISU capably managed this component (as described in more detail under "Community Response," below). Communication Unit The Communication Unit helped prepare communication strategies for other ES technical units and the interpersonal communication services for the community projects and in line ministries, including the communication strategy of PLS-Health, with international short-term technical assistance. The Unit worked closely with PLS-Communication mainly through mass media (newspaper, radio, and TV), reviewing messages before diffusion in local languages and providing resources for outreach and purchase of air time. In addition, it managed the free AIDS hotline in the ES headquarters, and the ES website. One of the successful tools tested and developed by this unit was the boite à image or image box, a set of images used by service providers to explain how HIV can be 12 Association à base communautaire (ABC), in French. 11 transmitted and how to protect against infection. As detailed later, the success in raising HIV-related awareness and behavior change including increasing condom use is mainly due to the concerted efforts of the Communication Unit and PLS-Communication in maximizing the use of different communication channels, with particular focus on radio, the main mass media channel used by much of the population including people in remote areas. Social Marketing Unit The Social Marketing Unit was responsible for developing the social marketing strategy for condom distribution and promotion, and for coordinating these activities with line ministries and community associations. Procurement of condoms was financed by several donors including UNFPA, World Bank, and the Global Fund, and was facilitated by UNFPA. The original intention was to outsource condom distribution to an NGO, but no NGO had the necessary capacity, so the ES decided, with donors, to assign this task to the Social Marketing Unit. (The alternative -- to identify the best available NGO and build its capacity to manage condom distribution -- was considered but discarded.) This was a major shift in the ES's mandate and scope of responsibilities, since it was meant to be a coordinating rather than an implementing body. This unit faced several challenges in starting up the condom distribution program, including projecting condom needs, establishing condom sales points, managing the supply chain effectively, ensuring an uninterrupted supply, and pricing condoms. As with other ES units, it took more than a year to "get it right." By end of 2008, about 2.6 million condoms had been distributed, up from only 0.21 million in 2004 (Figure 2), and 420 sales points had been established (Figure 3). This vigorous condom distribution program was key in meeting the increasing demand for condoms generated by other program interventions that increased condom utilization. Figur e 2: Cumulative Number of Condoms Distr ibuted (millions) 2.63 3 2.29 1.77 2 1.15 1 0.21 0 Total Number of Condoms ... Source: HIV/AIDS Executive Secretariat Statistics, Djibouti. 12 Figur e 3: Condom Sales Points Source: HIV/AIDS Executive Secretariat, Djibouti. In order to sustain these efforts, the Social Marketing Unit helped establish a new NGO, the Djiboutian Association for Social Marketing (Association de Marketing Social Djiboutienne; AMASOD), which started selling its own condom brand in 2008, Prudence, at a government-subsidized price of DF30 for a strip of three condoms. Monitoring and Evaluation The Monitoring and Evaluation Unit (MEU) is responsible for establishing and monitoring the NACP performance indicators, and for supervising implementation of various surveys. The initial performance indicators were numerous and very difficult to collect and monitor. The MEU had technical assistance from WHO and World Bank financing, and together they significantly reduced the number of indicators. In 2005, the World Bank and Global Fund teams worked with the MEU to harmonize performance indicators, 13 further reducing the number and bringing them into conformity with the M&E manual on HIV/AIDS, Malaria and Tuberculosis prepared by UNAIDS partners. The MEU regularly collected data on the indicators and generated periodic (mostly quarterly) M&E reports that helped the ES and partners assess program implementation progress and adjust the interventions accordingly. NACP Management Structure Changes The initial NACP structure was complex, and had strengths and weaknesses. On the one hand, it reflected the comprehensiveness and inclusiveness needed to respond to HIV as well as the political will to develop and implement a national plan, but on the other, it did not take into account the population's limited awareness of the nature of the epidemic, the weak institutional capacity, and the lack of qualified human resources to manage a complex program. The high level of resources mobilized from the international community, and attending pressure (with some political influence) to implement the multisectoral component in all 11 ministries and sectors further strained program management, delayed implementation of several activities, and distracted the ES from other critical program components. Both the IC and ITC lacked representation from development partners and international NGOs working on HIV, and therefore did not benefit from their technical input or international knowledge and experience. The CCMI did not have representation from various line ministries, which weakened the multisectoral response. Although the CCMI is currently the leading national body for coordinating the NACP, it does not provide a forum for government agencies to consult among themselves without the influence of development partners. In 2009, the NACP management structure was simplified by reducing the number of layers and giving the ES and its technical units a more central role in managing the program, consolidating the several advisory and coordinating committees, and integrating technical support into the implementing units. Figure 4 depicts the new structure. 14 Figur e 4: National HIV/AIDS Contr ol Pr ogram Institutional Str uctur e, 2009 Donor Support, Organization, and Harmonization The many partners working with the government to implement the NACP include the African Development Bank, French Development Agency (Agence Française de Développement), Global Fund, UNFPA, UNICEF, USAID, World Bank, World Food Program, and WHO. Prior to 2002, a few partners supported small-scale HIV initiatives, pilot projects, training, and studies. It was the World Bank that led the preparation of a comprehensive program that resulted in the development of the NASP and establishment of the NACP. The World Bank supported all program components focused primarily on prevention through a multisectoral response, including capacity building of the ES and of the Ministry of Health to implement care and treatment of PLHIV. The Global Fund capitalized on the existing program structure and expanded the scope of the program, particularly in scaling up antiretroviral (ARV) therapy and the DASAP. The French Development Agency upgraded facilities and trained laboratory technicians for the blood bank. The African Development Bank financed the construction of a blood transfusion center. USAID started STI case management by the syndromic approach in the health posts that it rehabilitated, equipped, and upgraded. Various United Nations agencies provided technical support in critical areas such as M&E (WHO), PMTCT (UNICEF), and condom procurement (UNFPA). UNAIDS used its mandate to convene donors and ensure the implementation of a well-coordinated strategy. Table 1 shows the various development partners supporting the NACP during 2002­ 2010. Curbing the AIDS Epidemic in Djibouti 15 Table 1: Support fr om Development Par tner s to the National HIV/AIDS Contr ol Pr ogr am, 2002­2010 2002 2003 2004 2005 2005 2006 2007 2008 2009 2010 WHO TA in M&E and Servicesa UNICEF TA in PMTCTa UNFPA TA in Reproductive Healtha World Food Program Food Supplya French Cooperation TA in Service Delivery French Physicians and Experts ($2 million) World Bank Multisectoral Project HIV/AIDS, TB, and Malaria Control Project ($12 million) Health and AIDS Project Health Sector Development ($1 million) Global Fund Multisectoral AIDS Project HIV/AIDS Grant ­ Round 4 ($11.9 million) Multisectoral AIDS Project HIV/AIDS Grant ­ Round 6 ($10.2 million) Multisectoral TB Project TB Control Grant ­ Round 6 ($3.6 million) French Development Agency Multisectoral AIDS Project AIDS and TB Project (5 million) a. The support from United Nations agencies is mostly in the form of technical assistance and its amount could not be determined. Curbing the AIDS Epidemic in Djibouti 16 The NACP exemplifies the "Three Ones" policy: one action framework that provides the basis for coordinating the work of all partners, one national AIDS coordination authority with a broad based multisectoral mandate, and one M&E system. 13 The NACP national framework was the first of the Three Ones to be established. The second was greatly strengthened when the ES became the principal beneficiary of Round 4 of the Global Fund, in 2005. The last of the Three Ones was established in late 2005 when the World Bank and Global Fund harmonized their M&E indicators. The Global Fund grant was linked to performance results, which was new for Djibouti. The World Bank and the Global Fund harmonized their performance indicators, and M&E was further strengthened by WHO. The harmonized M&E performance indicators have served as a national reference list for all partners and as a systematic management tool. The indicators are used to monitor implementation and results and impacts of programs, identify problems, and guide adjustments in interventions. For example, the Centrale d'Achat des Médicaments et Matériel Essentiels (Central Purchasing Agency for Essential Medicines and Supplies; CAMME) consulted the new national list of indicators to determine how many patients were under treatment and how the number had been changing as the basis for estimating drug supply needs and placing timely orders. Also, when the M&E system showed an increase in the ARV dropout rate, "treatment companion" cadres were created and a VCT referral tracking system for the CBAs to ensure that patients were continuing their treatment. In 2008, UNAIDS and UNICEF assisted the NACP in evaluating progress of the NASP (2003­2007) and in developing a new strategy for 2008­2012. HEALTH RESPONSE The National AIDS Program unit of the Ministry of Health was renamed PLS-Health in 2003, similar to other PLS units established in different ministries, when the ES became responsible for coordinating the NACP, The general objective of PLS-Health was to provide prevention and case management of HIV/AIDS and STI. Specifically, it aimed to: · Promote preventive activities for HIV and other STI · Develop access to VCT services for HIV · Improve case management of STI in health centers · Decrease PMTCT and improve access to adapted case management for children exposed to or infected by HIV · Consolidate and expand comprehensive case management of PLHIV, including antiretroviral treatment · Strengthen case management of TB/HIV coinfection by developing collaborative activities between PLS-Health and the National Program to Fight Tuberculosis (Programme National de Lutte contre la Tuberculose) 13 http://data.unaids.org/UNA-docs/Three-Ones_KeyPrinciples_en.pdf. 17 · Prevent HIV transmission in clinical settings · Strengthen epidemiological surveillance and follow-up of HIV and other STI · Strengthen the management and coordination capacity of PLS-Health at the Ministry of Health. At the outset, the scope of work and intensity of activities were underestimated. During the first year, PLS-Health suffered significant delays in implementing its activities due to structural problems, such as lack of staff and decision-making authority, and some confusion about its role vis-à-vis the ES. There were delays in hiring its own accountant, and the PLS-Health Coordinator did not have enough authority as most decisions were centralized in the Ministry of Health and the ES. It was in late 2006 when these problems were addressed. These early problems led to delayed technical reviews from the Directorate of Prevention and Hygiene, which has technical oversight over all national PLSs of line ministries. Given the delays and the realization that the mandate of PLS-Health covered too many areas, the ES started to prioritize its interventions. Strategically, it was important to focus on the most critical segment of the population and to start from the existing capacity, rather than wait for the full capacity of PLS-Health to be developed. Epidemiologically, the spread of HIV is mainly driven by sex work: CSWs, casual sex workers (girls in difficulty, out of school girls, waitresses in bars, etc,), refugees, and cross-border populations and their clients (dockers, truckers, young males, men in uniform, etc), Accordingly, the strategy needed to target the following four population groups: (i) infected people who know about their infection; (ii) infected people who do not know their status (and will therefore continue their risky behavior); (iii) a group that is not infected but at high risk of infection (because of contact with the first two groups) known as the "bridging population"; and ­ with less priority ­ (iv) a group that is not infected and at lower risk of infection (Figure 5). Figur e 5: HIV/AIDS Tr ansmission Dynamics 18 For the first group, the focus was on providing ARV therapy and promoting condom use. For the second group, the priority was to quickly establish VCT services and encourage people in this group to know their status and promote condom use to limit HIV transmission to their contacts (the third, high-risk group). For the third group, the key strategy is increasing knowledge of modes of transmission and how to protect themselves. For the fourth group, increasing awareness about modes of transmission was the focus. The lack of ARV drugs was a challenge to the program. The World Bank project focused primarily on prevention, screening, and detection of HIV, but made no provision for treatment. It was argued, however, that people would be discouraged from being tested if those who tested positive were not offered treatment, which would undermine the program. In the end, it was agreed to start treatment on a pilot basis and ARV for an initial number of 200 patients was bought. In fact, this was among the very few early World Bank multisectoral AIDS projects that supported treatment, an approach that became later an integral part of such programs in Africa. From 2004, PLS-Health prioritized its interventions in this sequence: provision of VCT services, integrated case management including treatment, and HIV surveillance. In 2005, PLS-Health had three permanent technical committees: VCT Committee, ARV College (Collège des ARV) for treatment and case management, and a Committee for Surveillance. In addition, an M&E specialist was recruited in early 2007. These efforts were supported by extensive technical assistance from different donors including the World Bank, Global Fund, French Development Agency, WHO, and UNICEF. Voluntary Counseling and Testing The starting point for VCT was adding it to the services of Centre Younis Toussaint, the center for diagnosis and treatment of STI. In 2002, with UNICEF support, voluntary testing started to be offered on a small-scale, to pregnant women in four health centers providing antenatal services, adding limited HIV testing capacity. Laboratory testing was limited to Centre Younis Toussaint and Centre Paul Faure. With the establishment of the ES, starting in 2004, protocols and guidelines for VCT services were prepared through technical assistance financed by the World Bank. Internationally recognized protocols were adapted with contributions from national and in-country experts. Several training courses built skills of a new cadre of counselors and service providers. Specifically, the trainees included midwives to provide counseling for pregnant women and to prevent transmission to their babies, and laboratory technicians to provide quality testing. The protocols and training were critical for effective VCT activities and follow-up. The training was later extended to include psychosocial companions (accompagnateurs psychosociaux) to provide psychological assistance for persons exposed, infected, and affected by HIV. The mandate of Centre Younis Toussaint -- originally to provide ambulatory treatment of STIs and opportunistic diseases -- expanded in 2005 to become the national reference center for VCT services. New staff were recruited, including internationally recruited 19 experts, a national reference psychologist, national counselors, and even receptionists who were trained to provide counseling on the correct use of condoms. After the ARV pilot program was established, Centre Younis Toussaint became one of the main ambulatory centers for AIDS treatment in the country. As of end-2004, VCT services had been extended to eight centers: three in Djibouti-Ville and five in the district medical and hospital centers, which provided in total about 1,460 tests. A UNICEF survey in 2005 revealed that use of VCT services was starting to pick up, but uptake was still low among youth, only 30 percent of whom were tested and only 70 percent of those tested picked up their test results (DISED and UNICEF 2005). A national strategy document on VCT prepared in 2006 emphasized strengthening and scaling up VCT in services that were recognized as entry points (STI screening and treatment, TB services, prenatal care and family planning, and inpatient department for adults and pediatric services). This included providing VCT in community health centers and district medical and hospital centers. This markedly improved use of VCT services. It was noted, however, that VCT services seemed to be used at a late stage of infection: among those who were tested for HIV before 2007, 80 percent came at stage 3­4 as per the WHO classification, leading to lower survival rates under treatment. In early 2008, the ES designed and implemented a new strategy of "active tracked referrals," which focused on high-risk groups referred by peers, in order to increase early VCT. The network of HIV testing laboratories was equipped with two ELISA machines, one mini VIDAS, two viral charge counters; and six CD4 counters. However, the supporting technical team still needs further strengthening to cope with the increasing load. By end-2008, VCT services were provided at 28 facilities and the number of VCT tests had more than quadrupled (Figure 6). In 2008, 6,063 persons -- about 85 percent of those tested -- actually picked up their test results, and only 23 people who tested positive did not pick up their results. Figur e 6: Gr owth of Voluntar y Counseling and Testing, 2004­2008 8000 7158 6114 6000 5321 4000 2687 2000 1461 352 752 724 732 182 0 2004 2005 2006 2007 2008 Total number of HIV/AIDS Tests Total number of HIV/AIDS Positive Test Results Source: Data compiled by authors from PLS-Health Statistics, Ministry of Health, Djibouti. 20 Integrated Care and Treatment The main focus of the NACP is prevention, which is implemented through behavior change communication activities to create demand for VCT services, among other things. However, VCT services and treatment are mutually reinforcing. In fact, the NACP would not have achieved its current results if those who tested HIV-positive were not assured of confidential care and treatment. In turn, awareness of access to care and treatment encourages the population to be tested and to find out their HIV status. AIDS Management and Treatment The NACP, with support from the World Bank project and other partners, implemented a one-year treatment pilot that included provision of ARV free of charge to seropositive persons needing treatment. The treatment pilot was implemented in three public health facilities: Centre Younis Toussaint under the Ministry of Health, a health center under the Ministry of Defense, and the OPS under the Ministry of Employment. This required strengthening the health infrastructure to be able to offer ARV therapy, including provision of test kits, drugs, laboratory equipment, medical supplies and consumables. This pilot served to assess the feasibility and cost of expanding the treatment program to the whole country. The World Bank's decision to finance a treatment pilot was facilitated by the Global Fund's approval (in 2004 under Round 4) of the government proposal for a grant to finance the NACP AIDS program, including scaling up ARV therapy. AIDS treatment is complex, so existing international protocols were reviewed and a national ARV therapy protocol was developed with technical support from French physicians financed by the French cooperation program. To address multiple clinical aspects of HIV/AIDS cases, an integrated HIV/AIDS case management guide was developed that included treatment of opportunistic infections. 14 The complexity of ARV therapy and the need for regular monitoring of patients' progress and clinical response required the establishment of a body to manage the process. Accordingly, an ARV Clinical Committee (the ARV College), was formed. It was composed of four leading French and Djiboutian doctors, whose role was to treat the initial cohort of patients, examine the medical files of patients, approve their treatment protocol, and review progress every two weeks. This committee was also responsible for training other physicians on the clinical protocols, a critical step for scaling up treatment. By 2008, the program had trained 29 "referral physicians" (médecins référents). By end-2008, 27 facilities, including five in the districts, provided ARV therapy with most equipment and consumables available, and covered almost 70 percent of the country with access and resources to treat PLHIV. Public/state-owned facilities are both branded and generic ARV drugs to reduce cost. Four facilities treated 80 percent of the patients: 14 The case management guide is mainly about ARV therapy. But since an immunodeficiency condition may lead to various types of diseases and complications that may affect any body system, case management needs to be developed based on the response of various patients, resistance to ARV therapy, and the types of opportunistic infections and complications that develop in a given population. 21 32 percent at Centre Younis Toussaint; 23 percent at Peltier General Hospital; 17 percent at Centre Paul Faure; and 8 percent at Bouffard Army Hospital. This posed a challenge to the program -- if a trained medical practitioner manages fewer than 10 PLHIV per year, skills erode. In response, a yearly accreditation system was established, coupled with training, to ensure that all ARV referral physicians had the required experience and expertise. Initially, patients were started on ARV therapy if their T-cell count was below 350 cells per mm3; this was lowered in 2006 to 200 cells per mm3. About one-fourth (26 percent) of those diagnosed as HIV positive are not clinically required to start ARV therapy. The number of PLHIV started on ARV treatment increased from 192 in 2004 to 1,293 in 2008. Table 2 summarizes trends in ARV therapy from 2004 to 2008. Patients' visits to public facilities equipped for ARV therapy were affected both by fear of stigma from neighbors and by former care-seeking habits. For example, persons in uniform preferred to be treated in other facilities rather than their designated facilities, and wealthier patients preferred to be treated in private clinics or abroad. Table 2: Antir etr ovir al Ther apy, 2004­2008 Category 2004 2005 2006 2007 2008 No. % No. % No. % No. % No. % Patients who started ARV 193 192 340 311 257 (annual) Patients who started ARV 193 385 725 1,036 1,293 (cumulative) Patients who died while 34 17.6 61 15.8 110 15.2 162 15.6 188 14.5 under ARV therapy Patients who stopped ARV 7 3.6 5 1.3 22 3.0 8 0.8 43 3.3 therapy Patients who dropped out 3 1.6 26 6.8 46 6.3 94 9.1 133 10.3 (cumulative) Patients transferred abroad 0 - 0 - 13 1.80 10 1.0 28 2.30 for treatment Patients under PMTCT 1 5 22 82 110 treatment (cumulative) Patients who continue ARV 149 93.7 288 88.9 512 83.3 680 77.8 816 73.8 therapy (cumulative) Source: Data compiled by the authors from PLS-Health Statistics, Ministry of Health, Djibouti. Treatment compliance in government-operated facilities was reasonable and the survival rate at 12 months was 67.8 percent in 2008. There was a high dropout rate, though, where several PLHIV discontinued treatment for various reasons, including social and financial. The program introduced two new cadres of service providers to address this defect: · Treatment companions (accompagnateurs thérapeutiques). These are social workers trained to assist HIV patients in complying with ARV therapy and in attending scheduled laboratory tests, as well as in changing their diet and lifestyle. 22 · Psychosocial companions (accompagnateurs psychosociaux). These are social workers trained for 10 days in counseling and in basic psychological support to assist HIV patients and their family members to cope with the disease and its associated psychological distress. They meet weekly with a psychologist in a support group and undergo a six-day periodic refresher course. They also provide the same support to TB patients (Box 2). Box 2 : Psychosocial Companions Function: Psychosocial companions are companions to PLHIV; they are the liaison between the health facility and the community. They are not health workers but community workers. Location: They are based in the health facility where they have a working place and permanent presence. However, they perform activities in the community in a defined geographic area. Role: - Monitoring PLHIV and their families within the community - Providing psychological support to PLHIV and their families - Making home visits - When necessary, accompanying PLHIV to the health center, hospital, social services, etc. - Informing and educating PLHIV and their families about HIV, ways of prevention, care, and basic care/hygiene services available - Supporting treatment compliance - Serving as a liaison between the support center and other structures, and the community, etc. - Facilitating comprehensive care for PLHIV Management of Sexually Transmitted Infections A baseline survey was carried out in September 2003 in the sentinel sites to assess the STI burden and develop an appropriate control strategy. The STI Control Program developed the following objectives: · To interrupt the transmission of sexually transmitted infections · To prevent the development of disease and its complications · To reduce the risk of HIV infection · To strengthen the health system for the treatment of STIs. STI case management using the "syndromic approach" was introduced in 2004 in Centre Younis Toussaint, expanded to a few centers in Djibouti-Ville in 2005, and scaled up across the whole country in 2006 after an adequate supply of STI drugs was secured. Health centers then succeeded in making STI screening more "ordinary" and mainstream. PLS-Health planned to recruit and train assistants to work at each VCT site. The centers most used for STI screening and treatment are Centre Younis Toussaint national reference center, as well as Einguella, Khor Bourhan, and Ambouli centers. Most new 23 cases are women, and more than 45 percent are diagnosed with vaginal discharge. Among STI patients seen at Centre Younis Toussaint are people residing in other Djibouti-Ville neighborhoods, a possible indication that people might avoid the nearest center for fear of stigma. As depicted in Figure 7, the total number of STI cases diagnosed and treated has been quite volatile, decreasing from 3,370 in 2002 to 1,214 in 2005, increasing to 5,047 in 2007, then declining to 4,416 in 2008. This points to some challenges that STI management is facing, such as with systematically applying the syndromic algorithm, and underreporting of STI cases (there are discrepancies between STI drugs distributed and STI cases reported). Figur e 7: Total Number of STI Cases Diagnosed and Tr eated, 2002­2008 Total Number of STI cases Diagnosed and Treated 6000 5047 3751 4416 5000 4000 3370 2018 3000 2000 1214 1255 1000 0 2002 2003 2004 2005 2006 2007 2008 Total Number of STI cases Diagnosed and ... Source: Data compiled by Authors from PLS-Health Statistics, Ministry of Health, Djibouti. TB Management TB prevalence is among the highest in the world, and HIV coinfection was 13.8% in 2006. About one third of TB patients are from migrant populations. Until 2002, the TB program was supported by the French Cooperation Agency. The World Bank provided modest support to the program in 2004, mainly for TB/HIV coinfection, as part of its support to the NACP. In 2006, the TB program received substantial support from the Global Fund. In early 2007, the French Development Agency strengthened TB case management by assigning three technical specialists and equipping the centers with testing and laboratory products. Since 2004, the TB program has focused on strengthening the health system to be better able to treat TB as an HIV coinfection and treat opportunistic infections. It has also 24 supported a program to decentralize TB screening and treatment. The TB program established TB tracking (microscopy, culture, radiology) and treatment, follow-up examinations (microscopy, culture, imagery), and counseling. All district and community health centers in Djibouti-Ville were provided with basic kits and supplies, and benefited from skill building and training of service providers. TB detection and DOTS were made available in all health centers in Djibouti-Ville and the five district medical and hospital centers free of charge, regardless of patients' income, citizenship, or immigration status. This policy, in effect since before the ES was set up in 2003, may partially explain the high recorded prevalence of TB in Djibouti. The TB program recorded a drop in TB-positive cases with a detection rate falling below the threshold of 70% after 2000 and bordering on 50% in 2004 and 2005 (based on WHO estimated incidence). The treatment success rate often remained lower than 80% for the same period, with a lost to follow-up rate between 15% and 27%. There was a small improvement between 2004 and 2007 in TB program results. The success rate of treatment of new TB cases improved from 80% to 82%, and people lost to follow-up fell slightly from 16% to 15%. The modest results may be attributed to an incompletely applied DOTS strategy; difficult access to diagnosis and treatment monitoring for poor and remote populations; an insufficient coordination team for the TB control program; and limited financing before 2006 (Global Fund n.d.). From 2007, Global Fund support started to address these weaknesses. About two-thirds of the TB-positive cases have been diagnosed and treated at the Centre Paul Faure. Care of TB/HIV coinfection began in 2004 in Centre Paul Faure, the national reference center with a cross-reference system. This made it possible to detect and treat TB cases earlier and reduce the level of TB transmission among PLHIV and the general population. By addressing HIV/TB coinfection, the Djibouti NACP made notable improvements in comprehensive case management of TB patients, particularly in strengthening the technical platform for diagnosis and treatment. About half the coinfected patients put on ARV were considered multidrug resistant and given second-line treatment. The proportion of registered TB patients who received HIV counseling and testing increased from 2.5% in 2005 to 46% by end-2008. One of the 10 Global Fund grant targets for the end of 2008 was exceeded: the number of HIV-positive TB patients receiving cotrimoxazole preventive therapy reached 142% of the target (Global Fund 2009). From the program's inception, there was a steady reduction of TB/HIV coinfection, and HIV prevalence among new TB patients declined from 13.8 percent in 2006 to an estimated 11.8 percent in 2008, as depicted in Figure 8. 25 Figur e 8: HIV Pr evalence among new TB patients, 2002­2008 HIV Prevalence among new TB Patients 25 22 22 20 20 16.2 13.8 13.6 15 11.8 10 5 0 2002 2003 2004 2005 2006 2007 2008 HIV Prevalence among new TB Patients Source: Data for 2006 to 2008 were compiled by the authors from PLS-Health Statistics, Ministry of Health, Djibouti. Note: HIV prevalence data from 2002 to 2005 are estimates based on data collected from Centre Paul Faure and World Bank internal reports. Prevention of Mother-to-Child Transmission In 2003, UNICEF introduced a program for the Prevention of Mother-to-Child Transmission (PMTCT) in a few maternal health centers. With support from the ES, PMTCT was expanded to 19 centers in 2008. The PMTCT program has the following objectives: · To strengthen PMTCT coordination at the central level as well as in the districts/regions · To extend counseling and voluntary testing in all consultation services · To improve global case management of HIV-positive mothers and their children · To strengthen community communication and participation to prevent parent-to- child HIV transmission. The program has made good progress in VCT. For example, in 2008, of 10,739 pregnant women who attended their first prenatal care visit, 85 percent (9,129 women) were counseled, 76 percent (8,118 women) accepted the HIV test, and 75 percent (6,100 women) of those tested actually received the results. Of 127 pregnant women who tested HIV positive, only 32 pregnant women did not return to receive their results, or about 25 percent. Against this, the PMTCT program has made slow progress in prevention, as shown in Figure 9. From 2003 to 2008, the proportion of pregnant women who received ARV prophylaxis among those who tested HIV positive increased from 27.6 percent to 33.9 percent. The children born to HIV-positive mothers who received ARV prophylaxis fluctuated widely: 94.4 percent in 2003, 47.6 percent in 2004, and 97.3 percent in 2008. In 2008, about 71% of children born to mothers with HIV were sero-negative. 26 Figur e 9: Utilization of PMTCT Ser vices, 2003­2008 250 Total pregnant women who tested VIH+ 200 192 Total pregnant 183 women who 170 received the test results 150 137 141 Women who 127 delivered in a 112 medical setting 95 Infants born to 100 8485 85 HIV+ mothers 65 60 58 5252 55 55 45 4944 43 43 Women who 50 42 4242 37 36 received ARV 21 16 1920 2522 18 17 prophylaxis Infants who 0 received ARV 2003 2004 2005 2006 2007 2008 prophylaxis Source: Data compiled by Authors from PLS-Health Statistics, Ministry of Health, Djibouti. HIV Surveillance In 2003, sentinel surveillance was introduced by UNICEF as part of PMTCT in selected maternal health centers. The ES, with support from the World Bank, expanded the services to create a network of sentinel surveillance centers within prenatal care services and blood banks. With support from WHO, serosurveillance of pregnant women was relaunched in sentinel sites 15 in the districts of Djibouti, 16 and in Dikhil, Ali Sabieh, Obock, and Tadjourah. Every pregnant woman seen at prenatal care services is offered HIV testing; those who accept it are oriented to VCT services and those who test HIV positive are later referred to the PMTCT services available in the maternity hospitals of Djibouti-Ville, including that of the parastatal Social Protection Office (Office de Protection Sociale; OPS). PMTCT sites in maternity hospitals provide data that can be regularly collected and analyzed. The sentinel surveillance was expanded in 2006 with support from the WHO to include five groups: pregnant women, patients diagnosed with STI, TB patients, CSWs, and blood donors. The sentinel surveillance system produced valuable results. Compared to the seroprevalence surveys conducted in 2002 that found prevalence to be 2.5 percent, the sentinel surveillance system revealed that prevalence was steadily decreasing to 2.4 percent in 2005 and 1.9 percent in 2008. This strongly suggested the stabilization of the epidemic. Figure 10 depicts the trend in HIV prevalence among pregnant women measured through the sentinel surveillance system. 15 "Relaunched" because PMTCT sites initially managed by UNICEF were halted and relaunched under the WHO serosurveillance program. 16 Including the health center at Point Kilometrique 12 (PK12), the biggest rest stop for trucks in Djibouti. 27 Figur e 10: HIV Ser opr evalence in Djibouti, 2002­2008 3 2 2.5 2.4 2.2 2.1 1.9 1 0 2002 2005 2006 2007 2008 HIV Seroprevalence Source: Sentinel Surveillance System Statistics, PLS-Health, Ministry of Health, Djibouti. MULTISECTORAL RESPONSE Under this component, nine ministries, Parliament, and Regional Health Committees (Comités Régionaux de Santé; CRS) were involved in the response to HIV/AIDS. 17 The national army, defense ministry, and the OPS had some experience with HIV, but none of the remaining ministries, Parliament, or the CRS had, and needed extensive capacity building under the program. As noted earlier, each ministry established its own PLS. Each PLS prepared a yearly action plan that was approved by the MTCU of the ES. Funding was provided through the World Bank project and also by the Global Fund from 2007. Initially, the IC was the national body overseeing implementation; CCMI subsequently took over this role, de facto. The performance of the different ministries and sectors varied significantly, as discussed in paragraphs that follow. Ministry of Communication PLS-Communication facilitated media coverage of the NACP, particularly on Djiboutian Radio and Television. In addition to its own action plan as a targeted ministry, this PLS closely collaborates with the ES on programs for the general public. After consulting with line ministries, religious leaders, and selected community members, the ES Communication Unit provides the contents and scheduling of mass media messages and reports, and covers most expenses for mass media programming on the three major diseases. The PLS-Communication strategy focused mainly on radio as the main mass media channel used by many segments of the population including those in remote areas. Ministry of Youth and Sports PLS-Youth revamped its community development centers (centres de développement communautaire; CDCs), located in each neighborhood of the capital city and in the districts, which disseminate information to youth. PLS-Youth operationalized 14 CDCs and established Youth Information Points (or points information jeunes; PIJs) staffed with 12 male and female educators. The CDCs were equipped with modern audiovisual 17 Three ministries represented in the IC were not involved in program implementation. 28 materials and desktop computers with support from the World Bank and UNFPA. In addition to disseminating information on transmission methods and protection against HIV, the CDCs also distributed condoms, but mostly to young males as few young females frequented the CDCs. This indicated the need for more culturally appropriate venues for young girls, such as sewing lessons. Ministry of Education PLS-Education is one of the most active, providing support to 20 health clubs in schools in the capital city and the five districts. Despite scarce human resources (one coordinator, one deputy, and one accountant) all PLS-Education planned activities were completed, including a mini-survey in schools. Educational agendas and radio and TV educational programs were among its successful activities. A health education guide for teachers was also distributed. As with the umbrella NGOs and PLSs in other line ministries, PLS- Education faced initial difficulties in quantifying its target groups and assessing the size of its PVGs, and in collecting data in a timely manner from schools. By 2008, PLS-Education had many positive and visible results in increased HIV/AIDS awareness and continuing prevention activities in all secondary schools and at the university. PLS-Education became the "champion" of the HIV response by introducing a mandatory training module on HIV/AIDS into the high-school curriculum. Most schools opened health clubs. Each year, through the National Education Research and Information Production Center (CRIPEN), the Ministry of Education produces more than 4,000 school agendas for students containing awareness and prevention messages on HIV/AIDS, malaria, and tuberculosis. The CRIPEN school radio is in the vanguard of HIV/AIDS communication: it designs, produces, and disseminates messages, radio spots, plays, and sketches on the three diseases via Djibouti radio and television. Ministry of Defense PLS-Defense used its support from the ES mainly for prevention activities including the provision of VCT and treatment services through military clinics. In addition, PLS- Defense received funds from the U.S. Army. One of its key targeted field activities was the deployment of HIV "prevention caravans" along the country's borders for use by persons in uniform. Ministry of Muslim Affairs and Waqfs PLS-Waqfs prepared its action plans starting in 2005 but the Ministry delayed implementation. The plans included training religious leaders and imams. In 2007, with support from the ES Communication Unit, PLS-Waqfs organized an awareness and consensus workshop for 70 Djiboutian religious leaders and imams, facilitated by University of Al-Ahzar Islamic scholars and experts. The ES-led endeavor was successful in clarifying the role of religious leaders, and in reaching a consensus on appropriate HIV responses, including protecting women and children against the epidemic. By end-2008, 25 religious leaders had been trained as trainers, working with two consultants from Al- Ahzar University. They are training other imams and providing fatwas (religious opinions) in support of the NACP. 29 The Djibouti Declaration was adopted in 2007. It stated that Djiboutian religious leaders endorsed the use of condoms as one of the ways to prevent HIV and for birth spacing. It acknowledged the necessity to fight HIV/AIDS and recommended protection for couples with discordant HIV status and for birth spacing. The lifting of opposition, taboos, and rumors against the use of condoms led to a substantial increase in condom use. This was a key (and remarkable) intermediate result. Ministry of Justice and Penitentiary Affairs PLS-Justice organized several conferences and training both in-country and abroad for high-level technical staff from the Ministry of Justice and managers from the central prison of Gabode. With support from the ES technical coordination units, prevention activities that targeted staff and inmates were conducted in different institutions and at the central prison. The key achievement of PLS-Justice was leading the preparation of a law protecting PLHIV and other vulnerable groups, which was ratified by Parliament in late 2007, and implemented. Ministry of Women's Promotion, Social Affairs and Family Welfare PLS-Women, one of the more active PLSs, effectively implemented its annual action plans. Activities focused on organizing seminars for women aiming at increasing their awareness about modes of transmission and protection against HIV. Inspired by the Sister-to-Sister project (Box 3, below) initiated by CISU, PLS-Women secured additional funding for vocational training of women in precarious living conditions, which is one of the key achievements of this unit. Ministry of Employment and Solidarity, and Social Protection Office Most of the activities carried out under the Ministry of Employment were implemented by the OPS, and largely comprised clinical services. The OPS provided different clinical services, which were expanded through ES support to include PMTCT, STI screening and treatment, VCT, and ARV. Ministry of Equipment and Transportation PLS-Transport conducted HIV prevention activities for land transport workers, mainly along the Djibouti­Addis Ababa corridor, in collaboration with the "SafeTstop" project managed by Family Health International. Two road corridors and one railway corridor connect the port of Djibouti to Addis Ababa in Ethiopia. The most heavily used is the Djibouti­Galafi­Addis Ababa trade corridor, where migration, transport and CSW activity generate substantial HIV risk (Brushett and Osika 2005). This corridor is an 850 km two-way stretch with more than 27 towns and stops. About 1,000 truckers and assistants, mostly younger than 30, traverse daily and stop for meals, overnight stays, meet with sex workers in hotels and bars, and engage in unprotected sex. Three program interventions aimed at (i) increasing demand for HIV/AIDS-related services along the corridor through training field communicators and distributing flyers, 30 stickers, booklets, T-shirts, pocket diaries, and song cassettes; (ii) preventing HIV spread through condom distribution; and (iii) detecting HIV through referral to VCT centers. From 2004 to 2008, 10 rest stops for truckers were established. More than 650 dockers in the port of Djibouti and 2,690 truck drivers in major resting stops were provided with "safety kits" of HIV/AIDS information, condoms, and referrals to clinics located on the transport corridors that provide services for STI treatment and VCT for HIV. Several NGOs provide HIV/AIDS awareness and prevention activities in cafes and restaurants around PK12 (Point Kilometrique 12), which is the most used rest stop, including performance groups made up of AIDS orphans. There were also awareness sessions and condom distribution for urban taxi drivers. Figure 11 shows the location of condom sales points along the corridor. Figure 11: Condom Sales Points along the Djibouti­Addis Addis Ababa Corridor Source: HIV/AIDS Executive Secretariat, Djibouti. Parliament Members of Parliament, selected according to availability and interest in contributing to the HIV response, received specific HIV/AIDS advocacy training, which was instrumental in implementing advocacy sessions in the different communities and regions. This capacity building and engagement of Parliamentarians increased their awareness of the negative effects of stigma and discrimination, and encouraged their sponsorship for ratifying the law protecting PLHIV and other vulnerable groups. 31 Regional Health Committees As part of the government decentralization process, the Ministry of Health established decentralized regional health committees (CRS) in four of the five districts of Djibouti. In 2003, UNDP supported focal points in two districts, on a pilot basis, as part of the health response. In 2004, the World Bank continued this support and expanded it to two more districts to cover Ali Sabieh, Dikhil, Obock, and Tadjourah. Under the aegis of the ITC, the CRS mandate is to plan and coordinate activities to fight HIV/AIDS and STI in the districts. Activities conducted by the CRS were sporadic and poorly coordinated. The program evaluation conducted in 2007 concluded that only two CRS, in Dikhil and Tadjourah, were operational, which led to the cessation of funding of the CRS. COMMUNITY RESPONSE As explained earlier, NGOs in Djibouti in 2002 were very weak. It was therefore decided to run capacity-building activities before launching community interventions, and to implement the community response through a two-tier system of umbrella NGOs and CBAs. Umbrella NGOs The ES competitively selected six organizations to become umbrella NGOs. They are responsible for training local CBAs, assisting them with subproject preparation, and supervising and controlling the quality of their activities. The six umbrella NGOs and their geographic areas of activity are: · ADEPF: Agence Djiboutienne pour l'Equilibre et la Promotion Familiale (Dikhil and parts of Djibouti-Ville) · CAB: Coordination des associations de Balbala (all of Balbala) · CCAF: Cabinet Conseils Appui et Formation (Tadjourah and some areas of Djibouti-Ville) · CIS: Cabinet Ingénierie Sociale (Ali Sabieh and some areas of Djibouti-Ville) · UDC: Union pour le Développement et la Culture (Obock and some areas of Djibouti-Ville) · UNFD: Union Nationale des Femmes Djiboutiennes (Arta and some areas of Djibouti-Ville). The umbrella NGOs are either associations or consulting firms with experience in monitoring projects. Some, such as UNFD and ADEPF, are important structures. Established in 1994, ADEPF is affiliated with the International Planned Parenthood Federation and has more than 20 CBA members, 11 of which are funded by the World Bank. Other umbrella NGOs with headquarters in Djibouti-Ville have branches in districts for better monitoring of projects implemented at this level. As a whole, the umbrella NGOs had trained 239 peer educators (112 in Djibouti-Ville and 127 in districts) as of end-2008. 32 Community-based Associations These are a fixed feature of Djiboutian society, often established for cultural reasons and for social networking. They were identified as good channels for HIV prevention activities. At the start of the program, there were a dozen CBAs interested in HIV prevention, including one association for PLHIV. All CBAs had weak capability and minimal expertise in implementing communication for behavior change. Supporting institutions were deemed necessary to ensure quality and regular supervision. The CBAs that were sustained by umbrella NGOs were implementing preventive activities, such as awareness sessions and peer education for early detection of HIV by VCT. To launch the activities, the program conducted focus group discussions with vulnerable groups, such as CSWs and dockers, to identify attitudes and practices exposing them to infection. The program entered into simple contractual agreements with the NGOs and CBAs to implement community subprojects that included the following activities: providing care and support for PLHIV and their dependents (for example, nutrition for AIDS patients, income generating activities); and information and communication for behavior change activities to prevent HIV and destigmatize PLHIV. The program dealt with the challenges faced the community response in an innovative way, namely by "mapping" and "targeting." Based on a rigorous process, this model aimed to be a practical guide that enabled integration of empirical data from the literature and data collected in the field from target populations. Planning a targeted intervention involved 19 tasks, grouped in six phases. During the preparatory phase, the health problem, its main determinants, and the target population were identified, and an environmental analysis was conducted. The next five phases consisted of developing the main steps of the intervention by identifying the specificities of the target population; determining the program objective; developing the performance objectives; specifying the determinants on which the program would act; and developing the related learning objectives. The result was a map of segments of the population identified as PVGs, with geographic distribution. A matrix was developed to allocate target PVGs to each umbrella NGO, depending on their local presence and experience (Table 3). 33 Table 3: Pr ior ity Vulner able Gr oups per Umbr ella NGO, 2004­2008 ADEPF CAB CCAF CIS UDC UNFD Total Out-of-school youth 2,120 3,087 4,950 1,538 3,350 8,192 23,237 Young women in 1,566 1,347 4,062 1,257 700 2,550 11,482 difficulty CSWs 150 0 0 0 200 0 350 Dockers 0 0 900 440 1,350 0 2,690 Private sector workers 0 0 1,748 801 1,100 1,764 5,413 Khat users 1,217 324 0 227 652 567 2,987 Refugees 0 0 0 462 0 0 462 Nomads 228 0 0 258 0 0 486 Orphans and other vulnerable children 0 0 160 0 0 0 160 Truckers 450 0 0 0 200 0 650 Youth at school 0 0 0 0 544 0 544 Handicapped 0 0 0 0 150 0 150 ADEPF: Agence Djiboutienne pour l'Equilibre et la Promotion Familiale; CAB: Coordination des associations de Balbala; CCAF: Cabinet Conseils Appui et Formation; CIS: Cabinet Ingénierie Sociale; UDC: Union pour le Développement et la Culture; UNFD: Union Nationale des Femmes Djiboutiennes. Figure 12 shows a map of Djibouti with the geographic distribution of CBAs, reflecting the locations where community interventions and activities were implemented. Figur e 12: Distr ibution of Community-based Associations Source: HIV/AIDS Executive Secretariat, Djibouti. The other innovation was the introduction of the DASAP, which aims at targeting social support on the basis of a detailed impact analysis, looking at: 34 · social status and impact on the individual--social class, gender, age, literacy, marital status, employment status, segregation within the family, etc. · economic impact on the individual--income, employment, medical expenses, transportation, rent, etc. · social impact on the family--segregation of the entire family, rejection of other family members, loss of heritage rights, loss of children custody, etc. · economic impact on the family--loss of income, medical expenses, indirect expenses, loss of employment of other family members, etc. The DASAP decision pathways are schematized in Figure 13. Figure 13: Decision Pathways of the DASAP 35 By end-2008, CBAs were implementing 170 community projects. As a result of the CBAs' social work, 2,550 people received dry rations; 300 HIV-positive women benefited from psychosocial support; and 600 orphans and other vulnerable people received psychosocial support, nutrition, school fees and supplies, and financing of income-generating activities for their guardians. In addition, hundreds of members of target groups were directed to the nearest VCT centers, sometimes with a coupon or a referral chip. Sometimes a CBA member accompanied people who wanted to be tested. Referrals and recording of referral sources were not done systematically at first; a tracking system was established in 2008. A good example of a successful community project is currently operating at the PK12 truck stop. An NGO has trained peer educators to empower sex workers to persuade their clients to use condoms, runs condom social marketing campaigns and ensures the ready availability of condoms. Building on this success, USAID provided microcredit for vulnerable women in the area of PK12 to enable them to earn a livelihood in ways other than sex work, and to help alleviate poverty. Other projects include outreach to sex workers in Djibouti-Ville and to dock workers and truckers using peer educators, as well as distribution of food to affected needy families (El-Saharty and Ali 2006). The French Development Agency also had an active community mobilization intervention through the Community Development Fund management committee set up within the framework of the PK12 project. At the heart of the PK12 area, the "Sister-to-Sister" community Box 3 : Soeur-à-Soeur Project intervention project has been very Recognizing that women living in precarious active in increasing awareness among conditions and CSWs are most at risk, and that sex CSWs (Box 3). This project was work might be driving the epidemic, CISU prepared crucial, as about 53 percent of CSWs a special project titled "Sister-to-Sister" (Soeur-à- Soeur)--a name chosen to eliminate the tested positive for HIV, confirming stigmatizing label of sex trade, and recognizing that the assumption that sex work is an women may have little alternative to sex work to important driver of the HIV epidemic, fulfill their basic needs and those of their children. and calling for a program that The project goal was to reduce vulnerability to HIV addresses women's vulnerability that and other STIs among this marginalized group of may drive them into sex work. women. The strategy consisted in helping the women to understand the issues; giving them the Targeting sex workers through their skills and means to reduce their risk; and involving peers was a key intervention in them in implementing risk reduction in their groups. empowering sex workers and Activities designed and implemented by this enabling them to negotiate the use of marginalized community include: increased condoms with their clients. This, availability of condoms; condom distribution; sessions demonstrating condom use; group coupled with the increased access to education sessions to better understand the risks; condoms, led to the notable increase and advocacy sessions and focus group sessions to in condom use among this high risk make inventories of situations and common group, addressing the key driver of practices generating risks in their community. the epidemic. 36 RESULTS AND ENABLING FACTORS In about six years, Djibouti made considerable progress toward curbing the HIV/AIDS epidemic, starting almost from scratch. Between 2002 and 2008, the NACP, through its four components detailed above, achieved the following key results: · HIV prevalence among young pregnant women aged 15­24 was reduced from 2.7 percent to 1.9 percent, and among sentinel surveillance groups from 2.5 percent to 1.9 percent, suggesting the curbing of the epidemic. · HIV prevalence among TB patients was reduced from an estimated 22 percent to 11.8 percent. · Condom use during last intercourse outside marriage increased from 27 percent to 55 percent and reached 95 percent among CSWs. · Among the general population, awareness about HIV/AIDS increased to 95 percent and knowledge about transmission and prevention rose to 50 percent. In addition, 69 percent of respondents mentioned prayers, fasting and good Muslim behavior as methods of protection against HIV. · VCT services increased from 1,461 to 7,158 tests, an increase of more than fourfold. About 85 percent of women attending prenatal care services have accepted to be counseled and 76 percent to be tested for HIV. · ARV therapy was scaled up and 1,293 patients received treatment with ARV drugs, based on clinical eligibility criteria, with a survival rate at 12 months of 67.8 percent. · Survival trends under ARV therapy improved -- the proportion of patients on ARV therapy who died fell from 17.6 percent to 14.5 percent. · More than 3,620 persons with HIV (some with opportunistic infections) received comprehensive integrated care, including clinical, nutritional, social, and psychological care. · More than cumulative 2.6 million condoms were distributed, up from only 0.21 million condoms in 2004, through a widespread condom distribution network of 420 sales points. · Social support was provided to 80 percent of the families of PLHIV. Several enabling factors contributed to achieving the above results, including the following: 37 · Political commitment was key in mobilizing resources and engaging a multisectoral response. Getting support at the highest political level represented by the creation of the IC, under the prime minister, provided high visibility to the program. The CCMI, chaired by an active minister of health, succeeded in mobilizing multisectoral activities and donor support. · Engaging the community and religious leaders was critical in addressing cultural concerns and adapting messages to social values. Early on, the program mobilized community and religious leaders to cultivate their support and endorsement of its messages. Many trained religious leaders and imams were called on to conduct advocacy sessions. The Djibouti Declaration, in which religious leaders endorsed the use of condoms, was a remarkable breakthrough for a Moslem country in changing the attitude of many religious segments of the population. · Maximizing and diversifying the use of communication channels was the driver for increasing awareness and behavior change. Mass media campaigns started with television and quickly shifted to radio, which then became the main mass media channel. Also, the community response was very effective in targeting high risk and priority vulnerable groups with tailored messages. · The provision of an integrated package of medical and social services proved to be very effective. Lack of treatment compliance of many PLHIV, even though treatment was free, stemmed from psychological, social, and financial factors. The innovative DASAP was therefore introduced and implemented using NGOs to link treatment with nutritional, social, and psychological care, which has been provided to more than 3,620 PLHIV. As a result, compliance and survival rates improved greatly. Without the integrated package, the program might have not achieved these results. · In combating HIV/AIDS, access to ARV therapy is as important as prevention. Although prevention of HIV remains the most effective and cost-effective weapon against the epidemic, and ARV therapy is expensive, insight and closer consideration of epidemiologic and cultural aspects led to the conclusion that prevention (and particularly VCT services) would fail unless ARV therapy was offered simultaneously. As a result, the World Bank agreed with the ES to introduce ARV therapy, which was later scaled up by using a Global Fund grant. It was subsequently shown in service statistics that VCT utilization increased, particularly among the young and educated, when they were assured of access to affordable treatment. The improvement in HIV prevention would not have been achieved without the ARV therapy program. · The use of umbrella NGOs was instrumental in building capacity and supporting implementation of community projects. Despite two years' delay in implementing the community response, this approach proved very effective in supporting the work and activities of more than 170 CBAs, which in turn increased awareness about HIV/AIDS and utilization of services. 38 · Harmonization of donor support greatly facilitated effective implementation. The HIV/AIDS response is a long-haul effort that requires effective coordination among all stakeholders, particularly donors. The key donors that financed and supported the bulk of the program--the World Bank and Global Fund--worked together with the ES to implement the Three Ones policy, setting a successful example for using a single national authority, supporting a single national multisectoral plan, and tracking progress through a single national M&E system. For example, the introduction and piloting of the ARV therapy and the DASAP by the World Bank was scaled up by the Global Fund. That created a solid foundation for other donors to contribute, including UNICEF (PMTCT), WHO (M&E and surveillance), and the French Development Agency (AIDS/TB coinfection). In addition, the ES and donors focused on results and used the performance indicators to inform program implementation, ending unsuccessful activities, and expanding and reinforcing successful ones. · Institutional and management capacity building took time but paid off. The early delays in implementation were caused by the focus on strengthening the management capacity and qualified staffing of the newly established ES and developing detailed guidelines and procedures. However, this early investment by the World Bank later helped accelerate implementation and created a solid institutional platform that was used by other donors. This capacity was translated into successful adaptation to changing needs, such as the simplification of the NACP institutional structure and the decision by the ES to carry out condom distribution itself instead of contracting out the task to an NGO. REMAINING CHALLENGES Despite achieving many impressive results in a relatively short period, Djibouti still has a lot to do if it is to address some of the challenges, consolidate the gains, and sustain the NACP. Some of the challenges that face the NACP are summarized below. There is a need to develop alternative strategies to promote the use of VCT services among persons in uniform. The program established VCT and treatment clinics within the medical services of the armed forces and police. However, use of these services was limited because of concerns related to privacy and confidentiality. Therefore, NACP management may need to consider alternative strategies, such as integrating VCT services in other primary health care services used by the armed forces and police, or expanding testing to other public or private facilities. STI case management needs further strengthening. The total number of STI cases diagnosed and treated declined between 2002 and 2005, increased in 2006 and 2007, and declined in 2008. This fluctuation reflects challenges such as the lack of systematic application of the syndromic algorithm and underreporting of STI cases. A more in-depth 39 analysis is needed to identify the key constraints and redefine the implementation strategy, to improve program management of STIs. Social marketing needs to adopt a more rigorous and targeted strategy to increase condom use, particularly among high-risk groups. Despite the marked increase in condom use among CSWs, condom use among some other high-risk groups, such as persons in uniform and dockers, remains at 13 percent, which is extremely low. Also, condom use among the general population during the last intercourse outside marriage of 55 percent is no reason for complacency. Targeted messages to high-risk groups and the expansion of peer education should be adopted to address this challenge. Communication and awareness campaigns need to be strengthened. At only 50 percent knowledge among the general population about HIV transmission and prevention, the NACP is exactly half way from achieving universal knowledge about HIV transmission methods and prevention. Greater knowledge could increase use of program services and further curb the HIV/AIDS epidemic. The multisectoral response was very complex and difficult to implement at the beginning of the program and should continue to be focused on the few critical ministries and sectors. The multisectoral response was implemented in too many ministries, making it difficult to supervise and monitor. Also, the activities of several ministries were of limited impact. If the multisectoral response were confined to a few key relevant ministries, it would be more manageable and more possible to ensure that activities are implemented effectively. Even with additional resources, the NACP should confine the multisectoral interventions to three or four sectors that are effectively contributing to program results, such as education, youth, and communication. The NACP has been financed almost exclusively by donor funds, an approach that is not sustainable. The program benefited from financial resources from large donors, notably the World Bank, Global Fund, and French Development Agency, and from technical resources from donors such as UNICEF and WHO. These resources were needed to establish the program, build institutional and management capacity, and develop the various guidelines and systems. However, this is simply not sustainable. While the NACP will continue to benefit from some donors' support until 2012, the government of Djibouti has to develop a sustainability strategy to support the key program components from its own resources to ensure its viability. Otherwise, the country may experience a resurgence of the HIV/AIDS epidemic, given that many of the risk factors remain in place--such as the transport corridor and sex work. 40 REFERENCES Brushett, Stephen and John Stephen Osika. 2005. "Lessons Learned to Date from HIV/AIDS Transport Corridor Projects." World Bank, Washington, DC. Center for Research and Studies in Health (CREDES). 2003. "Epidémie à VIH/SIDA/IST en République de Djibouti." Volume 1. Paris. January. Conseil Santé. 2007. "Evaluation à mi-parcours du Projet de Lutte contre le VIH/SIDA, le Paludisme et la Tuberculose." République de Djibouti. Secrétariat Exécutif, Comité Intersectoriel de Lutte contre le SIDA. Rapport Final. May. Paris. DISED and UNICEF Djibouti. 2005. "Rapport de l'enquête CAP chez les jeunes." Djibouti. El-Saharty, Sameh and Omar Ali. 2006. "Djibouti's Effective, Well-coordinated Response to HIV". HIV/AIDS Getting Results, World Bank, Washington, DC. Global Fund to fight HIV/AIDS, Tuberculosis and Malaria (the Global Fund). n.d. "Djibouti Proposal Form--Sixth Call for Proposals: Support for Strengthening and Decentralization of the Tuberculosis Project." http://www.theglobalfund.org/grantdocuments/6DJBT_1306_0_full.pdf. ------. 2009. "Djibouti: Grant Performance Report (DJB-607-G03-T)." December 11, p. 20. http://www.theglobalfund.org/grantdocuments/6DJBT_1306_540_gpr.pdf. Government of Djibouti. 2003. "Décret n°2003-0049/PR/MEF/MS portant mise en place d'un cadre institutionnel de lutte contre le Sida, le Paludisme et la Tuberculose." 41 ANNEX 1: COUNTRY AND SECTOR BACKGROUND The Republic of Djibouti is a small country in the Horn of Africa with an estimated population of 830,000 (2008). The natural annual rate of population growth is around 2.4 percent. The country shares its borders with Eritrea, Ethiopia, and Somalia, and some points are less than 5­10 kilometers (km) from Yemen. The climate is desert-like and temperatures are high, averaging above 40°C between May and September. The Economy Djibouti's per capita income of $1,060 (2008) puts it among low-middle-income countries. However, it has some of the poorest social indicators in the world. Illiteracy, unemployment and poverty are high -- 60% of Djiboutians live below the poverty line. Two thirds of the population lives in the capital city Djibouti-Ville, most of the remaining population live in a handful of secondary towns. About 15 percent of the population is rural, subsisting on nomadic herding. The skilled workforce is limited. Djibouti is poorly endowed with natural resources (limited arable land, rainfall, and underground water). More than 80 percent of all food is imported from neighboring countries (Ethiopia and Yemen) or Europe. The country benefits from its location on the Red Sea's southern entrance. Politics Djibouti became independent on June 27, 1977. The population is 99 percent Muslim. There are two main ethnic groups, the Afars and the Issas, nomadic sheep-herding people. The Issas extend into Somalia and the Afars into Ethiopia and Eritrea. The power balance interplay between them is a key determinant of Djibouti's domestic political context and developments. Competition for power among ethnic groups led to a very damaging civil war in 1991­94. Djibouti's political system is relatively open, with an elected president, a Parliament, and a judiciary. The executive reflects a sharing of power between the two major ethnic groups. The official languages are Arabic and French; Somali and Afar are also spoken. The largest foreign populations are Somalis, Ethiopians, French, and American. Administratively, the decentralization process of the last decade created six regions in the country: one centered on the capital (Djibouti-Ville), which has its own status (a "collectivité territorial" subdivided into the three "communes" of Ras-Dika, Boulaos, and Balbala); and five other regions (or districts), namely Ali Sabieh, Arta, Dikhil, Obock, and Tadjourah. Health and the Health System Accurate health statistics are lacking. Overall, life expectancy was 54 years in 2008. Life expectancy at birth, particularly for females, is among the world's lowest. It has remained relatively stable in recent years, at around 55 years for females and 53 years for males. 18 18 Population Reference Bureau website: http://www.prb.org/Countries/Djibouti.aspx?p=1. Accessed December 26, 2009. 42 These figures are below the low-income country average of 60 years and 58 years for females and males, respectively, as estimated in 2008. Health expenditures were about 10­12 percent of the state's social spending, and 7 percent of gross domestic product (GDP) according to the Global Alliance for Vaccines and Immunisation in 2004. According to the National Health Accounts, public health expenditures amount to 54 percent of total health spending. Recurrent costs, salaries, and health benefits of health workers constitute a significant part of health spending. The health system includes public health services, funded by the Treasury, foreign aid, and partial cost recovery; health insurance, financed by employers' and employees' premiums; the private sector; and the health subsector for the army and police. Health care in public facilities is free of charge for all and the public health delivery system is split into three levels: primary health care, consisting of health centers and health posts; intermediary level, with five district medical and hospital centers; and tertiary level, with Peltier General Hospital, the Centre Paul Faure for Tuberculosis, the renovated Balbala Obstetrical Hospital, and the recently upgraded Dar El Hanan Maternity Hospital. Health providers in Djibouti consist of about 100 medical doctors -- about 1 per 6,600 inhabitants; 440 nurses -- about 1 per 500 inhabitants (there is one nurse in each community and health center); 53 midwives -- about 1 per 12,300 inhabitants; 9 dentists - - about 1 per 68,000 inhabitants; 8 pharmacists -- about 1 per 70,000 inhabitants; and 51 laboratory technicians -- about 1 per 13,200 inhabitants. Most health providers -- across all categories -- work in Djibouti-Ville where 65­75 percent of the population lives. Djibouti has 46 health facilities, including private clinics, of which 38 belong to the public health system (32 health centers and health posts, of which 20 are in rural areas). Overall health coverage and geographic accessibility to medical care are in theory relatively satisfactory compared to other sub-Saharan African countries. In reality, however, several factors undermine access: (i) the shortage of qualified health personnel and the poor geographic distribution of existing personnel (almost 90% percent of all doctors and 75 percent of all health personnel are assigned to Djibouti-Ville and the remaining 25 percent are mostly in the larger towns); (ii) the insufficiency of resources assigned to operations and training; and (iii) frequent drug stockouts. In addition, the health system is based on curative care, which hardly reaches the poor. Peltier General Hospital alone absorbs 40 percent of the human and financial resources of the sector. The main causes of death and visits to health facilities are communicable diseases, many of which can be prevented by improved hygiene and behavior change. Malaria, tuberculosis (TB), and HIV have health and economic impacts on households, the population, and the health system. Obstetric care from a trained provider during delivery is recognized as critical for reducing maternal and neonatal mortality. More than 70 percent of births in urban areas are assisted by a trained health professional, but only 3 percent of births in rural areas. 43 Djibouti's infant mortality rate of 67 per 1,000 live births in 2007 is considerably better than the Sub-Saharan African average of 100 per 1,000 live births, and the low- income-country average of 92 deaths per 1,000 live births. The child mortality rate is 94 per 1,000 live births. The maternal mortality ratio for 1998­2004 was 673 deaths per 100,000 live births (rounded, 7 deaths per 1,000 live births). Other data show the maternal mortality ratio for Djibouti for 1994­2000 to be 871 deaths per 100,000 live births (rounded, 9 deaths per 1,000 live births). As noted above, the primary factors underlying disparities in health outcomes include geography (that is, rural versus urban settings), level of educational attainment of adult household members, and socioeconomic status (wealthier individuals tend to have greater access to quality health care). Tuberculosis (TB) has a long history in Djibouti. It is associated with poverty, overcrowding, and poor hygiene. With 1,137 cases of TB per 100,000 inhabitants, Djibouti has the highest TB incidence in the world, far above the low-income-country average of 224. Over the last 10 years, Djibouti has recorded an average of 3,572 new cases of TB per year, peaking in 2000 with 4,121 diagnosed cases. However, about one third of cases come from neighboring countries, in particular from Ethiopia, which inflates the rate. Foreigners come to Djibouti because it offers more and better-quality services free of charge. Malaria has only been a problem in Djibouti since the late 1980s. Before 1973, when there was no urbanization, no irrigation, and an active attempt to control the vector during the rainy season, more than 80 percent of notified cases were from people entering Djibouti from neighboring countries. From 1973 to 1987, more Djibouti national cases appeared along the main transport lines linking Djibouti to neighboring countries, and after 1987, cases appeared in urban areas as thousands of refugees resettled in Djibouti. Since 1988, the number of reported cases of malaria has increased steadily. Uncontrolled urbanization with inappropriate water supplies, nonexistent wastewater evacuation systems, settlement of nomad populations in rural areas, increased irrigated areas, and frequent floods have all contributed. Djibouti currently records over 4,000 confirmed cases of malaria each year. WW W. WO RL D B A N K . O R G / A I DS For more information, please contact: The Global HIV/AIDS Program Health, Nutrition and Population Human Development Network World Bank Group 1818 H St. NW, Washington, DC 20433 Tel: +1 202 458 4946 Fax: +1 202 522 1252 wbglobalHIVAIDS@worldbank.org