Objectives: To determine retention in HIV care for individuals not yet eligible for antiretroviral therapy (ART) and to explore factors associated with retention in a rural public health HIV program.
Methods: HIV-infected adults (≥16 years) not yet eligible for ART, with CD4 cell count >200 cells per microliter from January 2007 to December 2007 were included in the analysis. Retention was defined by repeat CD4 count within 13 months. Factors associated with retention were assessed using logistic regression with clustering at clinic level.
Results: Four thousand two hundred twenty-three were included in the analysis (83.9% female). Overall retention was 44.9% with median time to return 201 days [interquartile range (IQR): 127-274]. Retention by initial CD4 count 201-350, 351-500, and >500 cells per microliter was 51.6% [95% confidence interval (CI): 49.1 to 54.0], 43.2% (95% CI: 40.5 to 45.9), and 34.9% (95% CI: 32.4 to 37.4), respectively. Compared with CD4 201-350 cells per microliter, higher initial CD4 count was significantly associated with lower odds of retention [CD4: 351-500 cells/μL adjusted odds ratio (aOR): 0.72, 95% CI: 0.62 to 0.84; CD4 >500 cells/μL aOR: 0.51, 95% CI: 0.44 to 0.60]. Male sex was independently associated with lower odds (aOR: 0.80, 95% CI: 0.67 to 0.96), and older age with higher odds of retention (for each additional year of age aOR: 1.03, 95% CI: 1.03 to 1.04).
Conclusions: Retention in HIV care before eligibility for ART is poor, particularly for younger individuals and those at an earlier stage of infection. Further work to optimize and evaluate care and monitoring strategies is required to realize the full benefits of the rapid expansion of HIV programs in sub-Saharan Africa.
From the *Africa Centre for Health and Population Studies, University of KwaZulu-Natal, South Africa; †Department of Clinical Research, London School of Hygiene and Tropical Medicine, London, United Kingdom; ‡UCL Department of Infection and Population Health, London, United Kingdom; §Department of Infectious Diseases, Imperial College, London, United Kingdom; and ‖UCL Institute of Child Health, London, United Kingdom.
Received for publication August 12, 2010; accepted November 16, 2010.
Supported by Wellcome Trust (grant numbers 050534 & 075393). The Hlabisa HIV Treatment and Care Programme receives support through the United States Agency for International Development (USAID) and the President's Emergency Plan (PEPFAR) under the terms of Award No. 674-A-00-08-00001-00.
The opinions expressed herein are those of the authors and do not necessarily reflect the view of the USAID or the United States Government.
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the article.
Data presented in part as poster at 5th IAS Conference on HIV Pathogenesis, Treatment and Prevention, 2009, Cape Town, South Africa. Abstract WEPED181.
Correspondence to: Dr. Richard J. Lessells, MBChB, Africa Centre for Health and Population Studies, PO Box 198, Mtubatuba, KwaZulu-Natal 3935, South Africa (e-mail: firstname.lastname@example.org).
South Africa is home to an estimated 5.7 million HIV-infected people, approximately 1 in 6 of the global HIV-infected population.1 Since 2004, South Africa has seen the scale-up of the largest public sector antiretroviral therapy (ART) program in the world, yet the number of new HIV infections per year is still considerably more than the number initiated on ART.2 The scale-up of HIV counseling and testing services has led to increased population testing rates.3-5 A considerable proportion of newly diagnosed individuals are not yet eligible for ART; in one study from Cape Town between 2001 and 2006, approximately 64% had CD4 cell count measurements above the threshold for eligibility at the time of diagnosis.4 People living with HIV who are not yet eligible for ART have received little attention as, in the early phase of antiretroviral rollout, the priority for HIV services and funding agencies has been on identifying and treating individuals in need of ART.6
Current WHO guidelines recommend clinical assessment and CD4 count monitoring every 6 months, to determine eligibility for ART as early as possible, and to prevent and treat HIV-related illnesses.7 Establishing ART eligibility in a timely fashion for individuals enrolled in care is important to reduce the early mortality on ART consistently reported from programs in sub-Saharan Africa (SSA).8-13 Retention in HIV care is also critical to facilitate integration of prevention strategies.14-16
Despite the recognized importance of retention, there are precious few data from SSA on retention in pre-ART care. In particular, there have been no published studies which explore factors determining pre-ART retention, and hence it remains unclear which groups might benefit most from any targeted supportive intervention. Here we study factors associated with pre-ART retention in a large decentralized HIV program, linked to a population demographic platform in rural South Africa.
Hlabisa HIV Treatment and Care Program
Hlabisa health subdistrict is situated in northern KwaZulu-Natal, South Africa, covering an area of 1430 km2, with approximately 228,000 individuals living largely in scattered homesteads in rural areas. The Hlabisa HIV Treatment and Care Program is a Department of Health initiative supported by the Africa Centre for Health and Population Studies (www.africacentre.com); details of the program have been reported previously.13,17 National ART guidelines are followed, which during the study period denoted ART eligibility in the presence of a WHO stage IV condition or CD4 cell count ≤200 cells per microliter.18
All clinics perform CD4 cell count testing, and tests are routinely done on the same day for any individual newly diagnosed with HIV infection. CD4 cell counts are performed at the National Health Laboratory Service laboratory at Hlabisa Hospital, using the Beckman Coulter EPICS XL flow cytometer (Beckman Coulter, Inc., Brea, CA), and patients are requested to return to clinic for results two weeks from the date of sample collection. Decisions about ART eligibility are usually made on the basis of a single CD4 count result, rarely confirmed in a repeat sample.
The model of care at the time of study for individuals not yet eligible for ART included individual counseling, with advice on healthy living, disclosure, partner notification and testing, transmission risk reduction measures, and family planning. All HIV-infected people regardless of disease stage were additionally invited to attend peer support groups at each clinic. Co-trimoxazole was indicated for individuals with CD4 count ≤200 cells per microliter or WHO stage III/IV. Isoniazid preventive therapy (IPT) was not implemented at a programmatic level at the time of study. All individuals were advised to return for repeat clinical assessment, including clinical staging and CD4 count measurement, 6 months later. Although guidelines stipulated repeat CD4 cell count at 12 months if CD4 count >500 cells per microliter, the actual practice varied and some clinics advised return after 6 months.19
For routine program monitoring and evaluation, all clinic and hospital attendances for CD4 measurement from January 1, 2007, were recorded.
Africa Centre Demographic Information System
A longitudinal demographic surveillance system has, since 2000, collected individual and household demographic data in the demographic surveillance area (DSA), within the Hlabisa health subdistrict, which includes approximately 11,000 households and 85,000 individuals.20 Data are collected 6 monthly on residency status of household members, births, marriages, deaths, and migrations. Data regarding socioeconomic status and employment are collected on an annual basis. Data are collated in the Africa Centre Demographic Information System (ACDIS). It is estimated that 30%-40% of people in the HIV Treatment and Care Program are resident in the surveillance area.
Individual records within the HIV Treatment and Care Program database and ACDIS were linked using the unique South African identity number in accordance with the Africa Centre data confidentiality protocols. Linkage was done to enable analysis of sociodemographic factors associated with retention and to determine vital status of individuals who did not return to care.
Patients were included in this analysis if they were as follows: had a first recorded CD4 cell count from a sample between January 1, 2007, and December 31, 2007; were ART naive; ≥16 years old at the time of CD4 test; and CD4 count result was >200 cells per microliter. Patients were excluded if they were as follows: had missing identity number and age/date of birth; and if ART was initiated after the initial CD4 count but before any subsequent CD4 count (it was assumed that these individuals initiated ART on the basis of clinical stage IV disease).
Retention was defined as repeat CD4 count within 13 months (395 days) of the initial test; this allowed for visits up to one year from collection of initial test result (as this was the time recommended for those with CD4 count >500 cells/μL). Time to retention was measured using the first repeat CD4 count within 13 months. CD4 counts within 14 days of the initial test (n = 101) were excluded from the analysis as it was highly likely that the patient had not received the result of the initial test before repeating the test. Further outcome measures were as follows: change in CD4 cell count per month (measured between initial test and first subsequent test); and progression to ART eligibility (CD4 ≤200 cells/μL) within 13 months. End time for all follow-up was January 30, 2009.
Additional analyses were performed with individuals linked to the demographic surveillance. Variables were chosen for the analysis either due to reported associations with retention in HIV treatment programs or postulated effect on retention.21,22 Residency status related to the defined living arrangements during the course of 13-month follow-up: nonresidents were members of households, but not ordinarily resident, within the DSA; in-migrants were initially nonresident but became resident during follow-up; and out-migrants were initially resident but became nonresident. Socioeconomic data were taken from information collected between July and December 2007 (91.6%) or between January and June 2006 (8.4%), whichever was closer in time to the initial CD4 measurement. Household economic status was determined using an asset wealth index23 and principal component analysis24; households were categorized into quintiles according to the wealth index.
Descriptive statistics were used for the baseline characteristics and overall retention. Proportions analysis stratified by age group, sex, and initial CD4 cell count was used to enable full understanding of retention patterns. Logistic regression with clustering at clinic level was used to explore factors associated with retention in care. Multilevel logistic regression models were used to estimate the independence of measured variables (ρ) at clinic level. The effect of missing data was assessed by addition of a “missing” category for each variable in the model and calculation of the log-likelihood P value. Median regression was used to determine factors associated with CD4 decline. STATA version 10.1 (StataCorp, College Station, TX) was used for all analyses.
Ethical approval was obtained from the University of KwaZulu-Natal for the retrospective analysis of anonymized data from the HIV Treatment and Care Program (BE066/07) and for the linkage of data from the HIV Treatment and Care Program to the ACDIS (E134/06). Approval was also granted by the Research Office of the KwaZulu-Natal Department of Health.
Ten thousand one hundred forty individuals had CD4 cell count results recorded between January 1, 2007, and December 31, 2007. Four thousand two hundred twenty-three (41.6%) were eligible for inclusion in the primary analyses and, of those, 930 (22.0%) were matched to ACDIS and were included in the additional analyses (Fig. 1).
Three thousand five hundred forty-three (83.9%) were female. Median age was 31 years [interquartile range (IQR): 25-38] for females and 37 years (IQR: 31-45) for males (P < 0.001). Median CD4 count was 407 cells per microliter (IQR: 301-565) for females and 365 cells per microliter (IQR 278-491) for males (P < 0.001). The distribution across predefined CD4 strata was as follows: 1605 (38.0%) CD4 201-350 cells per microliter; 1278 (30.3%) CD4 351-500 cells per microliter; 1340 (31.7%) CD4 >500 cells per microliter.
The individuals matched to ACDIS were similar to the unmatched individuals in terms of sex distribution and initial CD4 cell count but were marginally older (Table 1). For the matched individuals, the majority remained resident within the DSA for the 13-month period after the initial CD4 count. Most people (89.4% of those with data) lived within 5 km of the nearest primary health care clinic.
Retention in Care
Overall 1896 patients (44.9%) returned for a subsequent CD4 count within 13 months. Of these, 1371 (72.3%) returned only once and 525 (27.7%) returned on more than one occasion. The proportion retained in care was highest amongst the group with lower initial CD4 cell count: 51.6% for CD4 cell count 201-350 cells per microliter [95% confidence interval (CI) 49.1 to 54.0] versus 43.2% (95% CI 40.5 to 45.9) for CD4 cell count 351-500 cells per microliter and 34.9% (95% CI 32.4 to 37.4) for CD4 cell count >500 cells per microliter. Retention stratified by age and initial CD4 cell count is illustrated in Figure 2.
The median time to return was 201 days (IQR 127-274). The time was shortest for the group with lower initial CD4 count: 175 days (IQR 109-251) for CD4 201-350 cells per microliter versus 206 days (IQR 153-279) for CD4 351-500 cells per microliter and 230 days (IQR 162-310) for CD4 >500 cells per microliter (P < 0.001).
Change in CD4 Cell Count and Progression to ART Eligibility
The median decline in CD4 cell count between initial test and first subsequent test was 8.8 cells per microliter per month (IQR: −24.9 to +5.16), significantly greater with higher initial CD4 count: −5.2 cells per microliter per month for CD4 201-350 cells per microliter; −10.5 cells per microliter per month for CD4 351-500 cells per microliter (P < 0.01); and −18.1 cells per microliter per month for CD4 >500 cells per microliter (P < 0.01). In median regression, higher CD4 group and male sex were significantly associated with greater CD4 decline. There was no significant association with age.
Five hundred sixteen individuals (27.2% of all those who returned) progressed to CD4 ≤200 cells per microliter within 13 months and, of those, 390 (75.6%) were recorded to have subsequently initiated antiretroviral therapy.
Factors Associated With Retention
In multivariable analysis including all patients, higher initial CD4 count was associated with lower odds of retention (compared with CD4 201-350 cells/μL: adjusted odds ratio (aOR): 0.72 (95% CI: 0.62 to 0.84) for CD4 351-500 cells/μL; and aOR 0.51 (95% CI: 0.44 to 0.60) for CD4 >500 cells/μL]. Male sex was independently associated with lower odds of retention [aOR: 0.80 (95% CI: 0.67 to 0.96)]. With reference to age 16-25 years, older age was associated with increased likelihood of retention [aOR: 1.82 (95% CI: 1.55 to 2.14) for 26-35 years; aOR: 2.72 (95% CI: 2.25 to 3.28) for 36-45 years; aOR: 3.07 (95% CI: 2.43 to 3.89) for 46-55 years; and aOR: 1.89 (95% CI 1.27 to 2.82) for >55 years]. With age as a continuous variable, older age was also associated with increased retention [for each additional year of age, aOR: 1.03 (95% CI: 1.03 to 1.04)]. Positive matching to ACDIS was associated with higher odds of retention [aOR: 1.57 (95% CI: 1.32 to 1.87)].
In further analysis with the 930 individuals matched to ACDIS, education level, household wealth, and distance from nearest clinic were not significantly associated with retention in univariable analysis. The results of multivariable analysis are displayed in Table 2. Male sex, higher initial CD4 cell count, out-migration, full-time employment, and household size of greater than 10 members were all associated with lower likelihood of retention in care. With reference to age 16-25 years, older age (26-35 years, 36-45 years, and 46-55 years) was associated with increased retention; with age as a continuous variable, older age was also significantly associated with retention [for each additional year of age, aOR: 1.03 (95% CI 1.02-1.05)]. In-migration was also associated with increased retention.
In analyses stratified by sex, the associations with age, CD4 count, out-migration, employment, and household size remained significant for females. For males, the associations with CD4 count, employment, and out-migration were similar, although only the relationship with out-migration retained statistical significance due to smaller numbers, and nonresidency became significantly associated with lower retention (Table 3).
Outcomes for People Lost to Follow-Up
Four hundred thirty-two of the 930 (46.5%) individuals matched to ACDIS did not return within 13 months for CD4 testing (compared with 58.9% for the unmatched group, P < 0.001). Of these, 21 (4.9%) were reported to have died within the 13-month period and 72 (16.7%) migrated out of the demographic surveillance area. The remaining 339 individuals (78.5%) were documented to be alive and still a member of the demographic surveillance system at the end of the 13-month period after their initial CD4 cell count.
Retention in long-term HIV care both before and after the initiation of ART is important not only to reduce individual HIV-related mortality and morbidity but also as a means to deliver “positive prevention” interventions aimed at reducing ongoing transmission. It is of major concern, therefore, that in this large primary health care HIV program under the existing model of care for individuals not yet eligible for ART, fewer than 50% returned within 13 months for repeat CD4 cell count.
Retention in care after the initiation of ART has been the focus of much published work from SSA and is seen as a key indicator of program performance.25 Conversely, there has only been one small study focused specifically on retention in HIV care before eligibility for ART, which reported retention of only 31% at 12 months in an urban South African program.26 Pre-ART-monitoring strategies using CD4 counts have been shown in mathematical models to maximize the benefit of HIV treatment programs and to be cost-effective in a South African setting.27,28 However, the rates of retention reported here are much less than assumed in these models and should prompt their re-evaluation.
One plausible explanation for poor retention would be the lack of incentive for asymptomatic individuals to return for monitoring and previous work from our group has suggested that the majority return to care at the time of symptoms.29 The package of care for individuals not yet eligible for ART has been limited in this setting, with co-trimoxazole only for those with late symptomatic disease and until now no routine implementation of IPT. This is likely to have limited the effectiveness of programs as individuals will often return to care with opportunistic infections, possibly requiring hospitalization, and ART will continue to be initiated late with consequent sustained high mortality rates.30 Consistent with this, we have reported neither a significant change in the median CD4 count at ART initiation nor a reduction in the high early mortality rates in the first 4 years of the program.13
Recently updated national guidelines in South Africa recommend that individuals not yet eligible for ART are transferred to a “wellness program” for regular follow-up and repeat clinical assessment 6 monthly.31 This also incorporates the provision of IPT to individuals without evidence of tuberculosis disease.32 The evidence base to guide the framework of wellness programs is poor, and research is urgently required to determine optimal and cost-effective models of care.
In this study, gender affected both access to care and retention in care. The proportion of males in this pre-ART population was even lower than that seen in antiretroviral treatment cohorts and likely reflects the different entry points to HIV care, with a large number of asymptomatic females enrolled in HIV care through antenatal HIV testing.33 This might limit the generalizability of our findings to urban or work-based programs with higher proportions of males. Male sex and full-time employment were associated with lower rates of retention and highlight the need to explore health care utilization by men and to develop strategies to engage and retain men in HIV care, perhaps targeting work-based care.34 Factors shown elsewhere to be important determinants of loss to follow-up after ART initiation such as economic status and distance to treatment point were not shown to be significant in this context.21,35-37 This is perhaps explained by the relatively low cost of 6 monthly visits to the clinic compared with monthly visits on ART. The relationship between larger household size and lower likelihood of retention may relate to care commitments, which hinder clinic attendance or financial constraints from broader distribution of the household income.38
The decline in CD4 count for those retained in care was considerably greater than that from natural history research studies in South Africa; although the fact it was greater in those with higher initial CD4 counts was consistent.39,40 This large CD4 decline (equivalent to 105 cells/μL per year) may represent bias in that those with greater CD4 decline may have been more likely to return for follow-up due to symptomatic progression. Additionally, return visits may have been at the time of an intercurrent infection or other clinical episode, which might itself lower the CD4 count. However, these data support the hypothesis that retention in this setting is likely to be influenced more by symptomatic disease than by direct adherence to recommended monitoring strategies.
The contribution of mortality and migration to the high rate of loss to follow-up was relatively minor, and the majority of those lost to follow-up were alive and remained resident within the area. No detailed information regarding causes of death is available for the group lost to follow-up, and thus no conclusions can be drawn whether this mortality was HIV-related and how much the high burden of tuberculosis locally may have contributed.41 The relationship with migration is a complicated one. This rural area is characterized by high rates of circular migration related to urban employment and young adults who migrate to urban areas for employment often return to the family home when unwell.42 This would explain why retention was better for those individuals who were categorized as nonresident at the time of first test but became resident and the opposite relationship that those who became nonresident were less likely to have returned.
Ongoing high HIV incidence in this area despite significant scale-up of ART highlights the urgent need for improved integration of HIV care and prevention.43,44 Retention in long-term HIV care is important to enable the delivery of targeted biomedical and behavioral interventions aimed at reduction of onward transmission.14,15 It is of concern, therefore, that retention was particularly poor for younger people with higher CD4 counts, those who may be responsible for a significant proportion of transmission.45,46 Although there has been much recent interest in the concept of universal ART as prevention,47 the evidence to support this will likely take several years to accumulate,48 and it is imperative now that integrated care and prevention programs are prioritized and adequately resourced.
The main limitation of our study is that it is based on retrospective analysis of CD4 cell count data. There is emerging evidence that the loss to follow-up is considerable even between CD4 testing and collection of results.49 We were unable to quantify this, and our data should be interpreted as overall retention from the time of CD4 testing. Also we were unable to account for tests performed elsewhere and thus might have underestimated true retention. The proportion linked to the demographic surveillance platform (22%) was relatively low. Of those who initially attended 1 of the 6 clinics situated within the demographic surveillance area, 46% were successfully linked. Although mobility of patients within the subdistrict and drawing in of patients from outside the subdistrict might partly explain this, it is also possible that incomplete patient identifiers hampered the linkage process. The linked group had better retention than the unlinked individuals, which might partly reflect the fact that the demographic surveillance area is less rural and more developed and there is greater access to services than the rest of the subdistrict.
In conclusion, we have demonstrated that under existing models of public sector HIV care retention before eligibility for ART is poor, particularly for younger individuals with higher initial CD4 cell count. The next phase of HIV counseling and testing scale-up is likely to significantly increase the number of diagnosed HIV-infected individuals in care but not yet in need of ART.50 Trials to evaluate different models of pre-ART care or wellness programs, both facility based and community based, are an urgent priority. If the substantial benefits of the massive scale-up of HIV treatment and care programs are to be maintained, then we need to build an evidence base with which to inform the design of programs to offer comprehensive care throughout the continuum of HIV infection.
We thank Hilary Thulare (HIV Treatment and Care Programme leader); Colin Newell (senior database scientist); Samukelisiwe Dube (data quality officer); the Monitoring, Evaluation and Reporting team; and Makandwe Nyirenda (for assistance with the asset wealth index).
1. UNAIDS. Report on the Global HIV/AIDS Epidemic 2008
. Geneva, Switzerland: UNAIDS; 2008.
2. World Health Organization. Towards Universal Access. Scaling Up Priority HIV/AIDS Interventions in the Health Sector
. Geneva, Switzerland: WHO; 2008.
3. Shisana O, Rehle T, Simbayi LC, et al. South African National HIV Prevalence, Incidence, Behaviour and Communication Survey 2008: A Turning Tide Among Teenagers?
Cape Town, South Africa: Human Sciences Research Council; 2009.
4. April MD, Walensky RP, Chang Y, et al. HIV testing rates and outcomes in a South African community, 2001-2006: implications for expanded screening policies. J Acquir Immune Defic Syndr
5. Shorter MM, Ostermann J, Crump JA, et al. Characteristics of HIV voluntary counseling and testing clients before and during care and treatment scale-up in Moshi, Tanzania. J Acquir Immune Defic Syndr
6. Navario P. PEPFAR's biggest success is also its largest liability. Lancet
7. World Health Organization. Antiretroviral Therapy for HIV Infection in Adults and Adolescents: Recommendations for a Public Healh Approach-2006 Rev
. Geneva, Switzerland: WHO; 2006.
8. Lawn SD, Harries AD, Anglaret X, et al. Early mortality among adults accessing antiretroviral treatment programmes in sub-Saharan Africa. AIDS
9. Braitstein P, Brinkhof MW, Dabis F, et al. Mortality of HIV-1-infected patients in the first year of antiretroviral therapy: comparison between low-income and high-income countries. Lancet
10. Boulle A, Bock P, Osler M, et al. Antiretroviral therapy and early mortality in South Africa. Bull World Health Organ
11. Zachariah R, Fitzgerald M, Massaquoi M, et al. Risk factors for high early mortality in patients on antiretroviral treatment in a rural district of Malawi. AIDS
12. Castelnuovo B, Manabe YC, Kiragga A, et al. Cause-specific mortality and the contribution of immune reconstitution inflammatory syndrome in the first 3 years after antiretroviral therapy initiation in an urban African cohort. Clin Infect Dis
13. Mutevedzi PC, Lessells RJ, Heller T, et al. Scale-up of a decentralized HIV treatment programme in rural KwaZulu-Natal, South Africa: does rapid expansion affect patient outcomes? Bull World Health Organ
14. Remien RH, Berkman A, Myer L, et al. Integrating HIV care and HIV prevention: legal, policy and programmatic recommendations. AIDS
. 2008;22(Suppl 2):S57-S65.
15. World Health Organization. Essential Prevention and Care Interventions for Adults and Adolescents Living With HIV in Resource-Limited Settings
. Geneva, Switzerland: WHO; 2008.
16. Spaar A, Graber C, Dabis F, et al. Prioritising prevention strategies for patients in antiretroviral treatment programmes in resource-limited settings. AIDS Care
17. Houlihan CF, Bland RM, Mutevedzi PC, et al. Cohort Profile: Hlabisa HIV Treatment and Care Programme. Int J Epidemiol
. February 12, 2010 [Epub ahead of print].
18. Department of Health. National Antiretroviral Treatment Guidelines
. Pretoria, South Africa: Department of Health; 2004.
19. Department of Health. Operational Plan for Comprehensive HIV and AIDS Care, Management and Treatment for South Africa
. Pretoria, South Africa: Department of Health; 2003.
20. Tanser F, Hosegood V, Barnighausen T, et al. Cohort Profile: Africa Centre Demographic Information System (ACDIS) and population-based HIV survey. Int J Epidemiol
21. Cornell M, Myer L, Kaplan R, et al. The impact of gender and income on survival and retention in a South African antiretroviral therapy programme. Trop Med Int Health
22. Booysen F, De Wet K. Predictors of patient retention in the South African public sector ART programme [CDD146]. Presented at: 5th IAS Conference on HIV Pathogenesis, Treatment and Prevention; 2009; Cape Town, South Africa.
23. Rutstein S, Johnson K. The DHS Wealth Index
. Calverton, MD: ORC Macro; 2004.
24. Cooley W, Lohnes P. Multivariate Data Analysis
. New York, NY: John Wiley & Sons, Inc; 1971.
25. Rosen S, Fox MP, Gill CJ. Patient retention in antiretroviral therapy programs in sub-Saharan Africa: a systematic review. PLoS Med
26. Larson BA, Brennan A, McNamara L, et al. Early loss to follow up after enrolment in pre-ART care at a large public health clinic in Johannesburg, South Africa. Trop Med Int Health
. 2010;15(Suppl 1):43-47.
27. Hallett TB, Gregson S, Dube S, et al. The impact of monitoring HIV patients prior to treatment in resource-poor settings: insights from mathematical modelling. PLoS Med
28. Bendavid E, Young SD, Katzenstein DA, et al. Cost-effectiveness of HIV monitoring strategies in resource-limited settings: a southern African analysis. Arch Intern Med
29. Lessells RJ, Mutevedzi P, Cooke GS, et al. Characteristics of individuals attending a pilot pre-ART clinic in Hlabisa sub-district. Abstract 276. Presented at: 4th Southern African AIDS Conference; 2009; Durban, South Africa.
30. Ndiaye B, Ould-Kaci K, Salleron J, et al. Characteristics of and outcomes in HIV-infected patients who return to care after loss to follow-up. AIDS
31. Department of Health. Clinical Guidelines for the Management of HIV & AIDS in Adults and Adolescents. Pretoria, South Africa: Department of Health; 2010.
32. Department of Health. Guidelines for Tuberculosis Preventive Therapy Among HIV Infected Individuals in South Africa
. Pretoria, South Africa: Department of Health; 2010.
33. Braitstein P, Boulle A, Nash D, et al. Gender and the use of antiretroviral treatment in resource-constrained settings: findings from a multicenter collaboration. J Womens Health
34. Mills EJ, Ford N, Mugyenyi P. Expanding HIV care in Africa: making men matter. Lancet
35. Brinkhof MW, Dabis F, Myer L, et al. Early loss of HIV-infected patients on potent antiretroviral therapy programmes in lower-income countries. Bull World Health Organ
36. Maskew M, MacPhail P, Menezes C, et al. Lost to follow up: contributing factors and challenges in South African patients on antiretroviral therapy. S Afr Med J
37. Massaquoi M, Zachariah R, Manzi M, et al. Patient retention and attrition on antiretroviral treatment at district level in rural Malawi. Trans R Soc Trop Med Hyg
38. Goudge J, Gilson L, Russell S, et al. The household costs of health care in rural South Africa with free public primary care and hospital exemptions for the poor. Trop Med Int Health
39. Holmes CB, Wood R, Badri M, et al. CD4 decline and incidence of opportunistic infections in Cape Town, South Africa: implications for prophylaxis and treatment. J Acquir Immune Defic Syndr
40. Brumme Z, Wang B, Nair K, et al. Impact of select immunologic and virologic biomarkers on CD4 cell count decrease in patients with chronic HIV-1 subtype C infection: results from Sinikithemba Cohort, Durban, South Africa. Clin Infect Dis
41. Houlihan CF, Mutevedzi PC, Lessells RJ, et al. The tuberculosis challenge in a rural South African HIV programme. BMC Infect Dis
42. Welaga P, Hosegood V, Weiner R, et al. Coming home to die? The association between migration and mortality in rural South Africa. BMC Public Health
43. Barnighausen T, Tanser F, Gqwede Z, et al. High HIV incidence in a community with high HIV prevalence in rural South Africa: findings from a prospective population-based study. AIDS
44. Barnighausen T, Tanser F, Newell ML. Lack of a decline in HIV incidence in a rural community with high HIV prevalence in South Africa, 2003-2007. AIDS Res Hum Retroviruses
45. Hollingsworth TD, Anderson RM, Fraser C. HIV-1 transmission, by stage of infection. J Infect Dis
46. Wawer MJ, Gray RH, Sewankambo NK, et al. Rates of HIV-1 transmission per coital act, by stage of HIV-1 infection, in Rakai, Uganda. J Infect Dis
47. Granich RM, Gilks CF, Dye C, et al. Universal voluntary HIV testing with immediate antiretroviral therapy as a strategy for elimination of HIV transmission: a mathematical model. Lancet
48. Dabis F, Newell ML, Hirschel B. HIV drugs for treatment, and for prevention. Lancet
49. Larson BA, Brennan A, McNamara L, et al. Lost opportunities to complete CD4+ lymphocyte testing among patients who tested positive for HIV in South Africa. Bull World Health Organ
50. South African National AIDS Council Secretariat. The National HIV Counselling and Testing Campaign Strategy
. Pretoria, South Africa: SANAC; 2010.