Document of The World Bank Report No.: 28819 PROJECT PERFORMANCE ASSESSMENT REPORT BRAZIL FIRST AND SECOND AIDS AND STD CONTROL PROJECTS (LOAN 3659-BR AND 4392-BR) April 27,2004 Sector and Thematic Evaluation Group Operations Evaluation Department Currency Equivalents (as of July 1993) (as of March 14,2003) Currency Unit = Cruzeiro (Cr$) Currency Unit = Real US$l .OO= Cr$48.107 US$l .OO= R$3.48 Cr$ 1.OO= US$O.O2 R$l .OO= US$O.287 Abbreviations and Acronyns AIDS Acquired immunodeficiency syndrome ARV Anti Retroviral (drugs) ART Anti Retroviral Therapy AZT Azido-Deoxy Thymidine CAS Country Assistance Strategy CBO Community-based Organizations CDC United States Centers for Disease Control and Prevention CN National Coordination (AIDS Program) csw Commercial sex worker CTA Counseling and Testing Centers GOB Government of Brazil HAART Highly Active Antiretroviral Therapy HIV Human immunodeficiency virus ICR Implementation Completion Report IDU Intravenous drug user IEC Information, education and communication MDGs Millennium Development Goals MoH Ministry of Health MSM Men who have sex with men NASCP National AIDS and STD Control Program NGO Nongovernmental organization OED Operations Evaluation Department PAM Action and Target Plans for States and Municipalities (Planos de A@es e Metas) PI Protease inhibitor (component of HAART) POA Annual Operating Plan PPAR Project Performance Assessment Report PLWHA People living with HIV/AIDS PMTCT Prevention of mother to child transmission SIAIDS information System of AIDS Project SINAN Sistenza de In$mna~6es de Agravos de NotzJica@o (National Notification System for Communicable Diseases) SISCEL Control of Laboratory tests and examinations of the AIDSISTD program SICLOM Control of pharmaceutical procurement, stock & distribution information system STD Sexually transmitted disease sus Sistema Urzico de Saude (Unified Health System) UNAIDS United Nations AIDS Programme UNDP United Nations Development Programme UNDCP United Nations Drug Control Program UNESCO United Nations Educational Scientific and Cultural Organization WHO World Health Organization Fiscal Year Government: January 1 - December 3 1 Director-General, Operations Evaluation : Mr. Gregory K. Ingram Director, Operations Evaluation Department : Mr. Ajay Chhibber Manager, Sector and Thematic Evaluation Group : Mr. Alain Barbu Task Manager : Ms. Denise Vaillancourt OED Mission: Enhancing development effectiveness through excellence and independence in evaluation. I About this Report The Operations Evaluation Department assesses the programs and activities of the World Bank for two purposes: first, to ensure the integrity of the Bank's self-evaluation process and to verify that the Bank's work is producing the expected results, and second, to help develop improved directions, policies, and procedures through the dissemination of lessons drawn from experience. As part of this work, OED annually assesses about 25 percent of the Bank's lending operations. In selecting operations for assessment, preference is given to those that are innovative, large, or complex; those that are relevant to upcoming studies or country evaluations; those for which Executive Directors or Bank management have requested assessments; and those that are likely to generate important lessons. The projects, topics, and analytical approaches selected for assessment support larger evaluation studies A Project Performance Assessment Report (PPAR) is based on a review of the implementation Completion Report (a self-evaluation by the responsible Bank department) and fieldwork conducted by OED. To prepare PPARs, OED staff examine project files and other documents, interview operational staff, and in most cases visit the borrowing country for onsite discussions with project staff and beneficiaries. The PPAR thereby seeks to validate and augment the information provided in the ICR, as well as examine issues of special interest to broader OED studies. Each PPAR is subject to a peer review process and OED management approval. Once cleared internally, the PPAR is reviewed by the responsible Bank department and amended as necessary. The completed PPAR is then sent to the borrower for review; the borrowers' comments are attached to the document that is sent to the Bank's Board of Executive Directors. After an assessment report has been sent to the Board, it is disclosed to the public. About the OED Rating System The time-tested evaluation methods used by OED are suited to the broad range of the World Bank's work. The methods offer both rigor and a necessary level of flexibility to adapt to lending instrument, project design, or sectoral approach. OED evaluators all apply the same basic method to arrive at their project ratings. Following is the definition and rating scale used for each evaluation criterion (more information is available on the OED website: http://worldbank.org/oed/eta-mainpage. html). Relevance of Objectives: The extent to which the project's objectives are consistent with the country's current development priorities and with current Bank country and sectoral assistance strategies and corporate goals (expressed in Poverty Reduction Strategy Papers, Country Assistance Strategies, Sector Strategy Papers, Operational Policies). Possible ratings: High, Substantial, Modest, Negligible. Efficacy; The extent to which the project's objectives were achieved, or expected to be achieved, taking into account their relative importa rice. Possible rafings: High, S ubstantial, Modest, Negligibl e. Efficiency: The extent to which the project achieved, or is expected to achieve, a return higher than the opportunity cost of capital and benefits at least cost compared to alternatives. F)ossib/e ratings: High, Substantial, Modest, Negligible. This rating is not generally applied to adjustment operations. Sustainability; The resilience to risk of net benefits flows over time. possible ratings: Highly Likely, Likely, Unlikely, Highly Unlikely, Not Evaluable. hstitutional Development Impact: The extent to which a project improves the ability of a country or region to make more efficient, equitable and sustainable use of its human, financial, and natural resources through: (a) better definition, stability, transparency, enforceability, and predictability of institutional arrangements and/or (b) better alignment of the mission and capacity of an organization with its mandate, which derives from these institutional arrangements. Institutional Development Impact includes both intended and unintended effects of a project. possible rafings: High, Substantial, Modest, Negligible. Outcome: The extent to which the project's major relevant objectives were achieved, or are expected to be achieved, efficiently. Possible ratings: Highly Satisfactory, Satisfactory, Moderately Satisfactory, Moderately Unsatisfactory, Unsatisfactory, Highly Unsatisfactory. Bank Performance: The extent to which services provided by the Bank ensured quality at entry and supported implementation through appropriate supervision (including ensuring adequate transition arrangements for regular operation of the project). Possible ratings: Highly Satisfactory, Satisfactory, Unsatisfactory, Highly Unsatisfactory. Borrower Performance: The extent to which the borrower assumed ownership and responsibility to ensure quality of preparation and implementation, and complied with covenants and agreements, towards the achievement of development objectives and sustainability. Possible ratings: Highly Satisfactory, Satisfactory, Unsatisfactory, Highly Unsatisfactory. iii Contents Principal Ratings ~.~..................b........b...........~............~.............~..b........~...b......~~........~b..~~..~..b........~ V Preface .*...**.....**..*....*....**........*...*....*.*................*.bbbbbbbb*bbbbb...*.*.*..*...*...*.**..............*bb...b..*b.......**.. vii Summary ..**...**.......*....*...b*....*bbbbbb*..bbbbbb*bbbbbo.o.**bbb.bbbbbb.bbbb*bbbb*bbbb..*.bbbbb.bbbbb.bbbb..bbbbbb....**..b...*.,.*..b... ix 1. Background and Context ...~.........~~~~......~~~.~.......~..........b......~..b~b...~......~....~....b...b...b...~.~.......b.~1 2. Objectives and Design ............................................................................................................. 4 3. Implementation and Costs ...................................................................................................... 7 4. Outputs and Outcomes by Objective .......*.**......*...*...**....*b*..............*b...*.*....*...*......b.b...*..*.... 9 5. Ratings .................................................................................................................................... 23 6. Findings and Lessons ............................................................................................................ 28 7. Future Directions .............*..*......***....*......*........b.b..*..b**.............*....b.b....*.....**..................b.*b**. 30 References *.*...**.*......b***bbbbbbb.bbb..b*bbbbbbbb*bbbb***.*.......b.*.bb*b*.b*bb.*b.bb..b.b.....*.....bbb.bbbbbbbb*b.....**.........**..*..35 Annex A. Basic Data Sheet .....*.....*........*.****.......bbbbbbbb.b*bbbb...*...*..*....bbbbbbbbbbbbb*bbbbbbbbbbb*b.bb.bb*...**....37 Annex B. Persons and Organizations Consulted ............*..........*....*.........*..........*.*.......*.*....... 41 Annex C. HIV/AIDS Epidemic in Brazil 45 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Annex D. AIDS I and II: Presentation of Project Components ............................................... 49 Annex E. Project Costs and Financing ..................................................................................... 53 Annex F. Selected Data on Project Performance .*.***..*....**....*..*.**.......*.*...*.*.....*...**..**...*.*...*... 57 Annex G. Borrower Comments ...*.....***............*........*.......bb.....*.........*...........*....*.........b**b......... 63 Table 1. NGO Projects Supported by AIDS I and AIDS II ..................................................................... 8 Table 2. Summary OED Ratings of Outcome by Objective: AIDS I .............................................. 9 Table 3. Summary OED Ratings of Outcome by Objective: AIDS II ........................................... 13 Table 4. HIV Prevalence Estimates for Adults Aged 15-49, in Brazil, Year 2000 ....................... 15 Box 1. Issues and Challenges of Health System Development in Brazil .......*..*....*.**.*......*.........**. 2 Figure 1. Annual Sales and Unit Prices of Male Condoms in Brazil: 1992-2003 ......................... 18 This report was prepared by Denise Vaillancourt (Task Manager) who assessedthe project in April/May 2003. The reDort was edited bv William Hurlbut. and Pilar Barauero txovided administrative support. Principal Ratings ~&~~m-~~---~-~~----- - - -__---,-- I-~xx_^____-,,._-I_-XI-.XI_.......e,..,,xx-,~~"IIx"~-~,,~__x__I_x__-~--....,,,,~XXIXI-.d..A,,......-,-*,."-*,..a**am-,,e.eul-,s~A-.,_y_ema,......7.~,xx-,-.-"-..--~.-~~---..-I AIDS I A/OS ..._,---~",~~-,-~-,-,~~--"-,-~"~~-1 fiDS -AIDS 11** /CR /CR Review* PPAR PPAR Outcome Satisfactory Satisfactory Satisfactory Satisfactory Institutional Development Impact Substantial Substantial Substantial Substantial Sustainability Likely Likely Likely Likely Bank Performance Highly Satisfactory Satisfactory Satisfactory Satisfactory .--NYBorrower PerformanceIQ"II-P-l.memmm-&---M Satisfactory --MM-m -m----v---P Satisfactory Satisfactory Satisfactory *The Implementation Completion Report (ICR) is a self-evaluation by the responsible operational division of the Bank. The ICR review is an intermediate Operations Evaluation Department (OED) product that seek to independently verify the findings of the ICR. ** At the time of preparation of this report the Implementation Completion Report (ICR) on AIDS II was not available. Key Staff Responsible First AIDS and STD Control Project (AIDS I) Appraisal Maureen Lewis Alain Colliou Rainer B. Steckhan Completion Anabela Abreu Xavier Coll Gobind T. Nankani Second AIDS and STD Control Project (AIDS II) I......"Y~..,..a~--...,,,s,...,....,-",,,,-....^"xTdv...-,x.,-,-,,x_I.-,_,,".I(x,,-.,,-~.~-~II-."_,JYXW_YII,-"~~,-~,",,,,X--XI~~,-"X~.~,~~"~,,,~~.~,~-~~,~~-,-,~~"X.,--,X--~,t~~~"~.~"~~~~ Task Manager/Leader Division Chief/Sector Director Country I~X~~X--X-Y-.,..,+.a Director Appraisal Polly Jones Xavier Coil Gobind Nankani Completion Sandra Rosenhouse Evangeline Javier Vinod Thomas vii Preface This is the Project Performance Assessment Report (PPAR) for the First and Second AIDS and STD Control Projects (AIDS I and II) in Brazil. AIDS I was financed through IBRD Loan No. 3659 for $160 million with a government contribution of US$90 million equivalent and no donor cofinancing for a total cost of US$250 million equivalent. The loan was approved on November 9, 1993, became effective on June 1, 1994, and closed on June 30, 1998, as originally scheduled. AIDS II was financed through IBRD Loan No. 4392 in an amount of $165 million with government counterpart financing of US$135.0 million equivalent for a total cost of US$300.0 million equivalent. This second loan was approved on September 15, 1998, became effective on February 9, 1999, and closed on June 30,2003, six months after the original closing date. The loans were disbursed at 100 percent and 98 percent, respectively. The findings of this assessmentare based on an Operations Evaluation Department (OED) mission to Brazil, which took place in April/May 2003. This mission met in Brasilia with authorities and staff of the National HIV/AIDS Program and with key actors of the Federal District of Brasilia. It also visited five additional states (Pemambuco, Rio de Janeiro, Santa Catarina, Rio Grande do Sul and Sao Paulo), interviewing a wide range of actors (encompassing public sector and civil society actors and beneficiaries) and visiting selected facilities, research/academic institutions and community-based projects. Key documentary sources consulted include project appraisal documents on AIDS I and II and a completion report on AIDS I. (A completion report on AIDS II has not been issued at the writing of this review.) Supervision and evaluation reports, routine correspondence and other technical reports contained in the project' files were also consulted. This assessmentalso drew on documentation produced in Brazil (government and nongovernmental sectors alike), and on available Brazilian and international research on STDs and HIV/AIDS in Brazil. This PPAR is one of six conducted on the "first generation" of the Bank's HIV/AIDS projects as part of an ongoing OED evaluation of the Development Effectiveness of World Bank Assistance for Fighting the AIDS Epidemic. It was prepared in conjunction with a more extensive case study on the World Bank's support to Brazil in the fight against HIV/AIDS over the past lo-15 years, which encompassed policy advice and technical and financial support to Brazil's HIV/AIDS efforts. The OED mission gratefully acknowledges all those who made time for interviews and provided documents and information, including managers and staff of the Ministry of Health/National Program for STD and AIDS Control, and multiple actors and stakeholders of State and municipal programs for STD and AIDS Control in the Federal District of Brasilia, and in the States of Pemambuco, Rio de Janeiro, Santa Catarina, Rio Grande do Sul, and S%oPaulo. The mission is equally grateful to members of Brazil's research/academic community, representatives of nongovernmental and civil society organizations; staff of partner institutions and World Bank staff. This report draws heavily on the technical reports and inputs of mission members, Varun Gauri, Economist, Development Research Group, World Bank, and Chris Beyrer, Consultant Epidemiologist. The team is grateful to Francisco Inacio Bastos, Deputy Coordinator, Fiocruz AIDS Programme, Rio de Janeiro, for the extensive literature search and other technical inputs to this evaluation. Following standard OED procedures, copies of the draft PPAR were sent to the relevant government officials and agencies for their review and comments. A number of observations were made, which have been incorporated into the PPAR. Government's comments are presented in full in Annex G. ix Summary Shortly after the first AIDS caseswere diagnosed in Brazil in 1982 both civil society and state and federal governments took action to contain the growth of the HIV epidemic. By the time World Bank support for the fight against HIV/AIDS was solicited in 1992, Brazil already had a well-established National AIDS Program and National AIDS Commission, and almost a decadeof experience in the design and implementation of innovative prevention and human rights promotion activities carried out by a multiplicity of governmental and non-governmental actors. By that time as well, Brazil had already initiated treatment and care provision. In responseto Brazil's request the World Bank, through two projects, has provided technical and financial assistanceover the past decade to support: (a) prevention; (b) diagnosis, treatment, and care; (c) surveillance, research, monitoring and evaluation; and (d) institutional development The overall objectives of AIDS I (IBRD loan of US$160 million equivalent) and AIDS II (IBRD loan of US$165 million equivalent), approved in 1993 and 1998, respectively, were to reduce the incidence and transmission of HIV and STD infection and to strengthen program capacity in the areasof prevention, treatment, and management. These loans are now fully disbursed (at 100 percent and 98 percent, respectively) and a third loan of US$lOO million equivalent was approved in June 2003. While Brazil has implemented many innovative and theoretically cost-effective interventions, evidence of the actual cost-effectiveness and impact of its program efforts is lacking. In the absenceof adequatesystems for HIV and behavioral surveillance and for program monitoring and evaluation, confirmation is difficult of the achievements of Brazil's HIV/AIDS program, as well as of the World Bank's contribution thereto. Available data and information, combined with the culled views and perspectives of the myriad actors and beneficiaries of the Bank-financed projects, do point to successful outcomes of the Bank's support, as well as to some shortfalls. The loans have supported improvements and further expansion of state-of-the-art prevention activities, maintaining a focus on high-risk groups (men who have sex with men, injection drug users [IDUs] and commercial sex workers), expanding efforts to include at-risk populations (especially female sexual partners of IDUs and bisexual men, and persons with multiple sex partners), and attempting to maintain low prevalence rates among women, youth, rural populations and the poor. Sustained and innovative prevention interventions were extended to all regions in Brazil and may have contributed to the slowing or containment of the epidemic, reflected in a low prevalence rate of 0.65 percent among adults 15-49 years of age for year 2000 and a continued concentration of the epidemic among high-risk populations. Limited data indicate that during the life of theseprojects awarenessof the disease and of how to prevent its transmission has increased and healthier behaviors are being adopted by general populations, especially younger populations, and high risk groups. The Bank-financed prevention efforts have probably averted increased infection and sickness and kept the pool of HIV/AIDS patients under treatment and care lower than X would have been the case in the absenceof prevention, rendering the treatment and care efforts (described below) both more affordable and more feasible. AIDS I and II supported the expansion to all 27 statesand the quality enhancement of diagnosis, treatment and care services for persons with HIV, AIDS and STDs through: (a) the establishment and support of a network of 237 counseling and testing centers; (b) the development of an extensive network of specialized health services for HIV/AIDS patients, with much of the care shifted from expensive hospitalization to cheaper alternatives for inpatient and outpatient services; (c) the strengthening of Brazil's laboratory system for critical monitoring and follow-up of patients; (d) the development of software to manage and control the distribution and use of antiretroviral drugs (ARVs); and (e) training of tens of thousands of health care providers and other technical staff. While AIDS II did not finance the costs of ARVs, both AIDS I and II did support critical investments in treatment and care capacity, without which ARVs could not have been effectively delivered to patients. In 2003 about 140,000 AIDS patients were reported to be under active treatment in Brazil's program, and an additional 196,000 HIV-positive asyrnptomatic caseswere being monitored but not yet on treatment. Evidence suggeststhat AIDS morbidity, mortality, opportunistic infection rates and hospitalizations have sharply declined since 1997, the year following Brazil's decision to extend its policy of universal accessto health care to highly active antiretroviral therapy. Low drug resistance in Brazil (estimated at 6 percent) is an indicator of the quality of the treatment program. A continued concern, however, is that the majority of the estimated 600,000 Brazilians living with HIV in 2000 do not know that they are infected with the virus. Utilization of counseling and testing services is still modest. Furthermore, it is assumed that most of those with confirmed HIV would seek treatment and care that is available free of charge. Institutional Development. Both through innovative design and well-conceived conditionality, AIDS I and II contributed significantly to capacity development at the state and municipal levels. The projects nurtured and supported the strengthening and establishment of STD/AIDS coordination units in all 27 statesand in 150 municipalities and initiated a process of financial transfers to states and municipalities on the basis of targets and objectives contained in annual action plans prepared at the local level and negotiated and co-signed with the federal level. This process is now being consolidated and institutionalized. Bank assistancealso stimulated and nurtured public-private partnerships by making financing available for contracts with NGOs and the private sector to carry out critical components of prevention and care efforts, thus expanding considerably Program capacity and coverage, especially in reaching marginalized and hard-to-reach groups. This arrangement is being decentralized increasingly to statesand municipalities. Shortfalls of AIDS I and II. The projects were unsuccessful in setting up adequate surveillance systems for HIV and STDs and in strengthening program evaluation. There is no systematic monitoring of infection rates and behaviors of high risk groups or of high-risk sexual behavior among the general population. STDs are still considerably xi underreported and limited to clients of public STD services. Both projects supported ad hoc research on HIV infection and behavior of different populations in different geographical regions. However, the lack of a methodological framework for the design and conduct of these studies has obstructed the systematic tracking of trends over time. At the close of AIDS II the program was still without a viable system for monitoring and evaluating program performance and impact, despite the strong emphasis placed on M&E in each of the project designs, but important steps are now underway to strengthen and decentralize M&E capacity. The projects were also unsuccessful in assessingthe cost- effectiveness of program interventions. AIDS I and AIDS II had satisfactory outcomes overall. The institutional development impact of each project is rated substantial; and the sustainability of each project is likely. Both Bank and Borrower performance were satisfactory. Lessons and Challenges The gaps in data and trends on HIV infection rates and on behaviors among high-risk and general populations, attributable to inadequacies in the surveillance of these groups, make it difficult to track or model the evolution of the epidemic and behaviors over time and thus to evaluate and enhancethe effectiveness or impact of program interventions. Early, rigorous and sustained prevention efforts are likely to keep the pool of HIV and AIDS patients small relative to the counterfactual, and thus to reduce the burden on a country's health system, making treatment and care more feasible and affordable. The Bank supported Brazil to this effect, but Brazil's early responseon the prevention front, well before Bank intervention, was most critical. Even when financing a relatively modest shareof the overall costs of a well- established program, Bank support to HIV/AIDS efforts can make significant contributions to program expansion and to capacity building when it is strategically allocated. A researchprogram that is ad hoc and uncoordinated in its conception is likely to generate a large volume of information and data that will be difficult to interpret for measuring and improving program performance and impact. The publicly funded research agenda should be strategic both in the choice and prioritization of topics and in the standardization of methodology that would accommodate the analysis of data across studies and of trends over time. Project design and conditionality can be strategic and pivotal in initiating and supporting institutional innovations, which are capable of expanding and using more fully program implementation capacity including that of NGOs, decentralized levels of the program, and other development sectors. Gregory K. Ingram Director-General Operations Evaluation 1. Background and Context 11 Shortly after the first AIDS caseswere diagnosed in Brazil in 1982 both civil society and state and federal governments took action to contain the growth of the HIV epidemic. By the time World Bank support for the fight against HIV/AIDS was solicited in 1992, Brazil already had a well-established National AIDS Program and National AIDS Commission, and almost a decade of experience in the design and implementation of innovative prevention and human rights promotion activities carried out by a multiplicity of governmental and non-governmental actors. By that time as well, Brazil had already initiated treatment and care provision. In response to Brazil's request the World Bank, through two projects, has provided technical and financial assistanceover the past decade to support: (a) prevention; (b) diagnosis, treatment and care; (c) surveillance, research, monitoring and evaluation; and (d) institutional development. Financing provided under AIDS I amounted to about one-third of total public expenditures on HIV/AIDS/STDs in Brazil during the project period, net of expenditures on drugs; and that provided under AIDS II amounted to about 10 percent of total expenditure on HIV/AIDS during the life of the project, net of drugs (Ministry of Health, August 2002).Z 12 Economic and Political Context. Following a period of economic stagnation and high inflation during the 1980s and early 1990s and the impeachment in 1992 of President Collor, President Cardoso, during his two terms in office (1995-l 998 and 1999.2002), brought about considerable improvements in the economy and impressive gains in social indicators by the year 2000.3 However impressive, these gains mask considerable inequities across regions and acrossincome quintiles" (Gwatkin and others 2000). Newly elected President Lula da Silva is committed to fiscal discipline and targets agreedwith the International Monetary Fund in September 2002 and aims to: (a) consolidate hard-won stabilization gains; (b) reduce poverty and acute inequities; and (c) improve social indicators. 13 Since the early 1990s the World Bank's Country Assistance Strategies (CAS) have been overtly oriented around the central objective of poverty alleviation. The diminishing importance of Bank disbursements as a percentage of total debt flows to Brazil (from 16.5 in the 1980s to 6.4 in the 1990s) has led CASs to be more selective and more focused on the roots of poverty and on Brazil's poorest regions (Northeast). In this 1. Except as noted, this section is based on information contained in OED's Brazil Country Assistance Evaluation (2003). 2. Data on public expenditure for HIV/AIDS is not broken down by program component, making it impossible to assess the share of Bank's financing by component. 3. Under-five mortality rates decreased from 58 per thousand in 1990 to 39 in 2000 and infant mortality rates from 47.8 to 3 1.7 per thousand live births over the same period, attributable in great part to impressive immunization rates (99 percent in 1999). Maternal mortality at the end of the 1990s in Brazil is estimated at 160 per 100,000 live births, down from 260 in 1995. 4. For example, Demographic and Health Survey estimates for the period 1986-96 indicate that infant mortality rates for the lowest income quintile (83.2) are almost triple the level estimated for the highest income quintile (28.6). 2 t context health and education have been accorded high priority, along with the provision of other basic services and assetsto the poor. Box 1 summarizes the priorities and challenges of Brazil's health sector and the orientation of Bank support. Box 1. Issues and Challenges of Health System Development in Brazil The new Constitution of 1988 established health as a universal right and a duty of the State. It also created the framework for a gradual decentralization of health services. In 1990 the unified public health sys tern (SUS) was created. Under this new system, the social security health system was opened up to all citizens and service provision is consolidated into a single system and decentralized from the federal to state and municipal governments. This system also provides for a shared responsibility for public health matters among all levels of government, including disease control and prevention. Reform and refinement of this system are ongoing. OED's evaluation of Brazil's Country Assistance Strategy (2003) notes the need for irnprovement in the following areas: decentralization and regionalization of health care; allocative efficiency, cost containment and financial sustainability; reducing inequities in health status and in service access and reaching poor, marginalized populations; continued expansion of basic health care through the Family Health Program; human resources development and management; and the regulation of private health sector. The Bank's technical and financial assistance to Brazil's health sector seeks to support consolidation of the SUS, especially with regard to securing cost-effectiveness, affordability and sustainability. 14 HIV/AIDS Epidemic in Brazil. J 67 The following is a brief summary. A fuller description and some data on the HIV/AIDS epidemic are presented in Annex C. With the first AIDS casesdiagnosed in 1982, the HIV/AIDS epidemic in Brazil is now entering its third decade. HIV/AIDS in Brazil was first characterized by early (1980-85) and rapid spread of HIV-l among urban homosexual and bisexual men (MSM), and soon followed by substantial spreadin urban injecting drug users (IDU). The second decade of HIV/AIDS, roughly 1992-2002, was marked by increasing spreadof HIV among high- risk heterosexuals, including commercial sex workers (CSW), sexually-transmitted disease (STD) clinic attendees,and the wives and sex partners of male IDUs and bisexuals. The gender ratio in AIDS casereports has fallen from 25: 1 male to female in the earliest years of the epidemic to 1.7:1 in 2003. 15 AIDS mortality peaked at 12/100,000 in 1995, before the advent of highly active antiretroviral therapy (HAART), and has been declining ever since, reaching 6.3/100,000 by 2000. As of 2000 there were an estimated 600,000 Brazilians living with HIV infection, compared with PAHO's estimate of 750,000 in 19928. As of March 2002, 5. Unless otherwise indicated, data cited in this section are drawn from MoH/NASCP statistics. 6. Because AIDS case reporting in Brazil has been mandatory since 1986 and routine HIV surveillance is recent and still being strengthened, data on AIDS are much more plentiful and reliable. However, data on AIDS need to be interpreted with caution. Current AIDS cases reflect risky behaviors and infections that happened some 10 years previously. 7. Definitions of epidemiological terms used in this section: HIV incidence is the rate at which new HIV infections are acquired over a given period of time in a population. AIDS incidence is the rate of new AIDS diagnoses over a given period of time in a population. HIV prevalence is the proportion of persons with HIV infection in a given population at a specific moment in time. AIDS prevalence is the proportion of persons with full-blown AIDS in a population at a given time. 8. As part of the preparation of AIDS III, the Bank estimated that about 800,000 Brazilians were infected in 2002. The government is currently preparing an estimate for 2003. 3 237,588 AIDS casesand 110,65 1 deaths from AIDS had been reported since the beginning of the epidemic. Drawing on HIV surveillance data on pregnant women, the population prevalence of HIV among reproductive age adults has been estimated at 0.65 percent, (Szwarcwald (CL) 2002) well below the 2 percent rate of reproductive age adults, which UNAIDS has suggested for describing generalized epidemics . 16 Brazil's HIV epidemic thus remains concentrated, still largely limited to individuals and groups engaging in high-risk behaviors for HIV, including MSM, IDU, CSW, heterosexuals with multiple partners and female partners of IDU and bisexual men. Studies on the HIV prevalence of the highest risk groups reveal prevalence levels of 36.3 percent among IDUs, 10.8 percent among MSM and 6.5 percent among CSW (see Annex C, Table 2) for recent years in various samples, most of them non-nationa19. Trend data on HIV prevalence over time are not available for MSM and CSW. Analysis of trends among IDUs indicate a decline in HIV prevalence in selected urban areas.`* 17 The majority of reported AIDS caseswere causedby heterosexual transmission, still limited to high risk groups (partners of IDUs and bisexual men, CSW and people with multiple partners). Transmission through intravenous drug use rose from an estimated 18.5 percent of total reported AIDS casesin the 1980s to a high of 24.8 percent in 1992 and subsequently declined in the ensuing years to 11.4 percent in 2000. Transmission by transfusion has decreasedas a share of reported AIDS cases from 5.3 percent in the 1980s to 0.2 percent in 2000, while peri-natal transmission rose from 1.8 percent in the 1980s to 2.5 percent in 2000. 18 Incidence of AIDS in Brazil, as measured by year of diagnosis, increased from 8:2/100,000 in 1991 to a high of 18.2/100,000 in 1998 with a decline to 12.0 recorded in 2002, but patterns across regions vary. Again, trends in AIDS casesreflect levels and patterns of transmission that occurred some 10 years before. 19. Brazil's Response. Brazil's responseto the epidemic has been based on a three- pronged strategy: early and sustained prevention, the promotion and protection of human rights, and universal accessto treatment and care. In the early 1980s both civil society and government initiated action to contain the growth of the HIV epidemic. The first State AIDS program was established in Sao Paul0 in 1983 and two years later the number of established State programs increased to 11 of Brazil's 27 states. By 1986 an AIDS program was established under the Division of Sanitary Dermatology in the Ministry of Health and notification of AIDS casesbecame compulsory nationwide. The National AIDS Commission was established in 1988 under the leadership of the Ministry of Health and included representatives of Ministries of Labor, Justice, Education, the Order of Lawyers, the Federal Counsel of Medicine and other civil society organizations. By that year serological testing was required for all blood banks and Congress passeda law granting persons living with AIDS the rights guaranteedto workers with incapacitating or 9. As detailed in Annex C, Table 2, data must be interpreted with caution, due to quality of some studies and representativity of study samples. 10. Staff of Fiocruz AIDS Programme have informed the team of new, more sophisticated analyses of trends among lDUs in Rio, just being submitted for publication, and refer as well to aI published article on trends in Santos, highlighted in para. 4.15 (Mesquita et al.). 4 terminal illnesses. In 1989 the National AIDS Division, responsible for the National AIDSSTD Control Program (NASCP), was established within Ministry of Health. Also during the 198)Osnational, state and municipal governments and civil society (both for- profit and not- for-profit sectors) designed and launched campaigns to inform the general public, carried out targeted interventions aimed at changing the behavior of high-risk groups (MSM and IDUs), and undertook advocacy and human rights activities. 1.10 In 1991 MoH began the acquisition and free distribution of antiretroviral drugs (AZT monotherapy) to AIDS patients. In 1996, with the development of the first AIDS therapies with real benefits in survival and quality of life, Brazil's policy of universal accessto health care was extended to highly active antiretroviral therapy (HAART)! According to MoH statistics, in 2000, of the 600,000 people with HIV, some 93,500 were on antiretroviral therapy (ARV) and an additional 130,900 HIV-positive asymptomatic caseswith CD4 count above 200 were being monitored but were not on treatment. In 2003 about 140,000 patients were on ART with an additional 196,000 monitored but not yet on treatment. (NASCP Statistics). AIDS morbidity, mortality, opportunistic infection rates and hospitalizations have sharply declined since 1997. A continued concern, however, is that the majority of the estimated 600,000 Brazilians living with HIV do not know their status, have not had HIV testing, and a significant proportion are likely to be continuing behaviors that could maintain HIV spread. 2. Objectives and Design AIDS I 21 The two overall objectives of the First AIDS and STD Control Project (AIDS I) were to reduce the incidence and transmission of HIV and STDs and to strengthen public and private institutions responsible for STD and HIV/AIDS control. Its specific objectives were to put in place a set of institutions and activities that would: (a) define a minimal program for HIV and STD prevention and identify what interventions work best; (b) develop and adequately equip a technically competent cadre of health professionals to diagnose, treat and deliver social services to HIV/AIDS patients; (c) establish effective surveillance for HIV/AIDS and its corollary illnesses (e.g., STDs and tuberculosis); and (d) sustain project investments by the states and municipalities. 22 AIDS I was composed of four components: (a) prevention, including information, education and communication (IEC) programs aimed at the general population and at high-risk groups, expansion of capacity in diagnosis, counseling, testing and early intervention and condom procurement and distribution; (b) services for HIV and STD patients, including improvements in early diagnosis and treatment of STDs and integration of prevention and medical services with counseling both to improve and reduce the cost of care; (c) institutional development, including training of service Il. HAART is defined as triple agent therapy, with or without a protease inhibitor (PI). All patients in Brazil are now started on triple agent therapy. However, there are a small number (less than 10,000 overall) who were started on two drug regimens and had not been moved to three drug regimens by mid-year 2003. The OED case study on Brazil HIV/AIDS, currently under preparation, includes a description of the antiretroviral drug programs and regimens in Brazil. 5 providers, upgrading of laboratory services and technical assistance; and (d) surveillance, research and evaluation, including epidemiological surveillance, program evaluation and special studies on epidemiology, costs and economics of the disease. 23 Implementatiorz Arrangements. The National AIDS and STD Control Program (NASCP), under the Secretariat for Health Policy in the Ministry of Health (MOH), was responsible for project management and coordination. The head of the NASCP was designated as general project manager and was to be supported by an assistantproject manager and a team of experts recruited to satisfy all the fiduciary exigencies of the Bank. The project was designed to transfer resources to: (i) the federal level (primarily NASCP) to strengthen its capacity to design, pilot, lead, guide and support, contract out and, in some cases,implement a range of national program activities, encompassing prevention, care and treatment, surveillance, research, monitoring and evaluation; (ii) states and selected municipalities to support the design and implementation of their HIV/AIDS action plans; and (iii) non-governmental organizations and civil society organizations (NGOsKSOs), through contractual arrangements, for carrying out projects covering a range of activities, including: prevention, human rights, care and treatment. 24 All 27 states and 43 municipalities2 (the latter selected on the basis of needs and capacity criteria) were eligible to benefit from project support but could only accesssuch support on the condition that they would: establish, staff, and co-finance an HIV/AIDS and STD Control Coordination Unit within the health structure; sign agreementswith Ministry of Health on the terms and conditions of the transfers (a condition of effectiveness for the first five states; a condition of disbursement for the remaining states/municipalities); and prepare consolidated annual operations plans (POA) to be approved by the Bank (disbursements were contingent on inclusion of activities in POA). 25 Transfers to NGOs were contingent on the satisfaction of two sets of criteria, one l set to determine the adequacy of the NGO itself (structure, legal status, profile, organization and staffing, sources of financing and local level assessmentof its technical capacity and performance), the other to evaluate the quality of NGO proposals against technical, managerial, financial, epidemiological and geographic criteria. Proposals were reviewed and selected by a national panel. Those proposals that were considered promising, but in need of technical refinement, would receive technical support to raise them to a standard to be selected. 26 Both the direct transfer of project resources to statesand selected municipalities and the award and financing (through a competitive process) of contracts with NGOs for prevention and treatment/care activities were innovations in Brazil, conceived to accommodate a multi-sectoral approach to the fight against HIV/AIDS and to give greater responsibility to states and municipalities for the preparation and negotiation of POA conceived at the local level. 12. At the time of project preparation there were about 5,000 municipalities in Brazil. There are currently about 5,700. AIDS II 27. The overal 1objectives of the Second AIDS and STD Contro 1Project (AIDS II) were to reduce the incidence of HIV and STD infection and to expand accessto and improve the quality of diagnosis, treatment and care of persons with HIV, AIDS and STDs. Building on the first project, the specific objectives of this operation were to: (a) continue support of the Government's efforts; (b) broaden prevention activities; (c) strengthen evaluation capacity; (d) work towards program sustainability; and (e) identify the most cost-effective interventions. 28 AIDS II supported three components: (a) prevention of AIDS and STDs; (b) diagnosis, treatment and care for persons with HIV, AIDS and STDs; and (c) institutional strengthening, including epidemiological surveillance and monitoring and evaluation, among other interventions. The original articulation of components and activities for AIDS I and II, as drawn from the project documentation, is presented in Annex D for reference. 29 Design Modifications. While the general design of AIDS I was maintained for the second operation, certain modifications were introduced to addresschanges in the profile of the AIDS epidemic. In addition to continued efforts to support behavior change in high-risk groups (notably MSM, IDUs, and CSW), additional prevention interventions were expanded both to include new and broader target groups (e.g., students and women through reproductive health services) and to reach additional geographical areasin the country. 2.10 The design of AIDS II also took into account four program issues. First, in response to concerns about the affordability and sustainability of Government's decision to finance the full costs of treatment of HIV/AIDS patients in the context of its overall policy of universal and free accessto health care, the second operation did not provide for the financing of ARV drugs, but did include investments to promote alternatives to hospital care (day hospitals, home care, and other outpatient services) to both improve the quality and lower the costs of such care. Studies were programmed to assessthe costs and effectiveness of care and treatment. Second, the project envisaged the support and monitoring of NASCP efforts to reduce the cost of condoms to the public through tax reductions and other means. Third, in light of poor monitoring and evaluation performance under the first operation, AIDS II emphasized the establishment of a viable monitoring and evaluation system, most notably: (a) the definition of project objectives and indicators to measure impact; (b) the specification of the type of baseline information neededto track desired changes; and (c) agreement with the national program on key evaluation activities and their terms of reference. Fourth, it sought to decentralize program management and execution by devolving some responsibilities to statesand municipalities (e.g. the gradual transfer of review, selection and supervision responsibilities for projects executed by NGOs. 2.11 Relatively minor changes were made to the implementation arrangements for AIDS II. As a means of encouraging participation of smaller municipalities and of achieving more efficiency in the transfer, use and accounting of such funds, the participation of smaller municipalities was encouragedthrough the formation of consortia. 3. Implementation and Costs Implementation Experience 3.1 AIDS I. The institutional arrangements were respected throughout the project, with the three main implementers (federal level, states/municipalities, and NGOs) managing roughly equal sharesof project resources, according to NASCP reports. 32. Project conditions for the transfer of funds to states and municipalities were by and large fulfilled and POA were generally implemented in line with expectations, although the absorptive capacity of local-level programs was very much a function of political commitment at these levels. l3 Overall a total of about $115.8 million was transferred to local Governments under AIDS I during the period 1994-l 998.14Support to statesvaried with regard to their per capita amounts and with regard to the allocation per identified caseof AIDS. (See Table E-3 in Annex E.) 33. NGOs undertook a large shareof the prevention work financed under the project, in line with their acknowledged comparative advantage to reach and work effectively with high-risk groups. With the support of AIDS I 181 NGOs carried out 444 prevention projects and 140 treatment and care projects, for a total of some 584 projects or more than double the appraisal target of 261 (see Table 1 for details)." 34. The NGO liaison office within NASCP managed the NGO contracting. The monitoring of these projects was undertaken through the preparation by NGOs themselves of quarterly reports on implementation progress and through supervision visits undertaken by national program staff. To further consolidate successin contracting with NGOs, the mid-term review recommended: more attention to support of NGOs in their efforts to achieve financial sustainability of their activities, involvement of local level government in the selection and evaluation of NGO projects, and greater uniformity and continuity in reporting and monitoring. While progress has been made on all of these fronts during the second half of AIDS I and throughout AIDS II, the challenges of ensuring the financial sustainability of NGOs and of more rigorous and routine monitoring and evaluation of NGO activities were consistently raised in OED mission 13. For example, the city of Sao Paulo experienced major delays in the execution of its program and ultimately did not use its full allocation under the project, due to major changes in the structure of the health care delivery system, and also to low commitment of the city administration to AIDS control activities. A subsequent change in municipality leadership towards the end of AIDS I created an enabling environment for more intensified program effort. State-level staff interviewed reported that the new leadership engendered strong political support for the fight against HIV/AIDS and culminated in a well-resourced program, including both local funding and greater utilization of project funds. 14. World Bank completion report data. 15. Even with this doubling of the original target, there was capacity and demand for more projects. In the final year of implementation over 400 new proposals were received along with over 130 requests for extensions and supplemental financing. While the latter requests were fully accommodated, only 5 1 of new proposals were approved due to full utilization of NGO allocations and limited scope for reallocation in the project's final year. 8 interviews with federal and local level staff, as well as with NGO representatives. Initially the financial cap on NC0 projects was set at US$lOO,OOO,but was subsequently reduced to about US$55,000 equivalent when few if any NGOs could make use of these funds. AIDS II 35 AIDS II transferred Table 1. NGO Projects Supported by AIDS I and AIDS II resources for HIV/AIDS activities to all 27 states and the AIDS I * AIDS II ** number of municipalities ws--pp -....~,-.--- A-~~.~----~---.%~.~~ A. Prevention benefiting from direct transfers Injection drug users 15 171 more than tripled to 150. In Men who have sex with men 28 193 Commercial sex workers 54 285 December 2002, the Minister of Youth 73 192 Of which: Health signed a law establishing school children (17) federal transfers to states and Adolescents (56) children and adolescents municipalities to finance (192) Low income populations 14 160 HIWAIDSSTD activities. Women 40 213 Other 220 495 Under this law statesand Of which: municipalities must prepare and NGO sustainability ww IECXnstitutional development successfully negotiate action (118) Health and education Prof. (83) plans with targets (PAMs), Resettled populations (62) whose focus is on performance Prisoners (48) Truck drivers (26) benchmarks and targets, rather Indigenous populations (24 than inputs. This new law was Miners (10) Disabled (4 supportive of Brazil's Subtotal Prevention 444 1,709 decentralization policy and B. Treatment and Care 140 454 Total actual: 564 2,163 reflective of the design and implementation experience of 0 Source: Im AIDS I and II. In May 2003, 14 0 implementation Completion Report on AIDS I states and 158 municipalities had ** Source: Ministry of Health/NASCP 2002 approved PAMs. Because this *** Source: NASCP data provided during OED mission development happened close to the completion of AIDS II no funds under the project were channeled to statesunder this new mechanism. States continued to receive funds through the POA mechanism as designed under AIDS I. As shown in Table 4, both the number of NGOs and the number of NGO projects receiving support under AIDS II increased significantly over the levels achieved under AIDS I. Planned versus Actual Costs by Component16 36 ArDS I. The total actual cost of the project, US$250 million equivalent, was equal to that estimated at appraisal. Actual expenditures on prevention were 117 percent of the full costs estimated at appraisal, while actual expenditures on care and treatment were 89 percent of appraisal estimates. The cost of the institutional development component was about the same as appraisal estimates (102 percent). Expenditure on surveillance, research and evaluation fell far short of appraisal estimates (46 percent). 16. See Annex E for details 9 Data on public expenditure for HIV/AIDS is not broken down by program component (prevention, treatment and care, institutional development, and surveillance, monitoring and evaluation), making it impossible to assessthe share of Bank's financing by component. 37 AIDS II. The total actual cost of AIDS II was US$296.5 equivalent or 99 percent of appraisal estimates. The actual costs of the prevention and institutional strengthening components slightly exceeded full cost appraisal estimates (104 percent and 109 percent respectively), while the actual cost of the diagnosis, treatment and care component fell short (86 percent). The Bank's loan amount, which was 98 percent disbursed, amounted to about 10 percent of total expenditure on HIV/AIDS during the life of the project (Ministry of Health, August 2002) (See Table D-4 in Annex D for details.). 4. Outputs and Outcomes by Objective AIDS I Table 2. Summary OED Ratings of Outcome by Objective: AIDS I Importance" Efficacy** General Objectives To reduce the incidence and transmission of HIV and STDs High Not possible to evaluate To strengthen public and private institutions responsible for STD and Substantial Satisfactory HIV/AIDS control Specific Objectives To put in place a set of institutions and activities that would: Define a minimal program for HIV and STD prevention and identify High Moderately what interventions work best Satisfactory Through training develop an adequately equipped and technically Substantial Satisfactory competent cadre of health professionals to diagnose, treat and deliver social services to HIV/AIDS patients Establish effective surveillance for HIV/AIDS and its corollary illnesses High Unsatisfactory (e.g., STDs and TB) Sustain project investments by the states and municipalities Substantial Satisfactory 0 Importance is judged by the priority accorded to the objective in project design, based on both the project description and the amount of funds allocated. ** Efficacy is the extent to which the project's objectives were achieved, or expected to be achieved, taking into account their relative importance. 41. Because of a lack of adequatebaseline and end-of-project epidemiological and behavioral data, it is not possible to evaluate the outcome of the project against the general objective of reducing the incidence and transmission of HIV and STDs." The other project objectives are articulated in terms of outputs, rather than outcomes and thus can be assessed. The project succeededoverall in achieving the other general objective to strengthen public and private institutions responsible for STD and HIV/AIDS control. As is described in more detail below, the project supported the design, implementation and expansion of programs for prevention and for treatment and care. Of 17. The lack of data on trends in HIV prevalence and in behaviors among the general population and among high risk groups is a consequence of the project's failure to strengthen adequately surveillance and monitoring and evaluation, discussed later on in this section. 10 particular note is a significant investment in capacity building, including the training of 21,600 health professionals (target of 7,000) through the design and delivery of some 769 training courses (target of 550), and the provision of local and international technical assistance for program design and implementation (NASCP project reports). 42. The project was successful in defining and supporting a core program for HIV and STD prevention, but fell short of the objective of assessing which interventions work best with a view to guiding future decisions regarding resource allocation. AIDS I has made significant contribution to the establishment and successful implementation of sustained, innovative, state-of-the-art prevention activities in Brazil, including both information/awareness raising for the general public and behavioral change interventions aimed at high-risk groups. Both the State and civil society were particularly productive in reaching high-risk groups. For example, a combination of innovative public sector and NGO projects aimed at (a) preventing drug abuse and HIV infection among the general population and (b) reducing transmission of HIV among IDUs through education and needle exchange. Other examples of successful partnerships with civil society are found in project-supported efforts aimed at MSM and CSW, encompassing behavior change interventions and studies. Adolescents benefited from prevention efforts through partnerships with the Ministry of Education, the Roquete Pinto Foundation and Brazilian educational television. A major outcome of project support in terms of partnerships with the private sector was the creation of the Business Council (Conselho Empresarial), made up of 24 large businesses and over one million workers, which provides guidance to the private sector in the design and establishment of workplace interventions. 43 AIDS I invested in the design and delivery of 138 courses in prevention delivered to 8,673 trainees. The project also supported the design and implementation of seven national IEC campaigns, (one more than originally planned) to increase awarenessof HIV/AIDS and knowledge of how to prevent it. Other outreach/awarenesscreation activities supported included the establishment of an AIDS Hotline (contacted by about 1,000 callers per day) and a National Human Rights Network, made up of about 1,400 members with strong advocacy and outreach capacity. 44. From 1993 through March 1998 the project purchased and distributed 136 million condoms and supported social marketing activities. The national logistic system for condom distribution servesto control condom stocks, improve the flow to distributors and users and maintain a record of who is receiving condoms to project future needs. Efforts to reduce the import tax on condoms were initiated under AIDS I, and continued with greater successunder AIDS II. Figure 1 shows the decline in the unit prices of condoms over the life of the two projects. 45 Despite the lack of baseline data and a viable monitoring and evaluation system, available evidence does indicate that during the life of this project awarenessof the disease and of how to prevent its transmission did increase and healthier behaviors were adopted both by the general public and by high-risk groups. For example, the sale of domestic and imported condoms rose significantly from 70 million in 1993 to 228 million in 1997 (NASCP statistics). Program data from 1993 and 1996 show a moderate increase from 25 percent to 37 percent of males who report that they always use a condom and from 14 percent to 19 percent of males who report that they almost always use a condom. 11 In the absenceof more precise information on project and program performance, it is plausible to assume that project support of awarenessraising and targeted behavior change interventions and project achievements in improving condom availability did contribute to enhanced awarenessof the disease and risks of transmission and to the adoption of healthier behaviors documented in the above-mentioned surveys. While successful in the establishment and support of innovative and targeted prevention interventions, the project did not carry out studies and evaluations to assessthe cost- effectiveness of prevention activities, on which basis future strategic decisions and resource allocation were to have been decided. 46 The project was successful in strengthening capacity of the health system to diagnose, treat and deliver social services to HIV/AIDSpatients. At the project's outset HIV testing was very limited and carried out by blood banks with virtually no counseling services available. With the support of AIDS I a network of 104 counseling and testing centers (CTAs) was established acrossthe country covering all states except Acre (Ministry of Health/NASCP August 2002).** The project target was 125 CTAs. 47 Through an ambitious program of training of health providers, accreditation of existing hospitals and the creation of alternative services AIDS I succeededin establishing considerable capacity in Brazil to care for persons with HIV/AIDS, with much of the care shifted from expensive hospitalization to cheaper alternatives for inpatient and outpatient services. Specifically, the project fulfilled the target of establishing about 190 new treatment facilities for HIV/AIDS patients, although the mix of facilities developed was different than originally planned. Forty-nine day hospitals were established versus the project target of 100. While no public-financed home care services were originally planned under the project, 31 were established with project assistance. Sixty-six specialized outpatient care facilities were established exceeding the original target of 42 by 57%. Over 40 NGO projects established home-based care services falling short of the target of 50. In addition, while it was not included among the original project targets, 340 hospitals were accredited to provide HIV/AIDS treatment. 48 A logistical system for the management and control of ARVs was developed along with policies and guidelines for health care and other services for HIV patients. In addition, the project financed the design and delivery of 546 training courses in treatment to 10,985 health care professionals and community outreach workers across the country. While not specified in the original design, the project also financed anti-retroviral treatment for HIV+ pregnant women to prevent mother-to-child transmission of HIV and trained about 500 health professionals in the care of these patients 49 Project investments in training, equipment and other materials have also supported the establishment of a network of diagnostic and treatment services for STDs. The project supported the establishment of 700 STD clinical health services within already existing public health facilities, more than double the appraisal target (300). However, planned evaluation and research on effectiveness and cost effectiveness of different partner notification strategies, utilization and effectiveness of STD services, and 18. World Bank compl etion report data indicate that an additional 39 CTAs were in the process of being established at the time of the project' s closing. 12 feasibility of combining symptomatic and clinical testing in facilities was not carried out. Table F-l in Annex F provides statistics on STD case reports for the period 1996-2001, which are likely to be a fraction of the true total. 4.10 At the project's outset laboratories lacked the referral capacity to back up lower level laboratories and to ensure timely and quality services needed for treatment and care. Project investments supported the upgrading of the network of laboratory facilities with a view to improving diagnosis, treatment and research for HIV/AIDS and STDs. With project support 43 public laboratories began implementing viral load testing starting in 1997 and 60 laboratories throughout the country were upgraded to do T CD4+ lymphocyte testing. I9 The upgrading of the laboratory network has also enabled the monitoring of HIV subtypes and resistance to antiretroviral drugs. As a complement to support under other IBRD loans (REFORSUS project), AIDS I has helped establish an effective national blood quality control program, as evidenced in a significant reduction of the share of reported AIDS casescontracted through blood transfusions from 2.4 percent of all casesin 1992 to 0.2% in 2000 (Ministry of Health Epidemiological Bulletin, March 2002). 4.11 The project was unsuccessful in its initial attempts to set up HIV surveillance. Despite early attempts under the project to establish HIV surveillance, the system originally set up under the project was not viable given that follow-up collection samples and sample sizes were inadequate resulting in confidence intervals that were too high to be useful. The system was revamped in 1997 with project assistance and improvements included: a more careful selection of participating institutions, tighter and more central management, and more rigorous supervision. With these adjustments AIDS I ultimately did succeedin establishing 150 sentinel surveillance20sites (target of 80) in maternity clinics, emergency rooms and STD clinics. The project also financed the delivery of 72 courses in epidemiology to 1,598 health care professionals. At the project's closing in 1997, maternity and emergency room sentinel surveillance sites did manage to generate data on HIV prevalence in time for the project's final evaluation, notably: a prevalence of 1.8 percent among pregnant women (20-24 years) and 4.1% among emergency room patients (20-24 years). Sentinel surveillance at STD clinics was considered unreliable. The design of STD surveillance was also modified during project implementation, facilitating a more decentralized system of 72 state and municipal institutions (vs. 27 state facilities) selected on the basis of technical and geographical criteria. STD surveillance capacity established under the project generateddata on the clients seeking such services. These data are not representative of the population at large, or even of high-risk groups within the population; and they are under-reported. 4.12 AIDS I invested about US$6 million in 65 studies, of which 27 focused on prevention, risks and behaviors, 16 on diagnosis, treatment and care, 17 on the epidemiology and biology of the disease and 5 on other topics (costs, cost-effectiveness, program management and political economy). However, staff of state and municipal programs visited outside of Rio and Sao Paul0 interviewed by the OED mission noted the 19. Test designed to monitor the decline of the body's immune system. 20. A surveillance system in which a pre-arranged sample of reporting sources agrees to report all cases of one or more not ifiable conditions. 13 need for local-level researchthat would be more operationally relevant and useful than national level research and researchfocusing on the Southeast. Only a very small portion of research efforts and financing was devoted to the development of model refinements or to the study of costs and cost-effectiveness of prevention and treatment, which had been emphasized in the project design. Some of the research and studies undertaken with project support provided important information and insight on HIV and behavioral trends in the absenceof regular surveillance activity by the government. Most notably, AIDS I invested in a series of three studies of about 850,000 army recruits that provided critical data on HIWSTD prevalence and behavior. 4.13 Project support for ensuring the sustainability of investments at the state and municipality level was considerable. The project succeededin establishing state and municipal HIV/AIDS coordinating units, whose activities and recurrent costs were cofinanced by local government in line with decentralization policy. It is noteworthy that participation in the project required state/municipal financing of the personnel costs of coordinating units and cost sharing of selected activities. In addition, political and popular commitment at the state and municipal levels was very high overall thanks to NGO activism, the high visibility of the program and a participatory process for the planning and execution of local activities. An important exception was Sao Paul0 city, which did not, at that time, enjoy strong commitment of local leadership, and whose performance under the project suffered due to its inability to effectively absorb project funds. AIDS II Table 3. Summary OED Ratings of Outcome by Objective: AIDS II Imporfance* Efficacy** General Objectives To reduce the incidence of HIV and STD Infection High Not possible to evaluate To expand and improve the quality of diagnosis, treatment and care of Substantial Satisfactory persons with HIV, AIDS and STDs Specific Objectives To build on the first AlDSlSTD Control Project, especially to: Broaden further prevention activities (while maintaining a focus on high- High Satisfactory risk groups) Strengthen evaluation capacity High Unsatisfactory Work towards the sustainability of the program through devolution of Moderate Satisfactory some responsibilities to states and municipalities Identify the most cost-effective interventions Substantial Unsatisfactory *Importance is judged by the priority accorded to the objective in project design, based on both the project description and the amount of funds allocated. ** Efficacy is the extent to which project objectives were achieved or expected to be achieved, taking into account their importance. 4.14 In the absenceof adequatedata on epidemiological and behavioral trends it is not possible to evaluate the achievement of the project's general objective to reduce the incidence of HIV and STD infection.2' However, there are data from various studies that 21. This lack of data is a consequence of the project's failure to strengthen adequately monitoring and evaluation capacity, including epidemiological and behavioral surveillance, discussed later on in this section. 14 provide some information on trends in prevalence for different populations over different time periods. With regard to syphilis, prevalence in army conscripts is reported to have risen from 0.53 percent in 1996 to 1.8 percent in 2002 (Szwarcwald (CL) 2002), while prevalence among women who delivered babies in health facilities decreasedfrom 2.8 percent in 1998 to 1.7 percent in 2000 (SUS data). Available information on trends in HIV prevalence also shows a mixed picture. HIV prevalence is reported to have decreasedamong military conscripts from 0.2 percent in 1996 to 0.12 percent in the South, 0.19 percent in the North and Central West, and 0.08 percent in Rio and Sao Paul0 in 1998 (Szwarcwald (CL) 2002). HIV prevalence among pregnant/post-partum women of 20-24 years is reported to have increased during the period 1998 - 2002 from 0.3 percent to 0.6 percent (NASCP 2003). The proportion of children infected at birth has declined from 8.6 percent in 2000 to 7.6 percent in 2001 and 3.7 percent in 2002 (MoH statistics). 4.15 With regard to high-risk groups, while available data do indicate some encouraging signals and trends, they must be interpreted with caution as studies were not undertaken in the context of a single methodological framework nor did series of studies necessarily cover exactly the same target populations, or samples, in the same geographical areas(see Annex C, Table 2 for details). With these caveats, HIV prevalence seems to have decreasedin the IDU community from 52 percent in 1999 to 36.3 percent in 2001 (UFMGNASCP). Infection rates among this group in Rio fell to 8 percent by 2000 (Marins, et al. 2003). Prevalence rates of IDUs in Santos are reported to have risen from 63 percent in 1992 to 65 percent during the period 1994-96 and subsequently to have declined to 42 percent in 1999, asboth injection frequency and sharing of needles decreasedand smoking of crack cocaine increased (Mesquita et al.). While there are no data available on trends in HIV among CSWs and MSM over time, a study on prevalence among 3000 CSW indicated a low rate of 6.5 percent in 2001 (UNBRVASCP), and prevalence among the MSM community was estimated at an average of 10.8 percent (Sao Paulo SP, Projeto Bela Vista) during the period 1994-99. Overall, it can be noted that some 20 years after the first AIDS caseswere identified in the country, Brazil's HIV epidemic remains concentrated among a high-risk population, the prevalence of HIV having been reliably estimated by MoH at 0.65 percent of reproductive age adults. 4. I6 The project provided major support to expand and improve the quality of diagnosis, treatment and care ofpersons with HIV, AIDS and STDs. With the support of AIDS II the number of CTAs in Brazil more than doubled from 104 in 1999 to 237 in 2003, thus exceeding the project target to establish 100 new CTAs. The project also provided support to the functioning and strengthening of existing CTAs through the provision of training and materials. As shown in Chart F-2 in Annex F, there are now CTAs in every state in Brazil. The number of persons who were tested for HIV in CTAs annually rose from 120,468 in 1997 (basedon data from 1997 of 100 CTAs) to 271,056 in 2001 (based on reports from 115 of 174 CTAs). Despite this progress, there are still many HIV-positive people who have not been tested. 15 Table 4. HIV Prevalence Estimates for Adults Aged 15-49, in Brazil, Year 2000 Gender Prevalence Total infections Low range High range (40 0 Women 0.47 0.36, 0.58 N=l5,426 Men 0.84 0.65, 1.04 Estimated Total 0.65 0.51, 0.80 597,000 548,000 647,000 ~~~~~~,-~~~~~~-~~,~~~~~~~~~~-~~~~-~~-~~~-~~~*-~~-~~~~~--~~~ --+WbW-PW Source: Ministry ofHea/th/NASCP Program Statistics 4.17 The project also financed the significant expansion of low-cost, alternative, in- patient and out-patient treatment and care services. Thirty-five new day hospitals, 20 therapeutic home-based care services and 103 specialized ambulatory services were established and 47 conventional hospitals and 28 day hospitals were accredited. Thanks to this support (and that provided under AIDS I) Brazil boasts an extensive network of treatment and care facilities, including a total of 385 accredited conventional hospitals, 84 day hospitals (of which 77 are already accredited), 381 specialized ambulatory services and 58 therapeutic home-based care services. AIDS II also provided support and strengthening to already existing services/facilities in this network. In addition, a total of 369 civil society projects were financed to addressthe basic treatment and care needsof HIV/AIDS and STD patients. An array of initiatives were supported including: guest houses for the travel needs of patients, support groups, social reinsertion, income generation support, household support, psychosocial support, mental health services and assistancein assessingand adhering to ART. Some 126 projects financed the establishment of support homes for PLWHA. The number of patients under ARV treatment increased from 33,830 in May 1997 to 129,276 in May 2003 (Ministry of Health/NASCP 2003). The number of HIV-positive pregnant women under injecting AZT (Azido-Deoxy Thymidine) treatment increased from 1,472 in 1997 to 5,958 in 2002 (Ministry of HealthNASCP 2003). 4.18 Also supported were activities to strengthen the diagnosis and treatment of STDs and activities to reduce mother to child transmission of HIV/AIDS and congenital syphilis. Investments included the training of service providers and the development and dissemination of guidelines and protocols. STD case reports have increased since the Bank's involvement, with sustained increasesin reporting from 1997-2001 (see Table F-l in Annex F). However, the total reports for that five-year period of 239,357 remain a very small fraction of the likely total. 4.19 The project strengthened laboratory capacity in Brazil to improve the quality and timeliness of services to evaluate and monitor the viral and immune status of HIV/AIDS patients. HIV diagnosis was decentralized through some 70 projects to strengthen capacity in municipal and public health laboratories and to monitor drug resistance (of HIV patients to ARV, and of strains of gonorrhea to antibiotics). The project also supported the National System for Quality Control of Diagnostics in carrying out internal and external quality audits of some 521 facilities for HIV diagnosis, 357 for hepatitis diagnosis, 381 for CD4 and 264 for viral load testing. AIDS II also supported the work 16 of the Distance Education System for Blood Banks and Public Health Laboratories (TELELAB), set up with AIDS I assistance,to facilitate optimal use of this resource. Studies and other assistanceprovided under the project contributed as well to the strengthening and expansion of capacity to monitor drug resistance (of HIV patients to ARV, and of strains of gonorrhea to antibiotics). Project support was also provided in the form of technical and software assistanceand materials and supplies to strengthen the drugs and condom logistical management system established under AIDS I. AIDS II investments facilitated an increase in the number of public health laboratories that participate in external evaluations of diagnostic quality carried out by the National System of Quality Assurance for AIDS and STD lab tests, from 45 in 1998 to 123 in 2002 (Ministry of Health/NASCP 2003). 4.20 There is no data on accessof the poor and marginalized populations to diagnostic services and to ARV treatment, but both public sector and NGO personnel brought to the attention of OED's assessmentmission the need to provide social assistanceto these groups to encourage and support them in getting tested and in accessing and adhering to treatment protocols. Underutilization of these services by poor and marginal groups was not characterized by those interviewed by the mission as merely a factor of geographical access,but also an aspect of their marginalized status. The mission also noted a concerted effort on the part of CTAs visited to intensify their outreach work to promote and encourage a utilization of these services, especially among vulnerable and marginalized populations. 4.21 The mission noted indications of high quality of these services during field visits, including: technical capacity and strong client orientation of service providers, reliable supply of essential drugs and materials and available and good state of infrastructure and equipment. Another important indicator of service quality and effectiveness is the rate of ARV resistance. A recent study has documented Brazil's rate in 2003 at a low 6.6 percent, which is a reflection of rigorous patient management and follow-up and of strict monitoring and control of the distribution of ARV to patients and monitoring of patient adherence,which is built into management information systems set up with project assistance (Brindeiro 2003). Thus far, the possibility of behavior reversals due to Brazil's successful treatment program has not been monitored. 4.22 Survival after AIDS diagnosis has increased sharply from a mean of 5 months in 1989 to 58 months in 1997 and continues to increase (Marins et al. 2003 and Chequer et al. 1992 ). Since the advent of HAART, the number of hospital admissions per AIDS patient declined from 1.7 per year in 1996 to 0.28 per year in 2001 (SUS data). Brazil's experience with HAART replicates what has been seen in developed country world settings - once HAART is widely available, AIDS largely becomes an outpatient disease, and hospitalization rates, particularly for common opportunistic infections, decline sharply. MoH statistics show that during the period 1996 - 2002 implementation of Brazil's ARV policy has avoided (or delayed) 58,000 new casesof AIDS and 90,000 deaths due to AIDS. Hospitalization was reduced by 70%, with some 358,000 admissions avoided (1997-2001) and resource savings amounting to some US$2.2 billion (hospitals and outpatients) (See Table F-3, Annex F). The number of TB casesin patients with HIV/AIDS declined by 71.8 percent during this same period (MoH Statistics). These outcomes are attributable in part to Brazil's success in reducing the 17 costs of ART both through the establishment of local production capacity of some of these drugs and through rigorous negotiation of price reductions of other drugs with international drug companies (See Table F-4, Annex F, which shows average costs of ART between 1996 and 2002.). These outcomes can also be attributed in part to AIDS II investments in improving and expanding treatment and care services, as well as investments in strengthening the network of laboratories. 4.23 Building on the support of thefirst operation, AIDS II made significant contributions to the continued expansion and successful implementation of innovative, state-of-the-art prevention activities in Brazil. From 1998 to 2002, 20 mass media campaigns were carried out (3 times the number carried out under AIDS I) covering a range of topics aimed at general and specific audiences. Under AIDS II a substantial number of targeted interventions were implemented to prevent transmission among high- risk groups, as well as vulnerable groups, including truck drivers, prisoners, indigenous populations, poor and marginalized populations, adolescents and women. During the period 1999-2003 AIDS II financed 547 projects which covered an estimated 899,386 CSW, 631 projects reaching some 145,807 IDUs (an estimated 18.2 percent of that population), and 486 projects covering some 3,074,980 MSM reaching a reported coverage of 96 percent of that population. Coverage rates are in need of further substantiaton and should be interpreted with caution. These figures include both public sector and NGO-executed projects (See Table F-5, Annex F.). Overall, with project support the volume of prevention activities increased considerably, both those targeted at general populations and those targeted at high-risk groups; and prevention work was further broadened both geographically and in terms of intensified and new interventions aimed at other at risk groups, including, among others: rural, poor and indigenous populations, women, prisoners and street children. OED mission discussions with national, state and municipality program staff and with NGO representatives, combined with site visits undertaken by the mission, revealed that prevention programs are prioritized and carried out with strong commitment and concerted effort. IEC material was plentiful, appropriately targeted to different target groups, widely disseminated and easily available to others who seek information. From the mission's observation in states and project sites visited, prevention work is well grounded in good and trustful exchange with target groups, who are often involved in design, implementation and refinement of interventions. In this respect, program staff and NGOs as well as beneficiaries of prevention work interviewed stressedthe successof the project in securing respect and legitimacy for marginalized and stigmatized groups. 4.24 Public sector agencies that undertook prevention activities with project support include: Ministry of Education, Ministry of Justice (Special Secretariat for Human Rights), National Anti-Drugs Secretariat, and Ministry of Labor. Human rights promotion was also undertaken by both public and private sector agencies including capacity building, publications, direct and indirect support of people living with HIV/AIDS (PLWHA) and advocacy work with and through executive, legislative and judicial branches of government. 18 Figure 1. Annual Sales and Unit Prices of Male Condoms in Brazil: 1992-2003 0.8 427 0.6 0.2 0.1 0 I T 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Years m Male condomssold (right scale) -Reduction in mean price (left scale) Source: Ministry of HealthlNASCP Statistics 4.25 The project procured condoms that were made available to the public through targeted interventions, social marketing initiatives and distribution through the Ministry of Health. Male condoms distributed through the MoH more than tripled between 2000 and 2003 from 77 million to 270 million and those sold through the commercial sector increased from 228.4 million in 1997 to 427.1 million in 2003 (MoH/NASCP 2003). Table 8 shows that over the implementation periods of AIDS I and II the sale of condoms rose significantly and the mean price of condoms decreased. These trends are attributable in part to the efforts under these projects both to increase the supply and promote the use of condoms and to reduce the costs of condoms in to the Brazilian population through a number of measures, including reduction of duties and taxes. The project also supported operational research which documented high acceptability rates of female condoms, especially among men and women reachedby the public health services through community education efforts (Berquo et al. 1999). The results of this study served to fine-tune further project investment in the distribution, provision, targeting and social marketing of female condoms. Among the target groups are CSW, HIV-positive women, female drug users and women at risk for violence. 4.26 While a national survey of sexual behavior and perceptions did establish baseline data for 1998 (CEBRAP 2000), this survey was not repeated at the close of AIDS II, making it difficult to assessthe impact of these activities on behaviors of these group~.~~ However, it is plausible to assume that sustained prevention activities supported under AIDS II, building on those supported under AIDS I, contributed to Brazil's successin keeping its prevalence rate below 1% and in the continued concentration of the epidemic among high-risk groups. In addition, data generatedthrough studies (encompassing a 22. Government comments (presented in Annex G) do note that NASCP is commissioning a follow-up survey to be conducted in 2004 to enable the tracking of evolutions in this regard. 19 range of populations, timeframes and methodologies) do provide some indication that during the life of this project awarenessof the diseaseand of how to prevent its transmission has increased and healthier behaviors arebeing adopted both by the general population and by high-risk groups. Multiple studies on behaviors supported under the project have not been sufficiently systematic or uniform in their design and implementation. As such most of them only reveal levels for a particular population in a particular geographical areaover a specific time period, making it impossible to derive trends over time within population groups or to compare levels acrossregions. 4.27 With regard to generalpopulations, a population survey conducted in 2003 shows a significant increase in the proportion of persons who spontaneously cite the use of condoms as a method for protecting themselves, from 50 percent in 1998 to 88.9 percent in 2003 (CEBRAP 2000 and MSIBOPE 2003). The proportion of sexually active military conscripts that reported using a condom in their most recent sexual encounter was 68% in 2003, a slight increase from 1998 data that indicate proportions of 51.2 percent in the North and Central West, 66.6 percent in the South and 67.7 percent in Rio and Sao Paulo. In its final report on AIDS II MoH reports that use of condoms among the general population in all sexual encounters declined between 1998 and 2003 from 63.7 percent to 57.8 percent with casual partners and from 21.Oto 11.4 percent with long- term partners.23 4.28 A 2001 survey of schools carried out by MoH with support from UNESCO revealed that: 70 percent of the schools carried out prevention activities with students; 97% of all students had correct information on how AIDS was transmitted; and 90 percent of students with active sex lives changed their behavior regarding STD and AIDS after intensive exposure to school prevention activities. According to MoH statistics, between 1986 and 2003 use of condom in first sexual encounter among the general population increased from 4 percent to 55 percent. A 2003 poll reports that 65.2 percent of persons aged 14-19 reported using a condom in their first sexual encounter. The project was also successful in increasing the availability of condoms in the country and in lowering their costs to the public from US$O.57 per condom in 1997 to US$O.zS in 2001, due largely to reductions/elimination of duties and taxes (Figure 1). 4.29 Interventions targeted to high-riskgroups are shown to have had some impact, as derived from studies commissioned under the project. As detailed in Annex C, Table 2 these data must be interpreted with caution given issues of representativity of study samples (primarily participants). Between 1999 and 2001 condom use among IDUs increased from 42.1 percent to 62.9 percent and the proportion of IDUs ever tested for HIV increased from 52 percent to 66.4 percent. In addition, needle sharing among this group decreasedover this same time period from 70 percent to 59.4 percent (UFMGNASCP). A study of 3,000 CSW conducted in 2001 revealed that 73.8 percent of this group (composed of program participants) used condoms with their clients (vs. 60 percent of CSWs not participating in the program) but only 23.9 percent of them used condoms with their regular (non-paying) partners. Almost half of CSWs studied (49.2 23. MoH does caution in its report that this trend is questionable as the earlier study (CEBRAP 1999) documented behaviors of an urban-only population (which typically registers higher condom use) and the 2003 population poll included both urban and rural populations. 20 percent) have been tested for HIV vs. 36% of those not participating. (UNB/NASCP). Trends in CSW behavior over time are not available. Use of condoms among MSM in 2001 is reported to be high at 81 percent with regular partners and even higher at 95 percent with casual partners (Projeto Bela Vista). No baseline data are available to indicate levels of condom use among MSM at the start of AIDS I or AIDS II thus making it difficult to assesstrends over time. Testing history among MSM is reported to be significant, with 73% of urban men reporting having been tested in 2001 (Projeto Bela Vista). 4.30 The project was unsuccessful in strengthening evaluation capacity. As detailed below, project efforts to develop HIV and behavioral surveillance and to strengthen monitoring and evaluation did not culminate in the design and implementation of adequate systems for tracking program performance and impact. 4.3 1 Epidemiological surveillance. The project financed training, technical assistance and supervision to improve routine surveillance of AIDS and ST1 cases,including the refinement of case definitions both for AIDS (over and under 13 years of age) and for congenital syphilis and the upgrading of SINAN, the national AIDS case report database managed by the National Epidemiological Center. Project support to improve HIV surveillance included technical support and supervision of the sentinel surveillance sites, which monitor HIV prevalence in health services (maternity clinics, emergency rooms and STD clinics). AIDS II also financed the design and application of an instrument and method, and its incorporation into the SINAN, to guide the surveillance of HIV+ women and exposed newborns. In addition the project supported training and technical assistance in epidemiological surveillance aimed at professionals responsible for HIV and STD surveillance and staff working in CTAs. A number of studies were also financed under the project to supplement data and information on HIV, STD and Hepatitis B prevalence, and on behaviors covering different populations. 4.32 AIDS surveillance is well established, but it is HIV surveillance and incidence that are important in measuring and tracking the evolution of this epidemic and in measuring project impact. While Brazil's HIV surveillance system was strengthened under this project, it still falls short of a fully functional system that is able to generate a continuous stream of reliable and representative data that would permit the monitoring of the epidemic as it evolves in different areasof the country and among different population groups. During the three-year period 1997-l 999 sentinel surveillance data was reported twice a year from the three types of sentinel sites: maternities, emergency rooms and STD clinics. The numbers of sites reporting has fluctuated with each reporting period, making the data difficult to interpret and also making it difficult to determine the number of fully functional sentinel sites. In October 1999 reports were received from 39 STD clinics, 64 maternities and 46 emergency rooms. HIV testing was done on randomized patients of these clinics producing statistics per semester on around 7,000 STD patients, 12,000 pregnant women and 8,600 emergency room patients nationwide. In 2000 Brazil introduced significant changesto its surveillance system because it was perceived that hospital surveillance sites had become reference centers for HIV/AIDS and thus were providing biased estimates. The Brazilian surveillance system now selects a random sample of 150 hospital maternities which have over 500 deliveries per year to test women delivering in that hospital for a more representative measurement 21 of HIV prevalence. A different random sample of hospitals, that may include some of the same hospitals, is used for each measurement. Only two measurements have been taken since this change: one in 2000 with data from 128 hospital maternities (15,426 pregnant women tested) and from 25 STD clinics (4,636 STD patients tested) and one in 2003, for which data is not yet available. It was subsequently decided by the program to drop the STD clinic sentinel sites as the quality of the data from these sites was considered too unreliable. The Brazilians have opted for less frequent surveillance measurements becauseit is considered too expensive to conduct every year. Only half of project funds allocated to surveillance under AIDS II were used to this end. 4.33 At the end of AIDS II HIV surveillance is still limited to pregnant women and to military conscripts (the latter through annual surveys).24Data on pregnant women for the period 1997 through 1999 are difficult to interpret as the numbers of sites reporting fluctuated from one measurement exercise to the next. Methodology for monitoring HIV among pregnant women has changed in 2000 making it difficult to derive trends from the data stream from 1997-99, and measurement has not been sufficiently frequent since 2000. No routine surveillance of pregnant women was undertaken during 2001 and 2002 and the surveillance data on pregnant women for 2003 was not yet available at the writing of this report. Table F-5 in Annex F presents the summary report on Brazil's HIV surveillance efforts from 1997 through 2000. High priority, high-risk groups that are a primary focus of prevention efforts (MSM, IDUs, CSW) are not systematically included in routine HIV sentinel surveillance efforts. Special studies on HIV among these groups undertaken with project support have been carried out to supplement sentinel system data, but they lack a systematic methodology that would have allowed the tracking of trends over time or cross-regional analysis. With CDC support the NASCP initiated in 1998 efforts to strengthen its capacity to estimate the incidence of HIV based on cross-section studies of prevalence.25 4.34 Despite project support, the issues of STD surveillance, identified under AIDS I and II, persist: STDs are considerably underreported, data come primarily from clinics and represent only those who seek services rather than the general population or specific target groups. The syndromic approach to STD treatment26also undermines the accuracy of reporting. The uncertain validity of STD reporting in Brazil makes this a weak indicator for behavioral trends and a highly uncertain proxy for HIV trends. 4.35 Behavior Surveillance. With the exception of annual surveys of military conscripts, there is no surveillance system in Brazil for routinely collecting data on behaviors of segments of the general population or of high risk groups. AIDS II did support the establishment of baseline data for 1998 on sexual knowledge and behavior (CEBRAP 2000). However, this survey was not repeated at the close of AIDS II so as to 24. These data are now being supplemented by an information system established to support routing reporting of epidemiological and behavioral data on CTA clients by the network of CTAs. 25. This support involves the application of the Serologic Testing Algorithm for Recent HIV Seroconversion (STARHS) strategy, developed by the CDC, which allows for the determination of wh,ether people with newly reported cases of HIV were infected within the past year. 26. The syndromic approach to STD patient management bases diagnosis on a group of symptoms and treats for all diseases that could cause that syndrome. This approach is recommended by WHO in developing country settings as it provides for more accurate diagnosis without extensive lab tests and allows treatment with a single visit. 22 document trends in knowledge and behavior. NASCP has indicated to the team that this survey will be repeated in 2004 both to update data and to provide a baseline for AIDS III. In the absenceof a viable system for evaluating program performance and impact, the numerous studies and supervisions carried out by a multiplicity of actors, covering a range of methodologies, indicators, geographical areas,populations, implementers and timeframes. The results of these studies are virtually impossible to consolidate in order to provide a coherent overview of program performance or impact on behaviors. 4.36 Research. Building on support initiated under the first loan, AIDS II financed clinical and technological research through a competitive process. Research proposals were widely solicited from a range of qualified public and private institutions and vetted and approved by a National Research Committee. Of the 31 approved research proposals, only two focused on STD prevalence and only 1 was on health economics simply becausevery few were submitted. Regional breakdown of approved proposals is as follows: 45 percent for Southeast, 26 percent for South, 13 percent for Northeast, 10 percent for Central-West and 1 percent for North. There is no information available as of this writing on the full list of research financed under AIDS II, the main results of this research and whether/how research findings were applied for improved program performance and effectiveness. 4.37 A4onitoring and Evaluation. For the most part the project financed the activities envisaged in this subcomponent design, including the supervision of activities of states and municipalities and of NGOs, the undertaking of myriad specific studies and evaluations targeted at specific activities, executing agencies and target groups, and annual reporting on project performance. However, a number of important omissions obstructed the goal of systematic monitoring and evaluation of project performance and impact. First, the consolidation of indicators and preparation of a M&E plan were not carried out until the very end of the project. Second, over and above the lack of baseline data and data on trends in HIV and behavior noted above, there is still incomplete data (at national and local levels) on the sizes of target populations for program interventions and on trends in their effective coverage. Baseline studies were envisaged under the project but not carried out. 4.38 Preparation of state and municipal HIV/AIDS action plans (POAs) did encourage local HIV/AIDS coordination teams to establish some baseline data and to set annual targets as a part of their planning process, but these plans were not adequately monitored or evaluated for a number of reasons,most notably the lack of M&E capacity both in the NASCP and within state and municipal teams and the input orientation and administrative nature and treatment of theseplans. After a number of failed attempts a plan for the design and implementation of an M&E system was initiated at the very end of AIDS II in the context of preparing AIDS III, with technical and financial support from the Bank, CDC and other partners. Similarly, the establishment of a M&E unit in the NASCP, the recruitment of M&E staff to carry out these functions at the national and state levels, the provision of specialized technical assistanceand the design of basic and in-service training all were recently initiated with CDC support and in the context of preparing AIDS III. Up to that point, most staff in decentralized levels of the program did not have the skills to carry out effective M&E and none had the time. 23 4.39 Interviews with state and municipal coordination personnel and with NGOs indicated a very strong demand for better monitoring and evaluation that would accommodate their expresseddesire to know and improve the effectiveness and impact of their efforts. Supervision reporting by the Bank's team during the last few months of project implementation also notes the strong emphasis the incoming (in 2003) government places on demonstrating results, which provides incentive to the central unit to establish good M&E and to move toward a culture of using data for decision-making. 4.40 The project was successful in improving the sustainability of the program through devolution of some responsibilities to states and municipalities. It did so through the continued training, technical support and supervision of state and municipal coordination units to improve their capacity and successin designing and implementing their local-level action plans. Approximately 27,500 multidisciplinary professionals in the fields of health and education benefited from HIV/AIDS training delivered through a network of universities, professional training centers, NGOs and other educational institutions. Training encompassed a spectrum of skills building in the design and delivery of prevention and treatment activities and in program management. Fund-to- fund transfer mechanisms, described above, both ensure continued financing of state and municipality HIV/AIDS programs and build government interest and support through co- financing arrangements. The process of NGO subproject bidding and selection was also gradually decentralized to eight states under AIDS II, along with the responsibility to supervise such projects. Ten percent of the financing transferred through the fund-to- fund mechanism will be earmarked for NGOs. These actions are likely to be sustainable as they are coherent with decentralization policy, but will need to be further supported with capacity building efforts to ensurelocal level successin carrying out newly decentralized responsibilities. Interviews with state and municipal actors (both private and public) indicate that their AIDS mitigation efforts have been increasingly integrated into the government's overall poverty reduction efforts, especially at decentralized levels, but that concerted efforts are still required to improve interface of activities and priorities with local social assistanceprograms. 4.41 The project did not achieve the objective to identify the most cost-effective interventions. No researchwas undertaken to evaluate the cost-effectiveness of prevention activities. Some analysis of the costs and outcomes of Brazil's treatment and care program was undertaken and reveal substantial increases, as have been observed in developed country settings, in survival after AIDS diagnoses, and decreasesin mortality, morbidity, hospitalizations, opportunistic infection rates, and marked improvements in quality of life indicators. 5. Ratings 51. Outcome. Both AIDS I and AIDS II arejudged to have had satisfactory outcomes overall, as substantiated in the following paragraphs on their relevance, efficacy and effectiveness. 52 Relevance. The relevance of both AIDS I and AIDS II is rated substantial. The last full CAS of March 6, 2000 supports the Government's poverty alleviation objective 24 and includes health services for the poor as a key targeted intervention to this end. While not mentioned specifically in earlier CAS exercises, the fight against HIV/AIDS is given increased prominence in the most recent (May 20, 2002) CAS update. It is presented as one of the seven Millennium Development Goals (MDGs), vigorously embraced by both the Government of Brazil and the Bank, and consistent with their sharedpoverty alleviation agenda. As such, targets for the prevention and control of HIV/AIDS, aswell as for the other MDGs, are specified in the new CAS approved in December 2003. While the projects made considerable effort to extend their coverage of prevention and care services to emerging priority populations and to all regions, efforts to target women remain modest and the North and Northeast regions received less per capita support relative to other regions (see Tables F-7 and F-8 in Annex F). 53 Efficacy. Drawing on the assessmentof project performance by objective presented in Section 4, the overall efficacy of each of the two projects is rated as substantial. They have succeededin improving and further expanding prevention activities, while maintaining a focus on high-risk groups with state-of-the-art prevention interventions. Both projects have culminated in the considerable extension of a network of voluntary counseling and testing services and of treatment and care services (with a substantial shift to low-cost outpatient services) and in the development of crucial laboratory infrastructure needed to support the treatment program. 54 Research capacity was both utilized and strengthened with project assistance,but the research financed did not respond to project objectives to document the cost- effectiveness and impact of various interventions. While HIV surveillance of pregnant women improved with the establishment of surveillance sites and related training and support, data is lacking on HIV trends in high risk groups targeted under the project. Behavior trends among high risk groups and among the general population are not known as multiple studies using multiple methodologies, timeframes, and indicators are not amenable to comparison or documentation of trends. Project emphasis on monitoring and evaluation failed. Finally, establishment of coordination units at state and municipality levels, and of a system of transfers of financial support against negotiated annual plans have made significant headway in building capacity and in obtaining the sustainability of the program. 55 Efficiency. The efficiency of AIDS I and AIDS II was substantial. Project design supported government provision of public goods (epidemiological and behavioral surveillance, research, monitoring and evaluation), behavior change interventions targeted at high-risk and vulnerable groups, and early action where the epidemic is not widespread to avert an epidemic, consistent with efficiency in both epidemiological and public economics terms (World Bank 1997). Both projects also supported improved accessof patients to cost-effective health care through the development of low-cost, alternative services for inpatients and outpatients, and the full integration of such services into the SUS. The project might have been more forthcoming in extending coverage of prevention activities to high risk groups in low-prevalence geographical regions to reflect the good practice of early intervention to maintain low prevalence. The institutional arrangements for the projects (and the national program) were efficient overall as they drew significantly on the comparative advantages of the multiple partners in the fight against HIV/AIDS. The complementarity of roles and contributions of the 25 public and non-public sectors, respectively, is very strong. The devolution of program responsibilities to state and municipal levels is still in process, but taking a direction that will make for more efficient use of comparative advantagesof the various levels of the program, in sync with overall decentralization efforts. In this light there is scope for scaling back the size of NASCP and strengthening the staffing of local units. 56 Institutional Development Impact. Institutional development impact of both projects is rated substantial. The projects supported improvements to the technical and managerial capacity of the NASCP to further enhance its leadership, advocacy and coordination efforts as well as its continued efforts to develop and implement state-of-the art interventions for prevention, treatment and care, and to organize extensive capacity building efforts. The design and conditionality of AIDS I contributed significantly to the development of state and municipal capacity to prepare, negtotiate and implement local level operational plans.. AIDS II strengthened, consolidated and extended this capacity. The successful experience under both projects is being institutionalized through the new law, drafted by the NASCP and signed by the Minister of Health in December 2002, establishing a fund-to-fund transfer mechanism for qualifying states and municipalities that are affected by the epidemic. 57 Overall the projects have made a significant contribution to Brazil's ability to effectively mobilize and utilize human and financial resourcesto achieve national objectives for combating STDs and HIV/AIDS. The financing of technical assistance and of specialized contractual staff under both projects supplemented NASCP capacity thus enabling it to mount and coordinate an ambitious program. Both projects stimulated and nurtured public-private partnerships by making financing available for NASCP contracts with NGOs and CBOs to carry out critical components of the National Program, in line with their comparative advantages. This arrangement has been institutionalized and reinforced through capacity building of NGOs/CBOs in proposal-writing and contracting provided by NASCP and financed under both projects; and it is being decentralized increasingly to states and municipalities. Efforts are ongoing to consolidate staffing of NASCP and to strengthen the staffing of state and municipality coordination units commensurate with the increased decentralization of program responsibilities and with the new mechanism for fund-to-fund transfers for HIV/AIDS. 58 Sustainability. Both AIDS I and AIDS II are likely to be sustainable. Prospects for financial sustainability of project investments have improved considerably since the design of AIDS I, due to (a) considerable reductions in the costs of care and treatment due to (i) the establishment of lower-cost alternative care for inpatients and outpatients and (ii) efforts on the part of the Government to lower the costs of ARV; and (b) the significant investment in prevention efforts that may have slowed the rate of infections and thus of the pool of those requiring care and treatment. These achievements do not, however, deny the need for continued efforts: to contain costs, to strive for efficiency in allocation and implementation and to ensure adequatefinancing. Devolution of program responsibilities to states and municipalities will underpin sustainability, both from an institutional perspective and from a financial (cost-sharing) perspective. With the important exception of monitoring and evaluation, technical and economic resilience of the program is strong and are likely to continue given the strong capacity and leadership of the program. The program has enjoyed strong Government and popular support for 26 some 10 years and has thus been supported and its budget protected in the face of economic shocks and changesin leadership (including Ministers of Health and Head of State). Strong political constituencies and the Constitutional mandate for universal accessto ART provide strong assurancesfor the continued sustainability of the treatment program in particular. Many if not most NGOs working on HIV/AIDS in Brazil have relied heavily on transfers from the NASCP for funding under the project. There will be continued need for NGOs' critical contributions to HIV/AIDS control and thus for continued public financing of NGO contracts. 59 Bank Performance. With regard to the design of both projects Bank performance was satisfactory. The quality at entry of each project was adequate. Project designs were technically and economically sound and strategic in their focus. First they were designed to support the Government in ensuring that public good elements of the program continue to be accorded top priority (targeted prevention activities, research, surveillance, monitoring and evaluation) and in supporting care and treatment through studies and investments to contain costs of services and improve their quality and effectiveness. Second, Bank financing was conceived to stimulate innovation and reform, particularly: (i) the use of public monies to finance contracts with civil society to implement HIV/AIDS activities and (ii) the establishment of state and municipal coordination units and the transfer of financing to these entities to implement work program agreementsprepared by local-level actors. Preparation was undertaken in a context of strong Government ownership and elicited the inputs and perspectives of multiple stakeholders. The importance of baseline data and monitoring and evaluation were emphasized in both designs, but were not sufficiently pursued during implementation. More upstream support and guidance might have been given to the design of an M&E system and plan, and surveys should have been undertaken to establish epidemiological and behavioral baselines during preparation/launch stagesto enhance readiness for implementation and to set the stage for the setting and monitoring of targets. 5.10 The Bank's performance during implementation of AIDS I was mixed, but unsatisfactory. Only four full supervision missions were carried out during the four- year implementation period,z7half of the Bank's norm of two supervisions per year; and supervision undertaken was not sufficiently rigorous. None of the three weaknessesof project performance identified in this assessment(inadequate M&E, weak surveillance system, and lack of cost-effectiveness analysis of activities to inform future strategic and resource allocation decisions) were adequately addressed. Bank reports and follow-up communication with Government on the project launch reveal a strong focus on disbursement and procurement and do not addressthe importance of establishing baseline data for M&E. It was only in 1997, a year before the project's closing, that Bank management first mentioned in a routine follow-up letter to the Government the need to undertake evaluation activities during the next (remaining) 12 months. The mid-term review (MTR) was not fully exploited as an opportunity for renewed focus on the 27. While Bank data on AIDS I reports 9 supervision missions, a review of project status reports crosschecked with documents contained in project files documents only four full supervision missions taking place, respectively, on: 6/l 7/94, 6/l 6/95, 9126'96 and 5/21/97. Remaining missions were not full supervision missions. Rather they were small, often one-person, missions in Brazil on other business but mandated to spend a few days on very selective oversight of project performance, in line with their respective specialties (e.g., IEC, procurement). 27 project's original development objectives and to addressand correct weak performance on surveillance, monitoring and evaluation and cost-effectiveness analysis. The MTR report recommended that the Government consolidate indicators and prepare an evaluation plan (a recommendation also found in earlier supervision reports). Despite these shortcomings, supervision missions did provide technical oversight and sound advice in maintaining an emphasis on prevention activities, especially for high-risk groups, and in refining the content and coverage of prevention and treatment/care interventions. 5. I I The Bank 3 performance during supervision of AIDS II was also mixed, but satisfactory overall. During the four-year, four-month life of the project, seven supervision missions were undertaken for an average of one every seven months, in line with Bank coefficients. Supervisions for the most part were candid, proactive and specific in identifying issues and in recommending their resolution. Among other recommendations (on technical quality and coverage/equity issues), persistent weaknesses in surveillance and in monitoring and evaluation were raised and specific direction was provided during every supervision, both in Aide-memoires and in internal reporting. The MTR placed specific emphasis and provided operational advice on identifying the most cost-effective prevention interventions, strengthening M&E capacity, and strengthening epidemiological surveillance. However, baseline data were never established nor was program coverage documented or monitored. A breakthrough was made on the push for monitoring and evaluation during the last year of the project, and is attributable in great part to the leverage of AIDS III preparation. Staffing of missions included consultant epidemiologists who prepared technical reports advising on the nature of improvements needed,but there was insufficient technical advice and rigor in establishing a core program of HIV and behavior surveillance of the general population and of high-risk groups necessaryfor documenting trends and for tracking program performance and impact. Despite unsatisfactory performance on monitoring and evaluation, this element of the.project was rated as satisfactory throughout implementation, aswas achievement of development objectives. Had management rated the project components unsatisfactory where warranted, additional effort might have been leveraged to rectify weak aspectsof project performance. 5.12 Borrower Performatzce. The performance of the Borrower under AIDS I was satisfactory overall. During the design of AIDS I, Borrower commitment was very strong and it assumed its responsibility in undertaking a sound technical preparation and in ensuring the participation of a range of key stakeholders, encompassing key federal- level agencies, representatives from states and municipalities and from the NGO community. The Borrower's performance during implementation was also satisfactory. Counterpart obligations were met. Program/project management was well staffed and very efficient as evidenced by the following: (a) the project was executed within the original timeframe, (b) financial management was consistently satisfactory; and (c) the loan was fully disbursed. Contracts with NGOs and transfers to states and municipalities were managed and implemented in a satisfactory manner, although supervision of these executing agencies might have been stronger. Most quantitative project implementation targets were met or exceeded (number and targeting of prevention projects, number of NGO-executed projects, numbers of new services established/strengthened). The Borrower's performance fell short of expectations, however, with regard to its failure to 28 evaluate program/project impact, as well as the cost-effectiveness of prevention interventions. 5.13 Borrower performance under AIDS II was mixed, but satisfactory overall. Building on the experience of AIDS I, project preparation was technically sound overall and participatory. However, while many relevant lessons of experience under AIDS I were incorporated into the design of AIDS II, Borrower analysis was insufficient with regard to the above-mentioned failures (monitoring and evaluation, cost-effectiveness analysis, and surveillance) and resolve to addressand correct them. The Borrower's performance during AIDS II implementation was satisfactory overall. Counterpart obligations were met. Project management, NGO contracting, and transfers to statesand municipalities continued to be well executed by the Borrower. Most quantitative project implementation targets were met or exceededon all three project components. Unfortunately, the same failures of AIDS I persisted under AIDS II: weak monitoring and evaluation; weak epidemiological surveillance and no systematic behavioral surveillance; and virtually no cost-effectiveness analysis for informing priority-setting and resource allocation. With the strong capacity in Brazil for research and survey work in a multiplicity of disciplines (economic, epidemiology and sociology), these failures are likely attributable to the lower priority accorded these activities by the Borrower relative to prevention, treatment and care interventions. 6. Findings and Lessons 61 Failure to establish critical baseline data on HIV infection rates and on behaviors among high-risk and general population groups, and failure to undertake systematic surveillance to document trends in HIVand behaviors in both of these groups make it dif'jcult to track the evolution of the epidemic over time and to evaluate the effectiveness and impact of program interventions. While Brazil has implemented many innovative and theoretically cost-effective interventions, evidence of the actual costs, effectiveness and impact of its efforts is lacking. AIDS I and II have had limited successin strengthening Brazil's surveillance of HIV infection and of behaviors among priority populations. This has undermined Brazil's ability to evaluate the impact of its prevention efforts on behaviors and on the epidemiology of the disease. In the absenceof fully functional systems for routine monitoring of epidemiological and behavioral trends, Brazil has been forced to rely on indirect tools (ad hoc research and studies, often covering different populations, timeframes and geographical regions) to gain an appreciation of program performance and impact. 62 Inadequate data on the size and coverage of target populations for the prevention and treatment programs, respectively, especially at the state and local levels, undermine the efficiency and effectiveness of HIV/AIDS interventions. The size and coverage of high-risk groups for prevention are not adequately documented in Brazil making it difficult to establish viable targets or to estimate the'level of effort and resourcesneeded for effective intervention. While there are national estimates of the size of MSM and IDU populations, these are rudimentary at best and not meaningful at state and local levels. Baseline surveys were planned at the projects' outset, but never carried 29 out. By the same token, the effective coverage and accessof the treatment program (as a proportion of ARV-eligible persons) is not known. Baseline surveys, promotion of VCT services, improved outreach and integration of HIV/AIDS interventions with other health services are all critical in this regard. 63 Early, rigorous and sustained prevention efforts are likely to keep the pool of I&V and AIDS patients small relative to the counterfactual, and thus to reduce the burden on a country's health system and make treatment and care more feasible and affordable. It is plausible from available (albeit limited) data that prevention activities undertaken over the past 15 years in Brazil may have contributed to containing HIV infection rates and the number of new AIDS cases,thus averting a scenario of more rapid spread and the transition to a generalized epidemic. Over and above the avoidance of the devastating consequencesof HIV/AIDS on individuals, families, communities and on the country's development prospects overall, the number of HIV infections and AIDS cases averted (as reflected in declining incidence) has reduced significantly the costs and challenges of providing adequatequality and coverage of treatment and care. Simply put, the continued success and affordability of Brazil's treatment and care program depends critically on the success of its prevention efforts. 64 NGOs have a critical role to play in thefight against HIV/AIDS. In Brazil NGOs have proven highly effective in reaching stigmatized and marginalized groups making available to them both prevention and care services, for which they have a comparative advantage vis-a-vis public services. Their target populations constitute the three highest-risk groups that constitute the top priority for the prevention program: MSM, IDUs and CSW. They have also demonstrated their effectiveness in serving as an interface between vulnerable groups, such as the poor and women, and the public services with a view to improving accessand utilization. They also continue advocacy work. 65 Even when financing a relatively modest share of the overall costs of a well- l establishedprogram, Bank support to HIV/AIDS efforts can make significant contributions to program expansion and capacity building when it is strategically . allocated. While financial support provided under AIDS I and AIDS II amounted, respectively, to only about 30 percent and 10 percent of national expenditure on STDs and HIV/AIDS during their respective implementation periods, such support was channeled to the most cost-effective components, as indicated by evidence amassedon international experience evaluated thus far (World Bank 1997). AIDS I and II devoted significant financial and technical support to the design and delivery of prevention activities, including innovative, state-of-the-art interventions aimed at changing the behavior of high-risk groups, aswell asto the expansion and promotion of counseling and testing services. Bank support to improvements and expansion of treatment and care was selective and strategic, focusing on the establishment of lower-cost alternatives, including an increasing shift to outpatient care, and on improving service quality through capacity building of service providers and of the laboratory network. These inputs combined with Brazil's achievements in significantly lowering the cost of ARVs have lowered the cost of care and treatment and improved the quality of these services. 30 66 Bank support can bring legitimacy and leverage to controversial activities and to stigmatized segments of society. The strategic orientation of Bank's assistance:(a) has provided leverage in mobilizing additional national resources for HIV/AIDS activities at all levels of government, especially for controversial prevention interventions such as needle exchange programs; and (b) has been instrumental in securing respect and legitimacy for marginalized or stigmatized groups who were both implementers and beneficiaries of Bank support (e.g., commercial sex workers, prisoners, transgender persons and MSM, IDUs, people living with HIV/AIDS and their advocates). 67 A research program that is ad hoc and uncoordinated in its conception is likely to generate a large volume of information and data that will be difficult to interpret and usefor measuring and improving program performance and impact. While the researchprogram in Brazil has produced considerable information and data, it has not been possible to derive trends in the epidemic and in behaviors or insights on other aspects of program performance. Brazil's HIV/AIDS program would have benefited from a master plan for research, conceived to support its strategic objectives, both (a) in the choice and prioritization of topics for studies and operational research, and (b) in the standardization of methodology that would have accommodated the analysis of data across studies and of trends over time. 68 Project design and conditionality can be strategic in initiating and supporting institutional innovations which are capable of expanding and using more fully program implementation capacity, most notably: (a) use of public monies to further stimulate and mainstream the contracting of NGOs to work on prevention and care/treatment, in line with their comparative advantages;(b) the creation of state and municipality level HIV/AIDS and STD Coordination Units, the funding of locally developed work program proposals that would be the subject of formal agreements (contracts) between respective Coordination Units and the NASCP, and the co-financing of overhead costs and inputs by local governments; and (c) the funding of multisectoral responsesfor prevention and human rights promotion with Ministries of Education, Justice, Interior, among others. These innovations (described in paragraph 2.3-2.6)have served to solidify and expand partnerships with civil society and with other development sectors and local government and are being consolidated and institutionalized. 7. Future Directions 71 Over the past five years OED has distilled the experience of the World Bank in Biazil through a number of reviews that are relevant to the fight against HIV/AIDS.28 Taken together, these reviews point to a number of challenges for poverty alleviation and health sector development in Brazil over the coming years. They highlight the importance of: improving M&E to measure performance and impact and thus to improve effectiveness and outcome; reaching poor marginalized groups and regions (especially the North and Northeast); and sustaining and further improving public-private 28. These reviews include most notably: the Country Assistance Evaluation: Brazil of May 15, 2003, The Brazil Health System: Impact Evaluation Report of June 30, 1998, Brazil Maternal and Child Health of February 26, 2002, and NGOs in World Bank-Supported Proiects in Brazil: A Background Paper for OED's Study of NGOs in World Bank- Supported Projects of July 29, 1999. 31 partnerships. The decade of experience of Bank support to HIV/AIDS in Brazil resonates with themes and recommendations raised in the previous OED reports. With the imminent launch of a third Bank-financed loan for HIV/AIDS in Brazil, whose design has incorporated many of the lessons of the past decade,the following suggestions for further improving the development effectiveness of HIV/AIDS efforts in Brazil are offered. 72* First andforemost, Brazil should take the necessary steps to improve its capacity and incentives to monitor and evaluate program performance and impact and to document the cost-effectiveness of program efforts. * First, in line with UNAIDS guidelines (UNAIDS 2000 and 2002), Brazil should continue refining and expanding its surveillance efforts to enable the routine and systematic monitoring and analysis of trends in HIV infection as well as trends in behaviors among key populations. Second generation surveillance systems are designed to permit the analysis of behaviors both to anticipate and to explain trends in HIV infections. Given the current status of the epidemic in Brazil (still concentrated among populations with high risk behaviors, with heterosexual transmission being the primary and growing mode of transmission), the surveillance of HIV infection and of behaviors should focus on high-risk groups with more than 5 percent infection rates (MSM, IDU, CSW) and should also monitor trends in the general population, paying particular attention to links between thesetwo groups, particularly through bridge populations (female partners of IDUs and bisexuals and clients of CSW), and to consequent risks of rapid spread of the virus in the general population. HIV and behavioral surveillance of high-risk groups should be carried out on a regular basis and may be most efficiently executed at the decentralized level with a view to facilitating more in-depth assessmentof local-level issues and trends and improved responses. AIDS III is supporting Brazil in its efforts to exploit fully the voluminous data currently available through ad hoc studies and surveys and through the health information system, but this will not eliminate current and future needs for systematically collected data on HIV and on behavior of key populations. NASCP's plans to carry out a national sexual and risk behavior survey in 2004 to update the 1998 baseline survey (CEBRAP 2000) is an important step in this regard. Second, the time is ripe to design a monitoring and evaluation system and strategy that is decentralized and results-based: (a) the recently adopted fund-to-fund transfer mechanism for supporting HIV/AIDS in states and municipalities is based on the principle of results-based planning and management; (b) as noted earlier, local level public sector staff and NGOs have expresseda strong demand for good and meaningful M&E indicators, guidelines and training to enable them to track and improve their performance and impact; and (c) the new government is very committed to the design and implementation of a results-based approach to development programs. Such a system should establish baseline data on the size and current program coverage of target populations at the state and local levels to permit realistic target setting and effective monitoring and evaluation of program performance, as well as to assessthe level of effort and resources needed for a successful program. The monitoring and evaluation system should also accommodate the need for systematic evaluation of interventions implemented by NGOs against 32 program targets set at the national, state and local levels. Government comments (provided in Annex G) highlight recent achievements and ongoing efforts to redefine the M&E strategy for HIV/AIDS in favor of a more decentralized and results-based orientation, and note the continued support of CDC and the World Bank to this end. Of particular note are (i) the recent establishment of five sites of excellence in each of Brazil's five regions to foster and promote M&E activities at the decentralized level; and (ii) the development of postgraduate training in M&E of STD/HIV/AIDS designed to foster decentralization and institutionalization of M&E activities. 0 Third, the government should commission studies to document systematically the cost-effectiveness of program components and pilot interventions through (a) the design and active direction of a researchprogram that emphasizes cost-effectiveness and impact analysis; and (b) the solicitation and exploitation of considerable research capacity available in Brazil in the disciplines of economics, epidemiology (especially HIV incidence research) and sociology. Studies should prioritize the evaluation of the cost-effectiveness of current interventions aimed at MSM, IDUs and CSW. Such a program should be iterative in the sensethat it should both influence and monitor a continued fine-tuning of interventions and approachesthat will enhance cost- effectiveness and improve the efficiency of resource allocation. Cost-effectiveness analysis will also be useful to other countries looking to build their HIV/AIDS programs on the experience of Brazil's long-established and innovative programs. * Finally, rigorous monitoring and evaluation of Brazil's treatment program is critical, both becauseit will allow Brazil to track and improve the effectiveness and impact of its treatment and care program, and becausethe Brazilian experience, well documented, will provide guidance and insight to many other countries seeking to establish or strengthen national treatment programs. Monitoring and evaluation of the treatment program should encompass: behavior and compliance of those under treatment and their long-term effects on transmission, drug resistance, effective coverage, and equity in accessto services and in adherence. 73. In the meantime, in order to further consolidate and sustain Brazil's fight against HIV/AIDS and to achieve maximum impact, NASCP leadership must mobilize human andfinancial resources around the highestpriorities of the program, as evidenced by international experience to date. Government support to prevention should remain strong, ensuring that it is accorded highest priority. This will remain a challenge, especially as Brazil's treatment policy is mandated by the constitution and has a strong public constituency, whereas the many positive externalities of prevention are typically less appreciated by the public. Prevention efforts must continue to target behavior change in high-risk populations (MSM, IDUs, CSW) and expand efforts to better target emerging vulnerable groups (such as women, the poor, rural/interior populations). In addition, efforts to improve and expand surveillance, diagnosis and treatment of STDs, including enhancedoutreach and targeting, must be renewed and intensified as this package of activities constitutes a critical element of an effective prevention strategy. Sustained efforts are also neededto monitor and contain (if not reduce) the costs of care and treatment, as well as to ensure adequatefinancing of such services, all in the face of changing costs and technologies. By their very nature, public goods provision should be a priority for Government support, most particularly: 33 surveillance, research,monitoring and evaluation, capacity building and institutional development, and sustained support/development of the laboratory network. 74. In light of the trends of the epidemic, which indicate a spread to bridge populations, many of whom include women, poor populations, and rural/interior populations, equity will become an even more important objective of the national HIV/AIDS program as it attempts to reach these groups. This being said, there is also a need to continue and intensify targeted behavior change interventions aimed at highest- risk groups (MSM, IDUs and CSW). Both prevention and care and treatment interventions were relatively easy to mount when high-risk and affected populations were concentrated for the most part in and around the South and Southeast Regions and in urban areas. Both NGOs specialized in reaching high-risk target groups (MSM, IDUs, CSW) and health services responsible for monitoring and treating those affected by HIV and AIDS were more readily available and generally capable of providing needed services. This being said, only coverage of MSM is satisfactory at 96 percent. Considerable effort needs to be made to increase the coverage of IDUs (roughly estimated at 18 percent) and of CSW, whose effective coverage by the program is not known (Ministry of Health/NASCP draft final evaluation). Equity has been an important principle of the program from the outset and will become an increasingly challenging goal in the future. The emerging vulnerable groups are considerably harder to reach by nature of their locations and their status in society; and they do not typically seek out social support and services, even where they are available. More effort needs to be made to improve coverage of HIV/AIDS interventions especially in the Northeast (seeTables F-7 and F-8, Annex F). Furthermore, women and the poor should not be addressed, respectively, ashomogeneous groups given that subgroups within each of these categories are exposed to a spectrum of risks according to the multiplicity of situations and vulnerabilities of these subgroups. Operational researchand concerted efforts to effectively reach these groups will be a major challenge of the program. 75. Given the maturity of Brazil's HIV/AIDS program and its emerging challenges, consolidation and decentralization of the institutional framework for the figkt against HIV/AIDS and STDs should bepursued with a view to strengthening the capacity of state and municipality HIV/AIDS units to assume considerably increased responsibilities for strategic program management, oversight and implementation, as well asfor managing partnerships with civil society and local level government. Such activity might be underpinned by the undertaking of an assessmentof the overall institutional and organizational framework for the fight against HIV/AIDS (at all levels of the system) in light of decentralization and other reforms and in light of its emerging priorities. Such an assessmentwould be complemented by state/municipal-level assessmentsof opportunities for improved public-private partnerships. On the basis of these assessments,fine-tuning of the organization, mandates and staffing of different levels of the program, and of mechanisms for partnerships could render program efforts more efficient and cost-effective. Local level establishment of more operationally relevant baseline data on program coverage and performance is needed to facilitate both the setting of targets and the monitoring and evaluation of program performance. Instruments and processescan be improved to evolve more fully towards performance- based management. Pedagogical supervision and accountabilities need to be strengthened considerably. 34 76 Now, more than ever, efforts should be devoted to the integration of HIV/AIDS and STD control activities into the SUS, particularly special programs within the SUS, such asfamily health and reproductive health. It will be impossible to make significant headway in reaching the emerging target groups and vulnerable populations (women, the poor, rural/interior populations) of the program unless basic health services are more accessible to them and more fully utilized. This is indeed a formidable challenge, which will require improvements to the quality, coverage, equity and outreach of priority health programs and services. Improvements to HIV/AIDS work and to health system performance, especially primary health care and public health, can and should be simultaneous and synergistic. 77 NASCP should intens@ efforts to encourage afuller utilization of CTA l services now available in the country. Most people in Brazil who are infected with HIV are not aware of their seropositive status. An important contribution to prevention efforts is for people to be aware of their status and to take necessary action both to seek treatment and care and to prevent the transmission of infection to their partners. 35 References Berquo, E., R. Barbosa, and S. Kalckman, 1999. Aceitabilidade do Condom Feminino em Contextos Diversos. Relatiorio Final. CEBRAP, NEPO/UNICAMP e CN DSTlAids. Brindeiro RM, Dias RS, Sabion ED, et al.. Brazilian Network of HIV Drug Resistanse Surveillance (HIV -BresNet): A Survey of Chronically Infected Individuals. AIDS 2003: 17: l-7. CEBRAP, 2000. Comportamento Sexual da Popula@o Brasileira e Percepgoes do HIV/AIDS. Sao Paulo, Brasil. Chequer P., N. Hearst, ES. Hudes, E. Castilho, G. Rutherford, L. Loures, L. Rodrigues, and the Brazilian State AIDS Program Coordinators. "Determinants of Survival in Adult Brazilian AIDS Patients, 1982-1989." AIDS 1992; 6:483-487. Gwatkin, Davidson R., Shea Reutstein, Kiersten Johnson, Rphini P. Pande and Adam Wagstaff. 2000. "Socioeconomic Differences in Health, Nutrition and Population." World Bank. HNP Poverty Thematic Group, Washington, DC. Johnston, Timothy and Susan Stout. 1999. OED Report, Investing in Health: Development Effectiveness in the Health, Nutrition, and Population Sector. Report No. 21392. Marins JR, Jamal LF, Chen SY, Barros MB, Hudes ES, Barbosa AA, Chequer P, Teixeira PR, Hearst N. Dramatic improvement in survival among adult Brazilian AIDS patients. AIDS 2003, Jul25; 17(11):1675-82. Mesquita F., Kral A., Reingold A., Bueno R., Trigueiros D., Araujo P., Santos Metropolitan Region Collaborative Study Group. "Trends of HIV Infection among Injection Drug Users in Brazil in the 1990s: The Impact of Changes in Patterns of Drug Use." [pdf file 2,17OK] JAIDS Journal of Acquired Immune Deficiency Syndromes 28:298-302 (c) 2001 Lippincott Williams & Wilkins, Inc., Philadelphia Ministry of Health of Brazil. 2002. National AIDS Drug Policy, June.(MOH, Brasilia). Ministry of Health./NASCP, 2002. Relatorio de Implementa@io et Avaliagao-1998-2002. Accord0 de Emprestimo BIRD 4392/BR Projeto AIDS II. Brasilia, August 2002. Ministry of HealthNASCP, 2003. Relatorio de ImplementaQo e Avaliagao 1998 a 2003. Acordo de Emprestimo BIRD 4392/BR - Projeto AIDS II, Versa0 Preliminar, Brasilia- DF, August 2003. Szwarcwald CS. 2002. Oportunidades perdidas na detec@o precoce do HIV na gestagao: Resultados do Estudo Sentinela-Parturiente, Brasil, 2002. (Unpublished report, used with permission of the author.) UNAIDS. June 2000. National AIDS Progorammes: A Guide to Monitoring and Evaluation. Report No. UNAIDSOO. 17E. Geneva Switzerland. 36 UNAIDS. August 2002. Initiating Second Generation HIV Surveillance Systems: Practical Guidelines. Report No. UNAIDS/02.45E-WHOIHIV/2002.17. ISBN 92-9 173-2 19-2. World Bank. 1997. Confronting AIDS: Public Priorities in a Global Epidemic. The World Bank, Washington, D.C. Martha Ainsworth, Mead Over, Akhavan Dariush, Alexandre Abrantes, Renato Gusmao and Philip Musgrove. 1999. World Bank. 2002. Country Assistance Strategy Progress Report for the Federative Republic of Brazil. May 2002. Report 24182-BR. World Bank. 2000. Country Assistance Strategy for the Federative Republic of Brazil. Report 20160-BR World Bank. 2003. Project Appraisal Document: Brazil Third AIDS and STD Control Project. Report No. 25759-BR. World Bank. May 15, 2003. Country Assistance Evaluation: Brazil. Operations Evaluation Department. World Bank. 1998. Project Appraisal Document: Brazil Second AIDS and STD Control Project. Report No. 1833%BR. World Bank. 1998. Impact Evaluation Report. The Brazil Health System. OED Report No. 18142. 37 Appraisal Actual or Actual as % of estimate current estimate appraisal estimate Total project costs 250 250 100% Loan amount 160 159.8 100% Cofinancing Cancellation 0.2 w--m- ,~~~~~~-~~~~,~~~~-~--~-~~~-~~X-^-~ Cumulative Estimated and Actual Disbursements FY94 FY95 . FY96 FY97 FY98 FY99 Appraisal estimate (US$M) 17.7 53.2 88.7 124.3 160 Actual (US$M) 0 52.2 101.8 126.3 159.6 159.8 Actual as % of appraisal 98 115 102 100 100 Date of final disbursement: May 06, 1999 Proiect Dates -_..-._ Original Actual Board approval 11/09/1993 1 If0911 993 Signing 0311611994 03/I 6/l 994 Effectiveness 06101 /I 994 06/01/l 994 Closing date 06/30/l 998 06/30/l 998 FY92 FY93 FY94x FY95 FY96 FY97 FY98 Total Preappraisal 21.8 24.3 15.4 Appraisal & Negotiations 12.8 Supervision 11.3 7 10.3 15 7.2 50.8 Midterm review 6.5 6.5 Total 21.8 37.1 26.7 7 16.8 15 7.3 131.6 38 Mission Data (man th/year) persons field representeff Impl. Status Devt. Objs. Through appraisal Preparation 1992 4 21.18 D, A B, M Preparation 1992 3 18.9 0, B M Appraisal through Board approval Appraisal 1993 6 12.8 L B, D, M, M, M Post appraisal 1994 2 2.0 M, M Appraisal 1996 4 13.4 B, C, C, M Supervisions Supervision 1 1994 3 11.3 B, D, M 1 1 Supervision 2 1995 4 7.0 B, B, A K S HS Supervision 3 1996 4 6.5 B, M, M, N HS S Supervision 4 1996 1 1.2 F HS S Supervision 5 1996 1 2.0 G HS S Supervision 6 1996 1 0.6 F HS S Supervision 7 1997 3 0.8 H, B, K HS S Supervision 8 1997 3 0.8 H, B, K HS S Supervision 9 1998 3 7.2 B, H, K S S ~~~~~~~,~e)CU(~~~.~~~~~~~~~Y~~~~~~~~~~~~~~~~~~-~~~~~~ Mid-Term Review 1996 5 1,1,K, J, M a. 1 = Problem free; S = Satisfactory; HS = Highly satisfactory b. A = Public Health; B = Economist; C = Lawyer; D = Consultant; E = Operations Analyst; F - Procurement; G = IEC; H = Population Health Spl; I = Disease Control Spl; J = Project Spl; K = NGO Coordinator; L = Operations Officer, M = AIDSSTD Specialists; N = Disbursement 39 Total project costs 300 296.50 99% Loan amount 165 161.5 98% Cancellation 3.5 w-ii--- Cumulative ---`Y--x&X--P Estimated and Actual Disbursements FY99 FYOO FYO? FYOZ FYO3 FYo4 Appraisal estimate (US$M) 20.27 52.82 88.82 137.39 165.00 Actual (US$M) 20.0 55.6 88.5 137.2 154.0 161.5 Actual as % of appraisal 99 105 100 100 93 98 Date of final disbursement: August 26, 2003 Proiect Dates Original Actual Board approval 09/15/1998 0911511998 Signing 12/11/1998 12/l II1998 Effectiveness 02/09/l 999 02/09/l 999 Closing date 12/31/2002 0613012003 ,..,.v.,,,......e,..,ee*,.,.",d*....,..?,m...aAaGI-.--_u,,--_,,,.,__.",.~,.-.,,._..^.^,.__--,,,-_-~~.~~-+x~~,.-~*,,~.*,*,r,,_.-,,.Is,,,-,,e.-,,-.,,,x.,,-,._l,,,d^r...,.e..--*.I,I-I,---,.w*",-."-I-,~--~--.,,-x.l----.*---~-,~,~-----~,-~~-,.---~,~~-,,~,---.---~~~~,,~---,~---- Stage of Project Cycle Actual/Latest Estimate No. Staff weeks US$ (`000) Appraisal/Negotiation 135,266 Supervision 70.81 330,713 ICR 1.66 32,183 Total 73.47 498.162 40 Supervision O/6/26/1999 5 Task Team Leader (1); Cons. Diagnostic S S Treatment (1); Cons. Health Spec (1) Procurement Spec (1); Financial Specialist (1) 02/09/2000 7 Health Spec (1); S HIV AIDS Epidemiologist( 1); HIV AIDS Spec (1); Procurement Spl (1); NGO Spec (1); Financial Spec (1); Epidemiologist-SIDALAC (1) 09/24/2000 3 Public Health Spec (1); HIV AIDS Epidemiologist (1); HIV AIDS Spec (1) 05/09/200 1 4 Pub. Health spec (I ); Epidemiologist (1); Proj. Impl. Spec (1); Proc. Spec (1) 01/22/2002 2 Public Health Spl (1); Epidemiologist (1) 0911312002 9 Task Manager (1); Operations Analyst (1); CONS. Diagnostic, Treatment (I); Procurement Spec (1); Financial Mgmt Spec (1); Civil Society Spec (I); OPWUNAIDS (1); CDC-Monitoring/Eval. (I); CDC- Epidemiologist Survey (1) * Satisfactory 41 Annex B. Persons and Organizations Consulted Federal District of Brasilia National Program for STD and AIDS Control (NASCP): Paulo Roberto Teixeira, WHO (formerly Coordinator, NASCP) Alexandre Grangeiro, Coordinator, NASCP (formerly Deputy Coordinator, NASCP) Raldo Bonifacio, Deputy Coordinator Aristides Barbosa Junior, Coordinator, Monitoring and Evaluation Moises Francisco Baldo Taglietta, Coordinator, Planning Maria Alice Lipparelli Tironi, Planning Unit (ASPLAV) Rosemeire Munhoz., External Cooperation Unit (COOPEX) Denise Doneda, Director, Prevention and Education (PREV) Pamela Diaz Bermudez , Research Area (UEPVPesq) Didero Carlos Lopes, Finances Unit (UOF) Draurio Barreira, Epidemiology Unit (UEPI) Josue Lima, Diagnosis, Treatment and Care Unit (UDAT) Karen Bruck de Freitas, Civil Society and Human Rights Unit (SCDH) Mauro Siqueira, Communications Unit (ASCOM) Jackeline Fabiola E.F. de Souza, Human and Institutional Development Unit (UDHI) Carmen Dhalia, Epidemiologist (UEPI) Mauro Sanchez, Epidemiologist (UEPI) Leila Rossi, Director for MSM Prevention Programs (PREV) Paulo Junqueria Aguiar, Prevention Unit (PREV) Denise Serafim, Harm Reduction Coordinator (PREV) Omar Bravo, Consultant, Prison HIV/AIDS Programs State Program for STD and AIDS Control: Federal District of Brasilia Maria Josenilda Goncalves da Silva, Coordinator Suely Andrade, Health Coordinator Staff of State Coordination Managers/staff of Federal district HIV/AIDS facilities Civil Society Arco-Iris DF GAPA Brasilia State of Pernambuco State Program for STD and AIDS Control Francois Jose de Figueiroa, Coordinator Staff of State Coordination Managers/staff of State HIV/AIDS facilities (including visit to State Blood Bank) Municipal Programs for STD and AIDS Control, Municipality of Rio de Janeiro Municipal coordinators and coordination staff, including: D. Sair Pereira de Sena, Coordenadora, Olinda Maria Helena Gomes dos Santos, Coordenadora Municipal, Petrolina 42 Maria do Socorro Ratis, Gerente Administ., Paulista Francisco Leone Vay, Coordenador Municipal, Paulista Maria Candida Araujo de Oliveira, Coordenadora, Jaboatao dos Guarapes George C. Sa Barreto, Consultor Administ., Jaboatao dos Guarapes Scheyla M Silva Goncalices, Gerente, SAE Municipality coordination unit of Goncalores Managers/staff of selected municipal HIV/AIDS facilities Civil Society Roberto August0 de Brito, Rede National de Pessoas Vivendo corn HIV/AIDS Sec@o Pemambudo Vladimir Cardoso Ruis. Grupo de Trabahos Articulacao de Movimento Homosseseual do Recife et Area Metropolitani Iris de Fatima, Coordenadora, AMHOR Christiane/Lucielle, Aux de Coordenacao, Casa de Passagem Nanci Feijo de Melo, Coordenadora Geral, Associacao Pemambucana de Profissionais do Sex0 (APPS) Ivaldo Soles Alaide Alvias da Silmes, Presidente, ASAS: Associacao Solidaria Grupo Viva Rachid Josenita Dura Ciriaco, Socia Coordenadora, Crete Comunitaria Vivendo e Aprendendo Maristela Moraes, Asesora de ProjetosCoordenadora, Instituto PAPAVAssociacao Usuarios Alcohol e Drogas de Pemambuco State of Rio de Janeiro State Program for STD and AIDS Control Valdilea Goncalves Veloso dos Santos, Coordinator Denise Ribeiro Franqueira Pires, Deputy Coordinator Staff of State Coordination, including Luciana Kamel, Consultant, AIDS Program - State Health Department, Prevention Managers/staff of State HIV/AIDS facilities Programs for STD and AIDS Control Betina Durovni, Transmissible Diseases Department Head Municipal coordinator and coordination staff Research/Academic Community Francisco Inacio Bastos, Deputy Coordinator - Fiocruz AIDS Programme Celia Landman Szwarcwald, Senior Epidemiologist, Firocruz AIDS Programme Paulo Fieja Barroso, Federal University of Rio de Janeiro, Projeto Praca Onze Mauro Schechter, Professor of Infectious Diseases and Head of AIDS Research Laboratory, Federal University of Rio de Janeiro, Projeto Praca Onze Civil Society Maria Cristina Pimenta, ABIA Secretaria, FORUMONGRS Coordenador Geral, S.O.S. Vida Alex Uma dos Santos, Presidente, S.O.S. Vida Marisa Campany, Coordenadora, Movimento Mulheres em Sao Goncalo Centro Especial de Orientacao a Mulher - Zuzu Angel 43 Norma Maria Gomes, Tesoreira, Associa@o Carioca de Redu@o de Danos Edmundo, Coordenador, Adaps Flavio Lewe, Director Adjunto, Davida Gabriele Ceite, Diretora Executiva, Davida, Coordenadora Rede Begs. Profissionais do Sexo Roberto Pereira, Coordenador, Centro do Educa@o Sexual Forum de ONG/AIDS do Estado do Rio de Janeiro State of Santa Catarina State Program for STD and AIDS Control Ana Maria Henrique Martins Costa, Coordinator Staff of State Coordination Managers/staff of State HIV/AIDS facilities Municipal Programs for STD and AIDS Control Municipal coordinators and coordination staff (municipalities of Florian6polis and Itajai) Managers/staff of municipal HIV/AIDS facilities Civil Society Grupo de Apoio Regional Para Rehabilitaqao da AIDS, Lages, SC IINST Tuto Arco Iris, Florianopolis, SC Casa - Centro de Assessoria a Adolescencia, Florianopolis, SC State of Rio Grande do Sul Ministry of Health State Health Secretary State Program for STD and AIDS Control Geralda Maria Bauer Pereira Rigotti, Coordinator Staff of State Coordination Managers/staff of State HIV/AIDS facilities Municipal Programs for STD and AIDS Control Municipal coordinators and coordination staff, municipalities of: Porto Alegre Alvorada Uruguaiana Rio Grande Sao Leopold0 Caxias do Sul Managers/staff of municipal HIV/AIDS facilities Research/Academic Community Dr. Paulo Naud, Clinicas Hospital, University of Rio Grande do Sul, Port0 Alegre Civil Society Consorcio Intermunicipal de Saude, Regiao Centro do Estado, Santa Maria, RS Grupo de Apoio a Preven@o da Aids (GAPA) do Rio Grande do Sul. Port0 Alegre 44 State of S%oPaul0 State Program for STD and AIDS Control: State of S%oPaulo Oswald0 Yoshimi Tanaka, Deputy State Secretary of Health, Cabinet Secretary, Government of Sao Paul0 State Artur Olhovetchi Kalichman, Coordinator (and Head, Referral and Training Center) Maria Clara Giana, Deputy Coordinator Jo20 , Human Resources Head Christian0 Azevedo Marques, Head, Institute Adolf0 Lutz Staff of State Coordination, including: Anna Luiza Luns Gryschek, Nurse Technician Managers/staff of State HIV/AIDS facilities Municipal Programs for STD and AIDS Control Thomas Soderberg, Municipal Secretary of Health (Santos) Municipal coordinators and coordination staff (Sao Paulo and Santos) Managers/staff of municipal HIV/AIDS facilities Civil Society Jorge Beloqui, Grupo Pela VIDDA Forum ONG/AIDS SZio Paulo Forum ONG/AIDS ABC Grupo CORSA Lutando Pela Vida NASP - Nucleo de Apoio Solidiaro e Prevancao Associcao LAR Grupo de Incentive a Vida (GIV) Grupo de Acao Pela Cidadania Homosexual World Bank, Washington, D.C. Anabela Abreu, Sector Manager, SASHD (former Task Manager AIDS II) John Garrison, Senior Communications Officer, EXTIA (former Civil Society Specialist, LCC5A) Indermit Gill, Economic Adviser, PRMVP (former Lead Economist, LCCSA) Charles Griffin, Sector Director, SASHD (former Sector Director, LASHD) Theresa Jones, Lead Operations Officer, LCSHS (former Task Manager AIDS I and AIDS 10 Maureen A. Lewis, Sector Manager, HDNVP (former Task Manager AIDS I) Gobind Nankani, Vice President, PREM (former Country Director, Brazil) Sandra Rosenhouse, Senior Population and Health Specialist, LCSHH (Task Manager AIDS II and III) Joachim von Amsberg, Lead Economist, LCCSA (Brazil) Other Pedro Chequer, UNAIDS Country Program Adviser, Moscow (former Coordinator, NASCP) Michael Merson, former Head, Global Programme on AIDS William McFarland MD, PhD, Co-Director HIV/AIDS Statistics and Epidemiology Section, San Francisco Department of Health Dr. Theresa Diaz, Epidemiologist, Outgoing Chief of Party for Brazil, United States Centers for Disease Control 45 Annex C. HIV/AIDS Epide ic in Brazil. With the first AIDS casesdiagnosed in 1982, the HIV/AIDS epidemic in Brazil is now entering its third decade. HIV/AIDS in Brazil was first characterized by early (1980-85) and rapid spreadof HIV-l among urban homosexual and bisexual men (MSM), an epidemic similar in scope, spread and epidemiology to the epidemics in North America and Western Europe. This outbreak was concentrated in major coastal cities, including Sao Paulo and Rio de Janeiro. It was soon followed by substantial spread in urban injecting drug users (IDU), again in Sao Paul0 and Rio, but also in the southern States of Rio Grande Do Sul and Santa Catarina. By 1990, IDU spread accounted for approximately 20 percent of HIV infections in Brazil (World Bank 1993). Since both MSM and IDU outbreaks involved few women, the early gender ratio of infection was roughly 25: 1 male to female. The second decade of HIV/AIDS, roughly 1992-2002, was marked by increasing spread of HIV among high-risk heterosexuals, including commercial sex workers (CSW), sexually- transmitted disease (STD) clinic attendees,and the wives and sex partners of male IDUs or bisexuals. The gender ratio in AIDS casereports has fallen to 1.7:1 male to female in 2003, suggesting a feminization of the epidemic, as well as declines in infections among men that have occurred some 10 years ago. During this period AIDS morbidity and mortality rose, as those infected progressed clinically and developed opportunistic infections, most importantly HIV- associated tuberculosis. AIDS mortality peaked at 12/100,000 in 1995, before the advent of highly active antiretroviral therapy (HAART), and has been declining ever since, reaching 6.3/100,000 by 2000. As of 2000 there were an estimated 600,000 Brazilians living with HIV infection, down from PAHO's estimate of 750,000 in 199229.As of March 2002,237,588 AIDS cases and 1lo,65 1 deaths from AIDS had been reported since the beginning of the epidemic. In the absenceof adequateHIV surveillance data, the population prevalence of HIV has been reliably estimated at 0.65 percent through national probability sampling of pregnant women in 200230,well below the 2 percent rate UNAIDS has suggestedfor describing generalized epidemics (Szwarcwald (CL) 2002). Brazil's HIV epidemic in 2003 thus remains concentrated, still largely limited to individuals and groups engaging in high-risk behaviors for HIV, including MSM, IDU, CSW, heterosexuals with multiple partners and female partners of IDU and bisexual men. While transmission doubtless continues to occur in persons at risk (and an increasing proportion of persons at risk are women, especially partners of IDUs and bisexual men), there is no evidence of acceleration or increasing rates of new HIV infection. Rather HIV infections are stable or slightly declining, while HIV/AIDS survival times are significantly increasing. The modes of transmission of the epidemic, the prevalence and behaviors of high-risk groups, and the prevalence by region of the country have evolved since the early 1980s. From 29. As part of the preparation of AIDS III, the Bank estimated that about 800,000 Brazilians were infected in 2002. The government is currently preparing an estimate for 2003. 30. A study by Szwarcwald (CL) and Carvalho (2001) included 132 sentinel sites and anonymous blood collections from 15,426 pregnant women. Overall 0.47% of women (95% CI 0.36, 0.58) were HIV infected. Using a gender ratio derived from recent AIDS reporting, rates in men of reproductive age were estimated at 0.84% (95% CI 0.65, 1.04). This gave a combined population prevalence of 0.65% (95% CI 0.5 1,0.80) for the population aged 15-49 years nationwide. This well-done study is likely to be the best estimation available of the actual HIV infection rate in Brazil in 2000. Using wider (68%) confidence intervals and census data on the national population generates total numbers of HIV infections at 597,000 with a lower limit of 548,000 and an upper bound of 647,000. It is this study that generated the widely used figures of 0.65% HIV prevalence and the estimate of 600,000 people living with HIV in 2000, and there is little reason to think that these estimates are substantially lower or higher than the 68% confidence intervals. While the history of HIV/AIDS suggests complacency is always a danger, the characterization of a still concentrated epidemic seems, on the whole, correct. 46 1980 to 1990 sexual transmission accounted for 63.8 percent of AIDS cases and this has gradually increased to 74.4 percent of AIDS casesin 2000. Within this category, transmission through MSM decreasedby more than half from an estimated 47.4 percent of all AIDS cases during the 1980s to 20.3 percent in 2000, while the share of casesattributable to heterosexual transmission more than tripled from 16.4 percent to 54.1 percent. Heterosexual transmission is still limited to high risk groups (partners of IDUs and bisexual men and people with multiple partners). Transmission through intravenous drug use rose from an estimated 18.5 percent of total casesin the 1980s to a high of 24.8 percent in 1992 and subsequently declined in the ensuing years to 11.4 percent in 2000. Transmission by transfusion has decreasedas a share of reported AIDS casesfrom 5.3 percent in the 1980s to 0.2 percent in 2000 and perinatal transmission rose from 1.8 nercent in the 1980sto 2.5 nercent in 2000 (see Table 1). Cases by Mode of Transmission and Year of Diagnosis Cum u/a five Annual Cumulative ~xx~I.~x^.,I_,x---.-e--m,--, -MmIXX..- Total I, ..-,,.-..,..",,,,n,-,,..lz....,,,,..,,,,,as,,,,,.,,..,,,..,.,,,,,,,,,.,s,.,,,,,,,".,xx-,,..,,,.,.,,,..",,,,,..,,x,I 7980-90 7992 7994 7996 7998 2000 ,~~~I"-..,-+..xxl--~--~~,~ 7980-2002 Sexual 63.8 59.2 60.0 61.5 70.2 74.4 65.6 MSM 47.4 33.5 26.9 21.8 22.4 20.3 26.6 Hetero 16.4 25.7 33.1 39.7 47.8 54.1 39.0 IDU 18.5 24.8 21.4 17.9 133 11.4 17.2 Transfusion 5.3 2.4 1.8 1.7 0.2 0.2 0.5 Perinatal 1.8 2.8 3.4 4.0 3.3 2.5 3.0 Unknown 10.7 * 10.8 13.3 15.0 13.0 11.6 12.4 W%?%+*S.SP--~~~~.-~--~w~s~~-~mw100.0 Total 100.0 100.0 100.0 100.0 100.0 100 Source: Ministry of Health Epidemiological Bulletin, March 2002. Studies on the HIV prevalence of the highest risk groups reveal prevalence levels of 36.3 percent among IDUs, 10.8 percent among MSM and 6.5 percent among CSW (seeTable 2) for recent years in various samples, most of them non-national. As noted in Table 2, these data must be interpreted with caution. Trend data on HIV prevalence over time are not available for these groups. Incidence of AIDS in Brazil, as measured by year of diagnosis, increased from 8.2/100,000 in 1991 to a high of 18.2/100,000 in 1998 with a decline to 12.0 recorded in 2002, but patterns across regions vary. In 2002 the South surpassedthe Southeast as the region with the highest incidence; and incidence rates in the North and Northeast, while showing slight decline in 2000, are still of concern (See Table 3). Again, trends in AIDS casesreflect levels of transmission that occurred some 10 years before. 47 Table 2: HIV Infection Rates and Risk Indicators for High-Risk Populations (IDU, CSW, MSM) in Brazil, 1999-2002 Indicator IDU cswc3/ HIV prevalence 1999a/ 52% 1994-99 10.8% 2001 b/ 36.3% 2001 6.5% HCV prevalence 199960% 2001 56.4% 2001 4.5% Condom use 1999 42.1% clients, 73.8% 2001 d/ 2001 62.9% reg. partners 23.9% reg part. 81% casual 95% 2002e/ 70% use in all anal intercourse over last 6 months Ever HIV tested 199952% 49.2% 200073% 2001 66.4% 200269% Needle sharing 199970% 2001 59.4% Sample sizes N 1999=287 N 2001=3000 N 94-99= 1,082 N 2001=869 N2001=800 N2002=1,200 Source: Ministry of Health/NASCP 2003, compiled from the following studies: a/ Federal University of Minas Gerais and NASCP on 287 IDUs in five cities b/ Federal University of Minas Gerais and NASCP on 869 IDUs in seven cities c/ Federal University of Brasilia and NASCP on 3000 CSW in five cities d/ Opinion poll carried out by IBOPE on 800 MSM in seven capital cities e/ Opinion poll carried out by IBOPE on 1,200 MSM in 10 capital cities Note: All of these data need to be interpreted with caution. The first study on IDUs was of dubious quality and included only program users. The data on CSW condom use and HIV testing refer only to the CSWs who participated in the project. Use of condoms with clients by non-participating CSWs is less than that for participating CSWs (60% vs 74%) as is the percentage of those ever tested for HIV 36% vs 49%. Data on MSM for 1994-1999 relies on a very biased sample, given that more was considerable sample attrition. . _.,_.~I",rI..^I.....-I._1....X.~.. I.I...~...-.-."~.....G..&u- _.._"...-...-.---....v.,.x,._....-._xIIYY....Y.~~..,.x. ,~,..l,.._.."l...1_~...I..-.-- Cumulative 1 _.,.I..."...UsI..... .."-. .,..~..I.,."...._.."._.,_...___.."Yl..._ 79804990 . ....3....".11II,.......xx-..*...x._I_I..w+* uIy..~...*y"".~I..___..I..Y.,*~.~,l.,~x1_._1YX.._I...._y,.,._.,,,.,,._....._"a1 185 1.8 Northeast 1637 1I 1039 2.4 1 Center West 760 653 6.8 Southeast 20723 $,1i 11918 18.8 South 1831 i1\ 1523 6.8 Brazil 25172 j 15318 10.3 ~~~,~~~~~~~~~~~~~~~~~~~-,~~~~~~~~~~~,~~~~~-~~~~~~~~-~~~~~*~~~~~~~~-~ Source: Ministry of Health 2003 *Includes preliminary data through September 2003. \<> r . . . . . . . . . . . . . . . . . . . . . . . , . I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . " . " . . . i E&..."I1111~1.1".~x_l.~~. ..YxI..-yyI...--_"--~---. "-_l_xx,---.~--xxf,~h 2002 t)..*.~X.".~~..."I~..-IxxIx.~.-~.~~~--....-_\-......yI..~--x. Cases Incidence .I...I...,.~.~.~-.-,....-....".& 1 $ 606 4.5 1; iii 2383 4.9 j fi 957 7.9 i ! 11861 15.9 \ 1 5089 19.8 j ? 20897 12.0 49 Annex D. AIDS I and II: Presentation of Project Components AIDS and STD Control Project I - Components Prevention through: (4 information, education and communication (IEC) programs emphasizing targeted interventions to inform both the general population and, particularly, certain populations, of the risks of HIV/AIDS and means to prevent or treat it; @) expanding medical staff capacity for diagnosing HIV, AIDS and STDs through increasing sites and upgrading staff expertise; (c) broadening public sector capacity and capability in counseling and testing and early intervention; and (d) condom procurement and distribution. Services, including: (a> establishment of systems for identifying and treating STDs; and 03 integrating HIV/AIDS prevention and medical services with long-term counseling to improve access and delivery of services and to direct patients to effective, low cost substitutes for hospitalizations; Institutional Development to build capacity to control HIV/AIDS and STDS through: (a) training of (mostly existing) health workers in service delivery, quality control and monitoring of AIDS, for laboratory testing and quality control, and for counseling and treatment across the SUS; (b) upgrading State Reference Laboratories to a basic standard in each macro-region to support HIV/AIDS and STDs testing; and (c) technical assistance; Surveillance, Research and Evaluation through: (a) epidemiological surveillance; (b) program evaluation in IEC, counseling and testing, services and laboratory quality assurance; and 6) special studies, including epidemiological analysis and projections/ costs and cost effectiveness of interventions; and the economic impact of HIV/AIDS. AIDS and STD Control Project II - Components / Prevention of AIDS and STD ($128 million), including: (4 mass media campaigns to raise awareness and understanding of AIDS and STD transmission and promote safer practices; (b) promotion of safe practices through education, condom distribution and needle exchange programs; (c) targeted interventions among specific sub-groups at greater risk of contracting and/or spreading the HIV virus, including homosexual and bisexual men, intravenous drug-users, sex workers; adolescents, prisoners, truck drivers, and others; and preventive activities oriented to other vulnerable groups such as women and low income populations; (d) a National Human Rights Network promoting non-discrimination against persons with HIV/AIDS; (e) NGO executed sub-projects for AIDS and STD prevention, selected on a competitive basis; and (f) Counseling services and dial-in AIDS information (Pergunte AIDS). Diaqnosis, Treatment and Care for persons with HIV, AIDS and STD ($102 million) that would support:/improve: (a) the operation, standardization and quality control in diagnostic laboratories; implementation of nearly 100 new anonymous testing and counseling centers (CTAs) and maintenance of selected existing CTAs; support municipal public health laboratories through the purchase of equipment and supplies for viral load cd4 cell and other diagnostic testing; in complement to the investments being financed by the REFORSUS project in public labs, blood banks and control of transfusions; (b) care and treatment of persons with HIV, AIDS and STD including about 80 group homes (casas de apoio) for the care of AIDS patients and about 40 homes for orphans of AIDS; it would also support about 80 existing specialized care units, 30 existing day hospitals, 40 existing home care programs and AIDS in-patient care in about 100 hospitals; w strengthening of STD diagnosis and treatment services through: training in STD case management; licensing of about 10 national STD reference centers which will review and test norms and procedures and undertake special studies; and provision of re-agents, drugs, condoms, educational materials and other supplies; and (d) implementation of a centralized logistical control system for drugs and condoms, the implementation of a cost control system for HIV/AIDS care and the development of a reference system for gynecological care of HIV positive women. Institutional Strenothenina ($70 million) of Executing Agencies Responsible for AIDS and STD Control, in support of: (a) epidemiological surveillance including sentinel surveys for HIV/AIDS and expansion or upgrading of the HIV and STD case notification system, as well as a national study of STD transmission, HIV prevalence studies among specific risk groups, HIV sub-type studies and other epidemiological studies; (b) the National Reference Laboratories for HIV and STD including support for quality control in lab testing, expansion of the national network for STD drug resistance studies, implementation of a network to monitor HIV susceptibility to anti-retroviral therapy; (c) training activities, including training of CN-DSTIAIDS staff, specialized training of health workers involved in prevention and treatment among high-risk groups, training for NGO personnel in project development and reporting, as well as training in counseling and testing, epidemiology and laboratory diagnostics for AIDS and STD; (d) research, including studies of survival, case notification delay, opportunistic, epidemiology and projections, cost and impact, behavioral change, AIDS among indigenous populations and in mining camps and extractive reserves, a national st6dy of sexual behavior change, and others. The component would also support selection of, and support for, about 3 scientific centers of excellence to carry out long term multidisciplinary research; and monitoring and evaluation activities, including strengthening supervision and evaluation of CTAs, public health labs, and the syphilis case investigation network; evaluation of all state and municipal institutions undertaking project activities; evaluation of NGO projects including monitoring of 80 NGO projects, KAP studies on impact of preventive activities in the workplace; workshops for NGO participants on evaluation; preparation of monitoring and evaluation reports; project impact studies; and evaluation of interventions amona soecific oooulations. 50 Annex E. Project Costs and Financing Table E-l: AIDS I Project Cost by Component (US$ million) Component Appraisal Estimates* Actual Cost** Base Cost Full Cost Amount % of % of Base Full Prevention 93.9 102.7 120.0 128% 117% Services 75.7 84.6 75.0 99% 89% Institutional Dev 42.3 46.5 47.5 112% 102% Surveillance, Research And Evaluation 14.6 16.2 75. 5 1% 46% Phys. Contingencies 8.7 Price Contingencies 14.8 Total Project Cost 250.0 250.0 250.0 100% *Source: Proj ect Appraisal Document - PAD * *Source . Implementation Completion Report l Table E-2: AIDS II Project Cost by Component (US$ million) Component Appraisal Estimates * Actual Cost** Base Cost Full Cost Amount % of % of Base Full Prevention 119.0 128.0 132.92 112% 104% Diagnosis, Treat- ment, Care 94.0 102.0 87.23 93% 86% Institutional Strengthening 65.0 70.0 76.35 117% 109% Phys. Contingencies 0 Price Contingencies 22.0 Total Project Cost 300.0 300.0 296.50 99% * Source: Project Appraisal Document -- PAD ** Source Relatorio de Implementacao e Avaliacao 1998 a 2003, Acordo de Emprestimo BIRD l 4392/BR -`Projeto AIDS II, Versa0 Preliminar, Ministerio da Saude, Secretaria de Vigilancia em Saude, Programa National de DST e Aids, Brasilia-DF, agosto de 2003 51 Table E-3: Transfers to Local Government through AIDS I Nominal US Dollars Transfers Population Total per capita State 1994-1998 2000 (2000) Acre $690,902 557,882 1.24 Alagoas $830,654 2,827,856 0.29 Amapa $328,544 477,032 0.69 Amazonas $973,099 2,8 17,252 0.35 Bahia $4,641,070 13,085,769 0.35 Ceara $5,359,740 7,43 1,597 0.72 Distrito Federal $5,260,998 2,051,146 2.56 Espirito Santo $1,191,234 3,097,498 0.38 Goias $1,376,667 5,004,197 0.27 Maranhao $1,825,543 5,657,552 0.32 Mato Gross0 $1,567,443 2,505,245 0.62 Mato Gross0 do Sul $4,075,972 2,078,070 1.96 Minas Gerais $4,227,125 17,905,134 0.24 Para $2,350,5 17 6,195,965 0.38 Paraiba $1,281,765 3,444,794 0.37 Parana $7,053,120 9,564,643 0.74 Pernambuco $4,453,386 7,929,154 0.56 Piaui $2,683,067 2,843,428 0.94 Rio de Janeiro $14,732,832 14,392,106 1.02 Rio Grande do Norte $2,777,622 2,777,509 1.02 . Rio Grande so Sul $5,741,235 10,187,842 0.56 Rondonia $1,165,395 1,380,952 0.84 Roraima $1,224,480 324,397 3.77 Santa Catarina $6,081,919 5,357,864 1.14 Sao Paul0 $30,909,393 37,035,456 0.83 Sergipe $1,611,172 1,784,829 0.90 Tocantins $1,425,299 1,157,690 1.23 TotaVAverage 169,872,859 0.68 Source: Ministry of Health, NASCP 52 Table E-4: Project Financing Brazilian Government Investment in STDlAIDS (1997-2001) 600 500 400 I 1 300 200 100 0 1997 1998 1999 2000 2001 Year 53 Table E-5: Transfers to NGOs in AIDS I and AIDS II. bv state Per capita Per capita outreach spending spending (Ohof (Ohof North Amazonas 8 $30,387 $163,844 $194,231 2,8 17,252 46% 58% Acre 10 $34,185 $283,201 $3 17,386 557,882 379% 506% Amapa 6 $43,557 $31,166 $74,723 477,032 104% 65% Para 25 $302,882 $326,930 $629,8 12 6,195,965 68% 53% Rondonia 14 $19,435 $237,646 $257,08 1 1,380,952 124% 172% Roraima 3 $10,039 $43,845 $53,884 324,397 111% 135% Tocantins 7 $0 $86,279 $86,279 1,157,690 50% 74% Northeast Alagoas 6 $93,683 $52,444 $146,127 2,827,856 34% 19% Bahia 44 $351,337 $1,070,365 $1,421,702 13,085,76972% 82% Ceara 46 $244,589 $915,160 $1,159,749 7,431,597 104% 123% Maranhao 10 $44,948 $161,729 $206,677 5,657,552 24% 29% Paraiba 17 $241,374 $300,962 $542,336 3,444,794 105% 87% Pernanbuco 22 $250,710 $717,311 $968,02 1 7,929,154 81% 90% Piaui 10 $83,816 $107,149 $190,965 2,843,428 45% 38% Rio Grande do Norte 10 $66,188 $235,225 $301,413 2,777,509 72% 84% Sergipe 12 $9 1,130 $274,466 $365,596 1,784,829 137% 153% Center- West Mato Gross0 16 $34,869 $268,967 $303,836 2,505,245 81% 107% Mato Gross0 do Sul 31 $77,907 $678,72 1 $756,628 2,078,070 243% 326% Goias 21 $350,425 $327,932 $678,357 5,004,197 90% 65% Distrito Federal 35 $397,388 $589,357 $986,745 2,051,146 321% 287% Southeast Espirito Santo 14 $62,040 $133,872 $195,912 3,097,498 42% 43% Minas Gerais 32 $601,722 $554,654 $1,156,376 17,905,134 43% 31% Rio de Janeiro 124 $1,152,101$2,717,906 $3,870,007 14,392,106 179% 188% Sao Paul0 185 $2,594,280 $3,780,495 $6,374,775 37,035,456 115% 102% South Parana 45 $307,852 $1,110,555 $1,4 18,407 9,564,643 99% 116% Rio Grande do Sul 45 $608,063 $1,061,197 $1,669,260 10,187,842 109% 104% Santa Catarina 35 $345,975 $795,09 1 $1,141,066 5,357,864 142% 148% Brazil 830 $8,440,882 $17,026,468 $25,467,350 169,872,859 100% 100% 54 Annex F. Selected Data on Project Performance Table F-l: STD Case Reports by year. Brazil, 1996-2001. 1Cervicitis 1826 1241 1373 16.893 124.703 132.212 165.248 IHPV 1659 1389 Il.867 15.941 115.852 115.570 140.278 1 1Genital Herpes 144 194 1172 1790 12.131 12.204 15.435 I ISyphilis 12.710 Il.254 (2.395 111.496 115.968 113.138 146.961 1 1Genital Ulcers 169 149 1101 Il.083 12.696 12.861 16-.859 IUrethritis 11.165 1796 12.075 110.940 121.974 119.188 156.138 1 ICongenital syphilis 1376 12.025 13.868 14.366 14.487 13.316 118.438 1 ITOTAL 15.849 14.848 110.851 141.509 187.811 188.489 1239.357 ] Source: SINAN Table F-2: Number and Distribution of Counseling and Testing Centers (CTAs) in Brazil 2003 55 Table F-3: AIDS admissions in the SUS, Number of Admissions Avoided and Resource Savings, 1996-2001 1996 1997 1998 1999 2000 2001 No. of AIDS admissions registered 25.458 25.157 24.700 25.027 26.655 25.274 (*) in DATASUS Estimated number of patients attended by 15.390 31.140 43.823 57.604 69.447 90.400 SUS (year) Average number of admissions/patients 1,65419 0,81 0,56 0,43 0,38 0,28 by year. Estimated number of admissions in year, maintaining average 51.511 72.491 96.542 114.876 149.539 number of admissions for 1996 Estimated number of admissions avoided 26.354 47.791 71.54 88221 124.265 in year Estimated cost of admissions avoided 76.27 138.31 207.06 255.32 359.64 in year (US$million) (*) figures up to 30/l l/2001 Table F-4: Average Cost (US$) of ARV Therapy by Patient/Year Brazil, 1996 to 2002* 5000 4000 3000 2000 0 ! I I I I I I I 1996 1997 1998 1999 2000 2001 2002 Year 56 Table F-5: Targeted Interventions for Commercial Sex Workers, Injection drug users and Men who have Sex with Men: 1998-2003 Source: Ministry of HealthlNASCP 2003 57 Table F-6: HIV SEROPREVALENCE BY TYPE OF GROUP/SENTINEL SITE AND REPORTIG PERIOD BRAZIL 1997 - 2000 # Sites # Persons HIV-positive Confidence Interval of Reporting Period Group Tested 95% # % March 1997 STD patients 19 3392 143 4,22 [3,54 - 4,891 Pregnant women 32 6091 71 1,17 [0,90 - 1,441 Emergency room 27 5100 220 4,31 [3,76 - 4,871 patients October 1997 STD patients 36 6102 205 3,36 [ 2,91-3,811 Pregnant women 55 10234 89 0,87 [ 0,69- 1,051 Emergency room 44 7941 * 342 4,31 [ 3,86-4,751 patients March 1998 STD patients 38 7247 221 3,05 [ 2,65- 3,451 Pregnant women 55 10656 58 0,54 [ 0,40 - 0,681 Emergency room 44 8288 173 2,09 [1,78 - 2,40 ] patients Outubro STD patients 32 6327 179 2,83 [ 2,42 - 3,241 De 1998 Pregnant women 51 10218 61 0,60 [ 0,45 - 0,751 Emergency room 41 8214 217 2,64 [ 2,30 - 2,991 patients March 1999 STD patients 36 6901 121 1,75 [ 1,44 - 2,061 Pregnant women 61 12081 90 0,74 [ 0,59 - 0,903 Emergency room 45 8869 213 2,40 [ 2,08 -2,721 patients October 1999 STD patients 39 6711 198 2,95 [ 2,55 - 3,361 Pregnant women 64 11222 91 0,81 [ 0,64 - 0,981 Emergency room 46 8688 248 2,85 [2,50 - 3,20 ] patients 2"d Semester 2000 Pregnant women 128 15426 88 0,57 [ 0,45 - 0,691 STD patients 25 4636 124 2,67 [ 2,21 - 3,14 ] Source: Ministry of Health/NASCP Statistics 58 Table F-7: Brazil AIDS II: Distribution of Targeted Interventions Region Civil Society MSM Prison Truck Driver Projects Projects Projects Projects IMY- % US$ % # % # % # % dence Total share Proj share Proj share Proj share 2000 POP Center 98. 7 64. 9 59 13 5 85 1 4 l West Northeast 5.0 28 11.1 16 122 26 11 19 2 7 North 38 8 42 6 21 4 5 85 7 26 Southeast 18.7189 4215 3$411'6 5217 20961 4413 2513 4222 107 2637 South Total/ 12'4. 100 69'7. 100 472 100 59 100 27 100 Projects Women 3 NGO Projects for Indigenous Targeted to Projects Projects for Resettled Peoples Poor Adolescents ,-"Ix,-,,-.,,__,_.I-,,~~,______y,=x,-yxI,,,-~_yl--.~-- Populations I-.-.-I-x-.-,,,,--,-~~-~,s~-~--I-y----ylyI Projects Region # 0/ 0 # 0/ 0 # 0/ 0 0 Prq). 0 ProJ. # / 0 ProJ. # /0 share ProJ. share ProJ. share share share Central 12 8 na 7 18 16 21 29 15 34 West Northeast 36 22 na 23 28 25 23 32 2 5 North 9 6 na 6 4 4 13 18 17 39 Southeast 90 56 na 45 50 45 12 16 5 11 South 13 8 na 19 11 10 4 5 5 11 Total 160 100 na 100 111 100 73 100 44 100 59 .e.,,..,e,..,,,,..e,,,A-,,,,,..,e,,..,e,,,,,,.,S^2,population) -*,-,*,IX,,,..,,,.l*,..,*.aaz,.,,...,,,^.,IIX,^CI~,.,..,,..,,z,,,,-,,,4>.,-,*1-,,A,,,w,,,,H,,.,...,,..,,.,s,,,-,,,,,,,,,,.,,,-,.,,,w,,..,.,.%..,,,,,,,uI,,ua...-,-,.,,,..,,,I,x..,,~lr_"u,"*,...,,,..,,,,,-,,^,~x.,..,,x-~-y......,."e.,,,,...a,,,..,m,,,A...,,,,,.,,..-,,a,,,,X~~,,~,~,X_.~,X,X~.,II",NXY~X,X.XX,I,XXX Less than Equal to More than (within 3 percentage points) Center 9.8 7 C. Society: 9% MSM: 13% West Tr. Drivers: 4% Indigen: 34% Poor: 8% Adolesc.: 16% Women: 7% Resettled: 29% Prisoners: 8.5% Northeast 5.O 28 C. Sot: 16% MSM: 26% Resettled : 32% Prisoners: 19% Adolesc. 25% Tr. Drivers: 7% Poor: 22% Women: 23% Indigenous: 5% North 3.8 8 MSM : 4% Prisoners: 8.5% Tr. Drivers: 26% Adolesc. : 4% C. Sot: 6% Resettled: 18% Poor : 6% Indig: 39% Women : 6% Southeast 16.9 42% Tr. Drivers: 37% Prisoners: 42% C. Sot: 52% Resettled: 16% MSM: 44% Poor: 56% Indig: 11% Women: 45% Adolesc: 45% South Poor: 8% MSM: 13% Prisoners: 22% Adolesc: 10% c. sot: 17% Tr. Drivers: 26% Resettled: 5% Women: 19% Indig: 11% 60 Annex G. Borrower Comments Ministry of Health Health Surveillance Secretariat National STD/AIDS Control Program COMMENTS ON THE ASSESSMENT REPORT ON THE AIDS I AND AIDS II PROJECTS, PREPAIWD BY IBRD The IBRD Operations Evaluation Department submitted a draft Project Performance Assessment Report on the First and Second AIDS and STD Control Projects (IBRD Loans 3659- BR and 4392.BR) to the National STD/AIDS Control Program (NASCP), with the request that the National Program (NP) review the draft and provide feedback on the document. After a process of shared team work by the NP units, the following comments were produced and submitted. Epidemiological surveillance: In accordance with the UNAIDS guidelines, Brazil is required to develop Second Generation Surveillance by combining a set of strategies designed to ensure "information useful for action in public health." These strategies are: universal and compulsory notification of AIDS (since 1986) and of Seropositive Pregnant Women and Exposed Newborns (since 2000); sentinel surveillance of HIV and risky behaviors (by means of formation of sentinel groups and sentinel services); monitoring of HIV prevalence in more vulnerable populations; and performance of cross-section studies in the general population and among groups at high risk of infection, in addition to cross-linking of the national databases (National Notification System for Communicable Diseases, Mortality Data System and Hospital Data System, plus SICLOM [Control of pharmaceutical procurement, stock and distribution information system], SI-CTA [Information Systems - Testing and Counseling Centers] and SISCEL [Control of laboratory tests and examinations of the AIDS'STD program]). In the case of AIDS I, "150 sentinel surveillance sites (target: 80) were established in maternity clinics, emergency rooms [these sites did succeed in generating data on HIV prevalence in time for the evaluation of the AIDS I project] and STD clinics [not considered reliable]" (pp. 13-14, paragraph 4.11) and, in addition, "research and studies undertaken with project support provided important information and insight on HIV and behavioral trends" (p. 14, paragraph 4.12). In point of fact, the research was not only financed by the NASCP but actually formed its surveillance system. With regard to the AIDS II project, 2001 was the year in which we started setting up the Testing and Counseling Centers Information System for STD/AIDS (SI-CTA), with the aim of generating a continuous flow of reliable and representative data for serological and behavioral surveillance of HIV among pregnant women and some groups at high risk of infection. In 2003, during the time of the Bank mission, there were already data recorded in this system and it is not true that "at the close of AIDS II, surveillance is still limited to pregnant women and military conscripts, " that "no routine surveillance was undertaken during 2001 and 2002," and that "the data for 2003 is not yet available" (pp. 22-23, paragraph 4.33). 61 The incidence of AIDS is remaining stable, at around 25,000 new cases per year, or 15 new cases per 100,000 inhabitants five years ago. As regards HIV infection, a study of HIV prevalence among pregnant women, on a populational basis, performed in 2000 in 140 maternity clinics, indicated an HIV prevalence among pregnant women in Brazil of 0.61% (CI95% 0.48%- 0.73%). If this finding is compared with those of prevalence studies among pregnant women done in sentinel sites from March 1997 to October 1999, the stabilization of HIV infection in this population is also apparent (March/97 - 1.2%; October/97 - 0.8%; March/98 - 0.5%; October/98 - 0.6%; March/99 - 0.7%; October/99 - 0.8%). `3 can be noted that some 20 years after thefirst AIDS cases were identified in the country, Brazil's HIV epidemic remains concentrated among a high-risk population, the prevalence of HIV having been reliably estimated by MoH at 0.65 percent of reproductive age adults" (pp. 15-16, paragraph 4.15). With regard to the monitoring of the sexual behavior of the population in general, special mention should be made of the performance of a nation-wide research study in 1998 designed precisely to serve as a baseline for surveillance of the Brazilian population's sexual behavior in AIDS II. The repeating of this study in 2004 will enable an assessment of the trends of the Brazilian urban population's sexual behavior in the interval between the two studies, while also serving as a baseline for AIDS III, which will greatly facilitate the assessment of the effectiveness and impact of this project and of the Brazilian program in general. As regards "a routine of behavioral and HIV surveillance" among groups at high risk of infection (MSM, commercial sex workers (CSWs) and intravenous drug users (IDUs)), studies were done with CSWs (1 section) and IDUs (2 sections), while a study among STD carriers is in the final phase of execution and special studies (assigning priority to the MSM population) are to be carried out with project support. However, it is clear that Brazil "must give priority to and expand its surveillance efforts to enable routine and systematic monitoring and analysis of HIV infection trends, and of behavioral trends among target populations I' and that "the surveillance of HIV infections and of behaviors must focus on the high-risk groups" (MSM, IDUs, CSWs), without neglecting the monitoring of trends among the population in general, and paying special attention to the links between these two groups, particularly through the bridge populations (female partners of IDUs and bisexuals, and clients of CSWs). AIDS III is supporting Brazil in its efforts to fully explore the huge volume of data currently available from studies and research, but it is certain that this will not eliminate the present and future need to gather data on HIV and on the behavior of the target populations. We do not agree, however, that behavioral and HIV surveillance among high-risk groups should be done annually, or that it can be performed more efficiently at the state level. Finally, it must be emphasized regarding the actual collection of the information by the members of the World Bank mission, that with the evaluation being focused as it was on surveillance, monitoring and evaluation activities, one might have expected that the team and its technical managers would have been consulted in the information-gathering process. The National STD/AIDS Control Program is constantly seeking to refine its actions and intervene effectively to control the epidemic. The criticisms expressed by the members of the Bank mission will be taken into consideration and, where pertinent, will serve to improve the quality of the program; however, we cannot refrain from pointing out that a good part of them derive from incomplete analyses or information obtained from interlocutors insufficiently familiar with the STD and HIV/AIDS surveillance system operated by the Ministry of Health. 62 Monitoring and Evaluation: During the implementation phase of AIDS I, in the period between I994 and 1998, some individual cross-section studies of the KAPB (Knowledge, Attitudes, Practices and Behaviors) type were made which produced baseline information generating indicators of behaviors and testing and counseling services. In partnership with the SESLNacional, for instance, a population-based study was done covering industrial-sector workers and also young men conscripted into the Brazilian army. These studies made it possible to assess variables associated with condom use, risk perception and sexuality. In the case of the conscripts, the KAPB study was also correlated with serological evaluation, thereby marking the start of second generation surveillance in the country. During the same period, an opinion study was also carried out by Datafolha and CGT among urban workers, in order to evaluate the national campaigns and the knowledge, attitudes and risk perceptions of those workers. These studies provided the baseline for the HIV/AIDS prevention activities in the workplace. Regarding the CSWs and prospectors, two rapid assessments were conducted to ascertain the scope of the situation and of the vulnerability contexts of those populations, focusing on social groups, social mobilization, perceptions of risk and condom use. These studies used the tools derived from the ethnographic method. The methodology employed made intensive use of in- depth interviews, focus groups, document consultation and interviews with key informants. In the same period, a study involving serology and evaluation was made with IDUs, based on harm reduction centers. This study used the capture and recapture method and made it possible to determine the scale of the incidence of HIV among IDUs, as well as to obtain behavioral data in the five sites involved. In addition, the research also made it possible to establish the baseline for the IDU population. As regards the Testing and Counseling Centers (TCCs), the STD/AIDS program also produced, in the same period, a qualitative evaluation study of the actions connected with testing, counseling and the quality of the services. This research served to guide the restructuring of the TCCs and promoted changes in the approach adopted to the most vulnerable groups. The implementation period of AIDS II confirmed the need to carry out evaluation actions and to proceed with the preparation of the National Evaluation Plan, which envisioned formulation of indicators and their linking with the program management process. The intent of these initiatives was to move beyond the fragmentation stage and the residual nature attributed to the evaluative processes of the implementation period of AIDS I. A more integrated process was accordingly set in motion which combined the need to produce knowledge with the generation of data to assist the decision making process. It was in this phase that studies were produced on epidemiological trends and generation of baselines for different segments of the population while at the same time the use of the evaluative processes as management support instruments was consolidated. One of the most important studies in this period was the household survey on sexual behavior of the Brazilian population aged between 16 and 65 years, done by CEBRAP in December 1998 in the 169 microregions of the country. The main lines of action pursued in this period included the studies to evaluate the National STD/AIDS Program's intervention in populations with greater inherent vulnerability, such as CSWs, IDUs and adolescents, which were defined in conjunction with the World Bank in the AIDS II loan agreement. Another significant initiative in this period, within the scope of AIDS II, was the formulation of a set of indicators for monitoring and evaluation of the effectiveness of actions undertaken within the framework of the loan agreement. Also during this period, information technologies were devised for facilitating the monitoring and evaluation of actions to assist and treat persons living with HIV/AIDS, for example SISCEL and SIAIDS [Information System of AIDS Project], while recognizing that these systems present limitations, with a view to fostering their full implementation as evaluation tools. Another step worthy of 63 mention was the establishment of the evaluation program for external quality of laboratory examinations for monitoring the performance of the public health laboratories in connection with testing for HIV and viral load. From the organizational standpoint, the National STD/AIDS Program's structure for the implementation of evaluation actions was located in the Assessoria de Planejamento e AvailiacGo (ASPLAV - Planning and Evaluation Advisory Unit), which has assumed the primary function of reducing the degree of segmentation formerly prevailing in these processes. Notwithstanding the substantial advances achieved, efforts have been concentrated on planning and the search for a sustainable policy that will ensure continuity of the program's actions for dealing with the epidemic in Brazil. Redefinition of the Monitoring and Evaluation Strategy for the National STD/AIDS Program Recognizing the strategic importance of the recommendations made by the various World Bank missions during the implementation of the AIDS I and AIDS II projects, as of the beginning of 2003 the National STD/AIDS Program redefined its strategy for strengthening the monitoring and evaluation (M&E) of its activities--M&E were always present in some form in the program's activities, albeit not on a very systematized basis--with the creation of an Evaluation Area located in ASPLAV. The area team currently consists of two internal consultants and two external consultants, plus one evaluation expert and another expert in epidemiology and data analysis, while it is also able to draw on technical and financial support from the Global Aids Program of the Centers for Disease Control and Prevention (CDC-GAP). In the past year, the Evaluation Area developed the National Evaluation Plan (NEP) for the National STD/AIDS Program, which plan focuses basically on monitoring of impact and results. The strategic thrust of this plan presumes extreme diversity and complexity of the Brazilian program and sets two fundamental premises: (a) the commitment inherent in the institutionalization and decentralization of the evaluative processes, and (b) an integrational approach that renders viable evaluation of the program based on its real incorporation into the country's health system. In March of the present year, the National Public Health School (ENSP/FIOCRUZ), in partnership with the National STD/AIDS Program and CDC-GAP, launched the Postgraduate Course (Specialization and Master's conferring professional qualijkation) in Evaluation of Programs for Controlling Endemic Processes, with emphasis on STD/HIV/AIDS. This course is the chief strategy for decentralization and institutionalization of M&E activities. It aims to train teams for the STD/HIV/AIDS programs of the various levels of government, but assigning priority to the national level and to the technical staff of the five sites of excellence in evaluation3 l Recognition of the National STD/AIDS Program efforts in M&E The National STD/AIDS Program's efforts in establishing the process of institutionalizing and decentralizing M&E activities are already gaining recognition, as for 3 1. Five sites of excellence have been strategically selected that will serve to foster and promote M&E activities in each of Brazil's five regions. These sites are Manaus, the State of Pernambuco, the State of S%o Paulo, Curitiba and Campo Grande. 64 instance in a cableJ2 sent by the U.S. Embassy in Brazil, in which, under the heading "Experts praise, discuss next steps for CDC Aids Monitoring and Evaluation Program," it is stated that: "Throughout the course of the meetings and at its conclusion, the CDC representatives in attendance expressed their optimism about the M&E unit and its training programs, which they agreed showed great potential and represented a necessary evolution in Brazil's HIV/AIDS Control Program." Then the World Bank itself, following its last mission,33 declares that: "The mission was most satisfied with the progress achieved in establishing the monitoring and evaluation system for the program. The team has been strengthened and now there is a unit chief and a statistics expert, and it can draw on support as needed from FIOCRUZ and the collaboration of CDC. These partnerships have proven very fruitful.. ..The National Evaluation Plan has been refined to include, in addition to evaluation of results, the construction of a system for monitoring the establishment of the program and the creation of an evaluation culture. With the collaboration of the CDC the National Plan personnel are being trained in M&E and five centers of excellence are being established in two state and three municipal coordinating entities. . . . The results of the efforts of recent months are already being felt. The NP management units are focusing on results, the Terms of Reference for studies are including the participation of the M&E unit, the findings of the IBOPE survey were used in the design of the last national campaign on testing, the campaigns are starting to be evaluated, and the management units at the national and state level are requesting assistance in preparing monitoring tools."* EQUITY l The NP takes steps to reduce inequality of access to health for the various segments of the population with differing degrees of vulnerability. The challenge of expanding populational coverage to these segments is one of this program's commitments. l The actions to assist persons living with HIV/AIDS and STD carriers promoted by the National STD/AIDS Program are included in the unified public health system (SUS) and the directives and standards that regulate the establishment and implementation of the SUS are observed by means of constant agreements between its three levels of management, which operate under a directive to promote equity. l Access to health services and diagnosis and treatment inputs for these citizens, while not yet uniform throughout all of Brazil's vast national territory, is guaranteed by legal rules and instruments, and the NP is continually supporting the states with a view to expanding this access and the care and assistance network, and using new technologies with the aim of ensuring not just quick and easy access to care but also improving the quality of the care by making it more effective in managing and resolving problems. l The active and intense participation of organized civil society is spurring the NP to continually seek solutions that are better aligned with the SUS and with the needs of its users l Tables E-7 and E-8 of Annex E are not in the report submitted, so a better analysis is not possible. 32. Document generated by the Department of Science and Health and sent to all U.S. embassies and Washington, with a view to sharing successful experiences. 33. Quotation from the Aide Memoire of the Partial Supervision Mission of December 8- 12,2003. * [Translator's note: citation translated without reference to the original]. 65 Some errors noted: Preface, under vii, first paragraph: The amount of the AIDS II agreement.. .-was US$l65.0 million IBRD and US $135 million national counterpart funds, amounting to a total cost of US$300 million. The total funds released by IBRD were US$l61.5 million after deduction of the misprocurement in the AIDS I project in the amount of US$3.5 million from IBRD. The total amount applied in AIDS II was accordingly US$296.5 million. Page 4, under 1.10: In 2000, . . . ,135,OOO are on ART At the end of 200, about 93,500 patients were receiving ART and at the end of 2003 this number had grown to 140,000. Page 5, under 2.4: The number of municipalities in AIDS I was 43 and not 41.