H N P D I S C U S S I O N P A P E R Reaching The Poor Program Paper No. 6 South Africa: Who Goes to the Public Sector for Voluntary HIV/AIDS Counseling and Testing? Michael Thiede, Natasha Palmer, and Sandi Mbatsha October 2004 SOUTH AFRICA: WHO GOES TO THE PUBLIC SECTOR FOR VOLUNTARY HIV/AIDS COUNSELING AND TESTING? Michael Thiede, Natasha Palmer, and Sandi Mbatsha October 2004 Health, Nutrition and Population (HNP) Discussion Paper This series is produced by the Health, Nutrition, and Population Family (HNP) of the World Bank's Human Development Network. The papers in this series aim to provide a vehicle for publishing preliminary and unpolished results on HNP topics to encourage discussion and debate. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations or to members of its Board of Executive Directors or the countries they represent. Citation and the use of material presented in this series should take into account this provisional character. For free copies of papers in this series please contact the individual author(s) whose name appears on the paper. Enquiries about the series and submissions should be made directly to the Managing Editor. Submissions should have been previously reviewed and cleared by the sponsoring department, which will bear the cost of publication. No additional reviews will be undertaken after submission. The sponsoring department and author(s) bear full responsibility for the quality of the technical contents and presentation of material in the series. Since the material will be published as presented, authors should submit an electronic copy in a predefined format (available at www.worldbank.org/hnppublications on the Guide for Authors page). Drafts that do not meet minimum presentational standards may be returned to authors for more work before being accepted. The Managing Editor of the series is Joy de Beyer (jdebeyer@worldbank.org). The Editor in Chief for HNP publications is Alexander S. Preker (apreker@worldbank.org). For information regarding this and other World Bank publications, please contact the HNP Advisory Services at healthpop@worldbank.org (email), 202-473-2256 (telephone), or 202-522-3234 (fax). © 2004 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 All rights reserved. ii Health, Nutrition and Population (HNP) Discussion Paper South Africa: Who Goes to the Public Sector for Voluntary HIV/AIDS Counseling and Testing Michael Thiede a, Natasha Palmer b and Sandi Mbatsha a aHealth Economics Unit, University of Cape Town, South Africa b Health Economics and Financing Programme, London School of Hygiene and Tropical Medicine, London, England Paper prepared for the Program on Reaching the Poor with Effective Health, Nutrition, and Population Services, organized by the World Bank in cooperation with the William and Melinda Gates Foundation and the Governments of the Netherlands and Sweden. Abstract: This is a study of how well public voluntary counseling and testing (VCT) programs for HIV/AIDS reach poor people in township areas of Cape Town, South Africa. The study covered three public clinics, where lay counselors from local nongovernmental organizations provided counseling. A clinic nurse was responsible for testing and notifying patients of the results. Waiting room interviews of 540 patients included questions about the patients' possessions and housing conditions designed to assess economic status. This information was compared with comparable information for people in South Africa as a whole, and in South Africa's urban areas, collected through a large-scale household Demographic and Health Survey. The principal finding was a much higher use of VCT services by lower- than higher-income patients. Almost 75% of VCT patients came from the poorest 40% of South Africa's urban population, fewer than 10% of patients belonging to the urban population's highest 40%. VCT patients were also poorer on average than patients attending the clinics for other reasons. The study also included focus group discussions with residents of the townships where the clinics were located, designed to determine what factors influence use of the clinics for VCT. These suggest that an important reason for the predominance of poor people among clinic patients was the poor reputation of the services provided by the clinics. This led better-off people to seek care from other, more expensive available sources. Keywords: South Africa, Cape Town, health service inequality, HIV/AIDS, voluntary counseling and testing. Disclaimer: The findings, interpretations and conclusions expressed in the paper are entirely those of the authors, and do not represent the views of the World Bank, its Executive Directors, or the countries they represent. Correspondence Details: Michael Thiede Private Bag, Rondebosch 7701, Cape Town South Africa. mthiede@cormack.uct.ac.za Tel: (27-21) 406-6754 Fax: (27-21) 448-8152. iii iv Table of Contents Foreword..........................................................................................................................VII Acknowledgements...........................................................................................................IX HIV/AIDS in South Africa and Voluntary Counseling and Testing .................................. 1 Research Questions............................................................................................................. 1 Methods............................................................................................................................... 2 Nature and Sources of Data ................................................................................................ 3 Findings about the Distribution .......................................................................................... 4 Findings about Reasons for the Distribution ...................................................................... 8 Limitations........................................................................................................................ 11 Implications....................................................................................................................... 11 References......................................................................................................................... 13 List of Figures Figure 1: Township asset scores compared to urban South African DHS quintiles ....... 5 Figure 2: Patient asset scores compared to urban wealth quintiles ...........................6 Figure 3: Patient asset scores compared to township wealth quintiles........................7 v vi FOREWORD This discussion paper is one in a series presenting the initial results of work undertaken through the Reaching the Poor Program, organized by the World Bank in cooperation with the Gates Foundation and the Governments of Sweden and the Netherlands. The Program is an effort to begin finding ways to overcome social and economic disparities in the use of health, nutrition, and population (HNP) services. These disparities have become increasingly well documented in recent years. Thus far, however, there has been only limited effort to move beyond documentation to the action needed to alleviate the problem. The Program seeks to start rectifying this, by taking stock of recent efforts to reach the poor with HNP services. The objective is to determine what has and has not worked in order to guide the design of future efforts. The approach taken has been quantitative, drawing upon and adapting techniques developed over the past thirty years to measure which economic groups benefit most from developing country government expenditures. This discussion paper is one of eighteen case studies commissioned by the Program. The studies were selected by a professional peer review committee from among the approximately 150 applications received in response to an internationally-distributed request for proposals. An earlier version of the paper was presented in a February 2004 global conference organized by the Program; the present version will appear in a volume of Program papers scheduled for publication in 2005, Reaching the Poor with Effective Health, Nutrition, and Population Services: What Works, What Doesn't, and Why. Further information about the Reaching the Poor Program is available at the following sites: Program Overview: http://www1.worldbank.org/prem/poverty/health/rpp/overview.htm List of Papers Commissioned by the Program: http://www1.worldbank.org/prem/poverty/health/rpp/projectlist.htm Presentations at the Program Conference: http://www1.worldbank.org/prem/poverty/health/rpp/conference.htm . vii viii ACKNOWLEDGEMENTS The authors are grateful to the World Bank for publishing this report as an HNP Discussion Paper. ix x HIV/AIDS in South Africa and Voluntary Counseling and Testing HIV/AIDS poses a fundamental threat to global health. South Africa is one of the worst affected countries in the world with an estimated HIV prevalence of 11.4 percent (Shisana and Simbayi 2002). Studies show that socioeconomic status is the principal determinant for exposure to HIV/AIDS, with poverty and social inequalities leading cofactors in HIV transmission. (Gilbert and Walker 2002; Farmer 2001). For both prevention and awareness campaigns, reaching disadvantaged groups is therefore crucial. This chapter examines, and seeks to explore reasons for, any pattern observed in the socioeconomic status of individuals attending public sector clinics and having voluntary counseling and testing (VCT)1 in three township areas of Cape Town, South Africa. VCT for HIV/AIDS, combined with pre- and post-test counseling, has been promoted as a key motivating force for safer sexual behavior (UNAIDS 2002; Magongo et al. 2002). A critical component of any national strategy to limit transmission of HIV (Forsythe et al. 2002), VCT is a prerequisite for effective treatment, care and support services, including programs to reduce mother-to-child transmission, preventive therapy for TB, and administration of antiretrovirals. The evidence is growing that VCT can result in behavior change and improve the coping strategies of people with HIV, with reductions in reported risk behavior (Pronyk et al. 2002). Studies in Uganda (Nuwaha et al. 2002; Matovu et al. 2002) have explored reasons for having VCT. Factors influencing VCT uptake were attitudes toward the consequences of an HIV-positive result, influences from a sexual partner, cost, accessibility, awareness, and perceived risk of HIV infection. The features and quality of a counseling service also influence uptake by different groups. Matovu and colleagues (2002) suggest that features of VCT services that are valued by clients are: confidentiality, regularly available counseling (rather than a "one-off"), the possibility of receiving counseling without testing, nonresident counselors (for greater confidentiality), and counseling outside the health center (at a "neutral site" such as a community center). Research Questions This study attempts to ascertain the socioeconomic status of individuals accessing VCT at public sector clinics in South Africa and reasons for any unusual distribution of uptake. This is important to understand, because VCT will increasingly act as the entry point to a range of support and treatment services for people living with HIV. Any skewing of access to VCT in favor of certain socioeconomic groups will have implications for the equitable delivery of all these other services. 1VCT as part of antenatal care was explicitly excluded because this is part of a broader service. Moreover, the uptake of services within a package focusing on the prevention of mother-to-child transmission is based on a different set of motivations. 1 South Africa's health system offers free primary care at public sector clinics, alongside a thriving private sector. There is evidence that people from all socioeconomic groups use the private sector, believing that both the technical and interpersonal quality of care is superior (Schneider et al. 1999). In periurban settings such as where this research was conducted, a network of public sector primary care clinics offers care free of charge to the uninsured, more than 80 percent of the South African population. In urban South Africa, there is also a competitive market of private general practitioners (Chabikuli et al. 2002) and a growing number of commercial clinic chains (Palmer et al. 2003). VCT is free of charge in the public sector, and capacity to offer the service is slowly being established throughout the Western Cape province, although it is still rudimentary outside of the greater Cape Town area. Methods Three clinics with relatively well-established VCT programs in the greater Cape Town area were selected in consultation with the local Department of Health: · Masiphumelele Clinic is located in the settlement Masiphumelele (population approx. 20,000) that has developed over the last 10 years on the outskirts of Cape Town. · Khayelitsha Site B General Clinic is a big community clinic in Cape Town's largest township (with an estimated population of 500,000 people). · Langa Clinic serves the oldest township community within the Cape Town vicinity (population about 60,000 people). Langa Clinic has implemented an integrated program addressing the dual TB/HIV epidemic in cooperation with the ProTEST initiative. This initiative is coordinated by the World Health Organization/UNAIDS in collaboration with four countries in sub-Saharan Africa. The underlying principle of the approach is that early detection prevents ongoing transmission of TB and slows the progression of the HIV infection. A key feature of the ProTEST initiative is that every patient who comes to a pilot site for TB services receives counseling about TB and HIV and is encouraged to take an HIV test (Godfrey-Faussett et al. 2002). The model for VCT delivery is similar in all three clinics. Counseling is done by lay counselors from local nongovernmental organizations (NGOs). A nurse is responsible for testing and refers patients back to the counselor after informing them of their result. An interview administered in the waiting room of each clinic was used to assess access to VCT services by different socioeconomic groups. In-depth interviews and focus group discussions (FGDs) with staff, clinic users, and local community groups were conducted to explore barriers to access and attitudes to VCT. To establish socioeconomic status of the respondent, the waiting room questionnaire asked a series of brief, closed-ended questions about gender, race, education, employment status, sanitary and living conditions, and household assets. Many of these 2 latter questions were taken from the South African Demographic and Health Survey (DHS) in order to allow comparison with that data set during the analysis phase of the study. Some questions about knowledge around VCT were also asked. The waiting room interview was developed in English and then translated to Xhosa, and back translated to check the quality of the translation. It was administered by a single fieldworker at all three sites. Systematic sampling was used, with every fifth adult (above the age of 14) approached in the waiting room as well as in any specific VCT waiting area and asked to give informed consent to participate in the interviews. Interviews were split into two parts. The first was carried out at this initial phase. The second section was conducted once it had been established whether or not the person being interviewed would have VCT at that visit. For those who did attend VCT, reasons for electing to use the service were explored. For those who did not, more general questions were asked about sources of information about VCT and HIV. Qualitative methods were used for in-depth interviews and community and staff FGDs to explore attitudes to VCT uptake and provision. Sets of detailed interview guidelines were developed for in-depth interviews as well as for FGDs with clinic staff and community groups. These interviews and discussions were conducted both in English and Xhosa by different members of the research team. Participants for 15 extensive in-depth interviews were chosen in part from a subset of patients and in part randomly from the communities. Community focus group discussions were held with groups that were suggested by community members as representative (a community development group in Masiphumelele, a housing project group in Khayelitsha, and a church group in Langa). A further criterion for inclusion was knowledge about services at the local clinic, without the group members' necessarily being patients at the facility. Nature and Sources of Data Findings reported here are based on four sets of data: · Data from the waiting room surveys just described, on socioeconomic status of those individuals (1) attending clinics, (2) having VCT. A total of 540 waiting room interviews were conducted (50 at Masiphumelele, 270 in Khayelitsha, 220 in Langa). After data cleaning, the final sample included 525 patients, 208 of whom had attended the clinics for VCT. · Data from the South African Demographic and Health Survey (DHS) on socioeconomic status of urban households. To enable us to comment on any differences in socioeconomic status between those attending the clinic and having VCT and those within the catchment area of the clinic, a picture of the broader socioeconomic environment of the area was required. This was obtained by generating an asset index from South Africa's 1998 DHS. All urban households from the DHS were used to generate an asset index using principal component 3 analysis (n = 7,752). Household characteristics and assets included in the generation of the index were the particular household's main source of drinking water, the type of toilet facility, flooring and wall material of the dwelling, access to electricity, and a range of household valuables such as a radio, TV, or car. An asset score was thus given to all urban South African households included in the DHS. · Data on the Socioeconomic status of 507 DHS households in townships around Cape Town and Johannesburg. To compare the socioeconomic status of individuals attending township clinics with those living in similar areas, rather than the urban population of South Africa as a whole, a subset of the South African DHS was selected. As the universal sample of households included in the DHS from townships in Greater Cape Town or Johannesburg, 507 households were selected.2 These households formed the group with which we compared the socioeconomic status of individuals attending the clinics and having VCT. This comparison group was ranked into quintiles based on the asset scores generated from the whole DHS urban sample. The socioeconomic status of this study's 525 individuals attending services (including VCT) was then compared to the reference population of 507 township households. · Qualitative information from in-depth interviews and focus groups discussions. FGDs and in-depth interviews were taped, transcribed, and analyzed by two authors independently for key themes, drawing on fieldwork diaries as an additional source of information. Findings about the Distribution Of 525 people interviewed, 208 were attending the clinics for VCT. The findings compare the socioeconomic characteristics of three groups: · A universal sample of the 507 Cape Town and Johannesburg township households in the DHS (the reference population) · A systematic sample 525 people attending three clinics for any service · The 208 people of the 525 who were attending for VCT. Figure 1 illustrates how township households in the metropolitan areas of Johannesburg and Cape Town (n = 507) fit within socioeconomic quintiles of the sample of all urban households in the South African DHS. While the number of township households falling into the top quintile of all South African urban households is low at 8.1 percent, it is still higher than would be expected for this type of periurban setting. One explanation for this lies in the broad range of socioeconomic backgrounds and income levels that make up the top quintile in South Africa, a country with a highly unequal income distribution that is reflected by a Gini index of 59.3 (World Bank 2001). South Africa's apartheid history meant that black South Africans were restricted in where they could live. Therefore, townships today include areas of widely varying wealth. 2Johannesburg townships in the DHS, similar to those in Greater Cape Town periurban areas, were included to increase the sample size of reference households. 4 Figure 1: Township asset scores compared to urban South African DHS quintiles 35% 30% 28.8% 25% 25.4% 23.7% 20% Population 15% 14.0% Township 10% 8.1% % 5% 0% Low est Low asset Medium High asset Highest asset score score asset score score asset score DHS Quintiles (Urban Population) Sources: Demographic and Health Survey South Africa 1998. The percentage of people falling into the lowest wealth quintile is also lower than might be expected. Whereas the township areas in the two metropolitan areas under investigation are of key importance for the study of social and economic transformation and relative deprivation in South Africa, they do not represent the economically worst-off metropolitan neighborhoods in the country. A number of urban localities in other parts of South Africa are significantly more deprived. As can be expected, most of the township population falls into the middle three wealth quintiles with a peak at 28.8 percent in the central quintile. A first finding is therefore that the area in which this study was conducted is not worse- off overall than the South African urban population in terms of the relative distribution of household assets. A comparison taking into account both urban and rural households would reflect the fact that township households are on average considerably better-off than households in rural areas (Booysen 2002). The systematic sample of patients from three township clinics around Cape Town reveals a clear pattern when analyzed in terms of its asset scores. Figure 2 assigns the patient sample to South African urban wealth quintiles. The upper two wealth quintiles are underrepresented both among the patients visiting the clinics for general health services and people coming to the clinics for VCT. Only 1.9 percent of patients who came for general services and 1.0 percent of those who had VCT fell into the top quintile, 8.8 percent and 7.2 percent respectively into the second highest quintile. Within both subsamples the second lowest quintile was best represented (38.5 percent of general patients, 38.9 percent of VCT users). 5 Figure 2: Patient asset scores compared to urban wealth quintiles 50% 40% 38.9% 38.5% 35.6% s serU 30% 26.2% General Services icevreS 24.6% VCT Services 20% 17.3% % 10% 8.8%7.2% 1.9%1.0% 0% Lowest Low Medium High Highest asset asset asset asset asset score score score score score Urban Quintiles Source: Waiting room interviews and Demographic and Health Survey South Africa 1998. The most obvious differences between those attending these clinics for general services and those having VCT emerge in the lowest and the central quintiles. Nearly a quarter of general patients fall into the medium wealth quintile (24.6 percent), while the percentage of VCT users from this category is substantially lower (17.3 percent). However, there is a corresponding discrepancy at the lower end of the wealth scale: 26.2 percent of general patients are from the lowest wealth quintile, whereas more than a third of VCT users (35.6) represent this socioeconomic group. There is a significant relationship between township households' socioeconomic status and public sector VCT uptake (chi-squared for trend: 6.713; P = 0.00957). This result requires further investigation, given such pronounced differences among socioeconomic groups in utilization patterns of general and VCT services. The service-utilization pattern becomes more intriguing when regarded within the socioeconomic spectrum of the township environment. For the purpose of this analysis, the factor scores generated for each asset on the basis of the urban DHS sample were used to define the wealth quintiles for this subsample of township households from the DHS. Health services utilization across the wealth quintiles generated at the level of townships reveals a pattern in which public service utilization decreases with increasing wealth (figure 3). While this is true for both general health services and VCT services offered at the township clinics, the pattern is more pronounced for VCT services (chi-squared for trend: 4.802; P = 0.02843). Between general clinic utilization and the uptake of VCT, representation of service users from both top quintiles is low: 4.4 percent of general patients and 4.8 percent of VCT patients fall into the top quintile, while 9.8 percent and 9.6 percent, respectively, can be assigned to the second highest quintile. Between the service categories, notable differences exist in the socioeconomic structure of patients 6 from the lower three wealth categories. The most remarkable result is that more than half of the VCT patients (54.8 percent) fall into the lowest township wealth quintile. These results are unexpected because the services offered are targeted at the whole socioeconomic spectrum of public sector users. Figure 3: Patient asset scores compared to township wealth quintiles 60% 54.8% 50% 40% 38.8% Users General Services ce 30% 28.4% rvieS VCT Services 20.2% %20% 18.6% 10% 10.6% 9.8% 9.6% 4.4% 4.8% 0% Lowest Low asset Medium High asset Highest asset score score asset score score asset score Township Quintiles Source: Waiting room interviews and Demographic and Health Survey South Africa 1998. Overall, the findings show that socioeconomic groups are not evenly represented in clinic attendance, and this pattern is even more exaggerated when it comes to undergoing VCT. The least well-off quintiles among the township population take up the services more than the others. The waiting room survey gathered a wide range of data about people's knowledge around VCT and the reasons for VCT uptake. Of the multiple reasons people gave for having a test, the leading motives were TB (explicitly mentioned by 30.0 percent of patients) and sexually transmitted diseases (explicitly mentioned by 28.0 percent of people). In both cases, nearly all the patients had been referred by a health worker. People also mention a range of general symptoms of illness ("... because I am very sick," coughing, chest problems, loss of appetite, and similar complaints). Most of the younger age groups come for testing as a consequence of sexually transmitted diseases or merely wanting to know their status. In the older groups, TB emerges as the key reason for testing. Clinic health workers clearly play an important role in conveying information about HIV prevention services available at the clinic. Of those interviewed, 95 percent had received information from clinic health workers. The role of information transmitted via radio and TV also becomes apparent. The impact of TV information even reaches beyond those who live in a household that owns a TV. Posters at clinics in the study setting have reached more patients (63.7 percent of the sample) than campaigns that have tried to transmit information with the help of brochures and leaflets (only 14.8 percent of the 7 sample). Community health workers have reached 42.8 percent of the people interviewed at the clinics. Findings about Reasons for the Distribution The quantitative analysis above suggests that certain groups within the community are more likely to access public sector clinics and VCT. To complement these findings and to explore reasons that may explain them, FGDs and in-depth interviews were analyzed. Any discussion about voluntary counseling and testing is dominated by fears surrounding HIV/AIDS. The fear of an incurable disease is paired with the perceived risk of expulsion from the family or rejection by a partner. Lack of access to treatment was also expressed as a reason not to test: "[If I tested positive,] ... I would just feel like I am already dead because there is no cure for this." Woman, Khayelitsha "What makes people not come [to VCT services] is their background. Sometimes you get that their family do not accept a positive person. They see her as if she is someone who was misbehaving outside and then got positive. One is afraid to tell her husband because she is worried that the husband will divorce her. A mother who is not working is afraid of being left with the children to feed. ... Some people think when you touch them you going to make them positive." Woman, Langa "Positive people are not welcome in the family." Woman, Langa "People have a fear of knowing, they also say `Why must we test if the government does not treat us?' " Woman, Langa In staff focus groups, counselors observed that these general fears lead many people to delay attending VCT unless they experience symptoms that might be AIDS-related. A big impediment people face in their communities is the stigma that is associated with HIV/AIDS, leading people to shun anything that is HIV/AIDS-related. Resistance to outreach programs by counselors was expressed: "If [the clinic counselors] would go outside to the community, it would be worse. People do not want that counselors be seen who come to their door. It is better if the counselors stay there, so they can counsel those who go to the clinic." Woman, Khayelitsha "We also don't want [the counselors] really to come to our places." Man, Khayelitsha 8 "HIV/AIDS workshop is not very good because that name is scary." Man, Khayelitsha For the same reasons, people criticize the lack of anonymity in the clinics. "We are a small community. If you are seen [in the waiting area for VCT], there is some question mark above you. So people don't want to be seen there. They don't want to go local. HIV is often portrayed as misbehavior." Woman, Langa "There's a problem with the clinic: The room where counseling takes place. Everyone in the waiting room can see it and from the public entrance. If you go in there everybody knows what you are there for. There's the stigma." Woman, Langa Closely linked to the issue of stigma is the problem of confidentiality. People fear that their interaction with nurses and counselors is not confidential. Community members may travel to a clinic in a different part of town to have an HIV test. Some interviewees in the communities stated that they prefer to go to a private general practitioner for testing for reasons of confidentiality. "... if my counselor is my neighbor, I think that maybe she can tell people about me and my status. Therefore I decide to go and do the test in Salt River and not here in Langa, you understand?" Woman, Langa "... sometimes people they go to another clinic. We fear each other." Woman, Langa "We don't have any confidentiality here... for confidentiality we go to Wynberg." Woman, Khayelitsha "False Bay is not safe anymore because the health workers do go there and come back and tell, so the only place I see is Fishhoek clinic, and Wynberg. People prefer going to places far away from here." Woman, Masiphumelele Breaches of confidentiality reported in the communities may not actually reflect reality. Rumors about what happens at the clinics echo people's fears. "There is a particular chair in the clinic that people know if you are seen sitting in it, you know that is for people who tested positive." Woman, Masiphumelele Overall problems to do with waiting times and rudeness and favoritism of health workers were similar to those raised about primary care services generally in South Africa (Edgington et al. 2002). In a number of cases, a lack of trust toward health workers is clearly expressed, although this often appears to be based on expectation of what would happen or the experience of others rather than individuals' own experience: 9 "I don't trust anyone. Because I just hear from the TV that they treat some people with needles that have been used by an HIV person." Woman, Khayelitsha "The people at the clinic do not have a nice way of dealing with the issues in a sensitive way. ...[tells a story of someone else]...the health worker visited the house and shouted at her." Woman, Masiphumelele "I don't know if this test is voluntary." Woman, Khayelitsha "These health workers do not make people confident because they turn around the folders and check the status of the people." Woman, Masiphumelele The level of distrust seems to be more an expression of uneasiness and anxiety associated with HIV/AIDS than grounded in actual negative experiences with clinic staff and VCT services at the respective facilities. People interviewed who revealed their HIV-positive status generally expressed a different experience than others expected and reported kindness, support, and confidentiality from the clinic staff. "They treat them [those with HIV] good and advise us how to behave." Woman, Langa "The staff here care very much." HIV-positive woman, Langa Issues around information and the type of promotion of public VCT services were also discussed in the focus groups. Several participants mentioned the radio, leaflets at the clinic, and LoveLife, a media campaign promoting reproductive health. "The billboards are good but they must not be put on the freeway where you cannot see them." Woman, Langa "What about those who cannot read? They are people like us. How are they going to get this information?" Woman, Khayelitsha The most marginalized parts of the townships, informal settlements or squatter camps, were spoken of as those least likely to access services. People here were seen as high-risk groups for HIV but more inclined to call on traditional healers. Reasons mentioned for the lack of interaction between squatters and public primary health services are economic (people spend their time looking for money and food), cultural, language, and education related. "They in the informal [settlement] there, they suffer." Women, Khayelitsha "[HIV awareness should]... include people from the rural areas, who cannot read and also include our culture, the culture of the African people. According to our culture the elders are not allowed to talk with their children about sex." Man, Khayelitsha 10 Limitations As highlighted in the previous section, doing research into issues around HIV/AIDS is a highly sensitive task. Both the reluctance of people to be interviewed in depth and limitations on what they were willing to talk about in interviews hindered data collection. Due to the difficulties of recruiting people for interviews or focus groups, qualitative data are not drawn from a wide community-based sample but from specific groups. These groups' views may systematically differ from other members of the community. The second set of limitations concerns the use of the South African DHS for the comparison data. The approach had two drawbacks. Because the DHS was conducted in 1998, the data are relatively old, and asset ownership patterns within townships in South Africa may have changed. However, the only clear development in asset ownership affecting households across the board seems to relate to the possession of telephones due to the increase in the spread of mobile phones. The second weakness of the DHS data is the extent to which the asset ownership questions used in the DHS are appropriate for South African urban and periurban settings. Given the relatively high prevalence of many "basic" assets, a refined set of questions would probably have yielded superior results. Future work could benefit from focusing on more appropriate wealth indices for specific settings. The study design was highly focused and reveals a number of areas that would require further exploration. The study examined only relatively well-resourced township clinics in the public sector and thus highlights the need to know more about both what is happening in more rural areas and where else people might be going for VCT. Equally, our facility-based study design did not allow us to comment on the rate of uptake or exclusion from the service in the community as a whole. Larger, more comprehensive (and hence more expensive) research designs would be required to shed light on these important questions. Implications The population of townships in the Western Cape is relatively affluent compared to the rest of the South African DHS. From this group, public sector users are from the poorest quintiles. Furthermore, public sector users for VCT are still more concentrated in the poorest quintiles. Reasons for this are suggested by the qualitative findings, which consistently represent a perception that public sector VCT is deficient in confidentiality, a key quality dimension for this service. Health policymakers face a series of dilemmas related to access to services. The first is whether services reach the right population and in sufficient quantity. Related to this is the question of whether services are perceived to be of adequate quality and accessibility. Findings from this study suggest that VCT in public sector clinics reaches poor groups, 11 but this may be happening as a result of negative attitudes toward the service rather than positive ones. It is unlikely that the relatively wealthier groups choose not to have VCT at all, because studies from other countries show that uptake of VCT is positively associated with socioeconomic status and education (Kowalczyk et al. 2002). However, we do not know where individuals who do not use the public sector for VCT are going. Evidence abounds that many urban South Africans go to the private sector for general PHC services and that, for conditions of greater sensitivity, this is likely to be a very high percentage. For instance, more than 50 percent of patients with sexually transmitted infections attend the services of private general practitioners (Rispel et al. 1995; Wilkinson et al. 1998). We can therefore hypothesize that some groups, especially wealthier ones, go to the private sector and others may travel out of the area for VCT in more anonymous settings in the public sector. In the light of high private sector utilization throughout the world, the perception of public sector health services as being of poor quality is a problem of increasing concern for those who can afford no alternative. This undermines many areas of health service delivery in the public sector and causes impoverishment of vulnerable households that pay for private services instead of using the public sector. In this study, while findings suggested a progressive incidence of uptake of VCT, a more positive finding would have been to see an even distribution of VCT use across all socioeconomic quintiles, including the poorer groups. VCT recipients who choose not to use the public sector incur costs of either travel and time or private sector fees to access a service of tremendous public health importance that is being provided free. Measures to address this could include improving perceived confidentiality of services at public clinics, both by physical modifications such as changing waiting areas and room allocations, and by training health workers and helping to cultivate trust toward health workers at the clinics. This means both that the physical environment at the clinics must be designed in a way that assures confidentiality and that health workers and counselors must place higher value on patients' privacy. However, the reputation of services may keep many people away rather than actual shortcomings in the services offered. The problem may be one of suspicion and perception as much as of real breaches of confidentiality. This points to the need to address people's perceptions as well as features of the service. If many people are using the private sector, and are likely to continue doing so in the near future, the possibility of a voucher system for VCT could be explored if the private sector could offer good quality counseling and testing that is of greater acceptability and can be monitored for quality. Last, the study findings suggest a clear agenda for future research in this area. It is important to begin to understand the volume of exclusion from VCT in various service settings and areas. Research at community level into uptake of VCT and choice of provider (with costs incurred) would be one next step in advancing knowledge in this area. This should be coupled with more in-depth qualitative work to understand key barriers to access around VCT for different vulnerable groups. 12 References Booysen, F le R (2003). Measuring poverty with data from the Demographic and Health Survey (DHS): An application to South Africa. Journal for Studies in Economics and Econometrics, 26 (3), 53-70. Chabikuli N, Schneider H, Blaauw D, Zwi A, Brugha R. Quality and equity of private sector care for sexually transmitted diseases in South Africa. Health Policy and Planning, 17:40-46. Edgington M, Sekatane C and Goldstein S (2002). Patients' beliefs: do they affect tuberculosis control? A study in a rural district of South Africa. International Journal of Tuberculosis and Lung Disease, 6(12), 1075-1082. Farmer P (2001). Community based approaches to HIV treatment in resource-poor settings. Lancet, 358, 404-409. Forsythe S, Arthur G, Ngatia G, Mutemi R, Odhiambo J and Gilks C (2002). Assessing the cost and willingness to pay for voluntary HIV counseling and testing in Kenya. Health Policy and Planning, 17, 187-195. Gilbert L and Walker L (2002). Treading the path of least resistance: HIV/AIDS and social inequalities- a South African case study. Social Science and Medicine, 54, 1093- 1110. Godfrey-Faussett P, Maher D, Mukadi Y, Nunn P, Perriens J and Raviglione M (2002). How human immunodeficiency virus voluntary testing can contribute to TB control. WHO Bulletin, 80, 939-945. Kowalczyk J, Jolly P, Karita E, Nibarere J, Vyankandondera J and Salihu H (2002). Voluntary counseling and testing for HIV among pregnant women presenting in labor in Kigali, Rwanda. JAIDS, 31, 408-415. Magongo B, Magwaza S, Mathambo V and Makhanya N (2002). National report on the assessment of the public sector's voluntary counseling and testing programme Pretoria: Health Systems Trust. Matovu J, Kigozi G, Nalugoda F, Wabwire-Mangen F and Gray R (2002). The Rakai project counseling programme experience. Tropical Medicine and International Health, 7, 1064-1067. Nuwaha F, Kabatesi D, Muganwa M and Whalen C (2002). Factors influencing acceptability of voluntary counseling and testing in Bushenyi district of Uganda. East African Medical Journal, 79, 626-632. 13 Palmer N, Mills A, Gilson L, Wadee H and Schneider H (2003). A new face for private providers in developing countries: what implications for public health? WHO Bulletin, 81, 292-297. Pronyk P, Kim J, Makhubele M, Hargreaves J, Mohlala R and Hausler H (2002). Introduction of voluntary counseling and rapid testing for HIV in rural South Africa: from theory to practice. AIDS Care, 14, 859-865. Rispel, L., Beattie, A., Xaba, M., Cabral, J., Marawa, N., and Fonn, S. (1995). A Description and Evaluation of Primary Health Care Services Delivered by the Alexandra Health Centre and University Clinic Johannesburg: Centre for Health Policy. Schneider H, Blaauw D, Magongo B and Khumalo I (1999). STD care in the private sector. In: South African Health Review. Durban: Health Systems Trust, 83-94. Shisana O and Simbayi L (eds) (2002) Nelson Mandela/HSRC study of HIV/AIDS: South African national HIV prevalence, behavioural risks and mass media: household survey 2002. Cape Town: Human Sciences Research Council. UNAIDS (2002). HIV Voluntary Counseling and Testing: a gateway to prevention and care Geneva: UNAIDS. Wilkinson, D., Connolly, A., Harrison, A., Lurie, M., and Abdool Karim, A. (1998). Sexually transmitted disease syndromes in rural South Africa: Results from health facility surveillance. Sexually Transmitted Diseases, 25, 20-25. World Bank (2001). World Development Report 2000/01: Attacking poverty Oxford, UK: Oxford University Press. 14 About this series... This series is produced by the Health, Nutrition, and Population Family (HNP) of the World Bank's Human Development Network. The papers in this series aim to provide a vehicle for publishing preliminary and unpolished results on HNP topics to encourage discussion and debate. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations or to members of its Board of Executive Directors or the countries they represent. Citation and the use of material presented in this series should take into account this provisional character. For free copies of papers in this series please contact the individual authors whose name appears on the paper. Enquiries about the series and submissions should be made directly to the Managing Editor Joy de Beyer (jdebeyer@worldbank.org) or HNP Advisory Service (healthpop@worldbank.org, tel 202 473-2256, fax 202 522-3234). For more information, see also www.worldbank.org/ hnppublications. THE WORLD BANK 1818 H Street, NW Washington, DC USA 20433 Telephone: 202 477 1234 Facsimile: 202 477 6391 Internet: www.worldbank.org E-mail: feedback@worldbank.org