Prevention of mother-to-child transmission services has been available in South Africa for over 10 years. The uptake of these services has improved dramatically since the service was first introduced. In 2012, 98% of women were being tested for HIV during pregnancy and 92% of HIV-positive mothers were receiving ARV for treatment and/or prophylaxis.1 Regardless, HIV prevalence among pregnant women in South Africa remains high (30%).1 In the province of KwaZulu Natal (KZN), South Africa, where this study was conducted, prevalence among pregnant women is 37%, the highest in the country.1
Many of the steps of the PMTCT cascade are hampered by stigma and other factors that prevent women from accepting their status and accessing appropriate care and treatment.2 Challenges remain, in particular, with adherence to exclusive breastfeeding (EBF), ARV medication, and early infant diagnosis.3
This manuscript outlines the protocol of a randomized control trial designed to assess the impact of the feeding buddy (FB) strategy to improve adherence to the PMTCT guidelines.
Breastfeeding provides immunological support4 and promotes gut maturation,5 and thereby has potential to reduce infant mortality in resource-limited settings where formula feeding is not always a safe option.6,7 Although HIV can be transmitted through breast milk, evidence has shown that EBF, combined with the provision of ARV prophylaxis, significantly reduces the risk of vertical transmission of HIV.8,9 Yet, the prevalence of EBF in South Africa remains low at an estimated 7% at 6 months.10
A landmark study in KwaZulu-Natal Province (KZN), South Africa, found that the cumulative probability of HIV infection by 6 months of age was similar among the exclusively breastfed [0.194 (95% CI: 0.125 to 0.274)] and the never breastfed [0.194 (95% CI: 0.136 to 0.260)] groups, with the highest risk among those who were mixed feeders [0.261 (95% CI: 0.205 to 0.319)].11 Several additional studies substantiated these findings.12,13 Supporting women to EBF provides the safest and healthiest outcomes for women and their infants, therefore strategies to provide such support are important interventions that should be implemented.
When HIV-positive mothers receive antiretroviral treatment or prophylaxis and exclusively breastfeed, the risk of postnatal transmission at 6 months can be reduced to 0%–1%.14,15 Based on this clinical evidence, the World Health Organization (WHO) amended its PMTCT and infant feeding guidelines. The WHO states that the risks of HIV transmission through breastfeeding in resource-limited settings should be weighed against the mortality risks associated with not breastfeeding; the guidelines recommend EBF to 6 months and the introduction of complementary feeding thereafter with breastfeeding up to 12 months, with extended ARV prophylaxis for either infant or mother throughout the breastfeeding period.16,17 Breastfeeding should only stop once a nutritionally adequate and safe diet can be provided.
The South African PMTCT policy broadly adopted these guidelines in 2010 but stopped short of declaring breastfeeding the default feeding choice.2 After the Tshwane Declaration in 2011, breastfeeding was then clearly adopted as the feeding method of choice for HIV-exposed infants.18 Before 2011, free formula was provided by the clinics regardless of whether this was the safest infant feeding choice for HIV-positive women. These women were meant to meet certain AFASS criteria (accessibility, feasibility, affordability, sustainability, and safe) to ensure the safe provision of infant formula. But health care workers and mothers were confused about what constituted safe infant feeding practices. The Department of Health phased out the free formula in 2011 and women were encouraged to exclusively breastfeed their infants. The frequent changes in policy created confusion and a feeling of insecurity in mothers and health care workers.19 In January 2015, yet another policy change was introduced in South Africa with a change to WHO's Option B+.3 All pregnant women are now placed on lifelong ARV treatment, a single daily dose of triple ARV's (tenofovir, emtricitabine, and efavirenz) and daily nevirapine syrup is supplied to their infants from birth to 6 weeks.
Early Infant Diagnosis
Early infant diagnosis (EID) is critical for ensuring immediate care and treatment for infants who test positive for HIV. At the time of this study, EID testing was routinely done at 6 weeks. Early initiation of ARVs is critical for infant survival because a third of untreated HIV-positive infants die within a year of birth.20 The coverage of virological testing of HIV-exposed children at 2 months in South Africa in 2010 was 68%.21 Effective strategies for successful EID have been challenging in resource-limited settings. Many women are unemployed and have insufficient funds to travel to the clinics, the fear of receiving a positive diagnosis acted as a deterrent for taking their infants for testing, as did the lack of disclosure to other family members. Innovative approaches to address difficulties around EID are needed.
Stigma and Discrimination
Disclosure of an HIV status has led to isolation and exclusion of HIV-positive mothers.22 Stigma prevents women from accessing health services, contributes to their dropping out of PMTCT programs,23,24 and impacts their overall mental health which is likely to impact on infant health outcomes. Women interviewed in South African Health Care Facilities related the positive benefit of having psycho-social support in the form of a person to confide in and discuss issues with.25 Given the challenges facing women during the antenatal and postnatal periods (eg, disclosure of HIV status, stigma, health beliefs, and domestic/social issues), a support system is crucial to a woman's well-being.
Defining the Intervention
In this project, a feeding buddy (FB) was selected by the HIV-positive mother to accompany her on PMTCT counseling and clinic sessions and to provide ongoing and continuous support to adhere to the PMTCT guidelines. These include specifically adherence to: ARV treatment, EBF, and overcoming cultural practices linked to mixed feeding. Additionally, to promote infant testing and strategies to reduce stigma and discrimination, the mothers and their selected buddies received training on essential PMTCT and health behaviors and skills.26 The effects of a treatment buddy on ARV adherence have been previously assessed and linked to higher rates of adherence, but the effects of a FB who focused on several aspects of the PMTCT cascade have not evaluated yet.
Figure 1 Conceptual Model of Feeding Buddy intervention.
Both the mother and buddy received intensive training around key target health messages, which covered pregnancy, safe delivery, and postnatal care. This take-home booklet was produced specifically for the intervention, so that mothers and FBs would have these messages documented and available at hand. The training covered all aspects of breastfeeding, the importance of EBF, and safe complementary feeding. The importance of taking medication regularly for both mother and baby was stressed, as was monitoring the growth of the baby and the importance of immunization. In addition, the training covered illness and the danger signs for which children need urgent medical attention. The buddy also provided the psycho-social support that disclosing to a trusted person is believed to provide based on other studies. See Figure 1 Feeding buddy intervention conceptual model below.
Feeding Buddy Concept
Facility-based interventions are often inadequate to effect sustained behavioral changes that enhance PMTCT in the face of multiple contextual factors. Several studies in South Africa noted that training peer counselors or community health workers (CHW) to conduct home-based visits have had a positive influence on the uptake of PMTCT services, including a study in KZN27 and EBF studies,28–31 such as the PROMISE-EBF trial, where low-intensity, individual home-based counseling was shown to increase the rate of EBF in 3 sub-Saharan countries.31
ARV buddies have demonstrated a positive impact as a support mechanism in encouraging disclosure, helping HIV-infected individuals to resist stigma, restoring a sense of social connection,31 and fostering social goodwill.32 Researchers in the Free State province of South Africa highlighted the crucial role of community support as a key predictor of ARV adherence.32 Disclosure to partners has been linked to decreased anxiety, risky behavior, and an increase in informed choices and uptake of health services.33 The FB strategy is similar to the mother-to-mother approach,34 the key difference being that it does not only rely on other mothers to be buddies, but anyone that the mother chooses can be trained as a feeding buddy.
A pilot study carried out in the Eastern Cape Province of South Africa examined the acceptability of the FB strategy and found integrating feeding buddies into an existing PMTCT program could be an effective strategy for providing community and home-based support for HIV-positive mothers.35
In this study, the FB intervention was implemented as one component of The Window of Opportunity Project (WinOp), a comprehensive and integrated maternal and child health project, being implemented in KZN. The WinOp Project focuses on maternal and infant morbidity and mortality during pregnancy and the first 2 years of the child's life. The goal being to strengthen health and community institutional capacity through training and mentoring, thereby improving the provision of services, community education, generating demand for services, and health-promoting behaviors. Under WinOp, the FB intervention was implemented as part of its community engagement.
We hypothesized that mothers who chose a FB would have greater adherence to EBF and ARV regimens, they would be more likely to disclose their status, and they would be more likely to take their infant be tested for HIV at 6 weeks.
To test our hypothesis, the primary objectives were to determine the effect of a FB on adherence to EBF. The secondary objectives were to determine the effect of a FB on adherence to ART regimens, to determine the effect of a FB on adherence to infant HIV testing at 6 weeks, and to determine the effect of a FB on disclosure and stigma.
Quantitative Data Collection
This study has a cluster randomized controlled trial design and was registered with NIH clinical trials registry: NCT02162498. The study was conducted in the Umlalazi and uMhlatuze subdistricts of the Uthungulu District of KwaZulu-Natal Province in South Africa. Of the 27 clinics in these 2 districts, these 16 clinics were the only 16 found to be eligible for selection for this study in terms of having sufficient numbers of pregnant mothers and accessibility of location. They were then randomly assigned to the control or intervention arm. The clusters consisted of 8 selected and randomly assigned control clinics where the FB intervention was not in place and 8 selected and randomly assigned intervention clinics where the feeding buddy intervention had been implemented. The clinics were stratified on the basis of urban/rural/periurban locale. Clinics were then randomly allocated to either intervention or control with each strata.
The trial evaluated the effect of a voluntary PMTCT FB program on HIV-infected women's adherence to EBF. The trial enrolled HIV-infected women who intended to breastfeed and were between 26 and 34 weeks gestation. The proportion of women exclusively feeding at 5.5 months after delivery was the primary end-point of the trial.
In the study intervention clinics, quantitative interviews with mothers and their chosen buddies took place at enrolment during pregnancy and at routine postdelivery visits at day 3 and then week 6, 14, and 22. These visits were aligned with routine PMTCT, well-child, and immunization visits and were conducted by trained research assistants. Women in the control clinics, those that did not yet have a FB program, who consented to the study, were followed using the same evaluation schedule.
Baseline data collection included household characteristics, maternal factors, FB characteristics, previous feeding history, disclosure of HIV status, and stigma. At follow-up visits, data collection included maternal factors, infant status, ARV adherence, infant feeding, stigma, and disclosure of HIV status. A study visit tracking form was completed for each of the study's scheduled visits. Loss to follow-up and reasons for this were also recorded. Participants who missed scheduled visits were contacted twice by research staff before they were considered lost to follow-up.
Sample Size Estimation
It was assumed that in the control arm, 25% of women would be exclusively breastfeeding up to 5.5 months and that the FB intervention would increase this proportion to at least 44% in the intervention group. Sample size calculations indicated that 16 clusters (8 in each arm), with 41 women in each cluster (N = 656), were required to achieve an 80% power to detect this difference. Estimated power was calculated using the Satterthwaite approximate F test for 2-sample comparison of proportions with clustering implemented within Stata 11. Underlying statistical parameters were as follows: an alpha-type-one error of 5% (2-tailed) and an allowance of 20% of study participants lost to follow-up. A coefficient of variation of 0.25 was assumed to account for within-group variation between clusters. See Figure 2 below.
Study Population Inclusion and Exclusion Criteria
The quantitative study sample was drawn from these 2 subdistricts, which constituted 58% of the district population. Two subdistricts were purposively sampled because they accounted for 78% of the district population on ART and had high proportion of reported ANC coverage. Although they did not represent of the district as a whole, they did represent the populations with the greatest need for intervention and the greatest opportunity for study.
Site Assessments, Selection, and Randomization
Site visits were conducted on all 27 clinics in the 2 subdistricts, uMhlathuze and uMlalazi, and data were collected regarding antenatal clinic numbers and well-baby clinic numbers among other things. Clinics with below 10 first antenatal bookings per month and a very low percentage of follow-up visits for well-baby clinics were excluded from the randomization. Consideration was also given to the local infrastructure and demographics (eg, roads, population density, distribution of urban and rural areas) to enable the assessment to be completed in a feasible timeframe. The clinics were then stratified on the basis of urban/rural/periurban locale and randomization into 8 intervention and 8 control clinics was undertaken.
Qualitative Data Collection
Qualitative data were collected by trained research assistants using in-depth interviews conducted among a convenience sample of 10 PMTCT counselors and 10 Community Care Givers (CCGs) to capture their experiences in implementing the FB strategy (work load, satisfaction and challenges, limitations). Additionally, in-depth interviews were conducted with a random sample of 16 mothers from intervention clinics and 16 FB's and 16 from control clinics to document their differing perspectives and overall experiences around the challenges around breastfeeding; adhering to ARV medication; early infant diagnosis; stigma and having/being a feeding buddy and EBF. Interviews were conducted in the local language isiZulu and recorded for later translation into English before analysis would be commenced.
Pregnant women included in the study had to be 18 years or age or older; attending one of the study clinics and enrolled in the PMTCT program; were to be at least 26 but not more than 34 weeks pregnant; indicated their intention to deliver and remain within the study area for at least 6 months after delivery; intending to exclusively breastfeed; and had to provide written informed consent.
An additional eligibility criterion applied only to the mothers in the intervention clusters. They had to select a feeding buddy who agreed to attend training and participate in the evaluation. Mothers in the intervention arm were therefore preenrolled until they had selected their FB who attended at least one visit at which time the enrolment was confirmed.
EBF is defined as no other food or drink, including water, apart from breast milk (including expressed breast milk) with the exception of drops or syrups consisting of vitamins, mineral supplements or medicines. The number of months engaged in EBF was calculated from date of birth of the infant to the discontinuation of breastfeeding and introduction of anything other than breast milk.36
Adherence to ARV
Two separate outcomes were examined and included adherence to ART to prevent MTCT among HIV-positive mothers and adherence to infant ARV prophylaxis of infant. Adherence to ARV in each case was defined as taking more than 95% of prescribed dose.37 Adherence was assessed through self-report at study visits and measured from enrollment to final survey.
Exposure of Interest
The intervention group effectively received 2 separate interventions: having a feeding buddy, and training on breast-feeding. Exposure to the feeding buddy was measured by recording whether buddies attended orientation sessions, the number of antenatal and postnatal visits accompanied by the buddy, and the degree of contact between mother and buddy. Exposure to training was expressed as a numerical training score, calculated as a weighted total number of sessions received, with greater weights for early training, and repeated training. Training dates for each mother will be obtained from the records of the training organization.
A range of potential explanatory variables were also collected: a urban or rural setting; age of mothers and buddies; infant's gender; education of the mother; maternal employment; water source; toilet type; fuel type; number of people living in the household; number of children under 5 living in the house; the number of children supported by the mother; the number of rooms in the house; the transport mode to the clinic; transport time to the clinic. In addition, if the mother had a partner, was married, lived with the partner; the gestational ages of the infant, type of delivery, birth weight of the infant, the number of live births, frequency of CCG visits, and previous breastfeeding experience.
DATA COLLECTION AND MANAGEMENT
Trained research assistants conducted surveys and collected data using a mobile phone platform that transmitted all data confidentially to an internet-based research console. A data manager managed the research console and weekly quality assurance visits by a field supervisor were made to the clinics sites to troubleshoot any data queries. Refresher trainings were provided to the research assistants during regular team meetings to improve data quality.
Approval for the study was obtained by both the University of KwaZulu Natal's Biomedical Research Ethics Committee (BE001/13) and the PATH Research Ethics Committee (HS721). Approval was also obtained from local stakeholders, which included the Provincial Department of Health and the Manager of the Maternal and Child Health Program in KZN.
This study involved minimal risk to participants because no novel interventions were introduced. PMTCT guidelines state that routine HIV counseling should include encouraging the client to disclose their status to someone for support. Similarly, the PMTCT guidelines recommend that HIV-positive clients identify a “buddy” who can provide support.
At enrolment, all participants were assigned a unique study identification number, which was utilized on all nonclinic-based data collection forms in lieu of their names. All forms were kept locked in a secure filing cabinet. All computer records and databases were password-protected to limit their access to feeding buddy personnel and the research assistant consultant only.
The qualitative and quantitative work will be analyzed and results will be published separately, but early anecdotal information from the qualitative interviews suggests that multiple challenges remain for women within PMTCT programs. FBs played an encouraging, supportive role in the intervention group, which improved the participant's ability to accept and cope with their HIV status and further reduced feelings of social isolation.
The breastfeeding outcome, which constitutes interval-censored survival data, will be analyzed using Cox regression, assuming the event happens at interval midpoint, allowing for clustering at clinic level. Potential explanatory variables and interactions will be explored for inclusion into a multiple-variable model. Frailty models will also be fitted to assess unobserved heterogeneity in the data. Analysis will be done in Stata.
CHALLENGES AND LIMITATIONS
As an implementation science research study, the intervention and evaluation were led by separate groups. The intervention was led by the WinOp team and the evaluation was conducted by the research team from the University of KwaZulu Natal. The challenges experienced for the evaluation included delays in implementation of the feeding buddy program, which resulted in divergent timelines for control and intervention data collection. The delay in signing the memorandum of understanding between funders and the implementers delayed the appointment of trainers; the development of materials; and the roll out of the community mobilization and FB training. Additionally, there was lack of clear criteria to systematically determine when the FB program had been fully implemented and thus ready for accurate evaluation. Achieving adequate community awareness and support for the FB program may have been avoided with effective quality assurance checks, monitoring of the intervention, enhanced cooperation with the multiple local stakeholders which included policy makers and health facilities.
Although previous pilot data indicated buddies would attend clinic visits for training and data collection, reaching the buddies through clinic visits was a challenge. A revised community-based delivery design may have been more appropriate.
An inherent limitation of the study is the criterion to only enroll mothers in the intervention who actually chose a feeding buddy. There is therefore a possibility for selection bias whereby women who chose a feeding buddy were the more motivated women likely to disclose and follow instructions.
An important challenge of the study was adherence to critical time points for FB trainings for mothers and buddies; as a result, the true effect of key targeted messages per study protocol might not be measured accurately. Only 4 trainers were appointed to train both mothers and buddies in 8 clinics, which were spread over a wide area. Because of time constraints and difficulties with transport, the trainers were often unable to access the clinic at the time the mother and buddy arrived for their clinic visits. Ideally, a trainer should have been based at each of the 8 clinics, so she was readily available to train the mothers and buddies. Training should have been spread over pregnancy and completed by the early postnatal period. Often mothers and buddies ended up receiving all 4 training sessions at one time, sometimes after delivery thereby missing some key and timeous messages.
By supporting the current South Africa PMTCT guidelines to encourage EBF; adhere to ARV's; ensure early EID; encourage disclosure and build support systems, this intervention may help to diminish the damaging impact that stigma could potentially have on a mother's ability to safely feed her child and adhere to current recommendations. Safe infant-feeding benefits not only the infant, but also the entire family and community because its impacts on the long-term health of both the mother and her infant ensure the HIV-free survival of HIV-exposed infants.
Stigmatizing of HIV is destructive and addressing this has been a priority for HIV care and treatment programs. Despite the challenges experienced with the implementation, having a trained FB to provide not only social and emotional support but also to walk alongside them provide practical help, reminders, act as a mediator, and encourage disclosure could be a strategy that could be used in similar low resource settings with poor social support structures.
We acknowledge with appreciation the participation of the mothers and the buddies. We thank the Department of Nutrition, KwaZulu-Natal, the District Health Office in Uthungulu, and all the local clinics in the area for their cooperation and participation.
1. South Africa National Department of Health. The National Antenatal Sentinel HIV & Herpes Simplex Type-2 Prevalence Survey. Pretoria, South Africa: Department of Health; May 22, 2012.
2. Gourlay A, Birdthistle I, Mburu G, et al Barriers and facilitating factors to the uptake of antiretroviral drugs for prevention of mother-to-child-transmission of HIV in sub-Saharan Africa: a systematic review. J Int AIDS Soc. 2013;16:18588.
3. South Africa Department of Health. National Consolidated Guidelines for the Prevention of Mother-to-Child Transmission of HIV (PMTCT) and the Management of HIV in Children, Adolescents and Adults. Pretoria, South Africa: Department of Health; 2014.
4. Hanson LA, Ahlstedt S, Carlsson B, et al New knowledge in human milk immunoglobulin. Acta Paediatr Scand. 1978;67:577–582.
5. Udall JN, Colony P, Fritze L, et al Development of gastrointestinal mucosal barrier. II. The effect of natural versus artificial feeding on intestinal permeability to macromolecules. Pediatr Res. 1981;15:245–249.
6. Coutsoudis A, Goga A, Rollins N, et al Free formula milk for infants of HIV-infected women: blessing or curse? Health Policy Plan. 2002;17:154–160.
7. Morrison P, HIV and infant feeding: to breastfeed or not to breastfeed: the dilemma of competing risks. Part 1. Breastfeed Rev. 1999;7:5–13.
8. Ngoma MS, Misir A, Mutale W, et al Efficacy of WHO recommendation for continued breastfeeding
and maternal cART for prevention of perinatal and postnatal HIV transmission in Zambia. J Int AIDS Soc. 2015;18:19352. Available at: http://www.jiasociety.org/index.php/jias/article/view/19352/pdf_1. Accessed January 25, 2016.
9. Gartland MG, Chintu NT, Li MS, et al Field Effectiveness of combination antiretroviral prophylaxis for the prevention of mother-to-child transmission in rural Zambia. AIDS. 2013;27. Available at: http://http://www.ncbi.nlm.nih.gov
/pmc/articles/PMC3836017/pdf/nihms521144.pdf. Accessed January 25, 2016.
10. Shisana O, Labadarios D, Rehle T, et al SANHANES-1 Team. The South African National Health and Nutrition Examination Survey (SANHANES-1). Cape Town, South Africa: Human Sciences Research Council Press; 2013.
11. Coutsoudis A, Pillay K, Kuhn l, et al Method of feeding and transmission of HIV-1 from mothers to children by 15 months of age: prospective cohort study from Durban, South Africa. AIDS. 2001;15:379–387.
12. Iliff PJ, Piwoz EG, Tavengwa NV, et al the ZVITAMBO study group. Early exclusive breastfeeding
reduces the risk of postnatal HIV-1 transmission and increases HIV-free survival. AIDS. 2005;19:699–708.
13. Kuhn L, Sinkala M, Kansasa C, et al High uptake of exclusive breastfeeding
and reduced early post-natal HIV transmission. PLoS One. 2007;2:e1363.
14. Coovadia HM, Rollins NC, Bland RN, et al Mother-to-child transmission of HIV-1 infection during exclusive breastfeeding
in the first 6 months of life: an intervention cohort study. Lancet. 2007;369:1107–1116.
15. Kilewo C, Karlsson K, Ngarina M, et al Prevention of mother to child transmission of HIV-1 through breastfeeding
by treating mothers with triple antiretroviral therapy in Dar es Salaam, Tanzania: the Mitra Plus study. J Acquir Immune Defic Syndr. 2009;52:406–416.
16. Hudgens MG, Taha TE, Omer SB, et al Pooled individual data analysis of five randomized trials of infant nevirapine prophylaxis to prevent HIV-1 transmission through breast milk. Clin Infect Dis. 2013;56:131–139.
17. de Vincenzi I. Triple antiretroviral compared with zidovudine and single-dose nevirapine prophylaxis during pregnancy and breastfeeding
for prevention of mother-to-child transmission of HIV-1 (Kesho Bora study): a randomised controlled trial. Lancet Infect Dis. 2011;11:171–180.
18. Kuhn L, Kroon M. Breastfeeding
and the 2015 South Africa Guidelines for the prevention of mother-to-child transmission of HIV. S Afr J HIV Med. 2015;16:. Available at: . Accessed on October 3, 2015.
19. Doherty T, Chopra M, Jackson D, et al Effectiveness of the WHO/UNICEF guidelines on infant feeding for HIV-positive women: results from a prospective cohort study in South Africa. AIDS. 2007;21:1791–1797.
20. Newell ML, Coovadia H, Cortina Borja M, et al for Ghent International AIDS Society (IAS) working group on HIV infection in women and children. Mortality of infected and uninfected infants born to HIV-infected mothers in Africa: a pooled analysis. Lancet. 2004;364:1236–1243.
21. WHO. Towards Universal Access: Scaling up Priority HIV/AIDS
Interventions in the Health Sector ProgressReport. Geneva, Switzerland: World Health Organization; 2011.
22. Ostergard LR, Bula A. “They call our children “Nevirapine babies?””: A qualitative study about exclusive breastfeeding
among HIV positive mothers in Malawi. Afr J Reprod Health. 2010;14:213–222.
23. Thorsen VC, Sundby J, Martinson F. Potential initiators of HIV-related stigmatization: ethical and programmatic challenges for PMTCT programs. Dev World Bioeth. 2008;8:43–50.
24. dos Santos MML, Kruger P, Mellors SE, et al An exploratory survey measuring stigma and discrimination experienced by people living with HIV/AIDS
in South Africa: the People Living with HIV Stigma Index. BMC Public Health. 2014;14:80.
25. Sprague C, Chersich MF, Black V. Health system weaknesses constrain access to PMTCT and maternal HIV services in South Africa: a qualitative enquiry. AIDS Res Ther. 2011;8:10.
26. Pulerwitz J, Michaelis AP, Lippman SA, et al HIV-related stigma, service utilization, and status disclosure among truck drivers crossing the Southern borders in Brazil. AIDS Care. 2008;20:764–770.
27. Uwimana J, Zarowsky C, Hausler H, et al Engagement of non-government organisations and community care workers in collaborative TB/HIV activities including prevention of mother to child transmission in South Africa: opportunities and challenges. BMC Health Serv Res. 2012;12:233.
28. Desmond C, Bland RM, Boyce G, et al Scaling-up exclusive breastfeeding
support programmes: the example of KwaZulu-Natal. PLoS One. 2008;3:e2454.
29. Bland RM, Little K, Coovadia HM, et al Intervention to promote exclusive breast-feeding for the first 6 months of life in a high HIV prevalence area. AIDS. 2008;22:883–891.
30. Tylleskar T, Jackson D, Meda N, et al for POMISE-EBF Study Group. Exclusive breastfeeding
promotion by peer counsellors in sub-Saharan Africa (PROMISE-EBF): a cluster-randomised trial. Lancet. 2011;378:420–427.
31. O'Laughlin KN, Wyatt MA, Kaaya S, et al How treatment partners help: social analysis of an African adherence support intervention. AIDS Behav. 2012;16:1308–1315.
32. Ware NC, Idoko J, Kaaya S, et al Explaining adherence success in sub-Saharan Africa: an ethnographic study. Plos Med. 2009;6:e1000011.
33. Medley A, Garcia-Moreno C, Mc Gill S, et al Rates, barriers and outcomes of HIV serostatus disclosure among women in developing countries: implications for prevention of mother-to-child transmission programmes. Bull World Health Organ. 2004;82:299–307.
34. Shroufi A, Marara E, Saint-Sauveur JF, et al Mother to Mother peer support for women in prevention of mother to child transmission (PMTCT) programmes: a qualitative study. PLoS One. 2013;8:e64717.
35. Andersen J, Dana N, Hepfer B, et al Feeding buddies. A strategy to support safe infant feeding for HIV-positive mothers. J Hum Lact. 2013;29:90–93.
36. WHO/UNICEF. Infant and Young Child Feeding Counselling: An Integrated Course. Geneva, Switzerland: World Health Organization; 2006.
37. Unge C, Sodergard B, Marrone G, et al Long-term adherence to antiretroviral treatment
and program drop-out in a high-risk urban setting in sub-Saharan Africa: a prospective cohort study. PLoS One. 2010;5:e13613.