Nachega, Jean B MD, MPH*; Knowlton, Amy R ScD*; Deluca, Andrea MHS*; Schoeman, Jan H MA; Watkinson, Linda MS†; Efron, Anne MSN, MPH†; Chaisson, Richard E MD†; Maartens, Gary MBChB, FCP‡
South Africa has one of the worst HIV/AIDS epidemics in the world. It is estimated that more than 1700 people are infected each day and that there are a total of 5.4 million South Africans infected.1 A recent population-wide survey estimates that the prevalence of HIV in the general population is 11.4%.2 A 2004 antenatal clinic survey indicates that HIV prevalence among pregnant women is 29.5%.3 Projections suggest that without comprehensive treatment interventions, the mortality rate attributable to HIV/AIDS is likely to reach 800,000 deaths per year by 2010.4 Increasing access to highly active antiretroviral therapy (HAART), which has been shown to decrease morbidity and mortality, could alter this trend substantially.5-7
Recent international initiatives such as the US President's Emergency Plan for AIDS Relief (PEPFAR)8 and the United Nations Global Fund to Fight AIDS, Tuberculosis, and Malaria9 as well as other national and international initiatives have dramatically increased access to HAART in South Africa. Implementation of the operational plan for comprehensive HIV and AIDS care, management, and treatment,10 which provides HAART free of charge through the public sector, started in first quarter of the 2004. By June 2005, there were 138,000 patients (including children) on HAART, but an estimated 866,000 patients aged 0 to 49 years are still in need of treatment.11 In addition, the following psychosocial factors form part of the “treatment readiness” recommendations for patients who are to initiate HAART: (1) attend at least 3 preparation visits to an HIV clinic, (2) demonstrate no active alcohol or other substance abuse, (3) have no untreated active depression, (4) have disclosed their HIV status to at least 1 friend or family member or have joined an HIV support group, (5) have accepted their HIV-positive status and have insight into the consequences of HIV infection and the role of antiretroviral therapy before initiating HAART.
Although these recommendations are logical and practical, few context-specific models of behavioral interventions have been developed to promote HAART adherence among populations in limited-resource settings. Therefore, exploring these psychosocial variables before designing innovative approaches to behavioral interventions is needed for this population so as to promote adherence interventions successfully and enable patients to enjoy the full health benefits of HAART.
Prior research has shown that peer support can influence patients' preventative health behavior and complement the services of health care providers. Indeed, peer advocacy models have been highly successful in HIV prevention interventions.12,13 Research studies specific to southern Africa have documented comparable results for peer approaches to intervention. In Malawi, the presence of peer educators was found to reduce risky behavior in commercial sex workers.14 In Mozambique, peers acting as health educators in a prison setting significantly increased knowledge about HIV in the intervention population.15 Community-based peer group HIV prevention programs were also used in Botswana, and the study results indicated that such groups were a sustainable and inexpensive way to influence knowledge and attitudes about HIV.16
There are few comparable studies about medication adherence health behavior. Research on the characteristics of the health advocates and patients, their interpersonal relationships, and the role of those relationships in HAART adherence would contribute greatly to an understanding of effective adherence intervention models for this population. Informal caregiving and social influence theories suggest that taking advantage of existing supportive relationships may lead to sustained positive changes in adherence behavior.17 It is therefore essential to determine what effective support means for those on HAART and to understand the barriers, material and social, that decrease adherence.
In this study, we aimed to investigate patient and provider perceptions about support strategies that would encourage the highest rates of HAART adherence.
We conducted an exploratory qualitative pilot study in the Western Cape Province of South Africa from July 1 to August 30, 2004. The data were collected using focus group discussions of HIV-infected patients and key informant interviews with health providers. The goal was to gather information from patients (insiders) and health care providers (outsiders) involved in HAART programs to determine the characteristics of effective treatment support and to learn more about social and material barriers to HAART adherence.
Study Setting and Participants
We enrolled participants (patients and health care providers) working at or attending HIV primary care settings: 3 clinics in Khayelitsha and 1 clinic in Gugulethu. Nationally and internationally funded HIV treatment projects had been offering HAART to patients free of charge in these medically underserved periurban townships of the city of Cape Town for more than 1.5 years before the government-funded rollout of the HIV treatment program. Additional study participants were recruited from the G. F. Jooste HIV clinic, which is a secondary-level health care facility located within the township of Manenberg that has a high antenatal HIV prevalence of 25%. This is one of the official government HIV treatment rollout sites located in the Western Cape Province, and it began providing HAART free of charge to eligible patients in December 2003.
Criteria for eligibility in the study focus groups were HIV infection and age of 18 years or older. Patients were identified and then given an appointment to attend the focus group sessions. Participants in the focus group were given $5 to compensate them for their time and to cover their transport expenses. Health care providers with at least 5 years of experience in community-based management of HIV treatment programs were eligible to participate in the key informant interviews. All study participants signed written informed consent forms in their preferred language.
Focus Group Discussions
We conducted 2 focus groups with 6 HIV-infected adults (N = 12), each of which was facilitated by a research nurse and a scribe-note taker fluent in the Xhosa and Afrikaans languages. Patients were randomly selected by going to the HIV outpatient clinic on random days and asking for volunteer participants, who were selected regardless of disease severity, being on HAART or not, and length of treatment experience on HAART. Each focus group lasted 90 minutes, and the printed reports were prepared from the field notes immediately after the sessions. Permission was not given by all participants to audiotape-record the sessions. Questions for focus group participants covered topics related to norms and practices of health support networks, health-seeking behaviors, medical and social problems encountered, experiences taking HAART, and barriers to HAART adherence.
Individual in-depth interviews involved health care providers and were conducted in English by a South African social scientist and a physician trained in qualitative interviewing techniques. The interviews were audiotape-recorded after permission was obtained from participants. The main interview question was: “What support do you think people on HAART need?” Each interview session was conducted in a private room at the HAART clinic or at a convenient place for the study participant (business or home office) and lasted approximately 1 hour.
All in-depth interview discussions were transcribed verbatim from audiotapes. Field notes were translated and transcribed for focus groups. Transcripts were reviewed for completeness and accuracy by reviewing audiotapes where applicable. The transcripts were used for content analysis to identify main themes and were then coded for retrieval and analysis. Selected quotations were used to illustrate the themes.
This study was approved by the University of Cape Town Research Ethics Committee and the Johns Hopkins Medicine Institutional Review Board.
Description of Study Participants
Focus group discussion participants were almost all female (11 of 12 participants) and aged 22 to 42 years. All were black African and spoke Xhosa or Afrikaans as their main language. Almost all were heterosexual (11 of 12 partipants); 1 participant reported being homosexual. All 12 reported taking HAART, with treatment duration ranging from 3 to 24 months.
Seven key informants were interviewed, including 3 women and 4 men. All were health care professionals: 5 doctors working in community-based primary-level HAART clinics within the G. F. Jooste HIV Clinic catchment area, 1 nurse-social scientist with extensive experience in developing community-based tuberculosis (TB) treatment adherence models, and 1 nurse-community HAART adherence counselor.
The data collected from the key informants and from the focus groups indicate what participants identified as the critical forms and sources of support that facilitate adherence and the social and material barriers that they believe can impede adherence.
Based on participant reports, a model emerged of a patient's treatment support gradually shifting from clinic to home/community sources over time and from a focus on survival and physical health needs to the building of a social support network.
Short-Term Support: Focus on the Physical Condition
During the first 3 months of treatment, health providers said that they thought it was critical to describe to patients the treatment details and emphasize the need for excellent drug adherence. These health professionals all emphasized the importance of patients getting support from clinicians. This was supported by patients' reports. The support they received after diagnosis was important in inspiring them to take their HAART, and this short-term support came primarily from the clinic staff, such as treatment adherence counselors and social workers. Patients in the focus group said that education about illness management and disease progression was crucial after diagnosis if a person with HIV was going to reach high adherence.
The following comments reflect these perceptions:
▪ “Short term is a kind of physical support, they are usually not too well…they get confused [and] a substantial number, I would say about 10 percent, get worse when you start ARV, and ‘worse’ means confused, weak.” (medical professional A)
▪ “There are two different issues regarding support: support for patients who need to go on ARV and support for people on ARV…the key thing is information-education and ongoing counseling-this is very important…so that people understand what these drugs do. If they [understand] that, you are half way there.” (medical professional B)
▪ “The [medical and social] support that I got from my doctor but mainly by [my] treatment counselor and social worker helped me to not give up and keep taking my tablet when I was so weak that I thought I am going to die…” (female focus group discussion participant)
Long-Term Support: Focus on Empowered Living
After the initial clinically focused support phase, health care professionals described movement into a maintenance phase of empowered living, which they defined as having the necessary motivation, support, and knowledge to sustain excellent adherence. Once the patient reaches a point of high adherence to HAART and the physical condition improves, patients and health care professionals identified the need for further psychologic support. This type of support was explored in these interviews by discussing the topic of treatment supporters defined as family members or friend(s) within the community who provide them support.
▪ “If clients have real good support, they adhere to the treatment. When they don't have enough support, they don't do really that good. Because [from] 0 to 3 months, it is no problem. They [are] still sick; they take their medication, no problem.” (medical professional B)
Patients, however, had mixed feelings about the need for intensive support at this point.
▪ “My treatment supporter is overwhelming me…I needed him when I was very sick…but now I am feeling well and I need him to back off a little bit…I need him to give me some space so I can be in charge with my life again.” (female focus group discussion participant)
The key informants in our study recognized that getting a patient beyond the first 3 months depends on how successfully the patient can shift his or her psychologic outlook from an initial obsession with survival, through clinical support, to a stage of empowered living, which is sustained primarily through social support.
Long-term support should include a discussion of the importance of continuing adherence even when the patient becomes healthier and counseling about renewed sexual activity (issues of safer sex and other sexually transmitted diseases) and issues surrounding family planning. Key informants suggested that treatment supporters can help in this process.
▪ “[There are] issues we don't talk about, because they feel well, they become sexual beings again…both me and my partner are on [ARV] drugs, our viral loads are suppressed. Do we still need to use condoms? So we [health care professionals] ultimately ignore the central component of our patients…so the [treatment supporters]…say to them, fine, today we are going to talk about [sexual activity]. We are going to talk about safe pregnancy. We are going to talk about safe drinking. We are going to talk about welfare.” (medical professional C)
Key Characteristics of Treatment Supporters
To identify potential treatment supporters, we asked focus group participants to identify social network members who had been supportive after their (the patients') disclosure of HIV infection and who could potentially assist them with their HAART adherence outside the clinic setting. Several participants identified health care professionals, whereas others cited social workers, church members, and family members as primary sources of support.
In addition, key informants were asked to identify the characteristics of an effective treatment supporter, “that is, a person who would observe treatment or support the patient so that he or she would take treatment on time, every day.” Proximity to the patient was identified as a critical characteristic by the key informants.
▪ “It is easy for them to see they didn't take, because they've got the pill boxes they put where the [treatment supporter] should see…for us [health care professionals], they can put the pills in the toilet and come [to the clinic] with empty pill boxes and say I am taking them all…they [treatment supporters] will encourage them, they will pick up when they're forgetting, because they are there, they will have a look because the pill boxes have days on them.” (medical professional C)
▪ “I choose her as my treatment supporter because she is the only friend that I trust and she lives across the street…she helps me not to forget taking my tablets by reminding me almost every day if I took my tablet at time.” (female focus group discussion participant)
The key informants also said that an effective treatment supporter must be involved on an emotional level and have compassion and the willingness to listen. Having authority was the most salient characteristic of an effective treatment supporter, however, as identified by key informants.
▪ “And I think they [treatment supporters] automatically have moral authority, they understand better what the other's doing and because they have more knowledge usually…yes, moral in a sense it gives someone the authority to tell me although it is my life and I do whatever I want and anyway I am going to die, and what the **** if I don't take those pills. There is someone who's got a right to tell me something and I will listen to them.” (female social scientist)
▪ …they [treatment supporters] are not a doctor or nurse, they support, they make sure they take their pills, when [the patient is] confused, they count the pills for them. And they give them to drink and make sure all the other…moral authority, moral authority- that was the word I looked for." (medical professional B)
Key informants identified family members (excluding partners in most instances) as being most likely to be chosen as treatment supporters, as illustrated in the following comments:
▪ “[A sexual partner] is not the first one to disclose to-no, definitely not. Well, you know the setting here, the sexual partner is not a constant one and it's an understatement and not a very prominent person. The family, certainly for women, the mother and the family are more important than the partner.” (medical professional A)
▪ “First, it's the mother [as the treatment supporter]; second, the sisters; third, the brothers; father comes maybe the last one, then it goes to the partner, sexual partner.” (medical professional A)
▪ “So, you need to have a, usually, a family link, or at least a generation gap or distance so that the one will listen to the other…there is this physical…family is the first choice because physical proximity is more important. If they live in the same house, usually if it is friends of family, it is extremely important because people were around all the time, so they can really support very closely.” (medical professional A)
One woman from the focus group identified having a family member who helps her to take her medication on time. Other focus group participants identified family members as supporters who help them to get medical care or medication. These responses identified family support as being predominantly financial or material (eg, money for food) or instrumental (eg, giving a ride to the clinic). Several focus group participants also identified family members as providing moral support by accompanying them to clinic visits, visiting them when they were hospitalized, or helping them by reminding them to take their HIV medication on time.
▪ “I choose my mom as my treatment supporter because she will not discriminate me against others and she will be understanding…she is the person I wanted to talk to the first time that I learned my HIV status…you need someone you trust, who can listen to you and who will not reject you…” (female focus group participant)
▪ “My girlfriend is the only person I disclosed to, and she helped to cope with disease and to keep taking my HIV medicine at the right schedules…” (male focus group discussion participant)
▪ “My brother helps me in taking my medication by calling me through my mobile phone on the times my medication is due, even by just having my mobile phone to show the message: please call me.” (female focus group discussion participant)
One woman from the focus group identified having a friend in the community who is also infected with HIV as source of mutual support, and they help each other not to forget to take their HIV medicines on time.
Social Barriers to Disclosure and Stigma
Disclosure of HIV serostatus was identified by several key informants as an important step in mobilizing effective adherence support. As a health care professional stated:
▪ “Because the minute you disclose, it means you have accepted your illness. And it means if you accepted, you are ready to commit yourself to lifelong therapy. If there is no disclosure, there will be problems somewhere that need to be sorted out. Then the [treatment counselor] must dig deeper-why [is there] no disclosure to at least 1 person in the family who will be your treatment supporter?” (medical professional B)
During the focus groups, however, participants expressed their reluctance toward HIV status disclosure. In one of the focus groups, 4 of the 6 participants provided the same reason for not disclosing, which was that family members who knew a patient's HIV status “would stand in the street and shout [the patient's HIV status] when drunk.” This was echoed by one of the key informants in the following comment:
▪ “We had clients who say, ‘I can't tell my mother because she drinks. And when she is drunk, she is going to scorn me. And she is going to spread rumors about me.’ So if they cannot disclose to members of the family, they cannot be seen taking their medication. Obviously, they've got to hide to take their medication…and that is going to impact negatively on the adherence.” (medical professional C)
▪ “Of course it is much better to disclosure to him so one has not to tip-toe around when taking treatment to avoid questions such as ‘what are you taking medications for?’ But I am not prepared to disclose to him since doing so may harm our relationship and he can dump me…” (female focus group participant)
Another patient told the story that when she disclosed her HIV serostatus, she lost the support of her family members. She thought her mother would provide financial support, along with emotional support; instead, her mother rejected her and has turned her siblings against her.
Money, Food, and Other Needs
We found that all participants identified social support and material needs that, if unattended to, could be barriers to HAART adherence. The material needs included food, transport, pill boxes, and monetary support. Other support needs included treatment for depression and alcohol abuse and assistance with resolving family conflicts.
▪ “On an individual level, treatment adherence sounds very easy. But if you talk about a primary health care system trying to do these things, it takes a very carefully designed protocol of care around adherence…how do you do that when you're resource-strapped, you have poor organizational skills in the health system, and when we are dealing with communities who are in some ways under-resourced both around understanding of the disease and access to resources in the system.” (female social scientist)
▪ “But let's focus on the nutritional support. [It is] very hard for people to take pills when they are hungry. And the challenge is how do you link that person to a support mechanism that would ensure that there is some food on that person's table at least once a day?” (medical professional C)
Several focus group members mentioned that employed family members often provided the necessary financial support for adherence, including bus fare or rides to clinics, healthy foods to eat, and clothes to keep warm during the cold season. Because nearly half of all South Africans are unemployed, government social grants are also a major concern for the focus group participants. All participants have to reapply annually for the grant, which is approximately US $110 per month and is often the only source of income for HIV-positive persons.
Treatment Supporter and Directly Observed Therapy
Two key informants, a medical professional and an adherence counselor, reported the use of directly observed therapy (DOT) strategies of adherence support using family members or friends of patients as treatment supporters.
▪ “…we always encourage [treatment supporters] to observe when [the patients are] taking [the medications]…the treatment supporter is requested-if person must take at 8:00 AM, can you go and look at 9:00 AM if s/he has taken the tablets? Or, in the evening, can you before you go to bed go and find out if still not there? But we always encourage them to do it within the hour.” (female HAART adherence counselor)
▪ “…we do it [DOT] when we definitely have possibility of a family member, or a very supportive neighbor, to do the job. Neighbor is, of course, not so good as family members. But family, if you can create this bonding, this common responsibility with family members, by far, is the best results.” (medical professional A)
The previous comments highlight the importance of several forms of HAART adherence support at the time of treatment initiation, when the patient is still physically weak. Furthermore, they show the importance of having a treatment support network with a trusted family or community member to whom the patient has been able to disclose and having financial support, including government disability grants, for facilitating HAART adherence.
These results suggest that adherence interventions should initially use maximum input from the health care system to re-establish the general physical condition of the patient, followed by a gradual shift to emphasize family/community support for the patient. At the initiation of antiretroviral therapy, patients often have opportunistic infections and an expectation of a shortened life span. The first 3 months of treatment can bring about marked physical improvements such as observable weight gain, fewer opportunistic infections, and overall improvement in the quality of life. The key informants in our study, however, recognized that getting a patient beyond the first 3 months depends on how successfully the patient can shift his or her psychologic outlook from an initial obsession with survival through clinical support to a stage of empowered living sustained primarily through social support.
In our study, patient participants emphasized the important role of social support not only in helping them to take their medication but in helping them to cope psychologically with the disease. Major concerns were expressed about disclosure of HIV serostatus, which can potentially pose a barrier to peer-assisted adherence models. This model requires that support be provided on all levels: by the family, by counselors, and by the wider social network, including those in the health care realm and in community organizations such as churches.
Peer-driven intervention models have been used to facilitate this type of “treatment supporter” role, with a different set of definitions according to the study setting.18 According to the theory of group-mediated social control, people who interact daily with the patient, including family members, friends, and neighbors, must be considered in any intervention because of interdependency.19 In our own study, this aspect of interdependency within existing social networks was featured in the nominations of desired characteristics of a potential treatment supporter.
Our study suggests that an effective treatment supporter should have “moral authority” with the patient. This concept of moral authority was reported by patients as well as by key informants and is defined here as “the person to whom patients give the power to influence their decision making about their health.” This characteristic is critical and seems to play a large role in a patient's decision to nominate someone as a treatment supporter. As one key informant suggested, this position of authority is more highly regarded than other channels of communication (eg, health care professionals); hence, an opportunity is created for treatment facilitation. This treatment supporter should be included in education sessions so as to overcome barriers created by misinformation. It is also important to note that moral authority might be considered a primary factor in choosing a treatment supporter, and our results suggest that this is a culturally relevant norm seen as more influential than medical professionals and should be regarded as one of the most important channels of communication and influence when designing treatment adherence programs.
A major social barrier to attaining excellent adherence is the stigma associated with HIV/AIDS in South Africa, because the fear of stigma prevents disclosure to family members and loved ones who could potentially fill the role of treatment supporter. Most of the patients in our focus groups refused to disclose to family members out of fear that the particular family member would become intoxicated with alcohol and publicly announce the patient's HIV status. This fear highlights the enduring stigma surrounding HIV/AIDS despite continual educational campaigns in the media and the availability of treatment. It also highlights the problem of alcohol abuse on a community-wide scale. Thus, a community with high levels of stigma and discrimination against people living with HIV/AIDS is likely to exert a negative effect on HAART adherence. In a prior cross-sectional study in Soweto, South Africa, we found that the odds of obtaining a >95% level of adherence decreased considerably with an increased fear of stigmatization (rejection, violence, or both) by the patient's sexual partner.20 Indeed, fear of stigmatization may cause patients to hide their antiretroviral drugs or skip doses when others are present. Thus, medical adherence support interventions should make disclosure to sexual partners a point of primary concern and encouragement even if a family member is chosen as a treatment supporter. Interventions that encourage disclosure to a sexual partner should also address the consequences that patients may face in disclosing, especially for women in regard to potential violence and abandonment.21 Most focus group participants and key informants identified the mother as the most effective supporter, followed by siblings, friends (HIV-infected or not), and, lastly, the sexual partner. Partners were also seen as less likely candidates for support because this population believes that having a relationship with a stable partner is somewhat rare. Those in stable relationships saw their partners as important sources of support, however.
In addition, our results suggest that basic needs must be met to facilitate high adherence. These needs range from basic financial and material resources, such as money for nutritional food, to emotional, instrumental, and informational support. Our results also suggest the need for services that address more complex societal issues such as depression, alcohol abuse, and family conflict. Because government grants are identified as a need by our focus group participants, a possible measure to promote excellent adherence would be to train treatment supporters to assist patients in seeking this assistance. This could be readily accomplished by providing the necessary referral information to the patient and treatment supporters.
DOT HAART is a treatment strategy modeled after an effective method for controlling TB. By having someone who has frequent contact with the patient, actually observing him or her taking the dose of medicine each day, DOT HAART is proposed to minimize nonadherence, prevent drug resistance, and stem HIV transmission.22 As highlighted by key informants, our results suggest not only that DOT HAART may be effective and feasible using family or friends as peer treatment supporters but that DOT HAART programs based outside the home, school, or work may turn out to be culturally acceptable, relatively inexpensive, and logistically feasible. In South Africa and Malawi, TB DOT programs that use a patient-nominated or community-based treatment supporter (family or community member) have been shown to be effective and feasible.23-25 In addition, the community-based DOT HAART model piloted in rural Haiti uses community health workers to deliver AIDS and TB services.26 In that situation, the authors conclude that DOT HAART is effective, reduces mortality, lessens AIDS-related stigma (because of the dramatic results of therapy), improves medical staff morale, boosts interest in HIV testing and counseling, and thus contributes to prevention. DOT HAART using a treatment supporter, however, has not been proved to be an independent causal component in outcomes, because this may be confounded by the constellation of services provided by programs in which this strategy is but a single component. Thus, there is an urgent need to validate the treatment supporter strategy as an independent tool for enhancing adherence. This requires controlled randomized trials of enhanced DOT HAART using treatment supporter versus standard of care.
Although informative, our study has several limitations. Given the fact that it was conducted in the Western Cape Province of South Africa with English-, Afrikaans-, and/or Xhosa-speaking participants only and with data collected mostly in a periurban township setting, our results may not be applicable to the whole of South Africa or sub-Saharan Africa. Similar studies exploring the feasibility of such peer-driven HAART adherence interventions in other demographic groups and in rural settings are needed.
In summary, our results suggest that the characteristics of treatment supporters and various relationship factors may influence a patient's adherence to antiretroviral therapy. Intervention approaches addressing these potential influences within existing patient-supporter ties may be a promising avenue to effecting behavior change in resource-limited settings.
The authors thank Drs. E. J. Mills, J. Dick, and S. Maman and M. Roche for critical reading of the manuscript and the patients and health care workers who participated in this research.
2. Connolly C, Shishana O, Colvin M, et al. Epidemiology of HIV in South Africa-results of a national, community-based survey. S Afr Med J. 2004;94:776-781.
4. Dorrington R, Bourne D, Bradshaw D, et al. HIV/AIDS data in South Africa. Lancet. 2002;360:1177.
5. Palella FJ, Delaney KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med. 1998;351:853-860.
6. Mocroft A, Vella S, Benfield TL, et al. Changing patterns of mortality across Europe in patients infected with HIV-1. EuroSIDA Study Group. Lancet. 1998;352:1725-1730.
7. Moore RD, Chaisson RE. Natural history of HIV infection in the era of combination antiretroviral therapy. AIDS. 1999;13:1933-1942.
12. Barbacci M, Repke JT, Chaisson RE. Routine prenatal screening for HIV infection. Lancet. 1991;337:709-711.
13. Fogarty L, Heilig C, Armstrong K, et al. Long-term effectiveness of a peer-based intervention to promote condom and contraception use among HIV-positive and at-risk women. Public Health Rep. 2002;116:103-119.
14. Walden VM, Mwangulube K, Mankhumola-Nkhoma P. Measuring the impact of a behaviour change intervention for commercial sex workers and their potential clients in Malawi. Health Educ Res. 1999;14:545-554.
15. Vaz R, Gloyd S, Trindade R. The effects of peer education on STD and AIDS knowledge among prisoners in Mozambique. Int J STD AIDS. 1996;1:51-54.
16. Norr K, Norr J, McElmurry B, et al. Impact of peer group education on HIV prevention among women in Botswana. Health Care Women Int. 2004;25:210-226.
17. Knowlton AR. Informal HIV caregiving in a vulnerable population: toward a network resource framework. Soc Sci Med. 2003;56:1307-1320.
18. Broadhead R, Heckathorn DD, Altice F, et al. Increasing drug users' adherence to HIV treatment: results of a peer-driven intervention feasibility study. Soc Sci Med. 2002;55:235-246.
19. Heckathorn DD. Collective sanctions and compliance with prescribed medications: a formal theory of group-mediated social control. Am Sociol Rev. 1990;55:366-384.
20. Nachega J, Stein D, Lehman D, et al. Adherence to antiretroviral therapy in HIV-infected adults in Soweto, South Africa. AIDS Res Hum Retroviruses 2004;20:1053-1056.
21. Maman S, Mbwambo J, Hogan N, et al. Women's barriers to HIV-1 testing and disclosure: challenges for HIV-1 voluntary counseling and testing. AIDS Care. 2001;5:595-603.
22. Lucas GM, Flexner CW, Moore RD. Directly administered antiretroviral therapy in the treatment of HIV infection: benefit or burden? AIDS Patient Care STDS. 2002;16:527-535.
23. Clarke M, Dick J, Zwarenstein M, et al. Lay health worker intervention with choice of DOT superior to standard TB care for farm dwellers in South Africa: a cluster randomized control trial. Int J Tuberc Lung Dis. 2005;9:673-679.
24. Manders AJ, Banerjee A, van den Borne HW, et al. Can guardians supervise TB treatment as well as health workers? A study on adherence during the intensive phase. Int J Tuberc Lung Dis. 2001;9:838-842.
25. Wilkinson D, Davies GR, Connolly C. Directly observed therapy for tuberculosis in rural South Africa, 1991 through 1994. Am J Public Health. 1996;86:1094-1097.
26. Farmer P, Leandre F, Mukherjee JS, et al. Community-based approaches to HIV treatment in resource-poor settings. Lancet. 2001;358:404-409.
© 2006 Lippincott Williams & Wilkins, Inc.