244 Human Development November 2004 Findings reports on ongoing operational, economic, and sector work carried out by the World Bank and its member governments in the Africa Region. It is published periodically by the Knowledge and Learning Center on behalf of the Region. The views expressed in Findings are those of the author/s and should not be attributed to the World Bank Group. IMPORTANT NOTICE: Effective November 2004, Findings and InfoBriefs will only be available electronically. Register your e-mail address at www.worldbank.org/afr/ findings to continue receiving the electronic version. Learning by Doing: Uganda's AIDS Control Project Empowers Local Managers by Joseph J. Valadez and Peter Nsubuga Surveillance systems in Uganda other. Diverse geographical, cul- detect that HIV prevalence de- tural or logistical conditions con- clined from 21.1 percent in 1991 tribute to this variation. Also, to 6.4 percent in 2001. The most some areas may have had HIV/ common explanations for this de- AIDS programs for several years, crease are that the population while in other areas programs are mobilized itself with the conse- recently established. Due to these quence that more people were substantial differences, local man- faithful to their partners, or ab- agers should be in a better posi- stained from sexual contact, and tion to make decisions about how used condoms during sexual inter- to make tactical changes to their course (Low-Beer et al 2003). Al- programs than administrators sta- though one might debate which of tioned in a national capital. The these behavior changes contrib- people who are responsible on a uted most to the apparent reduc- daily basis for providing services tion in HIV prevalence, no one are in the best position to analyze would claim that Uganda can now the challenges in their areas and become complacent about its HIV/ to decide on the tactical changes. AIDS programs. Quite the contrary. National HIV/AIDS Committees What is Community Learning? continue to have the responsibil- ity for both covering their popula- During 2003 the Global HIV/AIDS tions with the highest quality pre- Monitoring and Evaluation Support vention, care, support, and treat- Team (GAMET)--a unit of the World ment programs possible, and to Bank's Global HIV/AIDS Program-- improve them constantly. launched its Community Learning Findings While this mandate is clear, the activities to equip local managers process that managers can use to with tools to rapidly assess their continually enhance programs is program interventions and identify not well understood. What contrib- priority areas for improvement. utes to this challenge is that pro- GAMET offered training in tools gram quality can vary substantially that are suitable to be used at the from one area of a nation to an- local level by service providers, re- mission (UAC), with support from GAMET, was implemented to as- sess the current status of its pro- grams. By September 30 districts were identified for assistance. For the initial assessment, UACP iden- tified 19 of these districts to participate. By mid-October all the District HIV/AIDS Committees (DHACs) had been contacted and agreed to participate. They recruited Civil Society Organizations in each three more workshops. The county who identified at least 2 "Learning by Doing" approach was quire a small amount of time to people to be trained in LQAS. The being successful and teams were collect data, and produce informa- DHACs used the national census becoming increasingly empowered tion that is reliable and easy to in- to identify villages targeted for HIV/ to use Community Learning ap- terpret. GAMET promotes a "Learn- AIDS program support. The UACP proaches. ing By Doing" approach to program developed a set of six short ques- management. Such an approach is tionnaires to collect information Steering programs by making particularly needed to improve useful to program management, tactical change HIV/AIDS programs due to the im- and translated them into six local Once the county teams analyzed mense variation in conditions that languages. The six questionnaires their data, the information was exits in a nation. By empowering were developed to survey small aggregated to display the condition local managers to guide their own samples of: orphans, mothers of of the program within each of the programs, the assumption is that infants, youths, sexually active districts that participated in the solutions can be tailored to specific men and women, and people living LQAS assessment. The LQAS as- local conditions. with HIV/AIDS. The surveys were sessment sampled villages some of Central to the Community intended to provide information for which were targeted for HIV/AIDS Learning tool kit is a method called key categories of HIV/AIDS Pro- programs and others that were not Lot Quality Assurance Sampling. grams including: Prevention of targeted. This stratification per- LQAS is an old method, having Mother to Child Transmission, Vol- mitted comparisons. However, Fig- been developed in the 1920s untary Testing and Counseling, ure 1 only displays mothers living (Dodge and Romig 1944) as an in- Home-Care, Prevention, Behavior in villages targeted by their district dustrial quality control method to Change, Care and Support of for HIV/AIDS programs. The figure assess industrial batch production people affected and or infected with is known as a radar-chart. The dis- (Reinke 1988, Valadez 1986 and HIV. tricts are arranged as though they 1991, Valadez et al 2002 and 2003 The data collection in most cases were on the face of a compass. At Robertson et al 1997). During the took 5-days. A few locations re- the top of the figure are districts mid-1980s it was adapted to man- quired an additional day due to in the center of the country. Mov- age public health programs in de- problems of rain and difficult roads. ing clockwise one finds districts to veloping countries. After that, the teams returned to the training venue in Mukono Dis- the East, Southeast, South, South- The Ugandan experience trict to hand tabulate their data. west, West, Northwest, and North One week later they presented pre- of Uganda. The figure is very re- During September 2003 the Ugan- liminary findings to their own vealing. Firstly, the black line dan AIDS Control Project (UACP) on DHACs in their home districts. shows that large proportions of behalf of the Uganda AIDS Com- Soon after, the UACP organized women throughout Uganda report they visited a clinic at least once However, few pregnant women, during their pregnancy for Antena- except those living in the urban tal Care (ANC). This result suggests areas of Kampala and Wakiso are that women do have access to accepting to take an HIV test health facilities. However, the blue (white line). This result reminds line, which depicts the proportion us once again that knowledge of a of women delivering babies in risk does not necessarily mean health facilities, varies consider- that people will adopt behavior that ably from one part of the country to reduces the risk. The yellow line another. In the center and east- shows where women are coun- ern districts larger percentages of seled during their ANC to take an woman deliver in health facilities. HIV test. Once again it is clear that Understanding problems But as one moves towards the west, except in the urban areas, coun- smaller proportions of woman re- seling is not associated with Another component of Community port they delivered their babies in women taking a test. This is an- Learning is a diagnostic phase in health facilities. Notice how the other case in which local manag- which the UACP and the districts line for the western and northern ers need to consult systematically try to understand underlying rea- districts attenuates, moving closer with women in their areas to bet- sons for the problems they detect. to the center of the radar chart. A ter understand the barrier that Let us consider the above-men- result such as this one suggest make HIV testing less acceptable tioned result, namely, that preg- that managers in the western and to them. nant women are not taking an HIV northern portions of the county A final example concerns people test despite their knowledge of need to understand the barriers living with HIV/AIDS (PLWHA) (Fig- MTCT and counseling during ANC. constraining women to deliver in ure 3). The data are aggregated at The Uganda AIDS Control Project facilities. Possibly, community out- a national level. The LQAS data selected 9 districts throughout the reach programs by clinically show that 78.3 percent of PLWA country in which to carry out fo- trained providers would increase were ill during the last month and cus group discussions of mothers. the proportion of woman delivering 94.5 percent of those who were ill As of this writing 5 districts had with someone trained to provide an sought medical care. This is an been visited--information from all antiretroviral as well as emer- very positive evidence that treat- visits indicate similar conclusions. gency obstetric care if needed. This ment component of the Home Care Firstly, mothers seldom agree dur- is an interesting result since it is program is functioning as PLWHA ing one ANC visit alone to take a obvious women have sufficient ac- are counseled to see medical sup- test. Several were required. This cess to attend ANC. Local manag- port whenever they are ill. Despite meant that more effective promo- ers are in the best position to in- this positive sign the data also re- tion of regular ANC visits was vestigate and propose tactics to veal portions of the program that needed. Secondly, women reported address this problem. need improvement. The data show that they needed social support. Another example is also associ- that 41 percent of PLWHA are sexu- They were very concerned that if ated with a relatively new program: ally active. However, as the figure they were HIV positive they would namely, voluntary testing and reveals only 51 percent of men and have to disclose their status to counseling during ANC. Figure 2 48 percent of women always used their husbands. This, they feared, has several interesting character- a condom. Twenty-four percent of would result in them losing their istics. Firstly, districts in each por- men and 22 percent of women re- marriage. They worry primarily for tion of the country exhibit both ported they never used a condom. their unborn child; should the high and low levels of knowledge These data show that not only are mother be cast adrift, then the that HIV can be transmitted dur- some PLWHA placing others at risk child's health and quality of life ing delivery (Mother To Child but they are also exposing them- would be placed in jeopardy. Moth- Transmission, MTCT) (blue line). selves to re-infection. ers recommended that husbands be included in ANC counseling so that can be built upon. Lessons are Change in Reducing HIV: Is that they participate in both coun- derived from local people working Uganda Unique?" African Journal seling and testing simultaneously together to solve their own prob- of AIDS Research, 2(1), 2003: 9- with the women. This finding is lems while being provided with the 21. being studied in more detail in the support they need. If local manag- Robertson, S.E., M. Anker, A.J. districts and discussed locally. A ers can do this, we consider the pro- Roisin, N. Macklai, and K. similar finding resulted from inter- gram to be on the road to success. Engstrom. "The Lot Quality Tech- views of PLWHA. They revealed that nique: A Global Review of Appli- they did not use a condom so as This article is reproduced from cations in the Assessment of not to disclose to their partners they Development Outreach the World Health Services and Diseases were HIV+. With a regular partner Bank Institute, July 2004 and Surveillance," World Health Sta- they said it was not possible to use authored by Joseph J. Valadez, tistical Quarterly; 50, 1997:199- condoms unless they revealed PhD, MPH, ScD, Senior Monitoring 209. their status. The next phase will and Evaluation Specialist, Global Valadez, J.J. Lot Quality Accep- be a strategic thinking workshop HIV/AIDS Monitoring and Evalua- tance Sampling for Monitoring in which the district teams will tion Support Team in the Global Primary Health Care Coverage. come together and share their in- HIV/AIDS Program, The World Office of Health Systems. Pan terpretations of the results. Bank; and Peter Nsubuga, MD, American Health Organization Project Coordinator, Uganda AIDS (WHO), Report No. HSS-SNIS-29, Asuccessstory? Control Project, Kampala, Uganda. 1986. It may be tempting to look at The authors gratefully acknowl- Valadez, J.J. Assessing Child Sur- Uganda as a success story; how- edge the essential contribution of vival Programs in Developing Coun- ever, it may be a disservice to us David Kaweesa Kisitu (Evaluation tries: Testing Lot Quality Assurance all to do this. At this stage of the Officer, for the Uganda AIDS Control Sampling. Harvard University epidemic no one really knows what Project, Kampala) and Joy Mukaire Press, 1991. the propitious program model is (GAMET Consultant based in Valadez, J.J. and B.R. Devkota. "De- that results in a decline in HIV Kampala). Without their dedicated centralized Supervision of Com- prevalence. If we did, it would be efforts the LQAS activities in Uganda munity Health Program Using replicated globally. The point of this could not have been carried out. Also LQAS in Two Districts of South- article is that for now we define a their comments on early versions of ern Nepal," JRaJ Wyon, ed., Com- success story as one in which man- this paper were extremely helpful. munity-Based Health Care: Les- agers are empowered to scrutinize sons from Bangladesh to Boston. their programs and courageously References Management Sciences for make tactical changes in an at- Health, 2002: 26. tempt to address their own condi- Dodge, H.F. and H.G. Romig. Sam- Valadez, J.J., B. Weiss, C. Leburg, tions. The more managers are pling Inspection Tables: Single and and R. Davis. Assessing Commu- empowered to do just this, the Double Sampling. (second ed.), nity Health Programs: Using LQAS greater the likelihood that we will John Wiley & Sons, 1944. for Baseline Surveys and Regular be able to gather a set of lessons Low-Beer, D. and R. Stoneburner. Monitoring. Teaching-aids at Low "Behavior and Communication Cost, 2003.