Retention of people living with HIV (PLHIV) under pre-antiretroviral (pre-ARV) care is vital for early initiation of antiretroviral therapy (ART) and reduction of preventable morbidity and mortality associated with HIV/AIDS. The WHO recommends routine pre-ARV follow-up care for newly screened non–ART-eligible PLHIV at least once every 3 months.1,2 Unfortunately, in sub-Saharan Africa, up to 59% of PLHIV do not enroll or adhere to the continuum of comprehensive HIV care from HIV testing and counseling to pre-ARV care to ART.3–5
In Iganga district, eastern Uganda, over 90% of the 1400 newly diagnosed PLHIV between 2004 and 2007 either did not enroll or were lost to follow-up under pre-ARV care.5 There is little information on reasons for nonentry into care or loss to follow-up of PLHIV during pre-ARV care in the context of Uganda, where HIV and ART awareness are presumably high. Available literature captures determinants for uptake of voluntary counselling and testing, ART, and nonadherence to ART in low- and high-income countries.6–8 Generalizing this literature to pre-ARV care would however be exigent since the challenges, barriers, or practical policy implications for uptake and loss to follow-up of PLHIV during the pre-ARV period may be different because the health situation, care demands, and social–psychological contexts for PLHIV without AIDS and those with AIDS are different. In December 2008, we explored reasons for lost to follow-up of PLHIV under pre-ARV care by interviewing PLHIV, half of whom had dropped out of pre-ARV care. The recurrent theme for failure to enroll or adhere to routine pre-ARV care was inadequate posttest counseling about when and why return for care.5 We subsequently conducted a randomized controlled trial to assess whether an intervention providing posttest counseling by clinic staff, combined with repeated home counseling visits by community support agents (CSAs) could improve retention in pre-ARV. An interim analysis at 3 months showed that PLHIV who underwent counseling by staff combined with CSA home visits were 1.8 times more likely to take up pre-ARV care compared with those who received standard care [adjusted risk ratio (ARR), 1.8; 95% confidence interval (CI): 1.4 to 2.1].9 The current analysis investigates the effect of the intervention over a period of 24 months on the retention of PLHIV in pre-ARV care in Uganda.
This was a parallel group balanced (1:1) client randomized controlled trial conducted in 3 health facilities in Iganga district, eastern Uganda.
Newly screened (WHO stage 1 or 2) non–ART-eligible HIV-positive adult (aged >18 years) at the 3 recruitment centers at the start of the study, who were well oriented in time, place and person, not enrolled under the prevention of mother to child transmission program and resident in the district for the follow-up period were included.
Study Setting and Location
The study was conducted in Iganga district 115 km east of the capital Kampala between July 2009 and December 2012. The 3 recruitment centers included Iganga hospital and 2 other health centers (HCs) of Kiyunga and Busesa. The hospital serves a catchment area of approximately 600,000 people, while each of the other 2 HC IVs serve a population of approximately 100,000 people. The centers were purposively selected because unlike the other lower level units, they had a constant supply of test kits and active pre-ARV clinics and were self controlling in the procurement and access to drugs and supplies. The district has a total of 104 health facilities many of which often ran out of stock of HIV test kits and cotrimoxazole because they have to wait for the rare supplies from the National Medical Stores.
District HIV services are organized according to the Uganda national hierarchical referral system, where 73 of the units are HCs; 2 (HC II) offering only cotrimoxazole refills and home-based care services to PLHIV; 27 are HC III offering voluntary counselling and testing, pre-ARV and home-based care services to PLHIV; and 3 HC IV and 1 general hospital offer routine counseling and testing, home-based care, pre-ARV care, and ART services to PLHIV. Depending on the severity of the cases, the lowest (HC II) will refer to HC III or HC level IV. The district also has over 230 private drug shops/private clinics which often sell cotrimoxazole to PLHIV.10 Reports in the district indicate that PLHIV who do not attend regular pre-ARV care actually buy cotrimoxazole refills from these drug shops/clinics.5 Over 94% of the population are subsistence farmers and belong to the Basoga tribe.10 At the time of the study, the regional HIV prevalence was 6%, corresponding to 36,000 PLHIV11 with about 5400 (15%) of PLHIV in the district presumed to be eligible for ART,11 3700 PLHIV regularly attended pre-ARV care, and approximately 3000 patients were on ART.10
Description of the Intervention
We followed newly screened PLHIV to ascertain retention under pre-ARV care at the clients nearest HC. Participants were randomized into one of the 2 arms: The participants in the control arm received standard care (test results, cotrimoxazole prophylaxis, and posttest counseling on disclosure, positive living, and the importance of quarterly pre-ARV attendance). In addition to this standard care, participants in the intervention arm also received monthly home visits by a CSA for a 2-hour counseling session on positive living, status disclosure, and a reminder to go for quarterly pre-ARV care. CSAs are influential community volunteers or experienced ART clients. They were randomly recruited and sensitized for at least 3 weeks by the district 5 years back before this study for purposes of community health mobilization and promotion including encouraging PLHIV seek HIV care and are part of the district health system, The CSAs are primarily between 30 and 40 years of age, with a minimum of primary education. They are purely volunteers and do not receive any salaries or other emoluments from the district and are supervised by the in charges of their nearest HCs. The CSAs received orientation about this study and additional skills in counseling PLHIV to be retained under HIV care. The CSAs (20 in total) were also each provided with a new bicycle purchased at US $48 and an annual bicycle maintenance fee of US $24 per agent. Each CSA visited all the assigned 10 clients once a month for the 24-month follow-up period. At each visit, the CSA evaluated the impressions of the PLHIV on the last pre-ARV clinic visit service, counseled them on any negative perceptions on the service, and encouraged them to go back again on the appointment date. The CSA would also feedback the client's feelings about the service to the appropriate facility for the staffs' attention. All the identified clients nearest HCs were stocked with enough cotrimoxazole to avoid any drug stock outs during the entire study period.
The primary outcome measure was the proportion of newly detected and enrolled PLHIV in both arms who had attended their scheduled quarterly pre-ARV care visits at their nearest HC for at least 6 out of the anticipated 8 visits over a period of 24 months since enrollment vs those who had missed their scheduled pre-ARV appointment for 3 or more of the anticipated 8 visits [categorized as either retained or not retained in pre-ARV care (see Figure S1, Supplemental Digital Content, http://links.lww.com/QAI/A701)].
There is no standard definition of retention under pre-ART care. The pre-ART period usually requires tracking patients across long periods from testing to ART initiation which may take very many years. Several studies in sub-Saharan Africa have defined retention in care differently although many agree with a consistent follow-up. Most of these studies as well define their average retention rates as being between 30% and 59%.12–16 Six of 8 visits (75%) over a period of 24 months for this study was therefore chosen to give us the chance of not only looking at the distribution of coming back for care for each client over 2 years but also by natural judgment a higher cutoff point compared with other studies.
Participants who went to their nearest HC for the scheduled pre-ARV care appointment in both arms were also on each visit interviewed on HIV status disclosure and indicators of safe sexual behavior such as consistent use of condoms and being faithful to 1 sexual partner after knowing their serostatus by a trained research assistant stationed at the center.
Data were collected on baseline characteristics. For participants who went back for pre-ARV care, information was also collected on motivation for adhering to care, HIV status disclosure, cotrimoxazole side effects, consistent condom use, and number of sexual partners since knowing serostatus.
Data Collection Techniques
Baseline data were collected by a trained research assistant at each of the respective recruitment centers, whereas the follow-up data were collected by a trained research assistant at each of the agreed participants nearest HC using interviewer-administered questionnaires. Forty research assistants were trained on the study aim, design, and tools. The study tools were pretested using a pilot study at another HC offering ART in the area. Throughout the study period, random validity checks were conducted by the first and second author (M.L. and N.M.T.) to ensure compliance, consistency, range of values, and completeness.
Sample size was calculated based on the formula by Altman and Bland.17,18 At a power of 80%, estimated pre-ARV retention of 50% for the control arm, a target to detect a difference of absolute 14% in the intervention arm as significant at 5% level, a minimum of 193 participants (rounded up to 200) per arm was found adequate. The proportion of newly screened PLHIV normally differed between the different recruitment centers and is twice as high at Iganga hospital compared with Busesa and Kiyunga. Based on these proportions, 50% (200 participants 100 in each arm) were enrolled at Iganga hospital and 25% (100 participants 50 in each arm) were recruited at each of the other 2 recruitment centers (District Health Office 2009).
Randomization and Masking
In all 3 recruitment centers, participants were randomly assigned to either the control or the intervention arm. Treatment assignment was randomly generated in blocks of 4 based on computer-generated random number sequences provided by one of the statisticians in the study team (D.K.) who had no contact with the study participants. Assignment sequences were placed in opaque sealed envelopes and appropriate numbers sent to each of the recruitment centers. At each recruitment center, participant enrollment was done by an assigned laboratory technician at the HIV/STI clinic who carried out the HIV testing on clients that wanted voluntary HIV counseling and testing. After a client had tested HIV positive, the technician at the center asked the eligible client to pick and open an envelope from a batch of 4 for assignment to the appropriate arm. At each of the recruitment centers, this procedure continued until the required allocated numbers per arm had been realized. None of the study subjects, however, was informed of the details of the arm to which they had been assigned. To limit contamination between and among the participants, however, each arm at every recruitment center was put in different rooms at least 50 meters apart.
Stage 1 involved a comparison of baseline characteristics of the participants in the 2 arms. The χ2 and the T test were used to test for significance between the 2 arms for categorical and quantitative data, respectively. Stage 2 involved bivariate analysis to establish the relationship between the different independent variables and the outcome variable using log-binomial regression. The risk ratios (RRs) including 95% CIs comparing retention under pre-ARV care in intervention vs the control arm were computed. We also computed Mantel–Haenszel Risk Ratios (MHRRs) to adjust for differences in RRs for the different levels in each variable.19 Stage 3 was a multivariate analysis where different models assessing the independent effect of the intervention on retention under pre-ARV care (primary outcome) were examined by controlling for the different potential confounders. The multivariate analysis started by computing the risk ratio (RR) generated by the intervention alone on retention under pre-ARV care. Other models were then computed by cumulatively adding each of the independent variables that had been significant in stage 2 above until all the variables had been exhausted. The intervention was considered significant if the RR did not change significantly for the different generated models. MHRRs were also computed for the secondary outcomes.
This study was approved by the Makerere University School of Public Health Institutional Review Board (ref MUSPH 06012009), the Iganga district authorities, and the Uganda National Council for Science and Technology. All study participants signed informed consent before enrollment.
A total of 400 eligible newly diagnosed PLHIV were randomly assigned either to the intervention arm (N = 200) or the control arm (N = 200). A total of 198 participants in the intervention arm were each visited per month for the 24 months and 2 were visited for 22 of the anticipated 24 times. After 24 months, the participants receiving monthly CSA home visits were almost 3 times more likely to be retained in pre-ARV care compared with those receiving standard care (the control arm), 164/200 (82.0%) vs 67/200 (33.5%) (Fig. 1).
Participants in both arms were comparable regarding gender (P = 0.75), age (P = 0.48), education status, (P = 0.83), marriage status (P = 0.90), occupation (P = 0.62), distance to the HC providing pre-ARV care (P = 0.56), and household size (P = 0.23). There were, however, observed differences between the 2 arms regarding pre-ARV awareness (P = 0.01), religion (P = 0.02), and marital status (P = 0.02); although on further analysis, these differences did not significantly influence the effect of the intervention on the outcome (Table 1).
Both the bivariate and multivariable analyses were conducted using intention to treat with a sample size of 200 participants per arm. Participants in the intervention arm were more likely to return for care compared with those in the control arm with an overall percentage of 33.5% and 82.0% for the control and intervention arms, respectively, and an unadjusted RR of 2.5 (95% CI: 1.9 to 3.0). The effect of the intervention remained significant in stratified analysis. The MHRRs and their 95% CIs shown by the independent variables were recruitment center 2.5 (95% CI: 2.0 to 3.0), gender 2.4 (95% CI: 1.9 to 2.9), age group 2.5 (95% CI: 2.0 to 3.0), education 2.5 (95% CI: 1.9 to 3.0), marital status 2.4 (95% CI: 2.0 to 3.0), marriage status 2.7 (95% CI: 2.0 to 3.5), occupation 2.5 (95% CI: 2.1 to 3.0), religion 2.4 (95% CI: 2.0 to 2.7), and pre-ARV awareness 2.3 (95% CI: 1.9 to 2.9) (Table 2). The MHRRs show that the intervention was as effective regardless of the differences in baseline characteristics. A test of homogeneity of the combined MHRR showed no significant difference among the strata.
All models generated in the multivariate analysis did not significantly change the ARR or the 95% CI imposed by the intervention alone on the outcome ARR 2.5 (2.0 to 3.0) (Table 3).
As secondary outcomes, clients in the intervention arm were 1.7 times more likely to disclose their HIV status compared with clients in the control arm (RR, 1.7; 95% CI: 1.2 to 2.4), 1.8 times more likely to consistently have protected sex compared with those in the control arm (RR, 1.8; 95% CI: 1.4 to 2.2) and 1.2 times more likely to have sex with only 1 sexual partner compared with those in the control arm (RR, 1.2; 95% CI: 1.0 to 1.5) (Table 4).
This is the first randomized trial in Uganda to demonstrate that CSAs in resource-poor settings can almost triple the retention of PLHIV under the critical stage of pre-ARV care, improve HIV disclosure, and promote safer sexual behaviors. Our findings show that over the 24-month follow-up period, approximately 80% of participants in the intervention arm were retained under pre-ARV compared with 34% of the participants in the control arm (ARR, 2.5; 95% CI: 2.0 to 3.0). We also found that the repeated counseling performed by the CSAs increased the likelihood of HIV status disclosure to the next of kin by 70% (RR, 1.7; 95% CI: 1.2 to 2.4) and the likelihood of PLHIV consistently having protected sex by 80% (RR, 1.8; 95% CI: 1.4 to 2.2). Although borderline in significance, the counseling also increased the likelihood of being faithful to 1 sexual partner after knowing the serostatus by 20% (RR, 1.2; 95% CI: 1.0 to 1.5).
Our intervention retention rate (80%) is much higher than the comparable average rate of retention of PLHIV under pre-ARV care (45%–57%) in sub-Saharan Africa.16,20 The rate of retention for this study also favorably compares with other studies in Uganda, Rwanda, and Zambia, where community health workers were found to improve retention rates for PLHIV under HIV care.21,22 The increase in retention of PLHIV under pre-ARV care in the intervention arm over time could be explained by the persistent monthly home counseling visits offered by the CSAs to the PLHIV. This finding is supported by other studies where appropriate home-based care has been found to improve health seeking behavior for PLHIV.23–26 Thus by encouraging participants to go the nearest HC for pre-ARV care, the CSAs acted as primary care linkages between the participants and the health care system. This is especially important in health care systems where access to services is poor and the human resources for health are not enough to conduct health facility work as well as do community mobilization and health promotion programs in the communities.27 The importance of primary linkages as a tool for improving health seeking behavior and access to comprehensive HIV care has also previously been outlined by other studies in resource-poor settings.28,29
The study also shows that clients in the intervention arm were 2 times more likely to disclose their HIV status to members of the immediate family compared with those in the control arm. This finding is similar to others in Africa and other low-income countries where home-based or integrated community-based approaches to HIV care have been found to improve acceptance and disclosure of HIV status.30 Similarly, more participants in the intervention arm used condoms consistently and although borderline had 1 sexual partner compared with those in the control arm. These sociobehavioral changes for participants in the intervention arm could be explained by the monthly counseling support provided by the CSAs in the intervention arm. The amount of time dedicated to posttest counseling for disclosure and positive living with HIV is vital in enhancing positive outcomes. Indeed, UNAIDS recommends sufficient time for posttest counseling, especially for clients that are potentially vulnerable to nondisclosure.31 The relationship between home-based care or counseling support and positive HIV-related behavior has similarly been established by other studies in Africa and other resource-poor settings.32,33
Because of shortage of test kits in most of the units, we purposively selected only the large stocked units for recruitment. Although this could introduce a recruitment bias, we feel that it should have less impact on this study since appropriate randomization was done and also the primary outcome was retention in care over time at other peripheral units whose pivot of attendance was availability of cotrimoxazole that we provided.5 Most of the participants in either arm or the CSAs stay in similar sociocontextual settings that are not so demographically distant from each other. It is possible that some may be close relatives or friends as well. This could potentially result in information sharing and contamination among the study participants and/or the study staff. Because most of the secondary questions were self reported about sex behavior, there is as well a potential for social desirability bias. Genuine efforts were, however, made during pretrial training and recruitment to inform the participants and the staff to avoid discussing issues related to the intervention among them and to be honest with their responses. It is also unlikely that the participants could freely share information about the trial between themselves given the sensitivity of the topic and the high stigma still associated with HIV in the community.34
Interpretation and Generalizability
Appropriate randomization of participants was done (Table 1) and the association between the intervention and the outcome measure is least likely to have been affected by participant allocation bias, but the intervention itself. The trial has therefore scientifically shown that conducting repeated monthly visits by CSAs for counseling support more than doubles the likelihood of retaining PLHIV under care at least over 2 years. This is against the background that retention of PLHIV under HIV care is vital for timely initiation of ART, enhances achievement of an undetectable viral load, and ultimately reduces HIV-related morbidity and mortality as well as reducing the risk for HIV transmission for those under timely care.27 The trial further demonstrates that home counseling support by the CSAs also increases status disclosure by 70% and protected sex by 80% among the intervention participants. This intervention can therefore be a firm foundation for re-enforcing community programmes that encourage status disclosure and positive living both of which are very important pillars for comprehensive HIV care, especially in this era of HIV combination prevention.27,35 In similar settings that are able to consistently mobilize volunteer CSAs therefore, such an intervention would be applicable and sustainable at least over a 2-year period. The CSAs would thus supplement the efforts of the already overloaded human resources for health at ART clinics and make the health system more competent.
The trial thus demonstrates that use of CSAs in resource poor settings can more than double retention of PLHIV under care and also additionally improve status disclosure and other elements of positive living all of which are important pillars of HIV combination prevention. The authorities would, however, need to address other important system deficiencies such as stock-outs of HIV testing kits and cotrimoxazole, poor staff attitudes toward pre-ARV care, and staff confidentiality that are common in similar settings.36
1. WHO. Guideline Development Group Meeting on Cotrimoxazole Prophylaxis in the Context of HIV Meeting Report 19-21 November 2013. Geneva, Switzerland. 2013. Available at: www.who.int/hiv/pub/meetingreports/.../en/
. Accessed April 15, 2014.
2. WHO. Guidelines on Co-Trimoxazole Prophylaxis for HIV-Related Infections Among Children, Adolescents and Adults: Recommendations for a Public Health Approach. Geneva, Switzerland: WHO; 2006.
3. Geng EH, Bwana MB, Muyindike W, et al.. Failure to initiate antiretroviral therapy, loss to follow-up and mortality among HIV-infected patients during the pre-ART period in Uganda. J Acquir Immune Defic Syndr. 2013;63:e64–e71.
4. Lessells RJ, Mutevedzi PC, Cooke GS, et al.. Retention in HIV care for individuals not yet eligible for antiretroviral therapy: rural KwaZulu-Natal, South Africa. J Acquir Immune Defic Syndr. 2011;56:e79–e86.
5. Lubega M, Nsabagasani X, Tumwesigye NM, et al.. Policy and practice, lost in transition: reasons for high drop-out from pre-antiretroviral care in a resource-poor setting of eastern Uganda. Health Policy. 2010;95:153–158.
6. Kumarasamy N, Safren SA, Raminani SR, et al.. Barriers and facilitators to antiretroviral medication adherence among patients with HIV in Chennai, India: a qualitative study. AIDS Patient Care STDS. 2005;19:526–537.
7. Knodel J, Kespichayawattana J, Saengtienchai C, et al.. The role of parents and family members in ART treatment adherence: evidence from Thailand. Res Aging. 2010;32:19–39.
8. Byakika-Tusiime J, Oyugi JH, Tumwikirize WA, et al.. Adherence to HIV antiretroviral therapy in HIV+ Ugandan patients purchasing therapy. Int J STD AIDS. 2005;16:38–41.
9. Muhamadi L, Tumwesigye NM, Kadobera D, et al.. A single-blind randomized controlled trial to evaluate the effect of extended counseling on uptake of pre-antiretroviral care in Eastern Uganda. Trials. 2011;12:184.
10. DHO. Annual Progress Report. Kampala, Uganda: Un published; 2012.
11. MOH. Uganda AIDS Indicator Survey. Kampala, Uganda: Ministry of Health; 2011.
12. Fox MP, Larson B, Rosen S. Defining retention and attrition in pre-antiretroviral HIV care: proposals based on experience in Africa. Trop Med Int Health. 2012;17:1235–1244.
13. Tobias CR, Fox JE, Walter AW, et al.. Retention of people living with HIV/AIDS in oral health care. Public Health Rep. 2012;127(suppl 2):45–54.
14. Schoni-Affolter F, Keiser O, Mwango A, et al.. Estimating loss to follow-up in HIV-infected patients on antiretroviral therapy: the effect of the competing risk of death in Zambia and Switzerland. PLoS One. 2011;6:e27919.
15. Mugglin C, Estill J, Wandeler G, et al.. Loss to programme between HIV diagnosis and initiation of antiretroviral therapy in sub-Saharan Africa: systematic review and meta-analysis. Trop Med Int Health. 2012;17:1509–1520.
16. Rosen S, Fox MP. Retention in HIV care between testing and treatment in sub-Saharan Africa: a systematic review. PLoS Med. 2011;8:e1001056.
17. Altman D. Statistical Methods for Medical Researchers. Oxford, United Kingdom: Oxford press; 1995.
18. Bland JM. Sample size in guidelines trials. Fam Pract. 2000;17:S17–S20.
19. Breslow NE. Elementary methods of cohort analysis. Int J Epidemiol. 1984;13:112–115.
20. Kranzer K, Govindasamy D, Ford N, et al.. Quantifying and addressing losses along the continuum of care for people living with HIV infection in sub-Saharan Africa: a systematic review. J Int AIDS Soc. 2012;15:17383.
21. Torpey KE, Kabaso ME, Mutale LN, et al.. Adherence support workers: a way to address human resource constraints in antiretroviral treatment programs in the public health setting in Zambia. PLoS One. 2008;3:e2204.
22. Chang LW, Kagaayi J, Nakigozi G, et al.. Effect of peer health workers on AIDS care in Rakai, Uganda: a cluster-randomized trial. PLoS One. 2010;5:e10923.
23. MacPherson P, Lalloo DG, Webb EL, et al.. Effect of optional home initiation of HIV care following HIV self-testing on antiretroviral therapy initiation among adults in Malawi: a randomized clinical trial. JAMA. 2014;312:372–379.
24. le Roux IM, Tomlinson M, Harwood JM, et al.. Outcomes of home visits for pregnant township mothers and their infants in South Africa: a cluster randomised controlled trial. AIDS. 2013;27:1461–1471.
25. Jaffar S, Govender T, Garrib A, et al.. Antiretroviral treatment in resource-poor settings: public health research priorities. Trop Med Int Health. 2005;10:295–299.
26. Lyimo RA, de Bruin M, van den Boogaard J, et al.. Determinants of antiretroviral therapy adherence in northern Tanzania: a comprehensive picture from the patient perspective. BMC Public Health. 2012;12:716.
27. Rebolledo P, Kourbatova E, Rothenberg R, et al.. Factors associated with utilization of HAART amongst hard-to-reach HIV-infected individuals in Atlanta, Georgia. J AIDS HIV Res. 2011;3:63–70.
28. Losina E, Bassett IV, Giddy J, et al.. The “ART” of linkage: pre-treatment loss to care after HIV diagnosis at two PEPFAR sites in Durban, South Africa. PLoS One. 2010;5:e9538.
29. Bassett IV, Regan S, Chetty S, et al.. Who starts antiretroviral therapy in Durban, South Africa? Not everyone who should. AIDS. 2010;24(suppl 1):S37–S44.
30. Ncama BP. Acceptance and disclosure of HIV status through an integrated community/home-based care program in South Africa. Int Nurs Rev. 2007;54:391–397.
31. UNAIDS. Counselling and HIV/AIDS; UNAIDS Best Practice Collection. Geneva, Switzerland: UNAIDS; 1997.
32. Mulogo EM, Abdulaziz AS, Guerra R, et al.. Self reported risk reduction behavior associated with HIV counseling and testing: a comparative analysis of facility- and home-based models in rural Uganda. AIDS Care. 2013;25:835–842.
33. Njau B, Watt MH, Ostermann J, et al.. Perceived acceptability of home-based couples voluntary HIV counseling and testing in Northern Tanzania. AIDS Care. 2012;24:413–419.
34. Lubega M, Musenze IA, Joshua G, et al.. Sex inequality, high transport costs, and exposed clinic location: reasons for loss to follow-up of clients under prevention of mother-to-child HIV transmission in eastern Uganda: a qualitative study. Patient Prefer Adherence. 2013;7:447–454.
35. Kennedy CE, Medley AM, Sweat MD, et al.. Behavioural interventions for HIV positive prevention in developing countries: a systematic review and meta-analysis. Bull World Health Organ. 2010;88:615–623.
36. Muhamadi L, Nsabagasani X, Tumwesigye MN, et al.. Inadequate pre-antiretroviral care, stock-out of antiretroviral drugs and stigma: policy challenges/bottlenecks to the new WHO recommendations for earlier initiation of antiretroviral therapy (CD <350 cells/muL) in eastern Uganda. Health Policy. 2010;97:187–194.