Share this article on:

Differences in normal activities, job performance and symptom prevalence between patients not yet on antiretroviral therapy and patients initiating therapy in South Africa

Rosen, Sydneya; Ketlhapile, Mpefeb; Sanne, Ianb; DeSilva, Mary Bachmana

doi: 10.1097/01.aids.0000327634.92844.91
Section III: Antiretroviral Treatment

Objectives: Little is known about how antiretroviral therapy (ART) affects patients' economic activities in resource-constrained settings. This study examined the association of ART with functional impairment, symptom prevalence, and employment during the first 6 months on therapy.

Methods: Interviews were conducted with 453 patients receiving pre-ART care and 427 patients on ART for 1–6 months in South Africa.

Results: Pre-ART subjects were almost twice as likely to report being functionally impaired (unable to perform primary normal activity) in the previous 5-day work week [adjusted odds ratio (OR) 1.97; 95% confidence interval (CI) 1.46–2.66], with a mean of 1.61 impaired days for pre-ART subjects versus 0.87 days for ART subjects (P < 0.0001). Pre-ART subjects were more likely to report fatigue (OR 2.84; 95% CI 2.10–3.84), pain (OR 2.06; 95% CI 1.53–2.76), and nausea (OR 1.61; 95% CI 1.13–2.28) in the previous week and feeling unwell physically (OR 1.71; 95% CI 1.27–2.32), feeling depressed (OR 1.42; 95% CI 1.04–1.95) or resting (OR 1.52; 95% CI 1.12–2.07) on the previous day. The prevalence of pain, fatigue, skin problems, and resting exceeded 40% in both groups. Employed subjects on ART for 3–6 months reported 3 days less health-related absenteeism per month than those on ART less than 3 months (P = 0.0353).

Conclusion: Patients who have initiated ART reported significantly less functional impairment, fewer symptoms, and better work performance than those who had not yet started treatment. The prevalence of some impairment, symptoms, and absenteeism remained high, however.

From the aCenter for International Health and Development, Boston University, Boston, Massachusetts, USA

bHealth Economics Research Office, Wits Health Consortium, University of the Witwatersrand, Johannesburg, South Africa.

Correspondence to Sydney Rosen, Health Economics Research Office, Helen Joseph Hospital (Themba Lethu Wing), Perth Road, Westdene 2092, South Africa. E-mail:

Back to Top | Article Outline


In late 2003, the South African National Department of Health announced its Operational Plan for Comprehensive HIV and AIDS Care, Management and Treatment for South Africa [1]. The plan included provisions for both antiretroviral therapy (ART) for medically eligible patients and pre-ART care for those not yet eligible. By September 2006, an estimated 235 000 adult South Africans were receiving ART from public healthcare facilities, along with another 80 000 patients on ART at private and non-governmental clinics [2].

The effectiveness of ART in suppressing viral replication and restoring immune function in north American and European patients with HIV/AIDS is well documented [3,4]. An increasing body of evidence from local populations confirms that ART is achieving similarly good biomedical results in South Africa [5]. At one Cape Town clinic, for example, well over 90% of patients initiating ART over a 3-year period had undetectable viral loads at the end of the period [6].

What is less certain is the impact of ART on patients' social and economic wellbeing. Before starting treatment, South African adults with advanced HIV/AIDS disease report serious constraints in their ability to carry out their normal activities [7]. Once they have started ART, can patients return to work, care for their families, perform their normal daily activities, and maintain an acceptable level of wellbeing? With hundreds of thousands of working-aged adults dependent on a lifelong therapy, the answers to these questions will be an important determinant of the long-term success of treatment programmes in South Africa.

The few studies that have examined economic outcomes of ART in African populations report generally encouraging results. In a rural population in Kenya, the probability of an HIV-positive adult participating in any economic activities increased by 20% in the first 6 months on ART, and the number of hours worked rose by 35% [8]. Several studies have also reported reductions in worker absenteeism after ART initiation in various industries and countries [9–12]. Improvements have also been reported in relevant domains of quality of life studies. Patients receiving ART in rural Uganda reported large improvements in physical and mental health summary scores on the Medical Outcomes Study HIV Health Survey (MOS-HIV) within 3 months of starting treatment, with a more gradual improvement over the subsequent 9 months [13]. Similarly, the proportion of a cohort of patients receiving ART in Cape Town, South Africa that reported no problems in the mobility, self-care, usual activities, pain/discomfort, and anxiety/depression domains of the EQ-5D instrument increased steadily over the first 12 months of ART, with most improvement occurring almost immediately after treatment initiation [14]. Finally, differences in physical and emotional health based on EQ-5D scores between patients who had not yet started ART and those on ART for less than 6 months were also reported by public sector patients during clinic visits in the Free State Province of South Africa [15,16].

In this paper, we add to the limited literature cited above with a report of the baseline findings of a prospective cohort study on self-reported differences in functional impairment, symptom prevalence, and employment status between South African adults who have started ART and those not yet on ART at public and non-governmental organization sector clinics.

Back to Top | Article Outline


The study was conducted at three sites in South Africa. The first site, the Themba Lethu Clinic of Helen Joseph Hospital in Gauteng Province, is a large, urban, public referral hospital that had nearly 5000 adult patients on ART as of September 2006. The other two sites are non-governmental clinics. The Witkoppen Health and Welfare Centre is a full service, primary care clinic serving informal peri-urban settlements in Gauteng Province. The ACTS Clinic is a dedicated HIV/AIDS clinic in a rural setting in Mpumalanga Province. As of September 2006, these two sites had approximately 450 and 700 patients on ART, respectively. The public hospital is funded primarily by the provincial government, with some supplemental donor support; the two non-governmental organization clinics are funded primarily by donors, with additional support from patient fees and government.

All three sites follow South African national guidelines for providing HIV/AIDS care and treatment. Under these guidelines, ART is initiated once a patient's CD4 cell count falls to less than 200 cells/μl, or WHO stage IV disease is diagnosed [17]. Approximately two-thirds of patients are initiated on the national first-line regimen of stavudine, lamivudine, and efavirenz; for most of the rest, efavirenz is replaced with either nevirapine or lopinavir/ritonavir. Patients not yet eligible for treatment are monitored and receive counselling and vitamin supplements. Whereas most patients at all three sites are South African, migrants from neighbouring countries are also eligible for care. Medications and laboratory tests are free of charge to patients at all sites; the only direct cost of treatment to patients is the clinic visit fee, which is often waived for those who cannot afford it [18].

At each site, study subjects were identified each day from patients present at the clinic, either waiting in a queue or attending an adherence or wellness class or support group. We selected subjects from these groups (queue, class, or support group) using nth-name sampling and invited them to participate in the study. Adult patients who were HIV-positive but not yet medically eligible for ART and adult patients who had initiated ART less than 6 months before recruitment were eligible for enrollment. Those who provided written consent were enrolled in the cohort and administered a baseline questionnaire by a study interviewer. Questionnaires were administered during routine clinic visits, usually while the subject was waiting in line to see a doctor or pick up medications. Patients who appeared acutely ill during their visit or were otherwise unable to complete the interview were not approached and were replaced by the next patient on the list. Consent documents and questions were translated verbally into local languages by the interviewers as requested by individual subjects.

The questionnaire was designed for this study and focuses on subjects' economic activities. It contains questions taken from the South African census [19], the MOS-HIV [20], and the World Health Organization Health and Performance Questionnaire [21], as well as questions suggested by our and others' previous research on HIV/AIDS and labour productivity. We report here the baseline findings about subjects' level of physical impairment, health condition, employment, and ability to perform their normal daily activities. The costs to subjects of obtaining treatment have been reported in a previous paper [18].

Medical records of subjects who had already started ART at the time of the baseline interview were also reviewed. Data collected included dates of clinic visits, CD4 cell count(s), viral load, date of ART initiation, ART regimen, and other medications prescribed. Subjects who had been on ART for less than one month at the time of the baseline interview were excluded from the analysis presented here, as they were neither ‘pre-ART’ nor solidly ‘on ART’ when baseline data were collected. The medical records of pre-ART subjects were not consistently maintained by the study sites and so were not available for this analysis.

For the statistical analysis of differences between subject groups (e.g. pre-ART and ART), bivariate differences in categorical variables were first assessed using Mantel–Haenszel χ2 tests; differences in continuous variables were assessed using t-tests. Then multivariate logistic regression models were estimated, using the various measures of impairment (e.g. bodily pain) as binary outcomes, in separate models with antiretroviral status as the primary predictor (1–6 months on ART as reference) and adjusting for those characteristics with significant associations with antiretroviral status in bivariate analysis (site, sex, housing type, employment status, and probability of having disclosed HIV status to others), as well as age, marital status, and education, which were retained because of demographic importance. Results are expressed as odds ratios (OR) with 95% confidence intervals (CI). Finally, for the subanalyses of only those patients on ART, time on ART (1–3 months versus ≥3 months, with ≥3 months as reference) and starting CD4 cell count (<100 cells/μl and ≥100 cells/μl, with ≥100 cells/μl as reference) were used as primary predictors of impairment, again in separate models. All data were analysed using SAS software version 9.1 (SAS Institute, Cary, North Carolina, USA). The study was approved by the Institutional Review Board of the Boston University Medical Campus and the Human Research Ethics Committee (Medical) of the University of the Witwatersrand.

Back to Top | Article Outline


Study recruitment is illustrated in Fig. 1. A total of 1072 subjects were enrolled in the cohort and interviewed. Of the 107 who declined to enroll, 37% cited lack of time to participate in the interview; others wanted to think more about it, said they were not interested in research, or gave no reason. Most of the 128 ineligible patients were found to have been on ART for more than 6 months at the time of recruitment. Of the enrolled subjects, 192 had been on ART for less than one month and were excluded from this analysis, leaving a total of 880 subjects.

Fig. 1

Fig. 1

The characteristics of the sample and bivariate associations with ART status at the time of the baseline interview are shown in Table 1. In the sample as a whole, most subjects were women between the ages of 25 and 44 years. The high proportion of female subjects largely reflects the patient populations of the study sites and of South Africa more generally, where a large majority of ART recipients are women [22]. A majority of the subjects reported being married to one spouse or having one long-term partner. Almost all were literate, but only a third had completed secondary school. Just under half the sample reported income-generating employment, either in the formal or informal sector. Most of the others were either unemployed and seeking work or engaged in unpaid housework or childcare.

Table 1

Table 1

There were no differences between ART and pre-ART subjects in age, marital status, or education level. Slightly fewer subjects on ART were women. Subjects on ART were more likely to live in modern houses, reflecting the weighting of the sample towards the urban hospital site. ART subjects were also somewhat more likely to report their main activity as seeking work and less likely to work in the informal sector or be engaged primarily in unpaid housework or family care.

The primary activities of study subjects are shown in Table 1. Table 2 presents multivariate results for self-reported functional impairment in the previous 5-day work week according to ART status, with functional impairment defined as the inability to perform the primary activities indicated in Table 1 and subjects on ART for 1–6 months as the reference group. This definition of functional impairment is intended to capture all of the different primary activities that study subjects report, including the large number of subjects who are not formally employed, which includes those who are unemployed and seeking work, informally employed, engaged in unpaid housework or family care, or engaged in other activities, such as study or leisure.

Table 2

Table 2

Pre-ART subjects were nearly twice as likely as ART subjects to have suffered any impairment in the previous week (OR 1.97; 95% CI 1.46–2.66). As a result, when the entire sample is considered, the average number of days of impairment in the previous week is also nearly twice as high (1.61 versus 0.87 days, P < 0.0001). When the analysis is limited only to those reporting any impairment in the previous week, however, the average number of days impaired is large for both groups– more than half the previous 5-day work week– but does not differ significantly between the groups. Women in both groups were more likely than men to report any impairment (data not shown), with no difference by treatment status.

Table 2 also shows that pre-ART subjects were more likely to report most symptoms of ill health in the previous week or previous day than ART subjects, with OR ranging from 1.4 to 2.8. There was no difference in the prevalence of skin problems, which were reported by nearly half of both groups. The prevalence of pain and fatigue was also high in both groups. Half of the pre-ART subjects and a third of the ART subjects reported that they did not feel well on the day before the interview.

In the unadjusted analysis (not shown), pre-ART subjects were significantly more likely to be employed at the time of the baseline interview than were ART subjects (40 versus 28%). This difference disappeared in the adjusted analysis, as shown in Table 2. A total of 138 subjects reported having lost their jobs (voluntarily or involuntarily) in the 12 months preceding the baseline interview; ART subjects were significantly more likely than pre-ART subjects to attribute their job loss to their health.

For those currently employed, ART subjects reported better work performance, less difficulty doing their jobs, and lower absenteeism from work than did those who had not yet started treatment. The amount of health-related absenteeism reported by ART subjects– more than two and a half days in the preceding month– was, however, also high.

For subjects already on ART, we also considered the association between treatment duration and starting CD4 cell count and our main outcome indicators. In Table 3, subjects who had been on ART for 1–3 months at their baseline interview are compared with those who had been on ART for 3–6 months. There were no significant differences reported in functional impairment or symptoms. These results are consistent with those of other studies that have observed the greatest improvement in quality of life immediately after ART initiation, with slower improvement in later months [13,14].

Table 3

Table 3

There was, however, some difference in the employment outcomes based on the duration of treatment. Subjects on ART for more than 3 months reported much less absenteeism from work and better job performance than did those starting treatment more recently. The difference in absenteeism is large in magnitude, nearly 3 days per month. This may be a consequence of the reduction in fatigue noted above or reflect a time lag in coming up to speed at work after a prolonged period of illness.

Finally, we looked at whether a subject's CD4 cell count nearest to treatment initiation was associated with different outcomes in the first 6 months on ART. The picture presented by the results, as shown in Table 4, is ambiguous. There were no significant differences in most indicators, although several were close to significant. Those whose starting CD4 cell count exceeded 100 cells/μl reported fewer skin problems and better job performance than those who started treatment with CD4 cell counts below 100 cells/μl, although the difference in job performance was marginally significant. The reverse held for most other symptoms, however; the lower starting CD4 cell count group was less likely to report pain, nausea, or depression, although the results for nausea and depression are marginally significant. Two possible explanations for the apparently better condition of those who started with lower CD4 cell counts both involve perception. First, these subjects were relatively sicker before receiving treatment and thus may judge themselves to have improved more and feel better on treatment than did those who had fewer symptoms at treatment initiation. Second, it is also possible that those who wait to start treatment until their CD4 cell counts are very low either experience fewer symptoms of HIV disease or are troubled less by the symptoms.

Table 4

Table 4

Back to Top | Article Outline


In a population in which only a quarter of adults are formally employed, the ability to perform normal daily activities, whatever those may be, is a better indicator of the economic outcomes of treatment than is participation in the formal sector workforce. Among the HIV-positive South African adults described here, individuals not yet on ART report greater difficulty carrying out normal daily activities than do those who have started ART. Study subjects who had received ART for a median of 3 months reported 46% less functional impairment than did pre-ART subjects, better performance at work, and a substantially lower prevalence of pain, nausea, fatigue, feeling unwell, and feeling sad or depressed. After 3 months on ART, those who were employed worked an average of 2.9 days more per month than did those just starting treatment, and 1.8 days more than those not yet on treatment. In a typical month consisting of 22 working days, these differences of 17 and 10% in working time could have an important effect on labour productivity.

Despite this generally positive association between ART and patient economic activity, the prevalence of any functional impairment in the previous week (32%) and the average number of days impaired (0.9) among ART patients was high, as was the prevalence of pain (48%), fatigue (40%), skin problems (46%), feeling unwell (33%), feeling depressed (26%), and resting during the day (43%). There is certainly some level of these conditions in any population, including among people who are generally healthy, but we do not have relevant data from an HIV-negative cohort with which to compare these findings. The finding of nearly a full day of functional impairment in the previous week among patients on treatment for 3–6 months and more than one day of health-related work absenteeism in the previous month, should they persist as the time on treatment increases, are a cause for concern. Basic conditions of employment in South Africa allow for 30 days of paid sick leave per 3-year cycle [23]; the average of 1.3 days/month that we observed would exceed the maximum allowed, even if the worker should suffer no other illnesses or accidents requiring sick leave. Interestingly, some of the reported absenteeism may be an artifact of the treatment programmes themselves, as each visit to the clinic can be assumed to require a full day off work, as a result of travel and waiting time.

The study described here has several limitations. First, and most important, the results presented in this paper are cross-sectional, reflecting only the baseline dataset. As with any cross-sectional analysis, causation is not definite. Although there appear to be few relevant differences between pre-ART and ART subjects in the sample, it is possible that the two groups differ in ways that affect the results. A particular concern is the possibility that reasons for clinic visits, and thus patient condition on the day of the visit, may differ between pre-ART and ART patients, with pre-ART patients more likely to come to the clinic as a result of overt ill health and ART patients more likely to visit for routine care. We cannot estimate the magnitude of this bias but, given the wide range of reasons for visiting the study clinics– ranging from participation in group counselling and educational sessions to acute care– and the effort of study interviewers not to approach patients who appeared acutely ill at the time of the interview– we believe that it is modest. Other limitations will be addressed as the study, which is longitudinal in design, progresses; subjects are now being interviewed at least two times per year, during their routine clinic visits. A number of subjects who were pre-ART at baseline had initiated ART by their second interview, allowing for a future pre and post-analysis.

Second, as with any study that relies on self-reported interview data, recall error is likely. We deliberately chose very short recall periods (e.g. yesterday or last week) to minimize this risk, but for some results, such as work performance relative to when the subject was healthy, inaccurate or biased recall is a possibility. Third, there may also be a participation bias in the sample. Approximately 8% of the patients invited to participate in the study declined, and many of them cited lack of time as their reason. The proportion of men among those who declined was also slightly higher than in the study cohort. It is possible that those who refused had a higher rate of employment or other time-dependent obligations than did those who consented, although the overall low refusal rate will limit the impact of any bias this creates.

Finally, we are unclear as to the meaning of some of the unexpected results, such as the higher prevalence of some symptoms among those whose starting CD4 cell count was more than 100 cells/μl and the lower employment rate of those already on ART. We have offered possible explanations for these findings, but the data do not allow us to present definite explanations.

Despite these limitations, the data presented here indicate that patients who have initiated ART are more able to engage in economic activities than are those who have not yet started treatment. The large differences in functional impairment, symptom prevalence, and job performance between pre-ART subjects and those on ART are a cause for optimism about the economic outcomes of the national treatment programme. Further improvements are still needed, however, if ART patients are to be considered fully active contributors to the economy. The effort to improve these outcomes for patients will be strengthened by the longitudinal datasets that we and others are now collecting.

Back to Top | Article Outline


The authors would like to thank the participating clinics: the Themba Lethu Clinic of Helen Joseph Hospital and the Gauteng Department of Health, the Witkoppen Health and Welfare Centre, and the ACTS Clinic, for their participation in the study and support of this research. They are also grateful to the patients who are part of the research cohort. Finally, the authors wish to thank Professor Bruce Larson of Boston University for his comments on the manuscript. The opinions expressed herein are those of the authors and do not necessarily reflect the views of the funding agencies or participating clinics or patients.

Back to Top | Article Outline

Authors' Contributions

All authors contributed to designing the study, analysing the data, and editing the manuscript. S.R. drafted the manuscript.

Sponsorship: Funding for this study was provided by the US National Institutes of Health through National Institute of Allergies and Infectious Diseases grant no. PEPFAR 13 to the Wits Health Consortium and by the South Africa Mission of USAID through cooperative agreement no. 674-A-00-02-00018 to Right to Care.

Conflicts of interest: The authors have no competing interests. I.S. is the Programme Director of Right to Care, a non-governmental organization that supports the delivery of antiretroviral therapy at the study sites.

Back to Top | Article Outline


1. South African Department of Health. Operational plan for comprehensive HIV and AIDS care, management, and treatment for South Africa. Pretoria, South Africa: National Department of Health; 2003. Available from: Accessed: April 2008.
2. Kapp C. South Africans hope for a new era in HIV/AIDS policies. Lancet 2006; 368:1759–1760.
3. May MT, Sterne JA, Costagliola D, Sabin CA, Phillips AN, Justice AC, et al. HIV treatment response and prognosis in Europe and North America in the first decade of highly active antiretroviral therapy: a collaborative analysis. Lancet 2006; 368:451–458.
4. Hunt PW, Deeks SG, Rodriguez B, Valdez H, Shade SB, Abrams DI, et al. Continued CD4 cell count increases in HIV-infected adults experiencing 4 years of viral suppression on antiretroviral therapy. AIDS 2003; 17:1907–1915.
5. Coetzee D, Hildebrand K, Boulle A, Maartens G, Louis F, Labatala V, et al. Outcomes after two years of providing antiretroviral treatment in Khayelitsha, South Africa. AIDS 2004; 18:887–895.
6. Bekker LG, Myer L, Orrell C, Lawn S, Wood R. Rapid scale-up of a community-based HIV treatment service: programme performance over 3 consecutive years in Guguletu, South Africa. S Afr Med J 2006; 96:315–322.
7. Hughes J, Jelsma J, Maclean E, Darder M, Tinise X. The health-related quality of life of people living with HIV/AIDS. DisabilRehabil 2004; 26:371–376.
8. Thirumurthy H, Graff Zivin J, Goldstein M. The economic impact of AIDS treatment: labor supply in western Kenya. NBER Working Paper 1187. Cambridge, MA: National Bureau of Economic Research; 2005. Available from: Accessed: 2 February 2007.
9. Larson BA, Fox MP, Rosen S, Bii M, Sigei C, Shaffer D, et al. Early effects of antiretroviral therapy on work performance: preliminary results from a cohort study of Kenyan agricultural workers. AIDS 2008; 22:421–425.
10. Eholie SP, Nolan M, Gaumon AP, Mambo J, Kouame-Yeboute Y, Aka-Kakou R, et al. Antiretroviral treatment can be cost-saving for industry and life-saving for workers: a case study from Cote d'Ivoire's private sector. In: Economics of AIDS and access to HIV/AIDS care in developing countries, issues and challenges. Paris: Agence Nationale de Recherches sur le Sida; 2003.
11. Habyarimana J, Mbakile B, Pop-Eleches C. HIV/AIDS, ARV treatment and worker absenteeism: evidence from a large African firm. In: 10th BREAD Conference on Development Economics. Princeton, NJ, 27–28 April 2007. Available from: Accessed: 24 January 2008.
12. Muirhead D, Kumaranayake L, Hongoro C, Charalambous S, Grant A, Fielding K, et al. Health care costs, savings and productivity benefits resulting from a large employer sponsored ART program in South Africa. In: XVIth International AIDS Conference 2006. Toronto, Canada, 13–18 August 2006 [Abstract MOPE0674].
13. Stangl AL, Wamai N, Mermin J, Awor AC, Bunnell RE. Trends and predictors of quality of life among HIV-infected adults taking highly active antiretroviral therapy in rural Uganda. Aids Care 2007; 19:626–636.
14. Jelsma J, Maclean E, Hughes J, Tinise X, Darder M. An investigation into the health-related quality of life of individuals living with HIV who are receiving HAART. AIDS Care 2005; 17:579–588.
15. Wouters E, Meulemans H, Van Rensburg HCJ, Heunis JC, Mortelmans D. Short-term physical and emotional health outcomes of public sector ART in the Free State province of South Africa. Qual Life Res 2007; 16:1461–1471.
16. Louwagie GM, Bachmann MO, Meyer K, Booysen FR, Fairall LR, Heunis C. Highly active antiretroviral treatment and health related quality of life in South African adults with human immunodeficiency virus infection: a cross-sectional analytical study. BMC Public Health 2007; 7:244.
17. South African Department of Health. National antiretroviral treatment guidelines. Pretoria: National Department of Health; 2004. Available from: Accessed: February 2007.
18. Rosen S, Ketlhapile M, Sanne I, Bachman DeSilva M. Costs to patients of obtaining treatment for HIV/AIDS in South Africa. South Afr Med J 2007; 97:524–529.
19. Statistics South Africa. Census 2001. Pretoria, South Africa: Statistics South Africa; 2001. Available from: Accessed: December 2006.
20. Wu AW, Revicki DA, Jacobson D, Malitz FE. Evidence for reliability, validity and usefulness of the Medical Outcomes Study HIV Health Survey (MOS-HIV). Qual Life Res 1997; 6:481–493.
21. Kessler RC. World Health Organization health and performance questionnaire (HPQ): clinical trials baseline version. Geneva: World Health Organization; 2002. Available from: Accessed: 12 December 2006.
22. Pienaar D, Myer L, Cleary S, Coetzee D, Michaels D, Cloete K, et al. Models of care for antiretroviral service delivery. Cape Town, South Africa: University of Cape Town; 2006. Available from: Accessed: April 2008.
23. South African Department of Labour. Basic Conditions of Employment Act and Amendments. Pretoria: National Department of Labour; 2007. Available from: Accessed: February 2007.

antiretroviral therapy; economics; HIV/AIDS; South Africa

© 2008 Lippincott Williams & Wilkins, Inc.